[soundcloud] Fix Soundcloud test case

* no space after @ in decorator

.github/workflows/ci.yml

@@ -1,73 +1,468 @@
 name: CI
-on: [push, pull_request]
+
+env:
+  all-cpython-versions: 2.6, 2.7, 3.2, 3.3, 3.4, 3.5, 3.6, 3.7, 3.8, 3.9, 3.10, 3.11, 3.12
+  main-cpython-versions: 2.7, 3.2, 3.5, 3.9, 3.11
+  pypy-versions: pypy-2.7, pypy-3.6, pypy-3.7
+  cpython-versions: main
+  test-set: core
+  # Python beta version to be built using pyenv before setup-python support
+  # Must also be included in all-cpython-versions 
+  next: 3.13
+
+on:
+  push:
+    # push inputs aren't known to GitHub
+    inputs:
+      cpython-versions:
+        type: string
+        default: all
+      test-set:
+        type: string
+        default: core
+  pull_request:
+    # pull_request inputs aren't known to GitHub
+    inputs:
+      cpython-versions:
+        type: string
+        default: main
+      test-set:
+        type: string
+        default: both
+  workflow_dispatch:
+    inputs:
+      cpython-versions:
+        type: choice
+        description: CPython versions (main = 2.7, 3.2, 3.5, 3.9, 3.11)
+        options:
+          - all
+          - main
+        required: true
+        default: main
+      test-set:
+        type: choice
+        description: core, download
+        options:
+          - both
+          - core
+          - download
+        required: true
+        default: both
+
+permissions:
+  contents: read
+
 jobs:
+  select:
+    name: Select tests from inputs
+    runs-on: ubuntu-latest
+    outputs:
+      cpython-versions: ${{ steps.run.outputs.cpython-versions }}
+      test-set: ${{ steps.run.outputs.test-set }}
+      own-pip-versions: ${{ steps.run.outputs.own-pip-versions }}
+    steps:
+    # push and pull_request inputs aren't known to GitHub (pt3)
+    - name: Set push defaults
+      if: ${{ github.event_name == 'push' }}
+      env:
+        cpython-versions: all
+        test-set: core
+      run: |
+        echo "cpython-versions=${{env.cpython-versions}}" >> "$GITHUB_ENV"
+        echo "test_set=${{env.test_set}}" >> "$GITHUB_ENV"
+    - name: Get pull_request inputs
+      if: ${{ github.event_name == 'pull_request' }}
+      env:
+        cpython-versions: main
+        test-set: both
+      run: |
+        echo "cpython-versions=${{env.cpython-versions}}" >> "$GITHUB_ENV"
+        echo "test_set=${{env.test_set}}" >> "$GITHUB_ENV"
+    - name: Make version array
+      id: run
+      run: |
+        # Make a JSON Array from comma/space-separated string (no extra escaping)
+        json_list() { \
+          ret=""; IFS="${IFS},"; set -- $*; \
+          for a in "$@"; do \
+            ret=$(printf '%s"%s"' "${ret}${ret:+, }" "$a"); \
+          done; \
+          printf '[%s]' "$ret"; }
+        tests="${{ inputs.test-set || env.test-set }}"
+        [ $tests = both ] && tests="core download"
+        printf 'test-set=%s\n' "$(json_list $tests)" >> "$GITHUB_OUTPUT"
+        versions="${{ inputs.cpython-versions || env.cpython-versions }}"
+        if [ "$versions" = all ]; then \
+          versions="${{ env.all-cpython-versions }}"; else \
+          versions="${{ env.main-cpython-versions }}"; \
+        fi
+        printf 'cpython-versions=%s\n' \
+          "$(json_list ${versions}${versions:+, }${{ env.pypy-versions }})" >> "$GITHUB_OUTPUT"
+        # versions with a special get-pip.py in a per-version subdirectory
+        printf 'own-pip-versions=%s\n' \
+          "$(json_list 2.6, 2.7, 3.2, 3.3, 3.4, 3.5, 3.6)" >> "$GITHUB_OUTPUT"
+
   tests:
-    name: Tests
+    name: Run tests
+    needs: select
+    permissions:
+      contents: read
+      packages: write
     runs-on: ${{ matrix.os }}
+    env:
+      PIP: python -m pip
+      PIP_DISABLE_PIP_VERSION_CHECK: true
+      PIP_NO_PYTHON_VERSION_WARNING: true
     strategy:
       fail-fast: true
       matrix:
         os: [ubuntu-20.04]
-        # TODO: python 2.6
-        # TODO: restore support for 3.3, 3.4
-        python-version: [2.7, 3.5, 3.6, 3.7, 3.8, 3.9, pypy-2.7, pypy-3.6, pypy-3.7]
+        python-version: ${{ fromJSON(needs.select.outputs.cpython-versions) }}
         python-impl: [cpython]
-        ytdl-test-set: [core, download]
+        ytdl-test-set: ${{ fromJSON(needs.select.outputs.test-set) }}
         run-tests-ext: [sh]
         include:
-        # python 3.2 is only available on windows via setup-python
         - os: windows-2019
-          python-version: 3.2
+          python-version: 3.4
           python-impl: cpython
-          ytdl-test-set: core
+          ytdl-test-set: ${{ contains(needs.select.outputs.test-set, 'core') && 'core' || 'nocore' }}
           run-tests-ext: bat
         - os: windows-2019
-          python-version: 3.2
+          python-version: 3.4
           python-impl: cpython
-          ytdl-test-set: download
+          ytdl-test-set: ${{ contains(needs.select.outputs.test-set, 'download') && 'download'  || 'nodownload' }}
           run-tests-ext: bat
         # jython
         - os: ubuntu-20.04
+          python-version: 2.7
           python-impl: jython
-          ytdl-test-set: core
+          ytdl-test-set: ${{ contains(needs.select.outputs.test-set, 'core') && 'core' || 'nocore' }}
           run-tests-ext: sh
         - os: ubuntu-20.04
+          python-version: 2.7
           python-impl: jython
-          ytdl-test-set: download
+          ytdl-test-set: ${{ contains(needs.select.outputs.test-set, 'download') && 'download'  || 'nodownload' }}
           run-tests-ext: sh
     steps:
-    - uses: actions/checkout@v3
+    - name: Prepare Linux
+      if: ${{ startswith(matrix.os, 'ubuntu') }}
+      shell: bash
+      run: |
+        # apt in runner, if needed, may not be up-to-date
+        sudo apt-get update
+    - name: Checkout
+      uses: actions/checkout@v3
+    #-------- Python 3 -----
     - name: Set up supported Python ${{ matrix.python-version }}
-      uses: actions/setup-python@v4
-      if: ${{ matrix.python-impl == 'cpython' && ! contains(fromJSON('["3.3", "3.4"]'), matrix.python-version) }}
+      id: setup-python
+      if: ${{ matrix.python-impl == 'cpython' && matrix.python-version != '2.6' && matrix.python-version != '2.7' && matrix.python-version != env.next }}
+      # wrap broken actions/setup-python@v4
+      # NB may run apt-get install in Linux
+      uses: ytdl-org/setup-python@v1
+      env:
+        # Temporary workaround for Python 3.5 failures - May 2024
+        PIP_TRUSTED_HOST: "pypi.python.org pypi.org files.pythonhosted.org"
       with:
         python-version: ${{ matrix.python-version }}
+        cache-build: true
+        allow-build: info
+    - name: Locate supported Python ${{ matrix.python-version }}
+      if: ${{ env.pythonLocation }}
+      shell: bash
+      run: |
+        echo "PYTHONHOME=${pythonLocation}" >> "$GITHUB_ENV"
+        export expected="${{ steps.setup-python.outputs.python-path }}"
+        dirname() { printf '%s\n' \
+            'import os, sys' \
+            'print(os.path.dirname(sys.argv[1]))' \
+            | ${expected} - "$1"; }
+        expd="$(dirname "$expected")"
+        export python="$(command -v python)"
+        [ "$expd" = "$(dirname "$python")" ] || echo "PATH=$expd:${PATH}" >> "$GITHUB_ENV"
+        [ -x "$python" ] || printf '%s\n' \
+            'import os' \
+            'exp = os.environ["expected"]' \
+            'python = os.environ["python"]' \
+            'exps = os.path.split(exp)' \
+            'if python and (os.path.dirname(python) == exp[0]):' \
+            '    exit(0)' \
+            'exps[1] = "python" + os.path.splitext(exps[1])[1]' \
+            'python = os.path.join(*exps)' \
+            'try:' \
+            '    os.symlink(exp, python)' \
+            'except AttributeError:' \
+            '    os.rename(exp, python)' \
+            | ${expected} -
+        printf '%s\n' \
+            'import sys' \
+            'print(sys.path)' \
+            | ${expected} -
+    #-------- Python next (was 3.12) -
+    - name: Set up CPython 3.next environment
+      if: ${{ matrix.python-impl == 'cpython' && matrix.python-version == env.next }}
+      shell: bash
+      run: |
+        PYENV_ROOT=$HOME/.local/share/pyenv
+        echo "PYENV_ROOT=${PYENV_ROOT}" >> "$GITHUB_ENV"
+    - name: Cache Python 3.next 
+      id: cachenext
+      if: ${{ matrix.python-impl == 'cpython' && matrix.python-version == env.next }}
+      uses: actions/cache@v3
+      with:
+        key: python-${{ env.next }}
+        path: |
+          ${{ env.PYENV_ROOT }}
+    - name: Build and set up Python 3.next
+      if: ${{ matrix.python-impl == 'cpython' && matrix.python-version == env.next && ! steps.cachenext.outputs.cache-hit }}
+      # dl and build locally
+      shell: bash
+      run: |
+        # Install build environment
+        sudo apt-get install -y build-essential llvm libssl-dev tk-dev  \
+                      libncursesw5-dev libreadline-dev libsqlite3-dev   \
+                      libffi-dev xz-utils zlib1g-dev libbz2-dev liblzma-dev
+        # Download PyEnv from its GitHub repository.
+        export PYENV_ROOT=${{ env.PYENV_ROOT }}
+        export PATH=$PYENV_ROOT/bin:$PATH
+        git clone "https://github.com/pyenv/pyenv.git" "$PYENV_ROOT"
+        pyenv install ${{ env.next }}
+    - name: Locate Python 3.next
+      if: ${{ matrix.python-impl == 'cpython' && matrix.python-version == env.next }}
+      shell: bash
+      run: |
+        PYTHONHOME="$(echo "${{ env.PYENV_ROOT }}/versions/${{ env.next }}."*)"
+        test -n "$PYTHONHOME"
+        echo "PYTHONHOME=$PYTHONHOME" >> "$GITHUB_ENV"
+        echo "PATH=${PYTHONHOME}/bin:$PATH" >> "$GITHUB_ENV"
+    #-------- Python 2.7 --
+    - name: Set up Python 2.7
+      if: ${{ matrix.python-impl == 'cpython' && matrix.python-version == '2.7' }}
+      # install 2.7
+      shell: bash
+      run: |
+        sudo apt-get install -y python2 python-is-python2
+        echo "PYTHONHOME=/usr" >> "$GITHUB_ENV"
+    #-------- Python 2.6 --
+    - name: Set up Python 2.6 environment
+      if: ${{ matrix.python-impl == 'cpython' && matrix.python-version == '2.6' }}
+      shell: bash
+      run: |
+        openssl_name=openssl-1.0.2u
+        echo "openssl_name=${openssl_name}" >> "$GITHUB_ENV"
+        openssl_dir=$HOME/.local/opt/$openssl_name
+        echo "openssl_dir=${openssl_dir}" >> "$GITHUB_ENV"
+        PYENV_ROOT=$HOME/.local/share/pyenv
+        echo "PYENV_ROOT=${PYENV_ROOT}" >> "$GITHUB_ENV"
+        sudo apt-get install -y openssl ca-certificates
+    - name: Cache Python 2.6
+      id: cache26
+      if: ${{ matrix.python-version == '2.6' }}
+      uses: actions/cache@v3
+      with:
+        key: python-2.6.9
+        path: |
+          ${{ env.openssl_dir }}
+          ${{ env.PYENV_ROOT }}
+    - name: Build and set up Python 2.6
+      if: ${{ matrix.python-impl == 'cpython' && matrix.python-version == '2.6' && ! steps.cache26.outputs.cache-hit }}
+      # dl and build locally
+      shell: bash
+      run: |
+        # Install build environment
+        sudo apt-get install -y build-essential llvm libssl-dev tk-dev  \
+                      libncursesw5-dev libreadline-dev libsqlite3-dev   \
+                      libffi-dev xz-utils zlib1g-dev libbz2-dev liblzma-dev
+        # Download and install OpenSSL 1.0.2, back in time
+        openssl_name=${{ env.openssl_name }}
+        openssl_targz=${openssl_name}.tar.gz
+        openssl_dir=${{ env.openssl_dir }}
+        openssl_inc=$openssl_dir/include
+        openssl_lib=$openssl_dir/lib
+        openssl_ssl=$openssl_dir/ssl
+        curl -L "https://www.openssl.org/source/$openssl_targz" -o $openssl_targz
+        tar -xf $openssl_targz
+        ( cd $openssl_name; \
+          ./config --prefix=$openssl_dir --openssldir=${openssl_dir}/ssl \
+            --libdir=lib -Wl,-rpath=${openssl_dir}/lib shared zlib-dynamic && \
+          make && \
+          make install )
+        rm -rf $openssl_name
+        rmdir $openssl_ssl/certs && ln -s /etc/ssl/certs $openssl_ssl/certs
+        # Download PyEnv from its GitHub repository.
+        export PYENV_ROOT=${{ env.PYENV_ROOT }}
+        export PATH=$PYENV_ROOT/bin:$PATH
+        git clone "https://github.com/pyenv/pyenv.git" "$PYENV_ROOT"
+        # Prevent pyenv build trying (and failing) to update pip
+        export GET_PIP=get-pip-2.6.py
+        echo 'import sys; sys.exit(0)' > ${GET_PIP}
+        GET_PIP=$(realpath $GET_PIP)
+        # Build and install Python
+        export CFLAGS="-I$openssl_inc"
+        export LDFLAGS="-L$openssl_lib"
+        export LD_LIBRARY_PATH="$openssl_lib"
+        pyenv install 2.6.9
+    - name: Locate Python 2.6
+      if: ${{ matrix.python-impl == 'cpython' && matrix.python-version == '2.6' }}
+      shell: bash
+      run: |
+        PYTHONHOME="${{ env.PYENV_ROOT }}/versions/2.6.9"
+        echo "PYTHONHOME=$PYTHONHOME" >> "$GITHUB_ENV"
+        echo "PATH=${PYTHONHOME}/bin:$PATH" >> "$GITHUB_ENV"
+        echo "LD_LIBRARY_PATH=${{ env.openssl_dir }}/lib${LD_LIBRARY_PATH:+:}${LD_LIBRARY_PATH}" >> "$GITHUB_ENV"
+    #-------- Jython ------
     - name: Set up Java 8
       if: ${{ matrix.python-impl == 'jython' }}
-      uses: actions/setup-java@v2
+      uses: actions/setup-java@v3
       with:
         java-version: 8
         distribution: 'zulu'
+    - name: Setup Jython environment
+      if: ${{ matrix.python-impl == 'jython' }}
+      shell: bash
+      run: |
+        echo "JYTHON_ROOT=${HOME}/jython" >> "$GITHUB_ENV"
+        echo "PIP=pip" >> "$GITHUB_ENV"
+    - name: Cache Jython
+      id: cachejy
+      if: ${{ matrix.python-impl == 'jython' && matrix.python-version == '2.7' }}
+      uses: actions/cache@v3
+      with:
+        # 2.7.3 now available, may solve SNI issue
+        key: jython-2.7.1
+        path: |
+          ${{ env.JYTHON_ROOT }}
     - name: Install Jython
-      if: ${{ matrix.python-impl == 'jython' }}
+      if: ${{ matrix.python-impl == 'jython' && matrix.python-version == '2.7' && ! steps.cachejy.outputs.cache-hit }}
+      shell: bash
       run: |
-        wget https://repo1.maven.org/maven2/org/python/jython-installer/2.7.1/jython-installer-2.7.1.jar -O jython-installer.jar
-        java -jar jython-installer.jar -s -d "$HOME/jython"
-        echo "$HOME/jython/bin" >> $GITHUB_PATH
-    - name: Install nose
-      if: ${{ matrix.python-impl != 'jython' }}
-      run: pip install nose
-    - name: Install nose (Jython)
-      if: ${{ matrix.python-impl == 'jython' }}
-      # Working around deprecation of support for non-SNI clients at PyPI CDN (see https://status.python.org/incidents/hzmjhqsdjqgb)
+        JYTHON_ROOT="${{ env.JYTHON_ROOT }}"
+        curl -L "https://repo1.maven.org/maven2/org/python/jython-installer/2.7.1/jython-installer-2.7.1.jar" -o jython-installer.jar
+        java -jar jython-installer.jar -s -d "${JYTHON_ROOT}"
+        echo "${JYTHON_ROOT}/bin" >> "$GITHUB_PATH"
+    - name: Set up cached Jython
+      if: ${{ steps.cachejy.outputs.cache-hit }}
+      shell: bash
       run: |
-        wget https://files.pythonhosted.org/packages/99/4f/13fb671119e65c4dce97c60e67d3fd9e6f7f809f2b307e2611f4701205cb/nose-1.3.7-py2-none-any.whl
-        pip install nose-1.3.7-py2-none-any.whl
+        JYTHON_ROOT="${{ env.JYTHON_ROOT }}"
+        echo "${JYTHON_ROOT}/bin" >> $GITHUB_PATH
+    - name: Install supporting Python 2.7 if possible
+      if: ${{ steps.cachejy.outputs.cache-hit }}
+      shell: bash
+      run: |
+        sudo apt-get install -y python2.7 || true
+    #-------- pip ---------
+    - name: Set up supported Python ${{ matrix.python-version }} pip
+      if: ${{ (matrix.python-version != '3.2' && steps.setup-python.outputs.python-path) || matrix.python-version == '2.7' }}
+      # This step may run in either Linux or Windows
+      shell: bash
+      run: |
+        echo "$PATH"
+        echo "$PYTHONHOME"
+        # curl is available on both Windows and Linux, -L follows redirects, -O gets name
+        python -m ensurepip || python -m pip --version || { \
+          get_pip="${{ contains(needs.select.outputs.own-pip-versions, matrix.python-version) && format('{0}/', matrix.python-version) || '' }}"; \
+          curl -L -O "https://bootstrap.pypa.io/pip/${get_pip}get-pip.py"; \
+          python get-pip.py; }
+    - name: Set up Python 2.6 pip
+      if: ${{ matrix.python-version == '2.6' }}
+      shell: bash
+      run: |
+        python -m pip --version || { \
+          curl -L -O "https://bootstrap.pypa.io/pip/2.6/get-pip.py"; \
+          curl -L -O "https://files.pythonhosted.org/packages/ac/95/a05b56bb975efa78d3557efa36acaf9cf5d2fd0ee0062060493687432e03/pip-9.0.3-py2.py3-none-any.whl"; \
+          python get-pip.py --no-setuptools --no-wheel pip-9.0.3-py2.py3-none-any.whl; }
+        # work-around to invoke pip module on 2.6: https://bugs.python.org/issue2751
+        echo "PIP=python -m pip.__main__" >> "$GITHUB_ENV"
+    - name: Set up other Python ${{ matrix.python-version }} pip
+      if: ${{ matrix.python-version == '3.2' && steps.setup-python.outputs.python-path }}
+      shell: bash
+      run: |
+        python -m pip --version || { \
+          curl -L -O "https://bootstrap.pypa.io/pip/3.2/get-pip.py"; \
+          curl -L -O "https://files.pythonhosted.org/packages/b2/d0/cd115fe345dd6f07ec1c780020a7dfe74966fceeb171e0f20d1d4905b0b7/pip-7.1.2-py2.py3-none-any.whl"; \
+          python get-pip.py --no-setuptools --no-wheel pip-7.1.2-py2.py3-none-any.whl; }
+    #-------- unittest ----
+    - name: Upgrade Unittest for Python 2.6
+      if: ${{ matrix.python-version == '2.6' }}
+      shell: bash
+      run: |
+        # Work around deprecation of support for non-SNI clients at PyPI CDN (see https://status.python.org/incidents/hzmjhqsdjqgb)
+        $PIP -qq show unittest2 || { \
+          for u in "65/26/32b8464df2a97e6dd1b656ed26b2c194606c16fe163c695a992b36c11cdf/six-1.13.0-py2.py3-none-any.whl" \
+              "f2/94/3af39d34be01a24a6e65433d19e107099374224905f1e0cc6bbe1fd22a2f/argparse-1.4.0-py2.py3-none-any.whl" \
+              "c7/a3/c5da2a44c85bfbb6eebcfc1dde24933f8704441b98fdde6528f4831757a6/linecache2-1.0.0-py2.py3-none-any.whl" \
+              "17/0a/6ac05a3723017a967193456a2efa0aa9ac4b51456891af1e2353bb9de21e/traceback2-1.4.0-py2.py3-none-any.whl" \
+              "72/20/7f0f433060a962200b7272b8c12ba90ef5b903e218174301d0abfd523813/unittest2-1.1.0-py2.py3-none-any.whl"; do \
+            curl -L -O "https://files.pythonhosted.org/packages/${u}"; \
+            $PIP install ${u##*/}; \
+          done; }
+        # make tests use unittest2
+        for test in ./test/test_*.py ./test/helper.py; do
+          sed -r -i -e '/^import unittest$/s/test/test2 as unittest/' "$test"
+        done
+    #-------- nose --------
+    - name: Install nose for Python ${{ matrix.python-version }}
+      if: ${{ (matrix.python-version != '3.2' && steps.setup-python.outputs.python-path) || (matrix.python-impl == 'cpython' && (matrix.python-version == '2.7' || matrix.python-version == env.next)) }}
+      shell: bash
+      run: |
+        echo "$PATH"
+        echo "$PYTHONHOME"
+        # Use PyNose for recent Pythons instead of Nose
+        py3ver="${{ matrix.python-version }}"
+        py3ver=${py3ver#3.}
+        [ "$py3ver" != "${{ matrix.python-version }}" ] && py3ver=${py3ver%.*} || py3ver=0
+        [ "$py3ver" -ge 9 ] && nose=pynose || nose=nose
+        $PIP -qq show $nose || $PIP install $nose
+    - name: Install nose for other Python 2
+      if: ${{ matrix.python-impl == 'jython' || (matrix.python-impl == 'cpython' && matrix.python-version == '2.6') }}
+      shell: bash
+      run: |
+        # Work around deprecation of support for non-SNI clients at PyPI CDN (see https://status.python.org/incidents/hzmjhqsdjqgb)
+        $PIP -qq show nose || { \
+          curl -L -O "https://files.pythonhosted.org/packages/99/4f/13fb671119e65c4dce97c60e67d3fd9e6f7f809f2b307e2611f4701205cb/nose-1.3.7-py2-none-any.whl"; \
+          $PIP install nose-1.3.7-py2-none-any.whl; }
+    - name: Install nose for other Python 3
+      if: ${{ matrix.python-version == '3.2' && steps.setup-python.outputs.python-path }}
+      shell: bash
+      run: |
+        $PIP -qq show nose || { \
+          curl -L -O "https://files.pythonhosted.org/packages/15/d8/dd071918c040f50fa1cf80da16423af51ff8ce4a0f2399b7bf8de45ac3d9/nose-1.3.7-py3-none-any.whl"; \
+          $PIP install nose-1.3.7-py3-none-any.whl; }
+    - name: Set up nosetest test
+      if: ${{ contains(needs.select.outputs.test-set, matrix.ytdl-test-set ) }}
+      shell: bash
+      run: |
+        # set PYTHON_VER
+        PYTHON_VER=${{ matrix.python-version }}
+        [ "${PYTHON_VER#*-}" != "$PYTHON_VER" ] || PYTHON_VER="${{ matrix.python-impl }}-${PYTHON_VER}"
+        echo "PYTHON_VER=$PYTHON_VER" >> "$GITHUB_ENV"
+        echo "PYTHON_IMPL=${{ matrix.python-impl }}" >> "$GITHUB_ENV"
+        # define a test to validate the Python version used by nosetests
+        printf '%s\n' \
+          'from __future__ import unicode_literals' \
+          'import sys, os, platform' \
+          'try:' \
+          '    import unittest2 as unittest' \
+          'except ImportError:' \
+          '    import unittest' \
+          'class TestPython(unittest.TestCase):' \
+          '    def setUp(self):' \
+          '        self.ver = os.environ["PYTHON_VER"].split("-")' \
+          '    def test_python_ver(self):' \
+          '        self.assertEqual(["%d" % v for v in sys.version_info[:2]], self.ver[-1].split(".")[:2])' \
+          '        self.assertTrue(sys.version.startswith(self.ver[-1]))' \
+          '        self.assertIn(self.ver[0], ",".join((sys.version, platform.python_implementation())).lower())' \
+          '    def test_python_impl(self):' \
+          '        self.assertIn(platform.python_implementation().lower(), (os.environ["PYTHON_IMPL"], self.ver[0]))' \
+          > test/test_python.py
+    #-------- TESTS -------
     - name: Run tests
+      if: ${{ contains(needs.select.outputs.test-set, matrix.ytdl-test-set ) }}
       continue-on-error: ${{ matrix.ytdl-test-set == 'download' || matrix.python-impl == 'jython' }}
       env:
         YTDL_TEST_SET: ${{ matrix.ytdl-test-set }}
-      run: ./devscripts/run_tests.${{ matrix.run-tests-ext }}
+      run: |
+        ./devscripts/run_tests.${{ matrix.run-tests-ext }}
   flake8:
     name: Linter
     runs-on: ubuntu-latest
@@ -81,3 +476,4 @@ jobs:
       run: pip install flake8
     - name: Run flake8
       run: flake8 .
+

README.md

@@ -33,7 +33,7 @@ Windows users can [download an .exe file](https://yt-dl.org/latest/youtube-dl.ex
 You can also use pip:
 
     sudo -H pip install --upgrade youtube-dl
-    
+
 This command will update youtube-dl if you have already installed it. See the [pypi page](https://pypi.python.org/pypi/youtube_dl) for more information.
 
 macOS users can install youtube-dl with [Homebrew](https://brew.sh/):
@@ -563,7 +563,7 @@ The basic usage is not to set any template arguments when downloading a single f
  - `is_live` (boolean): Whether this video is a live stream or a fixed-length video
  - `start_time` (numeric): Time in seconds where the reproduction should start, as specified in the URL
  - `end_time` (numeric): Time in seconds where the reproduction should end, as specified in the URL
- - `format` (string): A human-readable description of the format 
+ - `format` (string): A human-readable description of the format
  - `format_id` (string): Format code specified by `--format`
  - `format_note` (string): Additional info about the format
  - `width` (numeric): Width of the video
@@ -675,7 +675,7 @@ The general syntax for format selection is `--format FORMAT` or shorter `-f FORM
 
 **tl;dr:** [navigate me to examples](#format-selection-examples).
 
-The simplest case is requesting a specific format, for example with `-f 22` you can download the format with format code equal to 22. You can get the list of available format codes for particular video using `--list-formats` or `-F`. Note that these format codes are extractor specific. 
+The simplest case is requesting a specific format, for example with `-f 22` you can download the format with format code equal to 22. You can get the list of available format codes for particular video using `--list-formats` or `-F`. Note that these format codes are extractor specific.
 
 You can also use a file extension (currently `3gp`, `aac`, `flv`, `m4a`, `mp3`, `mp4`, `ogg`, `wav`, `webm` are supported) to download the best quality format of a particular file extension served as a single file, e.g. `-f webm` will download the best quality format with the `webm` extension served as a single file.
 
@@ -760,7 +760,7 @@ Videos can be filtered by their upload date using the options `--date`, `--dateb
 
  - Absolute dates: Dates in the format `YYYYMMDD`.
  - Relative dates: Dates in the format `(now|today)[+-][0-9](day|week|month|year)(s)?`
- 
+
 Examples:
 
 ```bash
@@ -1000,6 +1000,8 @@ To run the test, simply invoke your favorite test runner, or execute a test file
     python test/test_download.py
     nosetests
 
+For Python versions 3.6 and later, you can use [pynose](https://pypi.org/project/pynose/) to implement `nosetests`. The original [nose](https://pypi.org/project/nose/) has not been upgraded for 3.10 and later.
+
 See item 6 of [new extractor tutorial](#adding-support-for-a-new-site) for how to run extractor specific test cases.
 
 If you want to create a build of youtube-dl yourself, you'll need
@@ -1091,7 +1093,7 @@ In any case, thank you very much for your contributions!
 
 ## youtube-dl coding conventions
 
-This section introduces a guide lines for writing idiomatic, robust and future-proof extractor code.
+This section introduces guidelines for writing idiomatic, robust and future-proof extractor code.
 
 Extractors are very fragile by nature since they depend on the layout of the source data provided by 3rd party media hosters out of your control and this layout tends to change. As an extractor implementer your task is not only to write code that will extract media links and metadata correctly but also to minimize dependency on the source's layout and even to make the code foresee potential future changes and be ready for that. This is important because it will allow the extractor not to break on minor layout changes thus keeping old youtube-dl versions working. Even though this breakage issue is easily fixed by emitting a new version of youtube-dl with a fix incorporated, all the previous versions become broken in all repositories and distros' packages that may not be so prompt in fetching the update from us. Needless to say, some non rolling release distros may never receive an update at all.
 
@@ -1114,7 +1116,7 @@ Say you have some source dictionary `meta` that you've fetched as JSON with HTTP
 ```python
 meta = self._download_json(url, video_id)
 ```
-    
+
 Assume at this point `meta`'s layout is:
 
 ```python
@@ -1158,7 +1160,7 @@ description = self._search_regex(
 ```
 
 On failure this code will silently continue the extraction with `description` set to `None`. That is useful for metafields that may or may not be present.
- 
+
 ### Provide fallbacks
 
 When extracting metadata try to do so from multiple sources. For example if `title` is present in several places, try extracting from at least some of them. This makes it more future-proof in case some of the sources become unavailable.
@@ -1206,7 +1208,7 @@ r'(id|ID)=(?P<id>\d+)'
 #### Make regular expressions relaxed and flexible
 
 When using regular expressions try to write them fuzzy, relaxed and flexible, skipping insignificant parts that are more likely to change, allowing both single and double quotes for quoted values and so on.
- 
+
 ##### Example
 
 Say you need to extract `title` from the following HTML code:
@@ -1230,7 +1232,7 @@ title = self._search_regex(
     webpage, 'title', group='title')
 ```
 
-Note how you tolerate potential changes in the `style` attribute's value or switch from using double quotes to single for `class` attribute: 
+Note how you tolerate potential changes in the `style` attribute's value or switch from using double quotes to single for `class` attribute:
 
 The code definitely should not look like:
 
@@ -1331,27 +1333,114 @@ Wrap all extracted numeric data into safe functions from [`youtube_dl/utils.py`]
 
 Use `url_or_none` for safe URL processing.
 
-Use `try_get` for safe metadata extraction from parsed JSON.
+Use `traverse_obj` for safe metadata extraction from parsed JSON.
 
-Use `unified_strdate` for uniform `upload_date` or any `YYYYMMDD` meta field extraction, `unified_timestamp` for uniform `timestamp` extraction, `parse_filesize` for `filesize` extraction, `parse_count` for count meta fields extraction, `parse_resolution`, `parse_duration` for `duration` extraction, `parse_age_limit` for `age_limit` extraction. 
+Use `unified_strdate` for uniform `upload_date` or any `YYYYMMDD` meta field extraction, `unified_timestamp` for uniform `timestamp` extraction, `parse_filesize` for `filesize` extraction, `parse_count` for count meta fields extraction, `parse_resolution`, `parse_duration` for `duration` extraction, `parse_age_limit` for `age_limit` extraction.
 
 Explore [`youtube_dl/utils.py`](https://github.com/ytdl-org/youtube-dl/blob/master/youtube_dl/utils.py) for more useful convenience functions.
 
 #### More examples
 
 ##### Safely extract optional description from parsed JSON
+
+When processing complex JSON, as often returned by site API requests or stashed in web pages for "hydration", you can use the `traverse_obj()` utility function to handle multiple fallback values and to ensure the expected type of metadata items. The function's docstring defines how the function works: also review usage in the codebase for more examples.
+
+In this example, a text `description`, or `None`, is pulled from the `.result.video[0].summary` member of the parsed JSON `response`, if available.
+
+```python
+description = traverse_obj(response, ('result', 'video', 0, 'summary', T(compat_str)))
+```
+`T(...)` is a shorthand for a set literal; if you hate people who still run Python 2.6, `T(type_or_transformation)` could be written as a set literal `{type_or_transformation}`.
+
+Some extractors use the older and less capable `try_get()` function in the same way.
+
 ```python
 description = try_get(response, lambda x: x['result']['video'][0]['summary'], compat_str)
 ```
 
 ##### Safely extract more optional metadata
+
+In this example, various optional metadata values are extracted from the `.result.video[0]` member of the parsed JSON `response`, which is expected to be a JS object, parsed into a `dict`, with no crash if that isn't so, or if any of the target values are missing or invalid.
+
 ```python
-video = try_get(response, lambda x: x['result']['video'][0], dict) or {}
+video = traverse_obj(response, ('result', 'video', 0, T(dict))) or {}
+# formerly:
+# video = try_get(response, lambda x: x['result']['video'][0], dict) or {}
 description = video.get('summary')
 duration = float_or_none(video.get('durationMs'), scale=1000)
 view_count = int_or_none(video.get('views'))
 ```
 
+#### Safely extract nested lists
+
+Suppose you've extracted JSON like this into a Python data structure named `media_json` using, say, the `_download_json()` or `_parse_json()` methods of `InfoExtractor`:
+```json
+{
+    "title": "Example video",
+    "comment": "try extracting this",
+    "media": [{
+        "type": "bad",
+        "size": 320,
+        "url": "https://some.cdn.site/bad.mp4"
+    }, {
+        "type": "streaming",
+        "url": "https://some.cdn.site/hls.m3u8"
+    }, {
+        "type": "super",
+        "size": 1280,
+        "url": "https://some.cdn.site/good.webm"
+    }],
+    "moreStuff": "more values",
+    ...
+}
+```
+
+Then extractor code like this can collect the various fields of the JSON:
+```python
+...
+from ..utils import (
+    determine_ext,
+    int_or_none,
+    T,
+    traverse_obj,
+    txt_or_none,
+    url_or_none,
+)
+...
+        ...
+        info_dict = {}
+        # extract title and description if valid and not empty
+        info_dict.update(traverse_obj(media_json, {
+            'title': ('title', T(txt_or_none)),
+            'description': ('comment', T(txt_or_none)),
+        }))
+
+        # extract any recognisable media formats
+        fmts = []
+        # traverse into "media" list, extract `dict`s with desired keys
+        for fmt in traverse_obj(media_json, ('media', Ellipsis, {
+                'format_id': ('type', T(txt_or_none)),
+                'url': ('url', T(url_or_none)),
+                'width': ('size', T(int_or_none)), })):
+            # bad `fmt` values were `None` and removed
+            if 'url' not in fmt:
+                continue
+            fmt_url = fmt['url']  # known to be valid URL
+            ext = determine_ext(fmt_url)
+            if ext == 'm3u8':
+                fmts.extend(self._extract_m3u8_formats(fmt_url, video_id, 'mp4', fatal=False))
+            else:
+                fmt['ext'] = ext
+                fmts.append(fmt)
+
+        # sort, raise if no formats
+        self._sort_formats(fmts)
+
+        info_dict['formats'] = fmts
+        ...
+```
+The extractor raises an exception rather than random crashes if the JSON structure changes so that no formats are found.
+
 # EMBEDDING YOUTUBE-DL
 
 youtube-dl makes the best effort to be a good command-line program, and thus should be callable from any programming language. If you encounter any problems parsing its output, feel free to [create a report](https://github.com/ytdl-org/youtube-dl/issues/new).

devscripts/__init__.py

@@ -0,0 +1 @@
+# Empty file needed to make devscripts.utils properly importable from outside

devscripts/bash-completion.py

@@ -5,8 +5,12 @@ import os
 from os.path import dirname as dirn
 import sys
 
-sys.path.insert(0, dirn(dirn((os.path.abspath(__file__)))))
+sys.path.insert(0, dirn(dirn(os.path.abspath(__file__))))
+
 import youtube_dl
+from youtube_dl.compat import compat_open as open
+
+from utils import read_file
 
 BASH_COMPLETION_FILE = "youtube-dl.bash-completion"
 BASH_COMPLETION_TEMPLATE = "devscripts/bash-completion.in"
@@ -18,9 +22,8 @@ def build_completion(opt_parser):
         for option in group.option_list:
             # for every long flag
             opts_flag.append(option.get_opt_string())
-    with open(BASH_COMPLETION_TEMPLATE) as f:
-        template = f.read()
-    with open(BASH_COMPLETION_FILE, "w") as f:
+    template = read_file(BASH_COMPLETION_TEMPLATE)
+    with open(BASH_COMPLETION_FILE, "w", encoding='utf-8') as f:
         # just using the special char
         filled_template = template.replace("{{flags}}", " ".join(opts_flag))
         f.write(filled_template)

devscripts/create-github-release.py

@@ -1,7 +1,6 @@
 #!/usr/bin/env python
 from __future__ import unicode_literals
 
-import io
 import json
 import mimetypes
 import netrc
@@ -10,7 +9,9 @@ import os
 import re
 import sys
 
-sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
+dirn = os.path.dirname
+
+sys.path.insert(0, dirn(dirn(os.path.abspath(__file__))))
 
 from youtube_dl.compat import (
     compat_basestring,
@@ -22,6 +23,7 @@ from youtube_dl.utils import (
     make_HTTPS_handler,
     sanitized_Request,
 )
+from utils import read_file
 
 
 class GitHubReleaser(object):
@@ -89,8 +91,7 @@ def main():
 
     changelog_file, version, build_path = args
 
-    with io.open(changelog_file, encoding='utf-8') as inf:
-        changelog = inf.read()
+    changelog = read_file(changelog_file)
 
     mobj = re.search(r'(?s)version %s\n{2}(.+?)\n{3}' % version, changelog)
     body = mobj.group(1) if mobj else ''

devscripts/fish-completion.py

@@ -6,10 +6,13 @@ import os
 from os.path import dirname as dirn
 import sys
 
-sys.path.insert(0, dirn(dirn((os.path.abspath(__file__)))))
+sys.path.insert(0, dirn(dirn(os.path.abspath(__file__))))
+
 import youtube_dl
 from youtube_dl.utils import shell_quote
 
+from utils import read_file, write_file
+
 FISH_COMPLETION_FILE = 'youtube-dl.fish'
 FISH_COMPLETION_TEMPLATE = 'devscripts/fish-completion.in'
 
@@ -38,11 +41,9 @@ def build_completion(opt_parser):
             complete_cmd.extend(EXTRA_ARGS.get(long_option, []))
             commands.append(shell_quote(complete_cmd))
 
-    with open(FISH_COMPLETION_TEMPLATE) as f:
-        template = f.read()
+    template = read_file(FISH_COMPLETION_TEMPLATE)
     filled_template = template.replace('{{commands}}', '\n'.join(commands))
-    with open(FISH_COMPLETION_FILE, 'w') as f:
-        f.write(filled_template)
+    write_file(FISH_COMPLETION_FILE, filled_template)
 
 
 parser = youtube_dl.parseOpts()[0]

devscripts/gh-pages/add-version.py

@@ -6,16 +6,21 @@ import sys
 import hashlib
 import os.path
 
+dirn = os.path.dirname
+
+sys.path.insert(0, dirn(dirn(dirn(os.path.abspath(__file__)))))
+
+from devscripts.utils import read_file, write_file
+from youtube_dl.compat import compat_open as open
 
 if len(sys.argv) <= 1:
     print('Specify the version number as parameter')
     sys.exit()
 version = sys.argv[1]
 
-with open('update/LATEST_VERSION', 'w') as f:
-    f.write(version)
+write_file('update/LATEST_VERSION', version)
 
-versions_info = json.load(open('update/versions.json'))
+versions_info = json.loads(read_file('update/versions.json'))
 if 'signature' in versions_info:
     del versions_info['signature']
 
@@ -39,5 +44,5 @@ for key, filename in filenames.items():
 versions_info['versions'][version] = new_version
 versions_info['latest'] = version
 
-with open('update/versions.json', 'w') as jsonf:
-    json.dump(versions_info, jsonf, indent=4, sort_keys=True)
+with open('update/versions.json', 'w', encoding='utf-8') as jsonf:
+    json.dumps(versions_info, jsonf, indent=4, sort_keys=True)

devscripts/gh-pages/generate-download.py

@@ -2,14 +2,21 @@
 from __future__ import unicode_literals
 
 import json
+import os.path
+import sys
 
-versions_info = json.load(open('update/versions.json'))
+dirn = os.path.dirname
+
+sys.path.insert(0, dirn(dirn((os.path.abspath(__file__)))))
+
+from utils import read_file, write_file
+
+versions_info = json.loads(read_file('update/versions.json'))
 version = versions_info['latest']
 version_dict = versions_info['versions'][version]
 
 # Read template page
-with open('download.html.in', 'r', encoding='utf-8') as tmplf:
-    template = tmplf.read()
+template = read_file('download.html.in')
 
 template = template.replace('@PROGRAM_VERSION@', version)
 template = template.replace('@PROGRAM_URL@', version_dict['bin'][0])
@@ -18,5 +25,5 @@ template = template.replace('@EXE_URL@', version_dict['exe'][0])
 template = template.replace('@EXE_SHA256SUM@', version_dict['exe'][1])
 template = template.replace('@TAR_URL@', version_dict['tar'][0])
 template = template.replace('@TAR_SHA256SUM@', version_dict['tar'][1])
-with open('download.html', 'w', encoding='utf-8') as dlf:
-    dlf.write(template)
+
+write_file('download.html', template)

devscripts/gh-pages/update-copyright.py

@@ -5,17 +5,22 @@ from __future__ import with_statement, unicode_literals
 
 import datetime
 import glob
-import io  # For Python 2 compatibility
 import os
 import re
+import sys
 
-year = str(datetime.datetime.now().year)
+dirn = os.path.dirname
+
+sys.path.insert(0, dirn(dirn(dirn(os.path.abspath(__file__)))))
+
+from devscripts.utils import read_file, write_file
+from youtube_dl import compat_str
+
+year = compat_str(datetime.datetime.now().year)
 for fn in glob.glob('*.html*'):
-    with io.open(fn, encoding='utf-8') as f:
-        content = f.read()
+    content = read_file(fn)
     newc = re.sub(r'(?P<copyright>Copyright © 2011-)(?P<year>[0-9]{4})', 'Copyright © 2011-' + year, content)
     if content != newc:
         tmpFn = fn + '.part'
-        with io.open(tmpFn, 'wt', encoding='utf-8') as outf:
-            outf.write(newc)
+        write_file(tmpFn, newc)
         os.rename(tmpFn, fn)

devscripts/gh-pages/update-feed.py

@@ -2,10 +2,16 @@
 from __future__ import unicode_literals
 
 import datetime
-import io
 import json
+import os.path
 import textwrap
+import sys
 
+dirn = os.path.dirname
+
+sys.path.insert(0, dirn(dirn(os.path.abspath(__file__))))
+
+from utils import write_file
 
 atom_template = textwrap.dedent("""\
     <?xml version="1.0" encoding="utf-8"?>
@@ -72,5 +78,4 @@ for v in versions:
 entries_str = textwrap.indent(''.join(entries), '\t')
 atom_template = atom_template.replace('@ENTRIES@', entries_str)
 
-with io.open('update/releases.atom', 'w', encoding='utf-8') as atom_file:
-    atom_file.write(atom_template)
+write_file('update/releases.atom', atom_template)

devscripts/gh-pages/update-sites.py

@@ -5,15 +5,17 @@ import sys
 import os
 import textwrap
 
+dirn = os.path.dirname
+
 # We must be able to import youtube_dl
-sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.dirname(os.path.abspath(__file__)))))
+sys.path.insert(0, dirn(dirn(dirn(os.path.abspath(__file__)))))
 
 import youtube_dl
+from devscripts.utils import read_file, write_file
 
 
 def main():
-    with open('supportedsites.html.in', 'r', encoding='utf-8') as tmplf:
-        template = tmplf.read()
+    template = read_file('supportedsites.html.in')
 
     ie_htmls = []
     for ie in youtube_dl.list_extractors(age_limit=None):
@@ -29,8 +31,7 @@ def main():
 
     template = template.replace('@SITES@', textwrap.indent('\n'.join(ie_htmls), '\t'))
 
-    with open('supportedsites.html', 'w', encoding='utf-8') as sitesf:
-        sitesf.write(template)
+    write_file('supportedsites.html', template)
 
 
 if __name__ == '__main__':

devscripts/make_contributing.py

@@ -1,10 +1,11 @@
 #!/usr/bin/env python
 from __future__ import unicode_literals
 
-import io
 import optparse
 import re
 
+from utils import read_file, write_file
+
 
 def main():
     parser = optparse.OptionParser(usage='%prog INFILE OUTFILE')
@@ -14,8 +15,7 @@ def main():
 
     infile, outfile = args
 
-    with io.open(infile, encoding='utf-8') as inf:
-        readme = inf.read()
+    readme = read_file(infile)
 
     bug_text = re.search(
         r'(?s)#\s*BUGS\s*[^\n]*\s*(.*?)#\s*COPYRIGHT', readme).group(1)
@@ -25,8 +25,7 @@ def main():
 
     out = bug_text + dev_text
 
-    with io.open(outfile, 'w', encoding='utf-8') as outf:
-        outf.write(out)
+    write_file(outfile, out)
 
 
 if __name__ == '__main__':

devscripts/make_issue_template.py

@@ -1,8 +1,11 @@
 #!/usr/bin/env python
 from __future__ import unicode_literals
 
-import io
 import optparse
+import os.path
+import sys
+
+from utils import read_file, read_version, write_file
 
 
 def main():
@@ -13,17 +16,11 @@ def main():
 
     infile, outfile = args
 
-    with io.open(infile, encoding='utf-8') as inf:
-        issue_template_tmpl = inf.read()
+    issue_template_tmpl = read_file(infile)
 
-    # Get the version from youtube_dl/version.py without importing the package
-    exec(compile(open('youtube_dl/version.py').read(),
-                 'youtube_dl/version.py', 'exec'))
+    out = issue_template_tmpl % {'version': read_version()}
 
-    out = issue_template_tmpl % {'version': locals()['__version__']}
-
-    with io.open(outfile, 'w', encoding='utf-8') as outf:
-        outf.write(out)
+    write_file(outfile, out)
 
 if __name__ == '__main__':
     main()

devscripts/make_lazy_extractors.py

@@ -1,35 +1,48 @@
 from __future__ import unicode_literals, print_function
 
 from inspect import getsource
-import io
 import os
 from os.path import dirname as dirn
+import re
 import sys
 
 print('WARNING: Lazy loading extractors is an experimental feature that may not always work', file=sys.stderr)
 
-sys.path.insert(0, dirn(dirn((os.path.abspath(__file__)))))
+sys.path.insert(0, dirn(dirn(os.path.abspath(__file__))))
 
 lazy_extractors_filename = sys.argv[1]
 if os.path.exists(lazy_extractors_filename):
     os.remove(lazy_extractors_filename)
 # Py2: may be confused by leftover lazy_extractors.pyc
-try:
-    os.remove(lazy_extractors_filename + 'c')
-except OSError:
-    pass
+if sys.version_info[0] < 3:
+    for c in ('c', 'o'):
+        try:
+            os.remove(lazy_extractors_filename + 'c')
+        except OSError:
+            pass
+
+from devscripts.utils import read_file, write_file
+from youtube_dl.compat import compat_register_utf8
+
+compat_register_utf8()
 
 from youtube_dl.extractor import _ALL_CLASSES
 from youtube_dl.extractor.common import InfoExtractor, SearchInfoExtractor
 
-with open('devscripts/lazy_load_template.py', 'rt') as f:
-    module_template = f.read()
+module_template = read_file('devscripts/lazy_load_template.py')
+
+
+def get_source(m):
+    return re.sub(r'(?m)^\s*#.*\n', '', getsource(m))
+
 
 module_contents = [
-    module_template + '\n' + getsource(InfoExtractor.suitable) + '\n',
+    module_template,
+    get_source(InfoExtractor.suitable),
+    get_source(InfoExtractor._match_valid_url) + '\n',
     'class LazyLoadSearchExtractor(LazyLoadExtractor):\n    pass\n',
     # needed for suitable() methods of Youtube extractor (see #28780)
-    'from youtube_dl.utils import parse_qs\n',
+    'from youtube_dl.utils import parse_qs, variadic\n',
 ]
 
 ie_template = '''
@@ -62,7 +75,7 @@ def build_lazy_ie(ie, name):
         valid_url=valid_url,
         module=ie.__module__)
     if ie.suitable.__func__ is not InfoExtractor.suitable.__func__:
-        s += '\n' + getsource(ie.suitable)
+        s += '\n' + get_source(ie.suitable)
     if hasattr(ie, '_make_valid_url'):
         # search extractors
         s += make_valid_template.format(valid_url=ie._make_valid_url())
@@ -102,7 +115,17 @@ for ie in ordered_cls:
 module_contents.append(
     '_ALL_CLASSES = [{0}]'.format(', '.join(names)))
 
-module_src = '\n'.join(module_contents) + '\n'
+module_src = '\n'.join(module_contents)
 
-with io.open(lazy_extractors_filename, 'wt', encoding='utf-8') as f:
-    f.write(module_src)
+write_file(lazy_extractors_filename, module_src + '\n')
+
+# work around JVM byte code module limit in Jython
+if sys.platform.startswith('java') and sys.version_info[:2] == (2, 7):
+    import subprocess
+    from youtube_dl.compat import compat_subprocess_get_DEVNULL
+    # if Python 2.7 is available, use it to compile the module for Jython
+    try:
+        # if Python 2.7 is available, use it to compile the module for Jython
+        subprocess.check_call(['python2.7', '-m', 'py_compile', lazy_extractors_filename], stdout=compat_subprocess_get_DEVNULL())
+    except Exception:
+        pass

devscripts/make_readme.py

@@ -1,8 +1,14 @@
 from __future__ import unicode_literals
 
-import io
-import sys
+import os.path
 import re
+import sys
+dirn = os.path.dirname
+
+sys.path.insert(0, dirn(dirn(os.path.abspath(__file__))))
+
+from utils import read_file
+from youtube_dl.compat import compat_open as open
 
 README_FILE = 'README.md'
 helptext = sys.stdin.read()
@@ -10,8 +16,7 @@ helptext = sys.stdin.read()
 if isinstance(helptext, bytes):
     helptext = helptext.decode('utf-8')
 
-with io.open(README_FILE, encoding='utf-8') as f:
-    oldreadme = f.read()
+oldreadme = read_file(README_FILE)
 
 header = oldreadme[:oldreadme.index('# OPTIONS')]
 footer = oldreadme[oldreadme.index('# CONFIGURATION'):]
@@ -20,7 +25,7 @@ options = helptext[helptext.index('  General Options:') + 19:]
 options = re.sub(r'(?m)^  (\w.+)$', r'## \1', options)
 options = '# OPTIONS\n' + options + '\n'
 
-with io.open(README_FILE, 'w', encoding='utf-8') as f:
+with open(README_FILE, 'w', encoding='utf-8') as f:
     f.write(header)
     f.write(options)
     f.write(footer)

devscripts/make_supportedsites.py

@@ -1,17 +1,19 @@
 #!/usr/bin/env python
 from __future__ import unicode_literals
 
-import io
 import optparse
-import os
+import os.path
 import sys
 
-
 # Import youtube_dl
-ROOT_DIR = os.path.join(os.path.dirname(__file__), '..')
-sys.path.insert(0, ROOT_DIR)
+dirn = os.path.dirname
+
+sys.path.insert(0, dirn(dirn(os.path.abspath(__file__))))
+
 import youtube_dl
 
+from utils import write_file
+
 
 def main():
     parser = optparse.OptionParser(usage='%prog OUTFILE.md')
@@ -38,8 +40,7 @@ def main():
         ' - ' + md + '\n'
         for md in gen_ies_md(ies))
 
-    with io.open(outfile, 'w', encoding='utf-8') as outf:
-        outf.write(out)
+    write_file(outfile, out)
 
 
 if __name__ == '__main__':

devscripts/prepare_manpage.py

@@ -1,13 +1,13 @@
 from __future__ import unicode_literals
 
-import io
 import optparse
 import os.path
 import re
 
+from utils import read_file, write_file
+
 ROOT_DIR = os.path.dirname(os.path.dirname(os.path.abspath(__file__)))
 README_FILE = os.path.join(ROOT_DIR, 'README.md')
-
 PREFIX = r'''%YOUTUBE-DL(1)
 
 # NAME
@@ -29,8 +29,7 @@ def main():
 
     outfile, = args
 
-    with io.open(README_FILE, encoding='utf-8') as f:
-        readme = f.read()
+    readme = read_file(README_FILE)
 
     readme = re.sub(r'(?s)^.*?(?=# DESCRIPTION)', '', readme)
     readme = re.sub(r'\s+youtube-dl \[OPTIONS\] URL \[URL\.\.\.\]', '', readme)
@@ -38,8 +37,7 @@ def main():
 
     readme = filter_options(readme)
 
-    with io.open(outfile, 'w', encoding='utf-8') as outf:
-        outf.write(readme)
+    write_file(outfile, readme)
 
 
 def filter_options(readme):

devscripts/utils.py

@@ -0,0 +1,62 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+import argparse
+import functools
+import os.path
+import subprocess
+import sys
+
+dirn = os.path.dirname
+
+sys.path.insert(0, dirn(dirn(os.path.abspath(__file__))))
+
+from youtube_dl.compat import (
+    compat_kwargs,
+    compat_open as open,
+)
+
+
+def read_file(fname):
+    with open(fname, encoding='utf-8') as f:
+        return f.read()
+
+
+def write_file(fname, content, mode='w'):
+    with open(fname, mode, encoding='utf-8') as f:
+        return f.write(content)
+
+
+def read_version(fname='youtube_dl/version.py'):
+    """Get the version without importing the package"""
+    exec(compile(read_file(fname), fname, 'exec'))
+    return locals()['__version__']
+
+
+def get_filename_args(has_infile=False, default_outfile=None):
+    parser = argparse.ArgumentParser()
+    if has_infile:
+        parser.add_argument('infile', help='Input file')
+    kwargs = {'nargs': '?', 'default': default_outfile} if default_outfile else {}
+    kwargs['help'] = 'Output file'
+    parser.add_argument('outfile', **compat_kwargs(kwargs))
+
+    opts = parser.parse_args()
+    if has_infile:
+        return opts.infile, opts.outfile
+    return opts.outfile
+
+
+def compose_functions(*functions):
+    return lambda x: functools.reduce(lambda y, f: f(y), functions, x)
+
+
+def run_process(*args, **kwargs):
+    kwargs.setdefault('text', True)
+    kwargs.setdefault('check', True)
+    kwargs.setdefault('capture_output', True)
+    if kwargs['text']:
+        kwargs.setdefault('encoding', 'utf-8')
+        kwargs.setdefault('errors', 'replace')
+        kwargs = compat_kwargs(kwargs)
+    return subprocess.run(args, **kwargs)

devscripts/zsh-completion.py

@@ -7,6 +7,8 @@ import sys
 
 sys.path.insert(0, dirn(dirn((os.path.abspath(__file__)))))
 import youtube_dl
+from utils import read_file, write_file
+
 
 ZSH_COMPLETION_FILE = "youtube-dl.zsh"
 ZSH_COMPLETION_TEMPLATE = "devscripts/zsh-completion.in"
@@ -34,15 +36,13 @@ def build_completion(opt_parser):
 
     flags = [opt.get_opt_string() for opt in opts]
 
-    with open(ZSH_COMPLETION_TEMPLATE) as f:
-        template = f.read()
+    template = read_file(ZSH_COMPLETION_TEMPLATE)
 
     template = template.replace("{{fileopts}}", "|".join(fileopts))
     template = template.replace("{{diropts}}", "|".join(diropts))
     template = template.replace("{{flags}}", " ".join(flags))
 
-    with open(ZSH_COMPLETION_FILE, "w") as f:
-        f.write(template)
+    write_file(ZSH_COMPLETION_FILE, template)
 
 
 parser = youtube_dl.parseOpts()[0]

test/helper.py

@@ -1,22 +1,24 @@
 from __future__ import unicode_literals
 
 import errno
-import io
 import hashlib
 import json
 import os.path
 import re
-import types
 import ssl
 import sys
+import types
+import unittest
 
 import youtube_dl.extractor
 from youtube_dl import YoutubeDL
 from youtube_dl.compat import (
+    compat_open as open,
     compat_os_name,
     compat_str,
 )
 from youtube_dl.utils import (
+    IDENTITY,
     preferredencoding,
     write_string,
 )
@@ -27,10 +29,10 @@ def get_params(override=None):
                                    "parameters.json")
     LOCAL_PARAMETERS_FILE = os.path.join(os.path.dirname(os.path.abspath(__file__)),
                                          "local_parameters.json")
-    with io.open(PARAMETERS_FILE, encoding='utf-8') as pf:
+    with open(PARAMETERS_FILE, encoding='utf-8') as pf:
         parameters = json.load(pf)
     if os.path.exists(LOCAL_PARAMETERS_FILE):
-        with io.open(LOCAL_PARAMETERS_FILE, encoding='utf-8') as pf:
+        with open(LOCAL_PARAMETERS_FILE, encoding='utf-8') as pf:
             parameters.update(json.load(pf))
     if override:
         parameters.update(override)
@@ -72,7 +74,8 @@ class FakeYDL(YoutubeDL):
     def to_screen(self, s, skip_eol=None):
         print(s)
 
-    def trouble(self, s, tb=None):
+    def trouble(self, *args, **kwargs):
+        s = args[0] if len(args) > 0 else kwargs.get('message', 'Missing message')
         raise Exception(s)
 
     def download(self, x):
@@ -139,7 +142,7 @@ def expect_value(self, got, expected, field):
         self.assertTrue(
             contains_str in got,
             'field %s (value: %r) should contain %r' % (field, got, contains_str))
-    elif isinstance(expected, compat_str) and re.match(r'^lambda \w+:', expected):
+    elif isinstance(expected, compat_str) and re.match(r'lambda \w+:', expected):
         fn = eval(expected)
         suite = expected.split(':', 1)[1].strip()
         self.assertTrue(
@@ -178,18 +181,18 @@ def expect_value(self, got, expected, field):
             op, _, expected_num = expected.partition(':')
             expected_num = int(expected_num)
             if op == 'mincount':
-                assert_func = assertGreaterEqual
+                assert_func = self.assertGreaterEqual
                 msg_tmpl = 'Expected %d items in field %s, but only got %d'
             elif op == 'maxcount':
-                assert_func = assertLessEqual
+                assert_func = self.assertLessEqual
                 msg_tmpl = 'Expected maximum %d items in field %s, but got %d'
             elif op == 'count':
-                assert_func = assertEqual
+                assert_func = self.assertEqual
                 msg_tmpl = 'Expected exactly %d items in field %s, but got %d'
             else:
                 assert False
             assert_func(
-                self, len(got), expected_num,
+                len(got), expected_num,
                 msg_tmpl % (expected_num, field, len(got)))
             return
         self.assertEqual(
@@ -259,27 +262,6 @@ def assertRegexpMatches(self, text, regexp, msg=None):
             self.assertTrue(m, msg)
 
 
-def assertGreaterEqual(self, got, expected, msg=None):
-    if not (got >= expected):
-        if msg is None:
-            msg = '%r not greater than or equal to %r' % (got, expected)
-        self.assertTrue(got >= expected, msg)
-
-
-def assertLessEqual(self, got, expected, msg=None):
-    if not (got <= expected):
-        if msg is None:
-            msg = '%r not less than or equal to %r' % (got, expected)
-        self.assertTrue(got <= expected, msg)
-
-
-def assertEqual(self, got, expected, msg=None):
-    if not (got == expected):
-        if msg is None:
-            msg = '%r not equal to %r' % (got, expected)
-        self.assertTrue(got == expected, msg)
-
-
 def expect_warnings(ydl, warnings_re):
     real_warning = ydl.report_warning
 
@@ -297,3 +279,7 @@ def http_server_port(httpd):
     else:
         sock = httpd.socket
     return sock.getsockname()[1]
+
+
+def expectedFailureIf(cond):
+    return unittest.expectedFailure if cond else IDENTITY

test/test_InfoExtractor.py

@@ -3,19 +3,37 @@
 from __future__ import unicode_literals
 
 # Allow direct execution
-import io
 import os
 import sys
 import unittest
+
 sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
 
-from test.helper import FakeYDL, expect_dict, expect_value, http_server_port
-from youtube_dl.compat import compat_etree_fromstring, compat_http_server
-from youtube_dl.extractor.common import InfoExtractor
-from youtube_dl.extractor import YoutubeIE, get_info_extractor
-from youtube_dl.utils import encode_data_uri, strip_jsonp, ExtractorError, RegexNotFoundError
 import threading
 
+from test.helper import (
+    expect_dict,
+    expect_value,
+    FakeYDL,
+    http_server_port,
+)
+from youtube_dl.compat import (
+    compat_etree_fromstring,
+    compat_http_server,
+    compat_open as open,
+)
+from youtube_dl.extractor.common import InfoExtractor
+from youtube_dl.extractor import (
+    get_info_extractor,
+    YoutubeIE,
+)
+from youtube_dl.utils import (
+    encode_data_uri,
+    ExtractorError,
+    RegexNotFoundError,
+    strip_jsonp,
+)
+
 
 TEAPOT_RESPONSE_STATUS = 418
 TEAPOT_RESPONSE_BODY = "<h1>418 I'm a teapot</h1>"
@@ -100,6 +118,74 @@ class TestInfoExtractor(unittest.TestCase):
         self.assertRaises(RegexNotFoundError, ie._html_search_meta, 'z', html, None, fatal=True)
         self.assertRaises(RegexNotFoundError, ie._html_search_meta, ('z', 'x'), html, None, fatal=True)
 
+    def test_search_nextjs_data(self):
+        html = '''
+<!DOCTYPE html>
+<html>
+<head>
+  <meta http-equiv="content-type" content=
+  "text/html; charset=utf-8">
+  <meta name="viewport" content="width=device-width">
+  <title>Test _search_nextjs_data()</title>
+</head>
+<body>
+  <div id="__next">
+    <div style="background-color:#17171E" class="FU" dir="ltr">
+      <div class="sc-93de261d-0 dyzzYE">
+        <div>
+          <header class="HD"></header>
+          <main class="MN">
+            <div style="height:0" class="HT0">
+              <div style="width:NaN%" data-testid=
+              "stream-container" class="WDN"></div>
+            </div>
+          </main>
+        </div>
+        <footer class="sc-6e5faf91-0 dEGaHS"></footer>
+      </div>
+    </div>
+  </div>
+  <script id="__NEXT_DATA__" type="application/json">
+  {"props":{"pageProps":{"video":{"id":"testid"}}}}
+  </script>
+</body>
+</html>
+'''
+        search = self.ie._search_nextjs_data(html, 'testID')
+        self.assertEqual(search['props']['pageProps']['video']['id'], 'testid')
+        search = self.ie._search_nextjs_data(
+            'no next.js data here, move along', 'testID', default={'status': 0})
+        self.assertEqual(search['status'], 0)
+
+    def test_search_nuxt_data(self):
+        html = '''
+<!DOCTYPE html>
+<html>
+<head>
+  <meta http-equiv="content-type" content=
+  "text/html; charset=utf-8">
+  <title>Nuxt.js Test Page</title>
+  <meta name="viewport" content=
+  "width=device-width, initial-scale=1">
+  <meta data-hid="robots" name="robots" content="all">
+</head>
+<body class="BD">
+  <div id="__layout">
+    <h1 class="H1">Example heading</h1>
+    <div class="IN">
+      <p>Decoy text</p>
+    </div>
+  </div>
+  <script>
+  window.__NUXT__=(function(a,b,c,d,e,f,g,h){return {decoy:" default",data:[{track:{id:f,title:g}}]}}(null,null,"c",null,null,"testid","Nuxt.js title",null));
+  </script>
+  <script src="/_nuxt/a12345b.js" defer="defer"></script>
+</body>
+</html>
+'''
+        search = self.ie._search_nuxt_data(html, 'testID')
+        self.assertEqual(search['track']['id'], 'testid')
+
     def test_search_json_ld_realworld(self):
         # https://github.com/ytdl-org/youtube-dl/issues/23306
         expect_dict(
@@ -348,6 +434,24 @@ class TestInfoExtractor(unittest.TestCase):
                 }],
             })
 
+        # from https://0000.studio/
+        # with type attribute but without extension in URL
+        expect_dict(
+            self,
+            self.ie._parse_html5_media_entries(
+                'https://0000.studio',
+                r'''
+                <video src="https://d1ggyt9m8pwf3g.cloudfront.net/protected/ap-northeast-1:1864af40-28d5-492b-b739-b32314b1a527/archive/clip/838db6a7-8973-4cd6-840d-8517e4093c92"
+                    controls="controls" type="video/mp4" preload="metadata" autoplay="autoplay" playsinline class="object-contain">
+                </video>
+                ''', None)[0],
+            {
+                'formats': [{
+                    'url': 'https://d1ggyt9m8pwf3g.cloudfront.net/protected/ap-northeast-1:1864af40-28d5-492b-b739-b32314b1a527/archive/clip/838db6a7-8973-4cd6-840d-8517e4093c92',
+                    'ext': 'mp4',
+                }],
+            })
+
     def test_extract_jwplayer_data_realworld(self):
         # from http://www.suffolk.edu/sjc/
         expect_dict(
@@ -801,8 +905,8 @@ jwplayer("mediaplayer").setup({"abouttext":"Visit Indie DB","aboutlink":"http:\/
         ]
 
         for m3u8_file, m3u8_url, expected_formats in _TEST_CASES:
-            with io.open('./test/testdata/m3u8/%s.m3u8' % m3u8_file,
-                         mode='r', encoding='utf-8') as f:
+            with open('./test/testdata/m3u8/%s.m3u8' % m3u8_file,
+                      mode='r', encoding='utf-8') as f:
                 formats = self.ie._parse_m3u8_formats(
                     f.read(), m3u8_url, ext='mp4')
                 self.ie._sort_formats(formats)
@@ -892,7 +996,8 @@ jwplayer("mediaplayer").setup({"abouttext":"Visit Indie DB","aboutlink":"http:\/
                     'tbr': 5997.485,
                     'width': 1920,
                     'height': 1080,
-                }]
+                }],
+                {},
             ), (
                 # https://github.com/ytdl-org/youtube-dl/pull/14844
                 'urls_only',
@@ -975,7 +1080,8 @@ jwplayer("mediaplayer").setup({"abouttext":"Visit Indie DB","aboutlink":"http:\/
                     'tbr': 4400,
                     'width': 1920,
                     'height': 1080,
-                }]
+                }],
+                {},
             ), (
                 # https://github.com/ytdl-org/youtube-dl/issues/20346
                 # Media considered unfragmented even though it contains
@@ -1021,18 +1127,185 @@ jwplayer("mediaplayer").setup({"abouttext":"Visit Indie DB","aboutlink":"http:\/
                     'width': 360,
                     'height': 360,
                     'fps': 30,
-                }]
+                }],
+                {},
+            ), (
+                # https://github.com/ytdl-org/youtube-dl/issues/30235
+                # Bento4 generated test mpd
+                # mp4dash --mpd-name=manifest.mpd --no-split --use-segment-list mediafiles
+                'url_and_range',
+                'http://unknown/manifest.mpd',  # mpd_url
+                'http://unknown/',  # mpd_base_url
+                [{
+                    'manifest_url': 'http://unknown/manifest.mpd',
+                    'fragment_base_url': 'http://unknown/',
+                    'ext': 'm4a',
+                    'format_id': 'audio-und-mp4a.40.2',
+                    'format_note': 'DASH audio',
+                    'container': 'm4a_dash',
+                    'protocol': 'http_dash_segments',
+                    'acodec': 'mp4a.40.2',
+                    'vcodec': 'none',
+                    'tbr': 98.808,
+                }, {
+                    'manifest_url': 'http://unknown/manifest.mpd',
+                    'fragment_base_url': 'http://unknown/',
+                    'ext': 'mp4',
+                    'format_id': 'video-avc1',
+                    'format_note': 'DASH video',
+                    'container': 'mp4_dash',
+                    'protocol': 'http_dash_segments',
+                    'acodec': 'none',
+                    'vcodec': 'avc1.4D401E',
+                    'tbr': 699.597,
+                    'width': 768,
+                    'height': 432
+                }],
+                {},
+            ), (
+                # https://github.com/ytdl-org/youtube-dl/issues/27575
+                # GPAC generated test mpd
+                # MP4Box -dash 10000 -single-file -out manifest.mpd mediafiles
+                'range_only',
+                'http://unknown/manifest.mpd',  # mpd_url
+                'http://unknown/',  # mpd_base_url
+                [{
+                    'manifest_url': 'http://unknown/manifest.mpd',
+                    'fragment_base_url': 'http://unknown/audio_dashinit.mp4',
+                    'ext': 'm4a',
+                    'format_id': '2',
+                    'format_note': 'DASH audio',
+                    'container': 'm4a_dash',
+                    'protocol': 'http_dash_segments',
+                    'acodec': 'mp4a.40.2',
+                    'vcodec': 'none',
+                    'tbr': 98.096,
+                }, {
+                    'manifest_url': 'http://unknown/manifest.mpd',
+                    'fragment_base_url': 'http://unknown/video_dashinit.mp4',
+                    'ext': 'mp4',
+                    'format_id': '1',
+                    'format_note': 'DASH video',
+                    'container': 'mp4_dash',
+                    'protocol': 'http_dash_segments',
+                    'acodec': 'none',
+                    'vcodec': 'avc1.4D401E',
+                    'tbr': 526.987,
+                    'width': 768,
+                    'height': 432
+                }],
+                {},
+            ), (
+                'subtitles',
+                'https://sdn-global-streaming-cache-3qsdn.akamaized.net/stream/3144/files/17/07/672975/3144-kZT4LWMQw6Rh7Kpd.ism/manifest.mpd',
+                'https://sdn-global-streaming-cache-3qsdn.akamaized.net/stream/3144/files/17/07/672975/3144-kZT4LWMQw6Rh7Kpd.ism/',
+                [{
+                    'format_id': 'audio=128001',
+                    'manifest_url': 'https://sdn-global-streaming-cache-3qsdn.akamaized.net/stream/3144/files/17/07/672975/3144-kZT4LWMQw6Rh7Kpd.ism/manifest.mpd',
+                    'ext': 'm4a',
+                    'tbr': 128.001,
+                    'asr': 48000,
+                    'format_note': 'DASH audio',
+                    'container': 'm4a_dash',
+                    'vcodec': 'none',
+                    'acodec': 'mp4a.40.2',
+                    'url': 'https://sdn-global-streaming-cache-3qsdn.akamaized.net/stream/3144/files/17/07/672975/3144-kZT4LWMQw6Rh7Kpd.ism/manifest.mpd',
+                    'fragment_base_url': 'https://sdn-global-streaming-cache-3qsdn.akamaized.net/stream/3144/files/17/07/672975/3144-kZT4LWMQw6Rh7Kpd.ism/dash/',
+                    'protocol': 'http_dash_segments',
+                }, {
+                    'format_id': 'video=100000',
+                    'manifest_url': 'https://sdn-global-streaming-cache-3qsdn.akamaized.net/stream/3144/files/17/07/672975/3144-kZT4LWMQw6Rh7Kpd.ism/manifest.mpd',
+                    'ext': 'mp4',
+                    'width': 336,
+                    'height': 144,
+                    'tbr': 100,
+                    'format_note': 'DASH video',
+                    'container': 'mp4_dash',
+                    'vcodec': 'avc1.4D401F',
+                    'acodec': 'none',
+                    'url': 'https://sdn-global-streaming-cache-3qsdn.akamaized.net/stream/3144/files/17/07/672975/3144-kZT4LWMQw6Rh7Kpd.ism/manifest.mpd',
+                    'fragment_base_url': 'https://sdn-global-streaming-cache-3qsdn.akamaized.net/stream/3144/files/17/07/672975/3144-kZT4LWMQw6Rh7Kpd.ism/dash/',
+                    'protocol': 'http_dash_segments',
+                }, {
+                    'format_id': 'video=326000',
+                    'manifest_url': 'https://sdn-global-streaming-cache-3qsdn.akamaized.net/stream/3144/files/17/07/672975/3144-kZT4LWMQw6Rh7Kpd.ism/manifest.mpd',
+                    'ext': 'mp4',
+                    'width': 562,
+                    'height': 240,
+                    'tbr': 326,
+                    'format_note': 'DASH video',
+                    'container': 'mp4_dash',
+                    'vcodec': 'avc1.4D401F',
+                    'acodec': 'none',
+                    'url': 'https://sdn-global-streaming-cache-3qsdn.akamaized.net/stream/3144/files/17/07/672975/3144-kZT4LWMQw6Rh7Kpd.ism/manifest.mpd',
+                    'fragment_base_url': 'https://sdn-global-streaming-cache-3qsdn.akamaized.net/stream/3144/files/17/07/672975/3144-kZT4LWMQw6Rh7Kpd.ism/dash/',
+                    'protocol': 'http_dash_segments',
+                }, {
+                    'format_id': 'video=698000',
+                    'manifest_url': 'https://sdn-global-streaming-cache-3qsdn.akamaized.net/stream/3144/files/17/07/672975/3144-kZT4LWMQw6Rh7Kpd.ism/manifest.mpd',
+                    'ext': 'mp4',
+                    'width': 844,
+                    'height': 360,
+                    'tbr': 698,
+                    'format_note': 'DASH video',
+                    'container': 'mp4_dash',
+                    'vcodec': 'avc1.4D401F',
+                    'acodec': 'none',
+                    'url': 'https://sdn-global-streaming-cache-3qsdn.akamaized.net/stream/3144/files/17/07/672975/3144-kZT4LWMQw6Rh7Kpd.ism/manifest.mpd',
+                    'fragment_base_url': 'https://sdn-global-streaming-cache-3qsdn.akamaized.net/stream/3144/files/17/07/672975/3144-kZT4LWMQw6Rh7Kpd.ism/dash/',
+                    'protocol': 'http_dash_segments',
+                }, {
+                    'format_id': 'video=1493000',
+                    'manifest_url': 'https://sdn-global-streaming-cache-3qsdn.akamaized.net/stream/3144/files/17/07/672975/3144-kZT4LWMQw6Rh7Kpd.ism/manifest.mpd',
+                    'ext': 'mp4',
+                    'width': 1126,
+                    'height': 480,
+                    'tbr': 1493,
+                    'format_note': 'DASH video',
+                    'container': 'mp4_dash',
+                    'vcodec': 'avc1.4D401F',
+                    'acodec': 'none',
+                    'url': 'https://sdn-global-streaming-cache-3qsdn.akamaized.net/stream/3144/files/17/07/672975/3144-kZT4LWMQw6Rh7Kpd.ism/manifest.mpd',
+                    'fragment_base_url': 'https://sdn-global-streaming-cache-3qsdn.akamaized.net/stream/3144/files/17/07/672975/3144-kZT4LWMQw6Rh7Kpd.ism/dash/',
+                    'protocol': 'http_dash_segments',
+                }, {
+                    'format_id': 'video=4482000',
+                    'manifest_url': 'https://sdn-global-streaming-cache-3qsdn.akamaized.net/stream/3144/files/17/07/672975/3144-kZT4LWMQw6Rh7Kpd.ism/manifest.mpd',
+                    'ext': 'mp4',
+                    'width': 1688,
+                    'height': 720,
+                    'tbr': 4482,
+                    'format_note': 'DASH video',
+                    'container': 'mp4_dash',
+                    'vcodec': 'avc1.4D401F',
+                    'acodec': 'none',
+                    'url': 'https://sdn-global-streaming-cache-3qsdn.akamaized.net/stream/3144/files/17/07/672975/3144-kZT4LWMQw6Rh7Kpd.ism/manifest.mpd',
+                    'fragment_base_url': 'https://sdn-global-streaming-cache-3qsdn.akamaized.net/stream/3144/files/17/07/672975/3144-kZT4LWMQw6Rh7Kpd.ism/dash/',
+                    'protocol': 'http_dash_segments',
+                }],
+                {
+                    'en': [
+                        {
+                            'ext': 'mp4',
+                            'manifest_url': 'https://sdn-global-streaming-cache-3qsdn.akamaized.net/stream/3144/files/17/07/672975/3144-kZT4LWMQw6Rh7Kpd.ism/manifest.mpd',
+                            'url': 'https://sdn-global-streaming-cache-3qsdn.akamaized.net/stream/3144/files/17/07/672975/3144-kZT4LWMQw6Rh7Kpd.ism/manifest.mpd',
+                            'fragment_base_url': 'https://sdn-global-streaming-cache-3qsdn.akamaized.net/stream/3144/files/17/07/672975/3144-kZT4LWMQw6Rh7Kpd.ism/dash/',
+                            'protocol': 'http_dash_segments',
+                        }
+                    ]
+                },
             )
         ]
 
-        for mpd_file, mpd_url, mpd_base_url, expected_formats in _TEST_CASES:
-            with io.open('./test/testdata/mpd/%s.mpd' % mpd_file,
-                         mode='r', encoding='utf-8') as f:
-                formats = self.ie._parse_mpd_formats(
+        for mpd_file, mpd_url, mpd_base_url, expected_formats, expected_subtitles in _TEST_CASES:
+            with open('./test/testdata/mpd/%s.mpd' % mpd_file,
+                      mode='r', encoding='utf-8') as f:
+                formats, subtitles = self.ie._parse_mpd_formats_and_subtitles(
                     compat_etree_fromstring(f.read().encode('utf-8')),
                     mpd_base_url=mpd_base_url, mpd_url=mpd_url)
                 self.ie._sort_formats(formats)
                 expect_value(self, formats, expected_formats, None)
+                expect_value(self, subtitles, expected_subtitles, None)
 
     def test_parse_f4m_formats(self):
         _TEST_CASES = [
@@ -1053,8 +1326,8 @@ jwplayer("mediaplayer").setup({"abouttext":"Visit Indie DB","aboutlink":"http:\/
         ]
 
         for f4m_file, f4m_url, expected_formats in _TEST_CASES:
-            with io.open('./test/testdata/f4m/%s.f4m' % f4m_file,
-                         mode='r', encoding='utf-8') as f:
+            with open('./test/testdata/f4m/%s.f4m' % f4m_file,
+                      mode='r', encoding='utf-8') as f:
                 formats = self.ie._parse_f4m_formats(
                     compat_etree_fromstring(f.read().encode('utf-8')),
                     f4m_url, None)
@@ -1101,8 +1374,8 @@ jwplayer("mediaplayer").setup({"abouttext":"Visit Indie DB","aboutlink":"http:\/
         ]
 
         for xspf_file, xspf_url, expected_entries in _TEST_CASES:
-            with io.open('./test/testdata/xspf/%s.xspf' % xspf_file,
-                         mode='r', encoding='utf-8') as f:
+            with open('./test/testdata/xspf/%s.xspf' % xspf_file,
+                      mode='r', encoding='utf-8') as f:
                 entries = self.ie._parse_xspf(
                     compat_etree_fromstring(f.read().encode('utf-8')),
                     xspf_file, xspf_url=xspf_url, xspf_base_url=xspf_url)

test/test_YoutubeDL.py

@@ -10,14 +10,31 @@ import unittest
 sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
 
 import copy
+import json
 
-from test.helper import FakeYDL, assertRegexpMatches
+from test.helper import (
+    FakeYDL,
+    assertRegexpMatches,
+    try_rm,
+)
 from youtube_dl import YoutubeDL
-from youtube_dl.compat import compat_str, compat_urllib_error
+from youtube_dl.compat import (
+    compat_http_cookiejar_Cookie,
+    compat_http_cookies_SimpleCookie,
+    compat_kwargs,
+    compat_open as open,
+    compat_str,
+    compat_urllib_error,
+)
+
 from youtube_dl.extractor import YoutubeIE
 from youtube_dl.extractor.common import InfoExtractor
 from youtube_dl.postprocessor.common import PostProcessor
-from youtube_dl.utils import ExtractorError, match_filter_func
+from youtube_dl.utils import (
+    ExtractorError,
+    match_filter_func,
+    traverse_obj,
+)
 
 TEST_URL = 'http://localhost/sample.mp4'
 
@@ -29,11 +46,14 @@ class YDL(FakeYDL):
         self.msgs = []
 
     def process_info(self, info_dict):
-        self.downloaded_info_dicts.append(info_dict)
+        self.downloaded_info_dicts.append(info_dict.copy())
 
     def to_screen(self, msg):
         self.msgs.append(msg)
 
+    def dl(self, *args, **kwargs):
+        assert False, 'Downloader must not be invoked for test_YoutubeDL'
+
 
 def _make_result(formats, **kwargs):
     res = {
@@ -42,8 +62,9 @@ def _make_result(formats, **kwargs):
         'title': 'testttitle',
         'extractor': 'testex',
         'extractor_key': 'TestEx',
+        'webpage_url': 'http://example.com/watch?v=shenanigans',
     }
-    res.update(**kwargs)
+    res.update(**compat_kwargs(kwargs))
     return res
 
 
@@ -681,12 +702,12 @@ class TestYoutubeDL(unittest.TestCase):
 
         class SimplePP(PostProcessor):
             def run(self, info):
-                with open(audiofile, 'wt') as f:
+                with open(audiofile, 'w') as f:
                     f.write('EXAMPLE')
                 return [info['filepath']], info
 
         def run_pp(params, PP):
-            with open(filename, 'wt') as f:
+            with open(filename, 'w') as f:
                 f.write('EXAMPLE')
             ydl = YoutubeDL(params)
             ydl.add_post_processor(PP())
@@ -705,7 +726,7 @@ class TestYoutubeDL(unittest.TestCase):
 
         class ModifierPP(PostProcessor):
             def run(self, info):
-                with open(info['filepath'], 'wt') as f:
+                with open(info['filepath'], 'w') as f:
                     f.write('MODIFIED')
                 return [], info
 
@@ -930,17 +951,11 @@ class TestYoutubeDL(unittest.TestCase):
     # Test case for https://github.com/ytdl-org/youtube-dl/issues/27064
     def test_ignoreerrors_for_playlist_with_url_transparent_iterable_entries(self):
 
-        class _YDL(YDL):
-            def __init__(self, *args, **kwargs):
-                super(_YDL, self).__init__(*args, **kwargs)
-
-            def trouble(self, s, tb=None):
-                pass
-
-        ydl = _YDL({
+        ydl = YDL({
             'format': 'extra',
             'ignoreerrors': True,
         })
+        ydl.trouble = lambda *_, **__: None
 
         class VideoIE(InfoExtractor):
             _VALID_URL = r'video:(?P<id>\d+)'
@@ -1017,5 +1032,160 @@ class TestYoutubeDL(unittest.TestCase):
         self.assertEqual(out_info['release_date'], '20210930')
 
 
+class TestYoutubeDLCookies(unittest.TestCase):
+
+    @staticmethod
+    def encode_cookie(cookie):
+        if not isinstance(cookie, dict):
+            cookie = vars(cookie)
+        for name, value in cookie.items():
+            yield name, compat_str(value)
+
+    @classmethod
+    def comparable_cookies(cls, cookies):
+        # Work around cookiejar cookies not being unicode strings
+        return sorted(map(tuple, map(sorted, map(cls.encode_cookie, cookies))))
+
+    def assertSameCookies(self, c1, c2, msg=None):
+        return self.assertEqual(
+            *map(self.comparable_cookies, (c1, c2)),
+            msg=msg)
+
+    def assertSameCookieStrings(self, c1, c2, msg=None):
+        return self.assertSameCookies(
+            *map(lambda c: compat_http_cookies_SimpleCookie(c).values(), (c1, c2)),
+            msg=msg)
+
+    def test_header_cookies(self):
+
+        ydl = FakeYDL()
+        ydl.report_warning = lambda *_, **__: None
+
+        def cookie(name, value, version=None, domain='', path='', secure=False, expires=None):
+            return compat_http_cookiejar_Cookie(
+                version or 0, name, value, None, False,
+                domain, bool(domain), bool(domain), path, bool(path),
+                secure, expires, False, None, None, rest={})
+
+        test_url, test_domain = (t % ('yt.dl',) for t in ('https://%s/test', '.%s'))
+
+        def test(encoded_cookies, cookies, headers=False, round_trip=None, error_re=None):
+            def _test():
+                ydl.cookiejar.clear()
+                ydl._load_cookies(encoded_cookies, autoscope=headers)
+                if headers:
+                    ydl._apply_header_cookies(test_url)
+                data = {'url': test_url}
+                ydl._calc_headers(data)
+                self.assertSameCookies(
+                    cookies, ydl.cookiejar,
+                    'Extracted cookiejar.Cookie is not the same')
+                if not headers:
+                    self.assertSameCookieStrings(
+                        data.get('cookies'), round_trip or encoded_cookies,
+                        msg='Cookie is not the same as round trip')
+                ydl.__dict__['_YoutubeDL__header_cookies'] = []
+
+            try:
+                _test()
+            except AssertionError:
+                raise
+            except Exception as e:
+                if not error_re:
+                    raise
+                assertRegexpMatches(self, e.args[0], error_re.join(('.*',) * 2))
+
+        test('test=value; Domain=' + test_domain, [cookie('test', 'value', domain=test_domain)])
+        test('test=value', [cookie('test', 'value')], error_re='Unscoped cookies are not allowed')
+        test('cookie1=value1; Domain={0}; Path=/test; cookie2=value2; Domain={0}; Path=/'.format(test_domain), [
+            cookie('cookie1', 'value1', domain=test_domain, path='/test'),
+            cookie('cookie2', 'value2', domain=test_domain, path='/')])
+        cookie_kw = compat_kwargs(
+            {'domain': test_domain, 'path': '/test', 'secure': True, 'expires': '9999999999', })
+        test('test=value; Domain={domain}; Path={path}; Secure; Expires={expires}'.format(**cookie_kw), [
+            cookie('test', 'value', **cookie_kw)])
+        test('test="value; "; path=/test; domain=' + test_domain, [
+            cookie('test', 'value; ', domain=test_domain, path='/test')],
+            round_trip='test="value\\073 "; Domain={0}; Path=/test'.format(test_domain))
+        test('name=; Domain=' + test_domain, [cookie('name', '', domain=test_domain)],
+             round_trip='name=""; Domain=' + test_domain)
+        test('test=value', [cookie('test', 'value', domain=test_domain)], headers=True)
+        test('cookie1=value; Domain={0}; cookie2=value'.format(test_domain), [],
+             headers=True, error_re='Invalid syntax')
+        ydl.report_warning = ydl.report_error
+        test('test=value', [], headers=True, error_re='Passing cookies as a header is a potential security risk')
+
+    def test_infojson_cookies(self):
+        TEST_FILE = 'test_infojson_cookies.info.json'
+        TEST_URL = 'https://example.com/example.mp4'
+        COOKIES = 'a=b; Domain=.example.com; c=d; Domain=.example.com'
+        COOKIE_HEADER = {'Cookie': 'a=b; c=d'}
+
+        ydl = FakeYDL()
+        ydl.process_info = lambda x: ydl._write_info_json('test', x, TEST_FILE)
+
+        def make_info(info_header_cookies=False, fmts_header_cookies=False, cookies_field=False):
+            fmt = {'url': TEST_URL}
+            if fmts_header_cookies:
+                fmt['http_headers'] = COOKIE_HEADER
+            if cookies_field:
+                fmt['cookies'] = COOKIES
+            return _make_result([fmt], http_headers=COOKIE_HEADER if info_header_cookies else None)
+
+        def test(initial_info, note):
+
+            def failure_msg(why):
+                return ' when '.join((why, note))
+
+            result = {}
+            result['processed'] = ydl.process_ie_result(initial_info)
+            self.assertTrue(ydl.cookiejar.get_cookies_for_url(TEST_URL),
+                            msg=failure_msg('No cookies set in cookiejar after initial process'))
+            ydl.cookiejar.clear()
+            with open(TEST_FILE) as infojson:
+                result['loaded'] = ydl.sanitize_info(json.load(infojson), True)
+            result['final'] = ydl.process_ie_result(result['loaded'].copy(), download=False)
+            self.assertTrue(ydl.cookiejar.get_cookies_for_url(TEST_URL),
+                            msg=failure_msg('No cookies set in cookiejar after final process'))
+            ydl.cookiejar.clear()
+            for key in ('processed', 'loaded', 'final'):
+                info = result[key]
+                self.assertIsNone(
+                    traverse_obj(info, ((None, ('formats', 0)), 'http_headers', 'Cookie'), casesense=False, get_all=False),
+                    msg=failure_msg('Cookie header not removed in {0} result'.format(key)))
+                self.assertSameCookieStrings(
+                    traverse_obj(info, ((None, ('formats', 0)), 'cookies'), get_all=False), COOKIES,
+                    msg=failure_msg('No cookies field found in {0} result'.format(key)))
+
+        test({'url': TEST_URL, 'http_headers': COOKIE_HEADER, 'id': '1', 'title': 'x'}, 'no formats field')
+        test(make_info(info_header_cookies=True), 'info_dict header cokies')
+        test(make_info(fmts_header_cookies=True), 'format header cookies')
+        test(make_info(info_header_cookies=True, fmts_header_cookies=True), 'info_dict and format header cookies')
+        test(make_info(info_header_cookies=True, fmts_header_cookies=True, cookies_field=True), 'all cookies fields')
+        test(make_info(cookies_field=True), 'cookies format field')
+        test({'url': TEST_URL, 'cookies': COOKIES, 'id': '1', 'title': 'x'}, 'info_dict cookies field only')
+
+        try_rm(TEST_FILE)
+
+    def test_add_headers_cookie(self):
+        def check_for_cookie_header(result):
+            return traverse_obj(result, ((None, ('formats', 0)), 'http_headers', 'Cookie'), casesense=False, get_all=False)
+
+        ydl = FakeYDL({'http_headers': {'Cookie': 'a=b'}})
+        ydl._apply_header_cookies(_make_result([])['webpage_url'])  # Scope to input webpage URL: .example.com
+
+        fmt = {'url': 'https://example.com/video.mp4'}
+        result = ydl.process_ie_result(_make_result([fmt]), download=False)
+        self.assertIsNone(check_for_cookie_header(result), msg='http_headers cookies in result info_dict')
+        self.assertEqual(result.get('cookies'), 'a=b; Domain=.example.com', msg='No cookies were set in cookies field')
+        self.assertIn('a=b', ydl.cookiejar.get_cookie_header(fmt['url']), msg='No cookies were set in cookiejar')
+
+        fmt = {'url': 'https://wrong.com/video.mp4'}
+        result = ydl.process_ie_result(_make_result([fmt]), download=False)
+        self.assertIsNone(check_for_cookie_header(result), msg='http_headers cookies for wrong domain')
+        self.assertFalse(result.get('cookies'), msg='Cookies set in cookies field for wrong domain')
+        self.assertFalse(ydl.cookiejar.get_cookie_header(fmt['url']), msg='Cookies set in cookiejar for wrong domain')
+
+
 if __name__ == '__main__':
     unittest.main()

test/test_YoutubeDLCookieJar.py

@@ -46,6 +46,20 @@ class TestYoutubeDLCookieJar(unittest.TestCase):
         # will be ignored
         self.assertFalse(cookiejar._cookies)
 
+    def test_get_cookie_header(self):
+        cookiejar = YoutubeDLCookieJar('./test/testdata/cookies/httponly_cookies.txt')
+        cookiejar.load(ignore_discard=True, ignore_expires=True)
+        header = cookiejar.get_cookie_header('https://www.foobar.foobar')
+        self.assertIn('HTTPONLY_COOKIE', header)
+
+    def test_get_cookies_for_url(self):
+        cookiejar = YoutubeDLCookieJar('./test/testdata/cookies/session_cookies.txt')
+        cookiejar.load(ignore_discard=True, ignore_expires=True)
+        cookies = cookiejar.get_cookies_for_url('https://www.foobar.foobar/')
+        self.assertEqual(len(cookies), 2)
+        cookies = cookiejar.get_cookies_for_url('https://foobar.foobar/')
+        self.assertFalse(cookies)
+
 
 if __name__ == '__main__':
     unittest.main()

test/test_compat.py

@@ -23,6 +23,7 @@ from youtube_dl.compat import (
     compat_urllib_parse_unquote,
     compat_urllib_parse_unquote_plus,
     compat_urllib_parse_urlencode,
+    compat_urllib_request,
 )
 
 
@@ -135,6 +136,19 @@ class TestCompat(unittest.TestCase):
         self.assertEqual(compat_casefold('\u03a3'), '\u03c3')
         self.assertEqual(compat_casefold('A\u0345\u03a3'), 'a\u03b9\u03c3')
 
+    def test_compat_urllib_request_Request(self):
+        self.assertEqual(
+            compat_urllib_request.Request('http://127.0.0.1', method='PUT').get_method(),
+            'PUT')
+
+        class PUTrequest(compat_urllib_request.Request):
+            def get_method(self):
+                return 'PUT'
+
+        self.assertEqual(
+            PUTrequest('http://127.0.0.1').get_method(),
+            'PUT')
+
 
 if __name__ == '__main__':
     unittest.main()

test/test_download.py

@@ -9,7 +9,6 @@ import unittest
 sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
 
 from test.helper import (
-    assertGreaterEqual,
     expect_warnings,
     get_params,
     gettestcases,
@@ -20,27 +19,35 @@ from test.helper import (
 
 
 import hashlib
-import io
 import json
 import socket
 
 import youtube_dl.YoutubeDL
 from youtube_dl.compat import (
     compat_http_client,
-    compat_urllib_error,
     compat_HTTPError,
+    compat_open as open,
+    compat_urllib_error,
 )
 from youtube_dl.utils import (
     DownloadError,
     ExtractorError,
     error_to_compat_str,
     format_bytes,
+    IDENTITY,
+    preferredencoding,
     UnavailableVideoError,
 )
 from youtube_dl.extractor import get_info_extractor
 
 RETRIES = 3
 
+# Some unittest APIs require actual str
+if not isinstance('TEST', str):
+    _encode_str = lambda s: s.encode(preferredencoding())
+else:
+    _encode_str = IDENTITY
+
 
 class YoutubeDL(youtube_dl.YoutubeDL):
     def __init__(self, *args, **kwargs):
@@ -101,7 +108,7 @@ def generator(test_case, tname):
 
         def print_skipping(reason):
             print('Skipping %s: %s' % (test_case['name'], reason))
-            self.skipTest(reason)
+            self.skipTest(_encode_str(reason))
 
         if not ie.working():
             print_skipping('IE marked as not _WORKING')
@@ -122,7 +129,10 @@ def generator(test_case, tname):
         params['outtmpl'] = tname + '_' + params['outtmpl']
         if is_playlist and 'playlist' not in test_case:
             params.setdefault('extract_flat', 'in_playlist')
-            params.setdefault('playlistend', test_case.get('playlist_mincount'))
+            params.setdefault('playlistend',
+                              test_case['playlist_maxcount'] + 1
+                              if test_case.get('playlist_maxcount')
+                              else test_case.get('playlist_mincount'))
             params.setdefault('skip_download', True)
 
         ydl = YoutubeDL(params, auto_init=False)
@@ -183,13 +193,19 @@ def generator(test_case, tname):
                 expect_info_dict(self, res_dict, test_case.get('info_dict', {}))
 
             if 'playlist_mincount' in test_case:
-                assertGreaterEqual(
-                    self,
+                self.assertGreaterEqual(
                     len(res_dict['entries']),
                     test_case['playlist_mincount'],
                     'Expected at least %d in playlist %s, but got only %d' % (
                         test_case['playlist_mincount'], test_case['url'],
                         len(res_dict['entries'])))
+            if 'playlist_maxcount' in test_case:
+                self.assertLessEqual(
+                    len(res_dict['entries']),
+                    test_case['playlist_maxcount'],
+                    'Expected at most %d in playlist %s, but got %d' % (
+                        test_case['playlist_maxcount'], test_case['url'],
+                        len(res_dict['entries'])))
             if 'playlist_count' in test_case:
                 self.assertEqual(
                     len(res_dict['entries']),
@@ -231,8 +247,8 @@ def generator(test_case, tname):
                         if params.get('test'):
                             expected_minsize = max(expected_minsize, 10000)
                         got_fsize = os.path.getsize(tc_filename)
-                        assertGreaterEqual(
-                            self, got_fsize, expected_minsize,
+                        self.assertGreaterEqual(
+                            got_fsize, expected_minsize,
                             'Expected %s to be at least %s, but it\'s only %s ' %
                             (tc_filename, format_bytes(expected_minsize),
                                 format_bytes(got_fsize)))
@@ -245,7 +261,7 @@ def generator(test_case, tname):
                 self.assertTrue(
                     os.path.exists(info_json_fn),
                     'Missing info file %s' % info_json_fn)
-                with io.open(info_json_fn, encoding='utf-8') as infof:
+                with open(info_json_fn, encoding='utf-8') as infof:
                     info_dict = json.load(infof)
                 expect_info_dict(self, info_dict, tc.get('info_dict', {}))
         finally:

test/test_downloader_external.py

@@ -12,20 +12,69 @@ sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
 
 from test.helper import (
     FakeLogger,
+    FakeYDL,
     http_server_port,
     try_rm,
 )
 from youtube_dl import YoutubeDL
-from youtube_dl.compat import compat_http_server
-from youtube_dl.utils import encodeFilename
-from youtube_dl.downloader.external import Aria2pFD
+from youtube_dl.compat import (
+    compat_contextlib_suppress,
+    compat_http_cookiejar_Cookie,
+    compat_http_server,
+    compat_kwargs,
+)
+from youtube_dl.utils import (
+    encodeFilename,
+    join_nonempty,
+)
+from youtube_dl.downloader.external import (
+    Aria2cFD,
+    Aria2pFD,
+    AxelFD,
+    CurlFD,
+    FFmpegFD,
+    HttpieFD,
+    WgetFD,
+)
+from youtube_dl.postprocessor import (
+    FFmpegPostProcessor,
+)
 import threading
 
-TEST_DIR = os.path.dirname(os.path.abspath(__file__))
-
-
 TEST_SIZE = 10 * 1024
 
+TEST_COOKIE = {
+    'version': 0,
+    'name': 'test',
+    'value': 'ytdlp',
+    'port': None,
+    'port_specified': False,
+    'domain': '.example.com',
+    'domain_specified': True,
+    'domain_initial_dot': False,
+    'path': '/',
+    'path_specified': True,
+    'secure': False,
+    'expires': None,
+    'discard': False,
+    'comment': None,
+    'comment_url': None,
+    'rest': {},
+}
+
+TEST_COOKIE_VALUE = join_nonempty('name', 'value', delim='=', from_dict=TEST_COOKIE)
+
+TEST_INFO = {'url': 'http://www.example.com/'}
+
+
+def cookiejar_Cookie(**cookie_args):
+    return compat_http_cookiejar_Cookie(**compat_kwargs(cookie_args))
+
+
+def ifExternalFDAvailable(externalFD):
+    return unittest.skipUnless(externalFD.available(),
+                               externalFD.get_basename() + ' not found')
+
 
 class HTTPTestRequestHandler(compat_http_server.BaseHTTPRequestHandler):
     def log_message(self, format, *args):
@@ -70,7 +119,7 @@ class HTTPTestRequestHandler(compat_http_server.BaseHTTPRequestHandler):
             assert False, 'unrecognised server path'
 
 
-@unittest.skipUnless(Aria2pFD.available(), 'aria2p module not found')
+@ifExternalFDAvailable(Aria2pFD)
 class TestAria2pFD(unittest.TestCase):
     def setUp(self):
         self.httpd = compat_http_server.HTTPServer(
@@ -111,5 +160,113 @@ class TestAria2pFD(unittest.TestCase):
         })
 
 
+@ifExternalFDAvailable(HttpieFD)
+class TestHttpieFD(unittest.TestCase):
+    def test_make_cmd(self):
+        with FakeYDL() as ydl:
+            downloader = HttpieFD(ydl, {})
+            self.assertEqual(
+                downloader._make_cmd('test', TEST_INFO),
+                ['http', '--download', '--output', 'test', 'http://www.example.com/'])
+
+            # Test cookie header is added
+            ydl.cookiejar.set_cookie(cookiejar_Cookie(**TEST_COOKIE))
+            self.assertEqual(
+                downloader._make_cmd('test', TEST_INFO),
+                ['http', '--download', '--output', 'test',
+                 'http://www.example.com/', 'Cookie:' + TEST_COOKIE_VALUE])
+
+
+@ifExternalFDAvailable(AxelFD)
+class TestAxelFD(unittest.TestCase):
+    def test_make_cmd(self):
+        with FakeYDL() as ydl:
+            downloader = AxelFD(ydl, {})
+            self.assertEqual(
+                downloader._make_cmd('test', TEST_INFO),
+                ['axel', '-o', 'test', '--', 'http://www.example.com/'])
+
+            # Test cookie header is added
+            ydl.cookiejar.set_cookie(cookiejar_Cookie(**TEST_COOKIE))
+            self.assertEqual(
+                downloader._make_cmd('test', TEST_INFO),
+                ['axel', '-o', 'test', '-H', 'Cookie: ' + TEST_COOKIE_VALUE,
+                 '--max-redirect=0', '--', 'http://www.example.com/'])
+
+
+@ifExternalFDAvailable(WgetFD)
+class TestWgetFD(unittest.TestCase):
+    def test_make_cmd(self):
+        with FakeYDL() as ydl:
+            downloader = WgetFD(ydl, {})
+            self.assertNotIn('--load-cookies', downloader._make_cmd('test', TEST_INFO))
+            # Test cookiejar tempfile arg is added
+            ydl.cookiejar.set_cookie(cookiejar_Cookie(**TEST_COOKIE))
+            self.assertIn('--load-cookies', downloader._make_cmd('test', TEST_INFO))
+
+
+@ifExternalFDAvailable(CurlFD)
+class TestCurlFD(unittest.TestCase):
+    def test_make_cmd(self):
+        with FakeYDL() as ydl:
+            downloader = CurlFD(ydl, {})
+            self.assertNotIn('--cookie', downloader._make_cmd('test', TEST_INFO))
+            # Test cookie header is added
+            ydl.cookiejar.set_cookie(cookiejar_Cookie(**TEST_COOKIE))
+            self.assertIn('--cookie', downloader._make_cmd('test', TEST_INFO))
+            self.assertIn(TEST_COOKIE_VALUE, downloader._make_cmd('test', TEST_INFO))
+
+
+@ifExternalFDAvailable(Aria2cFD)
+class TestAria2cFD(unittest.TestCase):
+    def test_make_cmd(self):
+        with FakeYDL() as ydl:
+            downloader = Aria2cFD(ydl, {})
+            downloader._make_cmd('test', TEST_INFO)
+            self.assertFalse(hasattr(downloader, '_cookies_tempfile'))
+
+            # Test cookiejar tempfile arg is added
+            ydl.cookiejar.set_cookie(cookiejar_Cookie(**TEST_COOKIE))
+            cmd = downloader._make_cmd('test', TEST_INFO)
+            self.assertIn('--load-cookies=%s' % downloader._cookies_tempfile, cmd)
+
+
+# Handle delegated availability
+def ifFFmpegFDAvailable(externalFD):
+    # raise SkipTest, or set False!
+    avail = ifExternalFDAvailable(externalFD) and False
+    with compat_contextlib_suppress(Exception):
+        avail = FFmpegPostProcessor(downloader=None).available
+    return unittest.skipUnless(
+        avail, externalFD.get_basename() + ' not found')
+
+
+@ifFFmpegFDAvailable(FFmpegFD)
+class TestFFmpegFD(unittest.TestCase):
+    _args = []
+
+    def _test_cmd(self, args):
+        self._args = args
+
+    def test_make_cmd(self):
+        with FakeYDL() as ydl:
+            downloader = FFmpegFD(ydl, {})
+            downloader._debug_cmd = self._test_cmd
+            info_dict = TEST_INFO.copy()
+            info_dict['ext'] = 'mp4'
+
+            downloader._call_downloader('test', info_dict)
+            self.assertEqual(self._args, [
+                'ffmpeg', '-y', '-i', 'http://www.example.com/',
+                '-c', 'copy', '-f', 'mp4', 'file:test'])
+
+            # Test cookies arg is added
+            ydl.cookiejar.set_cookie(cookiejar_Cookie(**TEST_COOKIE))
+            downloader._call_downloader('test', info_dict)
+            self.assertEqual(self._args, [
+                'ffmpeg', '-y', '-cookies', TEST_COOKIE_VALUE + '; path=/; domain=.example.com;\r\n',
+                '-i', 'http://www.example.com/', '-c', 'copy', '-f', 'mp4', 'file:test'])
+
+
 if __name__ == '__main__':
     unittest.main()

test/test_execution.py

@@ -8,47 +8,53 @@ import unittest
 import sys
 import os
 import subprocess
-sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
-
-from youtube_dl.utils import encodeArgument
 
 rootDir = os.path.dirname(os.path.dirname(os.path.abspath(__file__)))
 
+sys.path.insert(0, rootDir)
 
-try:
-    _DEV_NULL = subprocess.DEVNULL
-except AttributeError:
-    _DEV_NULL = open(os.devnull, 'wb')
+from youtube_dl.compat import compat_register_utf8, compat_subprocess_get_DEVNULL
+from youtube_dl.utils import encodeArgument
+
+compat_register_utf8()
+
+
+_DEV_NULL = compat_subprocess_get_DEVNULL()
 
 
 class TestExecution(unittest.TestCase):
+    def setUp(self):
+        self.module = 'youtube_dl'
+        if sys.version_info < (2, 7):
+            self.module += '.__main__'
+
     def test_import(self):
         subprocess.check_call([sys.executable, '-c', 'import youtube_dl'], cwd=rootDir)
 
     def test_module_exec(self):
-        if sys.version_info >= (2, 7):  # Python 2.6 doesn't support package execution
-            subprocess.check_call([sys.executable, '-m', 'youtube_dl', '--version'], cwd=rootDir, stdout=_DEV_NULL)
+        subprocess.check_call([sys.executable, '-m', self.module, '--version'], cwd=rootDir, stdout=_DEV_NULL)
 
     def test_main_exec(self):
-        subprocess.check_call([sys.executable, 'youtube_dl/__main__.py', '--version'], cwd=rootDir, stdout=_DEV_NULL)
+        subprocess.check_call([sys.executable, os.path.normpath('youtube_dl/__main__.py'), '--version'], cwd=rootDir, stdout=_DEV_NULL)
 
     def test_cmdline_umlauts(self):
+        os.environ['PYTHONIOENCODING'] = 'utf-8'
         p = subprocess.Popen(
-            [sys.executable, 'youtube_dl/__main__.py', encodeArgument('ä'), '--version'],
+            [sys.executable, '-m', self.module, encodeArgument('ä'), '--version'],
             cwd=rootDir, stdout=_DEV_NULL, stderr=subprocess.PIPE)
         _, stderr = p.communicate()
         self.assertFalse(stderr)
 
     def test_lazy_extractors(self):
-        lazy_extractors = 'youtube_dl/extractor/lazy_extractors.py'
+        lazy_extractors = os.path.normpath('youtube_dl/extractor/lazy_extractors.py')
         try:
-            subprocess.check_call([sys.executable, 'devscripts/make_lazy_extractors.py', lazy_extractors], cwd=rootDir, stdout=_DEV_NULL)
-            subprocess.check_call([sys.executable, 'test/test_all_urls.py'], cwd=rootDir, stdout=_DEV_NULL)
+            subprocess.check_call([sys.executable, os.path.normpath('devscripts/make_lazy_extractors.py'), lazy_extractors], cwd=rootDir, stdout=_DEV_NULL)
+            subprocess.check_call([sys.executable, os.path.normpath('test/test_all_urls.py')], cwd=rootDir, stdout=_DEV_NULL)
         finally:
-            for x in ['', 'c'] if sys.version_info[0] < 3 else ['']:
+            for x in ('', 'c') if sys.version_info[0] < 3 else ('',):
                 try:
                     os.remove(lazy_extractors + x)
-                except (IOError, OSError):
+                except OSError:
                     pass
 
 

test/test_http.py

@@ -8,33 +8,163 @@ import sys
 import unittest
 sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
 
+import contextlib
+import gzip
+import io
+import ssl
+import tempfile
+import threading
+import zlib
+
+# avoid deprecated alias assertRaisesRegexp
+if hasattr(unittest.TestCase, 'assertRaisesRegex'):
+    unittest.TestCase.assertRaisesRegexp = unittest.TestCase.assertRaisesRegex
+
+try:
+    import brotli
+except ImportError:
+    brotli = None
+try:
+    from urllib.request import pathname2url
+except ImportError:
+    from urllib import pathname2url
+
+from youtube_dl.compat import (
+    compat_http_cookiejar_Cookie,
+    compat_http_server,
+    compat_str as str,
+    compat_urllib_error,
+    compat_urllib_HTTPError,
+    compat_urllib_parse,
+    compat_urllib_request,
+)
+
+from youtube_dl.utils import (
+    sanitized_Request,
+    update_Request,
+    urlencode_postdata,
+)
+
 from test.helper import (
+    expectedFailureIf,
+    FakeYDL,
     FakeLogger,
     http_server_port,
 )
 from youtube_dl import YoutubeDL
-from youtube_dl.compat import compat_http_server, compat_urllib_request
-import ssl
-import threading
 
 TEST_DIR = os.path.dirname(os.path.abspath(__file__))
 
 
 class HTTPTestRequestHandler(compat_http_server.BaseHTTPRequestHandler):
+    protocol_version = 'HTTP/1.1'
+
+    # work-around old/new -style class inheritance
+    def super(self, meth_name, *args, **kwargs):
+        from types import MethodType
+        try:
+            super()
+            fn = lambda s, m, *a, **k: getattr(super(), m)(*a, **k)
+        except TypeError:
+            fn = lambda s, m, *a, **k: getattr(compat_http_server.BaseHTTPRequestHandler, m)(s, *a, **k)
+        self.super = MethodType(fn, self)
+        return self.super(meth_name, *args, **kwargs)
+
     def log_message(self, format, *args):
         pass
 
+    def _headers(self):
+        payload = str(self.headers).encode('utf-8')
+        self.send_response(200)
+        self.send_header('Content-Type', 'application/json')
+        self.send_header('Content-Length', str(len(payload)))
+        self.end_headers()
+        self.wfile.write(payload)
+
+    def _redirect(self):
+        self.send_response(int(self.path[len('/redirect_'):]))
+        self.send_header('Location', '/method')
+        self.send_header('Content-Length', '0')
+        self.end_headers()
+
+    def _method(self, method, payload=None):
+        self.send_response(200)
+        self.send_header('Content-Length', str(len(payload or '')))
+        self.send_header('Method', method)
+        self.end_headers()
+        if payload:
+            self.wfile.write(payload)
+
+    def _status(self, status):
+        payload = '<html>{0} NOT FOUND</html>'.format(status).encode('utf-8')
+        self.send_response(int(status))
+        self.send_header('Content-Type', 'text/html; charset=utf-8')
+        self.send_header('Content-Length', str(len(payload)))
+        self.end_headers()
+        self.wfile.write(payload)
+
+    def _read_data(self):
+        if 'Content-Length' in self.headers:
+            return self.rfile.read(int(self.headers['Content-Length']))
+
+    def _test_url(self, path, host='127.0.0.1', scheme='http', port=None):
+        return '{0}://{1}:{2}/{3}'.format(
+            scheme, host,
+            port if port is not None
+            else http_server_port(self.server), path)
+
+    def do_POST(self):
+        data = self._read_data()
+        if self.path.startswith('/redirect_'):
+            self._redirect()
+        elif self.path.startswith('/method'):
+            self._method('POST', data)
+        elif self.path.startswith('/headers'):
+            self._headers()
+        else:
+            self._status(404)
+
+    def do_HEAD(self):
+        if self.path.startswith('/redirect_'):
+            self._redirect()
+        elif self.path.startswith('/method'):
+            self._method('HEAD')
+        else:
+            self._status(404)
+
+    def do_PUT(self):
+        data = self._read_data()
+        if self.path.startswith('/redirect_'):
+            self._redirect()
+        elif self.path.startswith('/method'):
+            self._method('PUT', data)
+        else:
+            self._status(404)
+
     def do_GET(self):
+
+        def respond(payload=b'<html><video src="/vid.mp4" /></html>',
+                    payload_type='text/html; charset=utf-8',
+                    payload_encoding=None,
+                    resp_code=200):
+            self.send_response(resp_code)
+            self.send_header('Content-Type', payload_type)
+            if payload_encoding:
+                self.send_header('Content-Encoding', payload_encoding)
+            self.send_header('Content-Length', str(len(payload)))  # required for persistent connections
+            self.end_headers()
+            self.wfile.write(payload)
+
+        def gzip_compress(p):
+            buf = io.BytesIO()
+            with contextlib.closing(gzip.GzipFile(fileobj=buf, mode='wb')) as f:
+                f.write(p)
+            return buf.getvalue()
+
         if self.path == '/video.html':
-            self.send_response(200)
-            self.send_header('Content-Type', 'text/html; charset=utf-8')
-            self.end_headers()
-            self.wfile.write(b'<html><video src="/vid.mp4" /></html>')
+            respond()
         elif self.path == '/vid.mp4':
-            self.send_response(200)
-            self.send_header('Content-Type', 'video/mp4')
-            self.end_headers()
-            self.wfile.write(b'\x00\x00\x00\x00\x20\x66\x74[video]')
+            respond(b'\x00\x00\x00\x00\x20\x66\x74[video]', 'video/mp4')
         elif self.path == '/302':
             if sys.version_info[0] == 3:
                 # XXX: Python 3 http server does not allow non-ASCII header values
@@ -42,60 +172,336 @@ class HTTPTestRequestHandler(compat_http_server.BaseHTTPRequestHandler):
                 self.end_headers()
                 return
 
-            new_url = 'http://127.0.0.1:%d/中文.html' % http_server_port(self.server)
+            new_url = self._test_url('中文.html')
             self.send_response(302)
             self.send_header(b'Location', new_url.encode('utf-8'))
             self.end_headers()
         elif self.path == '/%E4%B8%AD%E6%96%87.html':
-            self.send_response(200)
-            self.send_header('Content-Type', 'text/html; charset=utf-8')
+            respond()
+        elif self.path == '/%c7%9f':
+            respond()
+        elif self.path == '/redirect_dotsegments':
+            self.send_response(301)
+            # redirect to /headers but with dot segments before
+            self.send_header('Location', '/a/b/./../../headers')
+            self.send_header('Content-Length', '0')
             self.end_headers()
-            self.wfile.write(b'<html><video src="/vid.mp4" /></html>')
+        elif self.path.startswith('/redirect_'):
+            self._redirect()
+        elif self.path.startswith('/method'):
+            self._method('GET')
+        elif self.path.startswith('/headers'):
+            self._headers()
+        elif self.path.startswith('/308-to-headers'):
+            self.send_response(308)
+            self.send_header('Location', '/headers')
+            self.send_header('Content-Length', '0')
+            self.end_headers()
+        elif self.path == '/trailing_garbage':
+            payload = b'<html><video src="/vid.mp4" /></html>'
+            compressed = gzip_compress(payload) + b'trailing garbage'
+            respond(compressed, payload_encoding='gzip')
+        elif self.path == '/302-non-ascii-redirect':
+            new_url = self._test_url('中文.html')
+            # actually respond with permanent redirect
+            self.send_response(301)
+            self.send_header('Location', new_url)
+            self.send_header('Content-Length', '0')
+            self.end_headers()
+        elif self.path == '/content-encoding':
+            encodings = self.headers.get('ytdl-encoding', '')
+            payload = b'<html><video src="/vid.mp4" /></html>'
+            for encoding in filter(None, (e.strip() for e in encodings.split(','))):
+                if encoding == 'br' and brotli:
+                    payload = brotli.compress(payload)
+                elif encoding == 'gzip':
+                    payload = gzip_compress(payload)
+                elif encoding == 'deflate':
+                    payload = zlib.compress(payload)
+                elif encoding == 'unsupported':
+                    payload = b'raw'
+                    break
+                else:
+                    self._status(415)
+                    return
+            respond(payload, payload_encoding=encodings)
+
         else:
-            assert False
+            self._status(404)
+
+    def send_header(self, keyword, value):
+        """
+        Forcibly allow HTTP server to send non percent-encoded non-ASCII characters in headers.
+        This is against what is defined in RFC 3986: but we need to test that we support this
+        since some sites incorrectly do this.
+        """
+        if keyword.lower() == 'connection':
+            return self.super('send_header', keyword, value)
+
+        if not hasattr(self, '_headers_buffer'):
+            self._headers_buffer = []
+
+        self._headers_buffer.append('{0}: {1}\r\n'.format(keyword, value).encode('utf-8'))
+
+    def end_headers(self):
+        if hasattr(self, '_headers_buffer'):
+            self.wfile.write(b''.join(self._headers_buffer))
+            self._headers_buffer = []
+        self.super('end_headers')
 
 
 class TestHTTP(unittest.TestCase):
+    # when does it make sense to check the SSL certificate?
+    _check_cert = (
+        sys.version_info >= (3, 2)
+        or (sys.version_info[0] == 2 and sys.version_info[1:] >= (7, 19)))
+
     def setUp(self):
-        self.httpd = compat_http_server.HTTPServer(
+        # HTTP server
+        self.http_httpd = compat_http_server.HTTPServer(
             ('127.0.0.1', 0), HTTPTestRequestHandler)
-        self.port = http_server_port(self.httpd)
-        self.server_thread = threading.Thread(target=self.httpd.serve_forever)
-        self.server_thread.daemon = True
-        self.server_thread.start()
+        self.http_port = http_server_port(self.http_httpd)
+
+        self.http_server_thread = threading.Thread(target=self.http_httpd.serve_forever)
+        self.http_server_thread.daemon = True
+        self.http_server_thread.start()
+
+        try:
+            from http.server import ThreadingHTTPServer
+        except ImportError:
+            try:
+                from socketserver import ThreadingMixIn
+            except ImportError:
+                from SocketServer import ThreadingMixIn
+
+            class ThreadingHTTPServer(ThreadingMixIn, compat_http_server.HTTPServer):
+                pass
+
+        # HTTPS server
+        certfn = os.path.join(TEST_DIR, 'testcert.pem')
+        self.https_httpd = ThreadingHTTPServer(
+            ('127.0.0.1', 0), HTTPTestRequestHandler)
+        try:
+            sslctx = ssl.SSLContext(ssl.PROTOCOL_SSLv23)
+            sslctx.verify_mode = ssl.CERT_NONE
+            sslctx.check_hostname = False
+            sslctx.load_cert_chain(certfn, None)
+            self.https_httpd.socket = sslctx.wrap_socket(
+                self.https_httpd.socket, server_side=True)
+        except AttributeError:
+            self.https_httpd.socket = ssl.wrap_socket(
+                self.https_httpd.socket, certfile=certfn, server_side=True)
+
+        self.https_port = http_server_port(self.https_httpd)
+        self.https_server_thread = threading.Thread(target=self.https_httpd.serve_forever)
+        self.https_server_thread.daemon = True
+        self.https_server_thread.start()
+
+    def tearDown(self):
+
+        def closer(svr):
+            def _closer():
+                svr.shutdown()
+                svr.server_close()
+            return _closer
+
+        shutdown_thread = threading.Thread(target=closer(self.http_httpd))
+        shutdown_thread.start()
+        self.http_server_thread.join(2.0)
+
+        shutdown_thread = threading.Thread(target=closer(self.https_httpd))
+        shutdown_thread.start()
+        self.https_server_thread.join(2.0)
+
+    def _test_url(self, path, host='127.0.0.1', scheme='http', port=None):
+        return '{0}://{1}:{2}/{3}'.format(
+            scheme, host,
+            port if port is not None
+            else self.https_port if scheme == 'https'
+            else self.http_port, path)
+
+    @unittest.skipUnless(_check_cert, 'No support for certificate check in SSL')
+    def test_nocheckcertificate(self):
+        with FakeYDL({'logger': FakeLogger()}) as ydl:
+            with self.assertRaises(compat_urllib_error.URLError):
+                ydl.urlopen(sanitized_Request(self._test_url('headers', scheme='https')))
+
+        with FakeYDL({'logger': FakeLogger(), 'nocheckcertificate': True}) as ydl:
+            r = ydl.urlopen(sanitized_Request(self._test_url('headers', scheme='https')))
+            self.assertEqual(r.getcode(), 200)
+            r.close()
+
+    def test_percent_encode(self):
+        with FakeYDL() as ydl:
+            # Unicode characters should be encoded with uppercase percent-encoding
+            res = ydl.urlopen(sanitized_Request(self._test_url('中文.html')))
+            self.assertEqual(res.getcode(), 200)
+            res.close()
+            # don't normalize existing percent encodings
+            res = ydl.urlopen(sanitized_Request(self._test_url('%c7%9f')))
+            self.assertEqual(res.getcode(), 200)
+            res.close()
 
     def test_unicode_path_redirection(self):
-        # XXX: Python 3 http server does not allow non-ASCII header values
-        if sys.version_info[0] == 3:
-            return
+        with FakeYDL() as ydl:
+            r = ydl.urlopen(sanitized_Request(self._test_url('302-non-ascii-redirect')))
+            self.assertEqual(r.url, self._test_url('%E4%B8%AD%E6%96%87.html'))
+            r.close()
 
-        ydl = YoutubeDL({'logger': FakeLogger()})
-        r = ydl.extract_info('http://127.0.0.1:%d/302' % self.port)
-        self.assertEqual(r['entries'][0]['url'], 'http://127.0.0.1:%d/vid.mp4' % self.port)
+    def test_redirect(self):
+        with FakeYDL() as ydl:
+            def do_req(redirect_status, method, check_no_content=False):
+                data = b'testdata' if method in ('POST', 'PUT') else None
+                res = ydl.urlopen(sanitized_Request(
+                    self._test_url('redirect_{0}'.format(redirect_status)),
+                    method=method, data=data))
+                if check_no_content:
+                    self.assertNotIn('Content-Type', res.headers)
+                return res.read().decode('utf-8'), res.headers.get('method', '')
+            # A 303 must either use GET or HEAD for subsequent request
+            self.assertEqual(do_req(303, 'POST'), ('', 'GET'))
+            self.assertEqual(do_req(303, 'HEAD'), ('', 'HEAD'))
 
+            self.assertEqual(do_req(303, 'PUT'), ('', 'GET'))
 
-class TestHTTPS(unittest.TestCase):
-    def setUp(self):
-        certfn = os.path.join(TEST_DIR, 'testcert.pem')
-        self.httpd = compat_http_server.HTTPServer(
-            ('127.0.0.1', 0), HTTPTestRequestHandler)
-        self.httpd.socket = ssl.wrap_socket(
-            self.httpd.socket, certfile=certfn, server_side=True)
-        self.port = http_server_port(self.httpd)
-        self.server_thread = threading.Thread(target=self.httpd.serve_forever)
-        self.server_thread.daemon = True
-        self.server_thread.start()
+            # 301 and 302 turn POST only into a GET, with no Content-Type
+            self.assertEqual(do_req(301, 'POST', True), ('', 'GET'))
+            self.assertEqual(do_req(301, 'HEAD'), ('', 'HEAD'))
+            self.assertEqual(do_req(302, 'POST', True), ('', 'GET'))
+            self.assertEqual(do_req(302, 'HEAD'), ('', 'HEAD'))
 
-    def test_nocheckcertificate(self):
-        if sys.version_info >= (2, 7, 9):  # No certificate checking anyways
-            ydl = YoutubeDL({'logger': FakeLogger()})
-            self.assertRaises(
-                Exception,
-                ydl.extract_info, 'https://127.0.0.1:%d/video.html' % self.port)
+            self.assertEqual(do_req(301, 'PUT'), ('testdata', 'PUT'))
+            self.assertEqual(do_req(302, 'PUT'), ('testdata', 'PUT'))
 
-        ydl = YoutubeDL({'logger': FakeLogger(), 'nocheckcertificate': True})
-        r = ydl.extract_info('https://127.0.0.1:%d/video.html' % self.port)
-        self.assertEqual(r['entries'][0]['url'], 'https://127.0.0.1:%d/vid.mp4' % self.port)
+            # 307 and 308 should not change method
+            for m in ('POST', 'PUT'):
+                self.assertEqual(do_req(307, m), ('testdata', m))
+                self.assertEqual(do_req(308, m), ('testdata', m))
+
+            self.assertEqual(do_req(307, 'HEAD'), ('', 'HEAD'))
+            self.assertEqual(do_req(308, 'HEAD'), ('', 'HEAD'))
+
+            # These should not redirect and instead raise an HTTPError
+            for code in (300, 304, 305, 306):
+                with self.assertRaises(compat_urllib_HTTPError):
+                    do_req(code, 'GET')
+
+    # Jython 2.7.1 times out for some reason
+    @expectedFailureIf(sys.platform.startswith('java') and sys.version_info < (2, 7, 2))
+    def test_content_type(self):
+        # https://github.com/yt-dlp/yt-dlp/commit/379a4f161d4ad3e40932dcf5aca6e6fb9715ab28
+        with FakeYDL({'nocheckcertificate': True}) as ydl:
+            # method should be auto-detected as POST
+            r = sanitized_Request(self._test_url('headers', scheme='https'), data=urlencode_postdata({'test': 'test'}))
+
+            headers = ydl.urlopen(r).read().decode('utf-8')
+            self.assertIn('Content-Type: application/x-www-form-urlencoded', headers)
+
+            # test http
+            r = sanitized_Request(self._test_url('headers'), data=urlencode_postdata({'test': 'test'}))
+            headers = ydl.urlopen(r).read().decode('utf-8')
+            self.assertIn('Content-Type: application/x-www-form-urlencoded', headers)
+
+    def test_update_req(self):
+        req = sanitized_Request('http://example.com')
+        assert req.data is None
+        assert req.get_method() == 'GET'
+        assert not req.has_header('Content-Type')
+        # Test that zero-byte payloads will be sent
+        req = update_Request(req, data=b'')
+        assert req.data == b''
+        assert req.get_method() == 'POST'
+        # yt-dl expects data to be encoded and Content-Type to be added by sender
+        # assert req.get_header('Content-Type') == 'application/x-www-form-urlencoded'
+
+    def test_cookiejar(self):
+        with FakeYDL() as ydl:
+            ydl.cookiejar.set_cookie(compat_http_cookiejar_Cookie(
+                0, 'test', 'ytdl', None, False, '127.0.0.1', True,
+                False, '/headers', True, False, None, False, None, None, {}))
+            data = ydl.urlopen(sanitized_Request(
+                self._test_url('headers'))).read().decode('utf-8')
+            self.assertIn('Cookie: test=ytdl', data)
+
+    def test_passed_cookie_header(self):
+        # We should accept a Cookie header being passed as in normal headers and handle it appropriately.
+        with FakeYDL() as ydl:
+            # Specified Cookie header should be used
+            res = ydl.urlopen(sanitized_Request(
+                self._test_url('headers'), headers={'Cookie': 'test=test'})).read().decode('utf-8')
+            self.assertIn('Cookie: test=test', res)
+
+            # Specified Cookie header should be removed on any redirect
+            res = ydl.urlopen(sanitized_Request(
+                self._test_url('308-to-headers'), headers={'Cookie': 'test=test'})).read().decode('utf-8')
+            self.assertNotIn('Cookie: test=test', res)
+
+            # Specified Cookie header should override global cookiejar for that request
+            ydl.cookiejar.set_cookie(compat_http_cookiejar_Cookie(
+                0, 'test', 'ytdlp', None, False, '127.0.0.1', True,
+                False, '/headers', True, False, None, False, None, None, {}))
+            data = ydl.urlopen(sanitized_Request(
+                self._test_url('headers'), headers={'Cookie': 'test=test'})).read().decode('utf-8')
+            self.assertNotIn('Cookie: test=ytdlp', data)
+            self.assertIn('Cookie: test=test', data)
+
+    def test_no_compression_compat_header(self):
+        with FakeYDL() as ydl:
+            data = ydl.urlopen(
+                sanitized_Request(
+                    self._test_url('headers'),
+                    headers={'Youtubedl-no-compression': True})).read()
+            self.assertIn(b'Accept-Encoding: identity', data)
+            self.assertNotIn(b'youtubedl-no-compression', data.lower())
+
+    def test_gzip_trailing_garbage(self):
+        # https://github.com/ytdl-org/youtube-dl/commit/aa3e950764337ef9800c936f4de89b31c00dfcf5
+        # https://github.com/ytdl-org/youtube-dl/commit/6f2ec15cee79d35dba065677cad9da7491ec6e6f
+        with FakeYDL() as ydl:
+            data = ydl.urlopen(sanitized_Request(self._test_url('trailing_garbage'))).read().decode('utf-8')
+            self.assertEqual(data, '<html><video src="/vid.mp4" /></html>')
+
+    def __test_compression(self, encoding):
+        with FakeYDL() as ydl:
+            res = ydl.urlopen(
+                sanitized_Request(
+                    self._test_url('content-encoding'),
+                    headers={'ytdl-encoding': encoding}))
+            # decoded encodings are removed: only check for valid decompressed data
+            self.assertEqual(res.read(), b'<html><video src="/vid.mp4" /></html>')
+
+    @unittest.skipUnless(brotli, 'brotli support is not installed')
+    def test_brotli(self):
+        self.__test_compression('br')
+
+    def test_deflate(self):
+        self.__test_compression('deflate')
+
+    def test_gzip(self):
+        self.__test_compression('gzip')
+
+    def test_multiple_encodings(self):
+        # https://www.rfc-editor.org/rfc/rfc9110.html#section-8.4
+        for pair in ('gzip,deflate', 'deflate, gzip', 'gzip, gzip', 'deflate, deflate'):
+            self.__test_compression(pair)
+
+    def test_unsupported_encoding(self):
+        # it should return the raw content
+        with FakeYDL() as ydl:
+            res = ydl.urlopen(
+                sanitized_Request(
+                    self._test_url('content-encoding'),
+                    headers={'ytdl-encoding': 'unsupported'}))
+            self.assertEqual(res.headers.get('Content-Encoding'), 'unsupported')
+            self.assertEqual(res.read(), b'raw')
+
+    def test_remove_dot_segments(self):
+        with FakeYDL() as ydl:
+            res = ydl.urlopen(sanitized_Request(self._test_url('a/b/./../../headers')))
+            self.assertEqual(compat_urllib_parse.urlparse(res.geturl()).path, '/headers')
+
+            res = ydl.urlopen(sanitized_Request(self._test_url('redirect_dotsegments')))
+            self.assertEqual(compat_urllib_parse.urlparse(res.geturl()).path, '/headers')
 
 
 def _build_proxy_handler(name):
@@ -109,7 +515,7 @@ def _build_proxy_handler(name):
             self.send_response(200)
             self.send_header('Content-Type', 'text/plain; charset=utf-8')
             self.end_headers()
-            self.wfile.write('{self.proxy_name}: {self.path}'.format(self=self).encode('utf-8'))
+            self.wfile.write('{0}: {1}'.format(self.proxy_name, self.path).encode('utf-8'))
     return HTTPTestRequestHandler
 
 
@@ -129,10 +535,30 @@ class TestProxy(unittest.TestCase):
         self.geo_proxy_thread.daemon = True
         self.geo_proxy_thread.start()
 
+    def tearDown(self):
+
+        def closer(svr):
+            def _closer():
+                svr.shutdown()
+                svr.server_close()
+            return _closer
+
+        shutdown_thread = threading.Thread(target=closer(self.proxy))
+        shutdown_thread.start()
+        self.proxy_thread.join(2.0)
+
+        shutdown_thread = threading.Thread(target=closer(self.geo_proxy))
+        shutdown_thread.start()
+        self.geo_proxy_thread.join(2.0)
+
+    def _test_proxy(self, host='127.0.0.1', port=None):
+        return '{0}:{1}'.format(
+            host, port if port is not None else self.port)
+
     def test_proxy(self):
-        geo_proxy = '127.0.0.1:{0}'.format(self.geo_port)
+        geo_proxy = self._test_proxy(port=self.geo_port)
         ydl = YoutubeDL({
-            'proxy': '127.0.0.1:{0}'.format(self.port),
+            'proxy': self._test_proxy(),
             'geo_verification_proxy': geo_proxy,
         })
         url = 'http://foo.com/bar'
@@ -146,7 +572,7 @@ class TestProxy(unittest.TestCase):
 
     def test_proxy_with_idn(self):
         ydl = YoutubeDL({
-            'proxy': '127.0.0.1:{0}'.format(self.port),
+            'proxy': self._test_proxy(),
         })
         url = 'http://中文.tw/'
         response = ydl.urlopen(url).read().decode('utf-8')
@@ -154,5 +580,25 @@ class TestProxy(unittest.TestCase):
         self.assertEqual(response, 'normal: http://xn--fiq228c.tw/')
 
 
+class TestFileURL(unittest.TestCase):
+    # See https://github.com/ytdl-org/youtube-dl/issues/8227
+    def test_file_urls(self):
+        tf = tempfile.NamedTemporaryFile(delete=False)
+        tf.write(b'foobar')
+        tf.close()
+        url = compat_urllib_parse.urljoin('file://', pathname2url(tf.name))
+        with FakeYDL() as ydl:
+            self.assertRaisesRegexp(
+                compat_urllib_error.URLError, 'file:// scheme is explicitly disabled in youtube-dl for security reasons', ydl.urlopen, url)
+        # not yet implemented
+        """
+        with FakeYDL({'enable_file_urls': True}) as ydl:
+            res = ydl.urlopen(url)
+            self.assertEqual(res.read(), b'foobar')
+            res.close()
+        """
+        os.unlink(tf.name)
+
+
 if __name__ == '__main__':
     unittest.main()

test/test_jsinterp.py

@@ -492,10 +492,12 @@ class TestJSInterpreter(unittest.TestCase):
         jsi = JSInterpreter('''
         function x() { let a=/,,[/,913,/](,)}/; "".replace(a, ""); return a; }
         ''')
-        attrs = set(('findall', 'finditer', 'flags', 'groupindex',
-                     'groups', 'match', 'pattern', 'scanner',
-                     'search', 'split', 'sub', 'subn'))
-        self.assertTrue(set(dir(jsi.call_function('x'))) > attrs)
+        attrs = set(('findall', 'finditer', 'match', 'scanner', 'search',
+                     'split', 'sub', 'subn'))
+        if sys.version_info >= (2, 7):
+            # documented for 2.6 but may not be found
+            attrs.update(('flags', 'groupindex', 'groups', 'pattern'))
+        self.assertSetEqual(set(dir(jsi.call_function('x'))) & attrs, attrs)
 
         jsi = JSInterpreter('''
         function x() { let a=/,,[/,913,/](,)}/i; return a; }
@@ -575,9 +577,11 @@ class TestJSInterpreter(unittest.TestCase):
     def test_unary_operators(self):
         jsi = JSInterpreter('function f(){return 2  -  - - 2;}')
         self.assertEqual(jsi.call_function('f'), 0)
-        # fails
-        # jsi = JSInterpreter('function f(){return 2 + - + - - 2;}')
-        # self.assertEqual(jsi.call_function('f'), 0)
+        jsi = JSInterpreter('function f(){return 2 + - + - - 2;}')
+        self.assertEqual(jsi.call_function('f'), 0)
+        # https://github.com/ytdl-org/youtube-dl/issues/32815
+        jsi = JSInterpreter('function f(){return 0  - 7 * - 6;}')
+        self.assertEqual(jsi.call_function('f'), 42)
 
     """ # fails so far
     def test_packed(self):

test/test_subtitles.py

@@ -295,6 +295,7 @@ class TestNRKSubtitles(BaseTestSubtitles):
     def test_allsubtitles(self):
         self.DL.params['writesubtitles'] = True
         self.DL.params['allsubtitles'] = True
+        self.DL.params['format'] = 'best/bestvideo'
         subtitles = self.getSubtitles()
         self.assertEqual(set(subtitles.keys()), set(['nb-ttv']))
         self.assertEqual(md5(subtitles['nb-ttv']), '67e06ff02d0deaf975e68f6cb8f6a149')

test/test_swfinterp.py

@@ -5,16 +5,18 @@ from __future__ import unicode_literals
 import os
 import sys
 import unittest
-sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
 
+dirn = os.path.dirname
+
+sys.path.insert(0, dirn(dirn(os.path.abspath(__file__))))
 
 import errno
-import io
 import json
 import re
 import subprocess
 
 from youtube_dl.swfinterp import SWFInterpreter
+from youtube_dl.compat import compat_open as open
 
 
 TEST_DIR = os.path.join(
@@ -43,7 +45,7 @@ def _make_testfunc(testfile):
                     '-static-link-runtime-shared-libraries', as_file])
             except OSError as ose:
                 if ose.errno == errno.ENOENT:
-                    print('mxmlc not found! Skipping test.')
+                    self.skipTest('mxmlc not found!')
                     return
                 raise
 
@@ -51,7 +53,7 @@ def _make_testfunc(testfile):
             swf_content = swf_f.read()
         swfi = SWFInterpreter(swf_content)
 
-        with io.open(as_file, 'r', encoding='utf-8') as as_f:
+        with open(as_file, 'r', encoding='utf-8') as as_f:
             as_content = as_f.read()
 
         def _find_spec(key):

test/test_traversal.py

@@ -0,0 +1,509 @@
+#!/usr/bin/env python
+# coding: utf-8
+from __future__ import unicode_literals
+
+# Allow direct execution
+import os
+import sys
+import unittest
+sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
+
+
+import re
+
+from youtube_dl.traversal import (
+    dict_get,
+    get_first,
+    T,
+    traverse_obj,
+)
+from youtube_dl.compat import (
+    compat_etree_fromstring,
+    compat_http_cookies,
+    compat_str,
+)
+from youtube_dl.utils import (
+    int_or_none,
+    str_or_none,
+)
+
+_TEST_DATA = {
+    100: 100,
+    1.2: 1.2,
+    'str': 'str',
+    'None': None,
+    '...': Ellipsis,
+    'urls': [
+        {'index': 0, 'url': 'https://www.example.com/0'},
+        {'index': 1, 'url': 'https://www.example.com/1'},
+    ],
+    'data': (
+        {'index': 2},
+        {'index': 3},
+    ),
+    'dict': {},
+}
+
+
+if sys.version_info < (3, 0):
+    class _TestCase(unittest.TestCase):
+
+        def assertCountEqual(self, *args, **kwargs):
+            return self.assertItemsEqual(*args, **kwargs)
+else:
+    _TestCase = unittest.TestCase
+
+
+class TestTraversal(_TestCase):
+    def assertMaybeCountEqual(self, *args, **kwargs):
+        if sys.version_info < (3, 7):
+            # random dict order
+            return self.assertCountEqual(*args, **kwargs)
+        else:
+            return self.assertEqual(*args, **kwargs)
+
+    def test_traverse_obj(self):
+        # instant compat
+        str = compat_str
+
+        # define a pukka Iterable
+        def iter_range(stop):
+            for from_ in range(stop):
+                yield from_
+
+        # Test base functionality
+        self.assertEqual(traverse_obj(_TEST_DATA, ('str',)), 'str',
+                         msg='allow tuple path')
+        self.assertEqual(traverse_obj(_TEST_DATA, ['str']), 'str',
+                         msg='allow list path')
+        self.assertEqual(traverse_obj(_TEST_DATA, (value for value in ("str",))), 'str',
+                         msg='allow iterable path')
+        self.assertEqual(traverse_obj(_TEST_DATA, 'str'), 'str',
+                         msg='single items should be treated as a path')
+        self.assertEqual(traverse_obj(_TEST_DATA, None), _TEST_DATA)
+        self.assertEqual(traverse_obj(_TEST_DATA, 100), 100)
+        self.assertEqual(traverse_obj(_TEST_DATA, 1.2), 1.2)
+
+        # Test Ellipsis behavior
+        self.assertCountEqual(traverse_obj(_TEST_DATA, Ellipsis),
+                              (item for item in _TEST_DATA.values() if item not in (None, {})),
+                              msg='`...` should give all non-discarded values')
+        self.assertCountEqual(traverse_obj(_TEST_DATA, ('urls', 0, Ellipsis)), _TEST_DATA['urls'][0].values(),
+                              msg='`...` selection for dicts should select all values')
+        self.assertEqual(traverse_obj(_TEST_DATA, (Ellipsis, Ellipsis, 'url')),
+                         ['https://www.example.com/0', 'https://www.example.com/1'],
+                         msg='nested `...` queries should work')
+        self.assertCountEqual(traverse_obj(_TEST_DATA, (Ellipsis, Ellipsis, 'index')), iter_range(4),
+                              msg='`...` query result should be flattened')
+        self.assertEqual(traverse_obj(iter(range(4)), Ellipsis), list(range(4)),
+                         msg='`...` should accept iterables')
+
+        # Test function as key
+        self.assertEqual(traverse_obj(_TEST_DATA, lambda x, y: x == 'urls' and isinstance(y, list)),
+                         [_TEST_DATA['urls']],
+                         msg='function as query key should perform a filter based on (key, value)')
+        self.assertCountEqual(traverse_obj(_TEST_DATA, lambda _, x: isinstance(x[0], str)), set(('str',)),
+                              msg='exceptions in the query function should be caught')
+        self.assertEqual(traverse_obj(iter(range(4)), lambda _, x: x % 2 == 0), [0, 2],
+                         msg='function key should accept iterables')
+        if __debug__:
+            with self.assertRaises(Exception, msg='Wrong function signature should raise in debug'):
+                traverse_obj(_TEST_DATA, lambda a: Ellipsis)
+            with self.assertRaises(Exception, msg='Wrong function signature should raise in debug'):
+                traverse_obj(_TEST_DATA, lambda a, b, c: Ellipsis)
+
+        # Test set as key (transformation/type, like `expected_type`)
+        self.assertEqual(traverse_obj(_TEST_DATA, (Ellipsis, T(str.upper), )), ['STR'],
+                         msg='Function in set should be a transformation')
+        self.assertEqual(traverse_obj(_TEST_DATA, ('fail', T(lambda _: 'const'))), 'const',
+                         msg='Function in set should always be called')
+        self.assertEqual(traverse_obj(_TEST_DATA, (Ellipsis, T(str))), ['str'],
+                         msg='Type in set should be a type filter')
+        self.assertMaybeCountEqual(traverse_obj(_TEST_DATA, (Ellipsis, T(str, int))), [100, 'str'],
+                                   msg='Multiple types in set should be a type filter')
+        self.assertEqual(traverse_obj(_TEST_DATA, T(dict)), _TEST_DATA,
+                         msg='A single set should be wrapped into a path')
+        self.assertEqual(traverse_obj(_TEST_DATA, (Ellipsis, T(str.upper))), ['STR'],
+                         msg='Transformation function should not raise')
+        self.assertMaybeCountEqual(traverse_obj(_TEST_DATA, (Ellipsis, T(str_or_none))),
+                                   [item for item in map(str_or_none, _TEST_DATA.values()) if item is not None],
+                                   msg='Function in set should be a transformation')
+        if __debug__:
+            with self.assertRaises(Exception, msg='Sets with length != 1 should raise in debug'):
+                traverse_obj(_TEST_DATA, set())
+            with self.assertRaises(Exception, msg='Sets with length != 1 should raise in debug'):
+                traverse_obj(_TEST_DATA, set((str.upper, str)))
+
+        # Test `slice` as a key
+        _SLICE_DATA = [0, 1, 2, 3, 4]
+        self.assertEqual(traverse_obj(_TEST_DATA, ('dict', slice(1))), None,
+                         msg='slice on a dictionary should not throw')
+        self.assertEqual(traverse_obj(_SLICE_DATA, slice(1)), _SLICE_DATA[:1],
+                         msg='slice key should apply slice to sequence')
+        self.assertEqual(traverse_obj(_SLICE_DATA, slice(1, 2)), _SLICE_DATA[1:2],
+                         msg='slice key should apply slice to sequence')
+        self.assertEqual(traverse_obj(_SLICE_DATA, slice(1, 4, 2)), _SLICE_DATA[1:4:2],
+                         msg='slice key should apply slice to sequence')
+
+        # Test alternative paths
+        self.assertEqual(traverse_obj(_TEST_DATA, 'fail', 'str'), 'str',
+                         msg='multiple `paths` should be treated as alternative paths')
+        self.assertEqual(traverse_obj(_TEST_DATA, 'str', 100), 'str',
+                         msg='alternatives should exit early')
+        self.assertEqual(traverse_obj(_TEST_DATA, 'fail', 'fail'), None,
+                         msg='alternatives should return `default` if exhausted')
+        self.assertEqual(traverse_obj(_TEST_DATA, (Ellipsis, 'fail'), 100), 100,
+                         msg='alternatives should track their own branching return')
+        self.assertEqual(traverse_obj(_TEST_DATA, ('dict', Ellipsis), ('data', Ellipsis)), list(_TEST_DATA['data']),
+                         msg='alternatives on empty objects should search further')
+
+        # Test branch and path nesting
+        self.assertEqual(traverse_obj(_TEST_DATA, ('urls', (3, 0), 'url')), ['https://www.example.com/0'],
+                         msg='tuple as key should be treated as branches')
+        self.assertEqual(traverse_obj(_TEST_DATA, ('urls', [3, 0], 'url')), ['https://www.example.com/0'],
+                         msg='list as key should be treated as branches')
+        self.assertEqual(traverse_obj(_TEST_DATA, ('urls', ((1, 'fail'), (0, 'url')))), ['https://www.example.com/0'],
+                         msg='double nesting in path should be treated as paths')
+        self.assertEqual(traverse_obj(['0', [1, 2]], [(0, 1), 0]), [1],
+                         msg='do not fail early on branching')
+        self.assertCountEqual(traverse_obj(_TEST_DATA, ('urls', ((1, ('fail', 'url')), (0, 'url')))),
+                              ['https://www.example.com/0', 'https://www.example.com/1'],
+                              msg='triple nesting in path should be treated as branches')
+        self.assertEqual(traverse_obj(_TEST_DATA, ('urls', ('fail', (Ellipsis, 'url')))),
+                         ['https://www.example.com/0', 'https://www.example.com/1'],
+                         msg='ellipsis as branch path start gets flattened')
+
+        # Test dictionary as key
+        self.assertEqual(traverse_obj(_TEST_DATA, {0: 100, 1: 1.2}), {0: 100, 1: 1.2},
+                         msg='dict key should result in a dict with the same keys')
+        self.assertEqual(traverse_obj(_TEST_DATA, {0: ('urls', 0, 'url')}),
+                         {0: 'https://www.example.com/0'},
+                         msg='dict key should allow paths')
+        self.assertEqual(traverse_obj(_TEST_DATA, {0: ('urls', (3, 0), 'url')}),
+                         {0: ['https://www.example.com/0']},
+                         msg='tuple in dict path should be treated as branches')
+        self.assertEqual(traverse_obj(_TEST_DATA, {0: ('urls', ((1, 'fail'), (0, 'url')))}),
+                         {0: ['https://www.example.com/0']},
+                         msg='double nesting in dict path should be treated as paths')
+        self.assertEqual(traverse_obj(_TEST_DATA, {0: ('urls', ((1, ('fail', 'url')), (0, 'url')))}),
+                         {0: ['https://www.example.com/1', 'https://www.example.com/0']},
+                         msg='triple nesting in dict path should be treated as branches')
+        self.assertEqual(traverse_obj(_TEST_DATA, {0: 'fail'}), {},
+                         msg='remove `None` values when top level dict key fails')
+        self.assertEqual(traverse_obj(_TEST_DATA, {0: 'fail'}, default=Ellipsis), {0: Ellipsis},
+                         msg='use `default` if key fails and `default`')
+        self.assertEqual(traverse_obj(_TEST_DATA, {0: 'dict'}), {},
+                         msg='remove empty values when dict key')
+        self.assertEqual(traverse_obj(_TEST_DATA, {0: 'dict'}, default=Ellipsis), {0: Ellipsis},
+                         msg='use `default` when dict key and a default')
+        self.assertEqual(traverse_obj(_TEST_DATA, {0: {0: 'fail'}}), {},
+                         msg='remove empty values when nested dict key fails')
+        self.assertEqual(traverse_obj(None, {0: 'fail'}), {},
+                         msg='default to dict if pruned')
+        self.assertEqual(traverse_obj(None, {0: 'fail'}, default=Ellipsis), {0: Ellipsis},
+                         msg='default to dict if pruned and default is given')
+        self.assertEqual(traverse_obj(_TEST_DATA, {0: {0: 'fail'}}, default=Ellipsis), {0: {0: Ellipsis}},
+                         msg='use nested `default` when nested dict key fails and `default`')
+        self.assertEqual(traverse_obj(_TEST_DATA, {0: ('dict', Ellipsis)}), {},
+                         msg='remove key if branch in dict key not successful')
+
+        # Testing default parameter behavior
+        _DEFAULT_DATA = {'None': None, 'int': 0, 'list': []}
+        self.assertEqual(traverse_obj(_DEFAULT_DATA, 'fail'), None,
+                         msg='default value should be `None`')
+        self.assertEqual(traverse_obj(_DEFAULT_DATA, 'fail', 'fail', default=Ellipsis), Ellipsis,
+                         msg='chained fails should result in default')
+        self.assertEqual(traverse_obj(_DEFAULT_DATA, 'None', 'int'), 0,
+                         msg='should not short cirquit on `None`')
+        self.assertEqual(traverse_obj(_DEFAULT_DATA, 'fail', default=1), 1,
+                         msg='invalid dict key should result in `default`')
+        self.assertEqual(traverse_obj(_DEFAULT_DATA, 'None', default=1), 1,
+                         msg='`None` is a deliberate sentinel and should become `default`')
+        self.assertEqual(traverse_obj(_DEFAULT_DATA, ('list', 10)), None,
+                         msg='`IndexError` should result in `default`')
+        self.assertEqual(traverse_obj(_DEFAULT_DATA, (Ellipsis, 'fail'), default=1), 1,
+                         msg='if branched but not successful return `default` if defined, not `[]`')
+        self.assertEqual(traverse_obj(_DEFAULT_DATA, (Ellipsis, 'fail'), default=None), None,
+                         msg='if branched but not successful return `default` even if `default` is `None`')
+        self.assertEqual(traverse_obj(_DEFAULT_DATA, (Ellipsis, 'fail')), [],
+                         msg='if branched but not successful return `[]`, not `default`')
+        self.assertEqual(traverse_obj(_DEFAULT_DATA, ('list', Ellipsis)), [],
+                         msg='if branched but object is empty return `[]`, not `default`')
+        self.assertEqual(traverse_obj(None, Ellipsis), [],
+                         msg='if branched but object is `None` return `[]`, not `default`')
+        self.assertEqual(traverse_obj({0: None}, (0, Ellipsis)), [],
+                         msg='if branched but state is `None` return `[]`, not `default`')
+
+        branching_paths = [
+            ('fail', Ellipsis),
+            (Ellipsis, 'fail'),
+            100 * ('fail',) + (Ellipsis,),
+            (Ellipsis,) + 100 * ('fail',),
+        ]
+        for branching_path in branching_paths:
+            self.assertEqual(traverse_obj({}, branching_path), [],
+                             msg='if branched but state is `None`, return `[]` (not `default`)')
+            self.assertEqual(traverse_obj({}, 'fail', branching_path), [],
+                             msg='if branching in last alternative and previous did not match, return `[]` (not `default`)')
+            self.assertEqual(traverse_obj({0: 'x'}, 0, branching_path), 'x',
+                             msg='if branching in last alternative and previous did match, return single value')
+            self.assertEqual(traverse_obj({0: 'x'}, branching_path, 0), 'x',
+                             msg='if branching in first alternative and non-branching path does match, return single value')
+            self.assertEqual(traverse_obj({}, branching_path, 'fail'), None,
+                             msg='if branching in first alternative and non-branching path does not match, return `default`')
+
+        # Testing expected_type behavior
+        _EXPECTED_TYPE_DATA = {'str': 'str', 'int': 0}
+        self.assertEqual(traverse_obj(_EXPECTED_TYPE_DATA, 'str', expected_type=str),
+                         'str', msg='accept matching `expected_type` type')
+        self.assertEqual(traverse_obj(_EXPECTED_TYPE_DATA, 'str', expected_type=int),
+                         None, msg='reject non-matching `expected_type` type')
+        self.assertEqual(traverse_obj(_EXPECTED_TYPE_DATA, 'int', expected_type=lambda x: str(x)),
+                         '0', msg='transform type using type function')
+        self.assertEqual(traverse_obj(_EXPECTED_TYPE_DATA, 'str', expected_type=lambda _: 1 / 0),
+                         None, msg='wrap expected_type function in try_call')
+        self.assertEqual(traverse_obj(_EXPECTED_TYPE_DATA, Ellipsis, expected_type=str),
+                         ['str'], msg='eliminate items that expected_type fails on')
+        self.assertEqual(traverse_obj(_TEST_DATA, {0: 100, 1: 1.2}, expected_type=int),
+                         {0: 100}, msg='type as expected_type should filter dict values')
+        self.assertEqual(traverse_obj(_TEST_DATA, {0: 100, 1: 1.2, 2: 'None'}, expected_type=str_or_none),
+                         {0: '100', 1: '1.2'}, msg='function as expected_type should transform dict values')
+        self.assertEqual(traverse_obj(_TEST_DATA, ({0: 1.2}, 0, set((int_or_none,))), expected_type=int),
+                         1, msg='expected_type should not filter non-final dict values')
+        self.assertEqual(traverse_obj(_TEST_DATA, {0: {0: 100, 1: 'str'}}, expected_type=int),
+                         {0: {0: 100}}, msg='expected_type should transform deep dict values')
+        self.assertEqual(traverse_obj(_TEST_DATA, [({0: '...'}, {0: '...'})], expected_type=type(Ellipsis)),
+                         [{0: Ellipsis}, {0: Ellipsis}], msg='expected_type should transform branched dict values')
+        self.assertEqual(traverse_obj({1: {3: 4}}, [(1, 2), 3], expected_type=int),
+                         [4], msg='expected_type regression for type matching in tuple branching')
+        self.assertEqual(traverse_obj(_TEST_DATA, ['data', Ellipsis], expected_type=int),
+                         [], msg='expected_type regression for type matching in dict result')
+
+        # Test get_all behavior
+        _GET_ALL_DATA = {'key': [0, 1, 2]}
+        self.assertEqual(traverse_obj(_GET_ALL_DATA, ('key', Ellipsis), get_all=False), 0,
+                         msg='if not `get_all`, return only first matching value')
+        self.assertEqual(traverse_obj(_GET_ALL_DATA, Ellipsis, get_all=False), [0, 1, 2],
+                         msg='do not overflatten if not `get_all`')
+
+        # Test casesense behavior
+        _CASESENSE_DATA = {
+            'KeY': 'value0',
+            0: {
+                'KeY': 'value1',
+                0: {'KeY': 'value2'},
+            },
+            # FULLWIDTH LATIN CAPITAL LETTER K
+            '\uff2bey': 'value3',
+        }
+        self.assertEqual(traverse_obj(_CASESENSE_DATA, 'key'), None,
+                         msg='dict keys should be case sensitive unless `casesense`')
+        self.assertEqual(traverse_obj(_CASESENSE_DATA, 'keY',
+                                      casesense=False), 'value0',
+                         msg='allow non matching key case if `casesense`')
+        self.assertEqual(traverse_obj(_CASESENSE_DATA, '\uff4bey',  # FULLWIDTH LATIN SMALL LETTER K
+                                      casesense=False), 'value3',
+                         msg='allow non matching Unicode key case if `casesense`')
+        self.assertEqual(traverse_obj(_CASESENSE_DATA, (0, ('keY',)),
+                                      casesense=False), ['value1'],
+                         msg='allow non matching key case in branch if `casesense`')
+        self.assertEqual(traverse_obj(_CASESENSE_DATA, (0, ((0, 'keY'),)),
+                                      casesense=False), ['value2'],
+                         msg='allow non matching key case in branch path if `casesense`')
+
+        # Test traverse_string behavior
+        _TRAVERSE_STRING_DATA = {'str': 'str', 1.2: 1.2}
+        self.assertEqual(traverse_obj(_TRAVERSE_STRING_DATA, ('str', 0)), None,
+                         msg='do not traverse into string if not `traverse_string`')
+        self.assertEqual(traverse_obj(_TRAVERSE_STRING_DATA, ('str', 0),
+                                      _traverse_string=True), 's',
+                         msg='traverse into string if `traverse_string`')
+        self.assertEqual(traverse_obj(_TRAVERSE_STRING_DATA, (1.2, 1),
+                                      _traverse_string=True), '.',
+                         msg='traverse into converted data if `traverse_string`')
+        self.assertEqual(traverse_obj(_TRAVERSE_STRING_DATA, ('str', Ellipsis),
+                                      _traverse_string=True), 'str',
+                         msg='`...` should result in string (same value) if `traverse_string`')
+        self.assertEqual(traverse_obj(_TRAVERSE_STRING_DATA, ('str', slice(0, None, 2)),
+                                      _traverse_string=True), 'sr',
+                         msg='`slice` should result in string if `traverse_string`')
+        self.assertEqual(traverse_obj(_TRAVERSE_STRING_DATA, ('str', lambda i, v: i or v == 's'),
+                                      _traverse_string=True), 'str',
+                         msg='function should result in string if `traverse_string`')
+        self.assertEqual(traverse_obj(_TRAVERSE_STRING_DATA, ('str', (0, 2)),
+                                      _traverse_string=True), ['s', 'r'],
+                         msg='branching should result in list if `traverse_string`')
+        self.assertEqual(traverse_obj({}, (0, Ellipsis), _traverse_string=True), [],
+                         msg='branching should result in list if `traverse_string`')
+        self.assertEqual(traverse_obj({}, (0, lambda x, y: True), _traverse_string=True), [],
+                         msg='branching should result in list if `traverse_string`')
+        self.assertEqual(traverse_obj({}, (0, slice(1)), _traverse_string=True), [],
+                         msg='branching should result in list if `traverse_string`')
+
+        # Test re.Match as input obj
+        mobj = re.match(r'^0(12)(?P<group>3)(4)?$', '0123')
+        self.assertEqual(traverse_obj(mobj, Ellipsis), [x for x in mobj.groups() if x is not None],
+                         msg='`...` on a `re.Match` should give its `groups()`')
+        self.assertEqual(traverse_obj(mobj, lambda k, _: k in (0, 2)), ['0123', '3'],
+                         msg='function on a `re.Match` should give groupno, value starting at 0')
+        self.assertEqual(traverse_obj(mobj, 'group'), '3',
+                         msg='str key on a `re.Match` should give group with that name')
+        self.assertEqual(traverse_obj(mobj, 2), '3',
+                         msg='int key on a `re.Match` should give group with that name')
+        self.assertEqual(traverse_obj(mobj, 'gRoUp', casesense=False), '3',
+                         msg='str key on a `re.Match` should respect casesense')
+        self.assertEqual(traverse_obj(mobj, 'fail'), None,
+                         msg='failing str key on a `re.Match` should return `default`')
+        self.assertEqual(traverse_obj(mobj, 'gRoUpS', casesense=False), None,
+                         msg='failing str key on a `re.Match` should return `default`')
+        self.assertEqual(traverse_obj(mobj, 8), None,
+                         msg='failing int key on a `re.Match` should return `default`')
+        self.assertEqual(traverse_obj(mobj, lambda k, _: k in (0, 'group')), ['0123', '3'],
+                         msg='function on a `re.Match` should give group name as well')
+
+        # Test xml.etree.ElementTree.Element as input obj
+        etree = compat_etree_fromstring('''<?xml version="1.0"?>
+        <data>
+            <country name="Liechtenstein">
+                <rank>1</rank>
+                <year>2008</year>
+                <gdppc>141100</gdppc>
+                <neighbor name="Austria" direction="E"/>
+                <neighbor name="Switzerland" direction="W"/>
+            </country>
+            <country name="Singapore">
+                <rank>4</rank>
+                <year>2011</year>
+                <gdppc>59900</gdppc>
+                <neighbor name="Malaysia" direction="N"/>
+            </country>
+            <country name="Panama">
+                <rank>68</rank>
+                <year>2011</year>
+                <gdppc>13600</gdppc>
+                <neighbor name="Costa Rica" direction="W"/>
+                <neighbor name="Colombia" direction="E"/>
+            </country>
+        </data>''')
+        self.assertEqual(traverse_obj(etree, ''), etree,
+                         msg='empty str key should return the element itself')
+        self.assertEqual(traverse_obj(etree, 'country'), list(etree),
+                         msg='str key should return all children with that tag name')
+        self.assertEqual(traverse_obj(etree, Ellipsis), list(etree),
+                         msg='`...` as key should return all children')
+        self.assertEqual(traverse_obj(etree, lambda _, x: x[0].text == '4'), [etree[1]],
+                         msg='function as key should get element as value')
+        self.assertEqual(traverse_obj(etree, lambda i, _: i == 1), [etree[1]],
+                         msg='function as key should get index as key')
+        self.assertEqual(traverse_obj(etree, 0), etree[0],
+                         msg='int key should return the nth child')
+        self.assertEqual(traverse_obj(etree, './/neighbor/@name'),
+                         ['Austria', 'Switzerland', 'Malaysia', 'Costa Rica', 'Colombia'],
+                         msg='`@<attribute>` at end of path should give that attribute')
+        self.assertEqual(traverse_obj(etree, '//neighbor/@fail'), [None, None, None, None, None],
+                         msg='`@<nonexistent>` at end of path should give `None`')
+        self.assertEqual(traverse_obj(etree, ('//neighbor/@', 2)), {'name': 'Malaysia', 'direction': 'N'},
+                         msg='`@` should give the full attribute dict')
+        self.assertEqual(traverse_obj(etree, '//year/text()'), ['2008', '2011', '2011'],
+                         msg='`text()` at end of path should give the inner text')
+        self.assertEqual(traverse_obj(etree, '//*[@direction]/@direction'), ['E', 'W', 'N', 'W', 'E'],
+                         msg='full python xpath features should be supported')
+        self.assertEqual(traverse_obj(etree, (0, '@name')), 'Liechtenstein',
+                         msg='special transformations should act on current element')
+        self.assertEqual(traverse_obj(etree, ('country', 0, Ellipsis, 'text()', T(int_or_none))), [1, 2008, 141100],
+                         msg='special transformations should act on current element')
+
+    def test_traversal_unbranching(self):
+        self.assertEqual(traverse_obj(_TEST_DATA, [(100, 1.2), all]), [100, 1.2],
+                         msg='`all` should give all results as list')
+        self.assertEqual(traverse_obj(_TEST_DATA, [(100, 1.2), any]), 100,
+                         msg='`any` should give the first result')
+        self.assertEqual(traverse_obj(_TEST_DATA, [100, all]), [100],
+                         msg='`all` should give list if non branching')
+        self.assertEqual(traverse_obj(_TEST_DATA, [100, any]), 100,
+                         msg='`any` should give single item if non branching')
+        self.assertEqual(traverse_obj(_TEST_DATA, [('dict', 'None', 100), all]), [100],
+                         msg='`all` should filter `None` and empty dict')
+        self.assertEqual(traverse_obj(_TEST_DATA, [('dict', 'None', 100), any]), 100,
+                         msg='`any` should filter `None` and empty dict')
+        self.assertEqual(traverse_obj(_TEST_DATA, [{
+            'all': [('dict', 'None', 100, 1.2), all],
+            'any': [('dict', 'None', 100, 1.2), any],
+        }]), {'all': [100, 1.2], 'any': 100},
+            msg='`all`/`any` should apply to each dict path separately')
+        self.assertEqual(traverse_obj(_TEST_DATA, [{
+            'all': [('dict', 'None', 100, 1.2), all],
+            'any': [('dict', 'None', 100, 1.2), any],
+        }], get_all=False), {'all': [100, 1.2], 'any': 100},
+            msg='`all`/`any` should apply to dict regardless of `get_all`')
+        self.assertIs(traverse_obj(_TEST_DATA, [('dict', 'None', 100, 1.2), all, T(float)]), None,
+                      msg='`all` should reset branching status')
+        self.assertIs(traverse_obj(_TEST_DATA, [('dict', 'None', 100, 1.2), any, T(float)]), None,
+                      msg='`any` should reset branching status')
+        self.assertEqual(traverse_obj(_TEST_DATA, [('dict', 'None', 100, 1.2), all, Ellipsis, T(float)]), [1.2],
+                         msg='`all` should allow further branching')
+        self.assertEqual(traverse_obj(_TEST_DATA, [('dict', 'None', 'urls', 'data'), any, Ellipsis, 'index']), [0, 1],
+                         msg='`any` should allow further branching')
+
+    def test_traversal_morsel(self):
+        values = {
+            'expires': 'a',
+            'path': 'b',
+            'comment': 'c',
+            'domain': 'd',
+            'max-age': 'e',
+            'secure': 'f',
+            'httponly': 'g',
+            'version': 'h',
+            'samesite': 'i',
+        }
+        # SameSite added in Py3.8, breaks .update for 3.5-3.7
+        if sys.version_info < (3, 8):
+            del values['samesite']
+        morsel = compat_http_cookies.Morsel()
+        morsel.set(str('item_key'), 'item_value', 'coded_value')
+        morsel.update(values)
+        values['key'] = str('item_key')
+        values['value'] = 'item_value'
+        values = dict((str(k), v) for k, v in values.items())
+        # make test pass even without ordered dict
+        value_set = set(values.values())
+
+        for key, value in values.items():
+            self.assertEqual(traverse_obj(morsel, key), value,
+                             msg='Morsel should provide access to all values')
+        self.assertEqual(set(traverse_obj(morsel, Ellipsis)), value_set,
+                         msg='`...` should yield all values')
+        self.assertEqual(set(traverse_obj(morsel, lambda k, v: True)), value_set,
+                         msg='function key should yield all values')
+        self.assertIs(traverse_obj(morsel, [(None,), any]), morsel,
+                      msg='Morsel should not be implicitly changed to dict on usage')
+
+    def test_get_first(self):
+        self.assertEqual(get_first([{'a': None}, {'a': 'spam'}], 'a'), 'spam')
+
+    def test_dict_get(self):
+        FALSE_VALUES = {
+            'none': None,
+            'false': False,
+            'zero': 0,
+            'empty_string': '',
+            'empty_list': [],
+        }
+        d = FALSE_VALUES.copy()
+        d['a'] = 42
+        self.assertEqual(dict_get(d, 'a'), 42)
+        self.assertEqual(dict_get(d, 'b'), None)
+        self.assertEqual(dict_get(d, 'b', 42), 42)
+        self.assertEqual(dict_get(d, ('a', )), 42)
+        self.assertEqual(dict_get(d, ('b', 'a', )), 42)
+        self.assertEqual(dict_get(d, ('b', 'c', 'a', 'd', )), 42)
+        self.assertEqual(dict_get(d, ('b', 'c', )), None)
+        self.assertEqual(dict_get(d, ('b', 'c', ), 42), 42)
+        for key, false_value in FALSE_VALUES.items():
+            self.assertEqual(dict_get(d, ('b', 'c', key, )), None)
+            self.assertEqual(dict_get(d, ('b', 'c', key, ), skip_false_values=False), false_value)
+
+
+if __name__ == '__main__':
+    unittest.main()

test/test_unicode_literals.py

@@ -2,19 +2,21 @@ from __future__ import unicode_literals
 
 # Allow direct execution
 import os
+import re
 import sys
 import unittest
-sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
 
-import io
-import re
+dirn = os.path.dirname
 
-rootDir = os.path.dirname(os.path.dirname(os.path.abspath(__file__)))
+rootDir = dirn(dirn(os.path.abspath(__file__)))
+
+sys.path.insert(0, rootDir)
 
 IGNORED_FILES = [
     'setup.py',  # http://bugs.python.org/issue13943
     'conf.py',
     'buildserver.py',
+    'get-pip.py',
 ]
 
 IGNORED_DIRS = [
@@ -23,6 +25,7 @@ IGNORED_DIRS = [
 ]
 
 from test.helper import assertRegexpMatches
+from youtube_dl.compat import compat_open as open
 
 
 class TestUnicodeLiterals(unittest.TestCase):
@@ -40,7 +43,7 @@ class TestUnicodeLiterals(unittest.TestCase):
                     continue
 
                 fn = os.path.join(dirpath, basename)
-                with io.open(fn, encoding='utf-8') as inf:
+                with open(fn, encoding='utf-8') as inf:
                     code = inf.read()
 
                 if "'" not in code and '"' not in code:

test/test_utils.py

@@ -14,13 +14,12 @@ sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
 import io
 import itertools
 import json
-import re
 import xml.etree.ElementTree
 
 from youtube_dl.utils import (
     age_restricted,
     args_to_str,
-    encode_base_n,
+    base_url,
     caesar,
     clean_html,
     clean_podcast_url,
@@ -28,11 +27,12 @@ from youtube_dl.utils import (
     DateRange,
     detect_exe_version,
     determine_ext,
-    dict_get,
+    encode_base_n,
     encode_compat_str,
     encodeFilename,
     escape_rfc3986,
     escape_url,
+    expand_path,
     extract_attributes,
     ExtractorError,
     find_xpath_attr,
@@ -42,7 +42,6 @@ from youtube_dl.utils import (
     get_element_by_attribute,
     get_elements_by_class,
     get_elements_by_attribute,
-    get_first,
     InAdvancePagedList,
     int_or_none,
     intlist_to_bytes,
@@ -51,6 +50,7 @@ from youtube_dl.utils import (
     js_to_json,
     LazyList,
     limit_length,
+    lowercase_escape,
     merge_dicts,
     mimetype2ext,
     month_by_name,
@@ -59,24 +59,26 @@ from youtube_dl.utils import (
     OnDemandPagedList,
     orderedSet,
     parse_age_limit,
+    parse_bitrate,
     parse_duration,
     parse_filesize,
+    parse_codecs,
     parse_count,
     parse_iso8601,
     parse_resolution,
-    parse_bitrate,
+    parse_qs,
     pkcs1pad,
-    read_batch_urls,
-    sanitize_filename,
-    sanitize_path,
-    sanitize_url,
-    expand_path,
     prepend_extension,
-    replace_extension,
+    read_batch_urls,
     remove_start,
     remove_end,
     remove_quotes,
+    replace_extension,
     rot47,
+    sanitize_filename,
+    sanitize_path,
+    sanitize_url,
+    sanitized_Request,
     shell_quote,
     smuggle_url,
     str_to_int,
@@ -84,17 +86,14 @@ from youtube_dl.utils import (
     strip_or_none,
     subtitles_filename,
     timeconvert,
-    traverse_obj,
     try_call,
     unescapeHTML,
     unified_strdate,
     unified_timestamp,
     unsmuggle_url,
     uppercase_escape,
-    lowercase_escape,
     url_basename,
     url_or_none,
-    base_url,
     urljoin,
     urlencode_postdata,
     urshift,
@@ -112,7 +111,7 @@ from youtube_dl.utils import (
     cli_option,
     cli_valueless_option,
     cli_bool_option,
-    parse_codecs,
+    YoutubeDLHandler,
 )
 from youtube_dl.compat import (
     compat_chr,
@@ -122,16 +121,11 @@ from youtube_dl.compat import (
     compat_setenv,
     compat_str,
     compat_urlparse,
-    compat_parse_qs,
 )
 
 
 class TestUtil(unittest.TestCase):
 
-    # yt-dlp shim
-    def assertCountEqual(self, expected, got, msg='count should be the same'):
-        return self.assertEqual(len(tuple(expected)), len(tuple(got)), msg=msg)
-
     def test_timeconvert(self):
         self.assertTrue(timeconvert('') is None)
         self.assertTrue(timeconvert('bougrg') is None)
@@ -252,6 +246,18 @@ class TestUtil(unittest.TestCase):
         self.assertEqual(sanitize_url('https://foo.bar'), 'https://foo.bar')
         self.assertEqual(sanitize_url('foo bar'), 'foo bar')
 
+    def test_sanitized_Request(self):
+        self.assertFalse(sanitized_Request('http://foo.bar').has_header('Authorization'))
+        self.assertFalse(sanitized_Request('http://:foo.bar').has_header('Authorization'))
+        self.assertEqual(sanitized_Request('http://@foo.bar').get_header('Authorization'),
+                         'Basic Og==')
+        self.assertEqual(sanitized_Request('http://:pass@foo.bar').get_header('Authorization'),
+                         'Basic OnBhc3M=')
+        self.assertEqual(sanitized_Request('http://user:@foo.bar').get_header('Authorization'),
+                         'Basic dXNlcjo=')
+        self.assertEqual(sanitized_Request('http://user:pass@foo.bar').get_header('Authorization'),
+                         'Basic dXNlcjpwYXNz')
+
     def test_expand_path(self):
         def env(var):
             return '%{0}%'.format(var) if sys.platform == 'win32' else '${0}'.format(var)
@@ -509,11 +515,14 @@ class TestUtil(unittest.TestCase):
         self.assertEqual(float_or_none(set()), None)
 
     def test_int_or_none(self):
+        self.assertEqual(int_or_none(42), 42)
         self.assertEqual(int_or_none('42'), 42)
         self.assertEqual(int_or_none(''), None)
         self.assertEqual(int_or_none(None), None)
         self.assertEqual(int_or_none([]), None)
         self.assertEqual(int_or_none(set()), None)
+        self.assertEqual(int_or_none('42', base=8), 34)
+        self.assertRaises(TypeError, int_or_none(42, base=8))
 
     def test_str_to_int(self):
         self.assertEqual(str_to_int('123,456'), 123456)
@@ -680,38 +689,36 @@ class TestUtil(unittest.TestCase):
         self.assertTrue(isinstance(data, bytes))
 
     def test_update_url_query(self):
-        def query_dict(url):
-            return compat_parse_qs(compat_urlparse.urlparse(url).query)
-        self.assertEqual(query_dict(update_url_query(
+        self.assertEqual(parse_qs(update_url_query(
             'http://example.com/path', {'quality': ['HD'], 'format': ['mp4']})),
-            query_dict('http://example.com/path?quality=HD&format=mp4'))
-        self.assertEqual(query_dict(update_url_query(
+            parse_qs('http://example.com/path?quality=HD&format=mp4'))
+        self.assertEqual(parse_qs(update_url_query(
             'http://example.com/path', {'system': ['LINUX', 'WINDOWS']})),
-            query_dict('http://example.com/path?system=LINUX&system=WINDOWS'))
-        self.assertEqual(query_dict(update_url_query(
+            parse_qs('http://example.com/path?system=LINUX&system=WINDOWS'))
+        self.assertEqual(parse_qs(update_url_query(
             'http://example.com/path', {'fields': 'id,formats,subtitles'})),
-            query_dict('http://example.com/path?fields=id,formats,subtitles'))
-        self.assertEqual(query_dict(update_url_query(
+            parse_qs('http://example.com/path?fields=id,formats,subtitles'))
+        self.assertEqual(parse_qs(update_url_query(
             'http://example.com/path', {'fields': ('id,formats,subtitles', 'thumbnails')})),
-            query_dict('http://example.com/path?fields=id,formats,subtitles&fields=thumbnails'))
-        self.assertEqual(query_dict(update_url_query(
+            parse_qs('http://example.com/path?fields=id,formats,subtitles&fields=thumbnails'))
+        self.assertEqual(parse_qs(update_url_query(
             'http://example.com/path?manifest=f4m', {'manifest': []})),
-            query_dict('http://example.com/path'))
-        self.assertEqual(query_dict(update_url_query(
+            parse_qs('http://example.com/path'))
+        self.assertEqual(parse_qs(update_url_query(
             'http://example.com/path?system=LINUX&system=WINDOWS', {'system': 'LINUX'})),
-            query_dict('http://example.com/path?system=LINUX'))
-        self.assertEqual(query_dict(update_url_query(
+            parse_qs('http://example.com/path?system=LINUX'))
+        self.assertEqual(parse_qs(update_url_query(
             'http://example.com/path', {'fields': b'id,formats,subtitles'})),
-            query_dict('http://example.com/path?fields=id,formats,subtitles'))
-        self.assertEqual(query_dict(update_url_query(
+            parse_qs('http://example.com/path?fields=id,formats,subtitles'))
+        self.assertEqual(parse_qs(update_url_query(
             'http://example.com/path', {'width': 1080, 'height': 720})),
-            query_dict('http://example.com/path?width=1080&height=720'))
-        self.assertEqual(query_dict(update_url_query(
+            parse_qs('http://example.com/path?width=1080&height=720'))
+        self.assertEqual(parse_qs(update_url_query(
             'http://example.com/path', {'bitrate': 5020.43})),
-            query_dict('http://example.com/path?bitrate=5020.43'))
-        self.assertEqual(query_dict(update_url_query(
+            parse_qs('http://example.com/path?bitrate=5020.43'))
+        self.assertEqual(parse_qs(update_url_query(
             'http://example.com/path', {'test': '第二行тест'})),
-            query_dict('http://example.com/path?test=%E7%AC%AC%E4%BA%8C%E8%A1%8C%D1%82%D0%B5%D1%81%D1%82'))
+            parse_qs('http://example.com/path?test=%E7%AC%AC%E4%BA%8C%E8%A1%8C%D1%82%D0%B5%D1%81%D1%82'))
 
     def test_multipart_encode(self):
         self.assertEqual(
@@ -723,28 +730,6 @@ class TestUtil(unittest.TestCase):
         self.assertRaises(
             ValueError, multipart_encode, {b'field': b'value'}, boundary='value')
 
-    def test_dict_get(self):
-        FALSE_VALUES = {
-            'none': None,
-            'false': False,
-            'zero': 0,
-            'empty_string': '',
-            'empty_list': [],
-        }
-        d = FALSE_VALUES.copy()
-        d['a'] = 42
-        self.assertEqual(dict_get(d, 'a'), 42)
-        self.assertEqual(dict_get(d, 'b'), None)
-        self.assertEqual(dict_get(d, 'b', 42), 42)
-        self.assertEqual(dict_get(d, ('a', )), 42)
-        self.assertEqual(dict_get(d, ('b', 'a', )), 42)
-        self.assertEqual(dict_get(d, ('b', 'c', 'a', 'd', )), 42)
-        self.assertEqual(dict_get(d, ('b', 'c', )), None)
-        self.assertEqual(dict_get(d, ('b', 'c', ), 42), 42)
-        for key, false_value in FALSE_VALUES.items():
-            self.assertEqual(dict_get(d, ('b', 'c', key, )), None)
-            self.assertEqual(dict_get(d, ('b', 'c', key, ), skip_false_values=False), false_value)
-
     def test_merge_dicts(self):
         self.assertEqual(merge_dicts({'a': 1}, {'b': 2}), {'a': 1, 'b': 2})
         self.assertEqual(merge_dicts({'a': 1}, {'a': 2}), {'a': 1})
@@ -903,6 +888,111 @@ class TestUtil(unittest.TestCase):
         )
         self.assertEqual(escape_url('http://vimeo.com/56015672#at=0'), 'http://vimeo.com/56015672#at=0')
 
+    def test_remove_dot_segments(self):
+
+        def remove_dot_segments(p):
+            q = '' if p.startswith('/') else '/'
+            p = 'http://example.com' + q + p
+            p = compat_urlparse.urlsplit(YoutubeDLHandler._fix_path(p)).path
+            return p[1:] if q else p
+
+        self.assertEqual(remove_dot_segments('/a/b/c/./../../g'), '/a/g')
+        self.assertEqual(remove_dot_segments('mid/content=5/../6'), 'mid/6')
+        self.assertEqual(remove_dot_segments('/ad/../cd'), '/cd')
+        self.assertEqual(remove_dot_segments('/ad/../cd/'), '/cd/')
+        self.assertEqual(remove_dot_segments('/..'), '/')
+        self.assertEqual(remove_dot_segments('/./'), '/')
+        self.assertEqual(remove_dot_segments('/./a'), '/a')
+        self.assertEqual(remove_dot_segments('/abc/./.././d/././e/.././f/./../../ghi'), '/ghi')
+        self.assertEqual(remove_dot_segments('/'), '/')
+        self.assertEqual(remove_dot_segments('/t'), '/t')
+        self.assertEqual(remove_dot_segments('t'), 't')
+        self.assertEqual(remove_dot_segments(''), '')
+        self.assertEqual(remove_dot_segments('/../a/b/c'), '/a/b/c')
+        self.assertEqual(remove_dot_segments('../a'), 'a')
+        self.assertEqual(remove_dot_segments('./a'), 'a')
+        self.assertEqual(remove_dot_segments('.'), '')
+        self.assertEqual(remove_dot_segments('////'), '////')
+
+    def test_js_to_json_vars_strings(self):
+        self.assertDictEqual(
+            json.loads(js_to_json(
+                '''{
+                    'null': a,
+                    'nullStr': b,
+                    'true': c,
+                    'trueStr': d,
+                    'false': e,
+                    'falseStr': f,
+                    'unresolvedVar': g,
+                }''',
+                {
+                    'a': 'null',
+                    'b': '"null"',
+                    'c': 'true',
+                    'd': '"true"',
+                    'e': 'false',
+                    'f': '"false"',
+                    'g': 'var',
+                }
+            )),
+            {
+                'null': None,
+                'nullStr': 'null',
+                'true': True,
+                'trueStr': 'true',
+                'false': False,
+                'falseStr': 'false',
+                'unresolvedVar': 'var'
+            }
+        )
+
+        self.assertDictEqual(
+            json.loads(js_to_json(
+                '''{
+                    'int': a,
+                    'intStr': b,
+                    'float': c,
+                    'floatStr': d,
+                }''',
+                {
+                    'a': '123',
+                    'b': '"123"',
+                    'c': '1.23',
+                    'd': '"1.23"',
+                }
+            )),
+            {
+                'int': 123,
+                'intStr': '123',
+                'float': 1.23,
+                'floatStr': '1.23',
+            }
+        )
+
+        self.assertDictEqual(
+            json.loads(js_to_json(
+                '''{
+                    'object': a,
+                    'objectStr': b,
+                    'array': c,
+                    'arrayStr': d,
+                }''',
+                {
+                    'a': '{}',
+                    'b': '"{}"',
+                    'c': '[]',
+                    'd': '"[]"',
+                }
+            )),
+            {
+                'object': {},
+                'objectStr': '{}',
+                'array': [],
+                'arrayStr': '[]',
+            }
+        )
+
     def test_js_to_json_realworld(self):
         inp = '''{
             'clip':{'provider':'pseudo'}
@@ -973,10 +1063,10 @@ class TestUtil(unittest.TestCase):
             !42: 42
         }''')
         self.assertEqual(json.loads(on), {
-            'a': 0,
-            'b': 1,
-            'c': 0,
-            'd': 42.42,
+            'a': True,
+            'b': False,
+            'c': False,
+            'd': True,
             'e': [],
             'f': "abc",
             'g': "",
@@ -1046,10 +1136,26 @@ class TestUtil(unittest.TestCase):
         on = js_to_json('{ "040": "040" }')
         self.assertEqual(json.loads(on), {'040': '040'})
 
+        on = js_to_json('[1,//{},\n2]')
+        self.assertEqual(json.loads(on), [1, 2])
+
+        on = js_to_json(r'"\^\$\#"')
+        self.assertEqual(json.loads(on), R'^$#', msg='Unnecessary escapes should be stripped')
+
+        on = js_to_json('\'"\\""\'')
+        self.assertEqual(json.loads(on), '"""', msg='Unnecessary quote escape should be escaped')
+
     def test_js_to_json_malformed(self):
         self.assertEqual(js_to_json('42a1'), '42"a1"')
         self.assertEqual(js_to_json('42a-1'), '42"a"-1')
 
+    def test_js_to_json_template_literal(self):
+        self.assertEqual(js_to_json('`Hello ${name}`', {'name': '"world"'}), '"Hello world"')
+        self.assertEqual(js_to_json('`${name}${name}`', {'name': '"X"'}), '"XX"')
+        self.assertEqual(js_to_json('`${name}${name}`', {'name': '5'}), '"55"')
+        self.assertEqual(js_to_json('`${name}"${name}"`', {'name': '5'}), '"5\\"5\\""')
+        self.assertEqual(js_to_json('`${name}`', {}), '"name"')
+
     def test_extract_attributes(self):
         self.assertEqual(extract_attributes('<e x="y">'), {'x': 'y'})
         self.assertEqual(extract_attributes("<e x='y'>"), {'x': 'y'})
@@ -1565,238 +1671,6 @@ Line 1
         self.assertEqual(variadic('spam', allowed_types=dict), 'spam')
         self.assertEqual(variadic('spam', allowed_types=[dict]), 'spam')
 
-    def test_traverse_obj(self):
-        _TEST_DATA = {
-            100: 100,
-            1.2: 1.2,
-            'str': 'str',
-            'None': None,
-            '...': Ellipsis,
-            'urls': [
-                {'index': 0, 'url': 'https://www.example.com/0'},
-                {'index': 1, 'url': 'https://www.example.com/1'},
-            ],
-            'data': (
-                {'index': 2},
-                {'index': 3},
-            ),
-            'dict': {},
-        }
-
-        # Test base functionality
-        self.assertEqual(traverse_obj(_TEST_DATA, ('str',)), 'str',
-                         msg='allow tuple path')
-        self.assertEqual(traverse_obj(_TEST_DATA, ['str']), 'str',
-                         msg='allow list path')
-        self.assertEqual(traverse_obj(_TEST_DATA, (value for value in ("str",))), 'str',
-                         msg='allow iterable path')
-        self.assertEqual(traverse_obj(_TEST_DATA, 'str'), 'str',
-                         msg='single items should be treated as a path')
-        self.assertEqual(traverse_obj(_TEST_DATA, None), _TEST_DATA)
-        self.assertEqual(traverse_obj(_TEST_DATA, 100), 100)
-        self.assertEqual(traverse_obj(_TEST_DATA, 1.2), 1.2)
-
-        # Test Ellipsis behavior
-        self.assertCountEqual(traverse_obj(_TEST_DATA, Ellipsis),
-                              (item for item in _TEST_DATA.values() if item is not None),
-                              msg='`...` should give all values except `None`')
-        self.assertCountEqual(traverse_obj(_TEST_DATA, ('urls', 0, Ellipsis)), _TEST_DATA['urls'][0].values(),
-                              msg='`...` selection for dicts should select all values')
-        self.assertEqual(traverse_obj(_TEST_DATA, (Ellipsis, Ellipsis, 'url')),
-                         ['https://www.example.com/0', 'https://www.example.com/1'],
-                         msg='nested `...` queries should work')
-        self.assertCountEqual(traverse_obj(_TEST_DATA, (Ellipsis, Ellipsis, 'index')), range(4),
-                              msg='`...` query result should be flattened')
-
-        # Test function as key
-        self.assertEqual(traverse_obj(_TEST_DATA, lambda x, y: x == 'urls' and isinstance(y, list)),
-                         [_TEST_DATA['urls']],
-                         msg='function as query key should perform a filter based on (key, value)')
-        self.assertCountEqual(traverse_obj(_TEST_DATA, lambda _, x: isinstance(x[0], compat_str)), {'str'},
-                              msg='exceptions in the query function should be caught')
-
-        # Test alternative paths
-        self.assertEqual(traverse_obj(_TEST_DATA, 'fail', 'str'), 'str',
-                         msg='multiple `paths` should be treated as alternative paths')
-        self.assertEqual(traverse_obj(_TEST_DATA, 'str', 100), 'str',
-                         msg='alternatives should exit early')
-        self.assertEqual(traverse_obj(_TEST_DATA, 'fail', 'fail'), None,
-                         msg='alternatives should return `default` if exhausted')
-        self.assertEqual(traverse_obj(_TEST_DATA, (Ellipsis, 'fail'), 100), 100,
-                         msg='alternatives should track their own branching return')
-        self.assertEqual(traverse_obj(_TEST_DATA, ('dict', Ellipsis), ('data', Ellipsis)), list(_TEST_DATA['data']),
-                         msg='alternatives on empty objects should search further')
-
-        # Test branch and path nesting
-        self.assertEqual(traverse_obj(_TEST_DATA, ('urls', (3, 0), 'url')), ['https://www.example.com/0'],
-                         msg='tuple as key should be treated as branches')
-        self.assertEqual(traverse_obj(_TEST_DATA, ('urls', [3, 0], 'url')), ['https://www.example.com/0'],
-                         msg='list as key should be treated as branches')
-        self.assertEqual(traverse_obj(_TEST_DATA, ('urls', ((1, 'fail'), (0, 'url')))), ['https://www.example.com/0'],
-                         msg='double nesting in path should be treated as paths')
-        self.assertEqual(traverse_obj(['0', [1, 2]], [(0, 1), 0]), [1],
-                         msg='do not fail early on branching')
-        self.assertCountEqual(traverse_obj(_TEST_DATA, ('urls', ((1, ('fail', 'url')), (0, 'url')))),
-                              ['https://www.example.com/0', 'https://www.example.com/1'],
-                              msg='triple nesting in path should be treated as branches')
-        self.assertEqual(traverse_obj(_TEST_DATA, ('urls', ('fail', (Ellipsis, 'url')))),
-                         ['https://www.example.com/0', 'https://www.example.com/1'],
-                         msg='ellipsis as branch path start gets flattened')
-
-        # Test dictionary as key
-        self.assertEqual(traverse_obj(_TEST_DATA, {0: 100, 1: 1.2}), {0: 100, 1: 1.2},
-                         msg='dict key should result in a dict with the same keys')
-        self.assertEqual(traverse_obj(_TEST_DATA, {0: ('urls', 0, 'url')}),
-                         {0: 'https://www.example.com/0'},
-                         msg='dict key should allow paths')
-        self.assertEqual(traverse_obj(_TEST_DATA, {0: ('urls', (3, 0), 'url')}),
-                         {0: ['https://www.example.com/0']},
-                         msg='tuple in dict path should be treated as branches')
-        self.assertEqual(traverse_obj(_TEST_DATA, {0: ('urls', ((1, 'fail'), (0, 'url')))}),
-                         {0: ['https://www.example.com/0']},
-                         msg='double nesting in dict path should be treated as paths')
-        self.assertEqual(traverse_obj(_TEST_DATA, {0: ('urls', ((1, ('fail', 'url')), (0, 'url')))}),
-                         {0: ['https://www.example.com/1', 'https://www.example.com/0']},
-                         msg='triple nesting in dict path should be treated as branches')
-        self.assertEqual(traverse_obj(_TEST_DATA, {0: 'fail'}), {},
-                         msg='remove `None` values when dict key')
-        self.assertEqual(traverse_obj(_TEST_DATA, {0: 'fail'}, default=Ellipsis), {0: Ellipsis},
-                         msg='do not remove `None` values if `default`')
-        self.assertEqual(traverse_obj(_TEST_DATA, {0: 'dict'}), {0: {}},
-                         msg='do not remove empty values when dict key')
-        self.assertEqual(traverse_obj(_TEST_DATA, {0: 'dict'}, default=Ellipsis), {0: {}},
-                         msg='do not remove empty values when dict key and a default')
-        self.assertEqual(traverse_obj(_TEST_DATA, {0: ('dict', Ellipsis)}), {0: []},
-                         msg='if branch in dict key not successful, return `[]`')
-
-        # Testing default parameter behavior
-        _DEFAULT_DATA = {'None': None, 'int': 0, 'list': []}
-        self.assertEqual(traverse_obj(_DEFAULT_DATA, 'fail'), None,
-                         msg='default value should be `None`')
-        self.assertEqual(traverse_obj(_DEFAULT_DATA, 'fail', 'fail', default=Ellipsis), Ellipsis,
-                         msg='chained fails should result in default')
-        self.assertEqual(traverse_obj(_DEFAULT_DATA, 'None', 'int'), 0,
-                         msg='should not short cirquit on `None`')
-        self.assertEqual(traverse_obj(_DEFAULT_DATA, 'fail', default=1), 1,
-                         msg='invalid dict key should result in `default`')
-        self.assertEqual(traverse_obj(_DEFAULT_DATA, 'None', default=1), 1,
-                         msg='`None` is a deliberate sentinel and should become `default`')
-        self.assertEqual(traverse_obj(_DEFAULT_DATA, ('list', 10)), None,
-                         msg='`IndexError` should result in `default`')
-        self.assertEqual(traverse_obj(_DEFAULT_DATA, (Ellipsis, 'fail'), default=1), 1,
-                         msg='if branched but not successful return `default` if defined, not `[]`')
-        self.assertEqual(traverse_obj(_DEFAULT_DATA, (Ellipsis, 'fail'), default=None), None,
-                         msg='if branched but not successful return `default` even if `default` is `None`')
-        self.assertEqual(traverse_obj(_DEFAULT_DATA, (Ellipsis, 'fail')), [],
-                         msg='if branched but not successful return `[]`, not `default`')
-        self.assertEqual(traverse_obj(_DEFAULT_DATA, ('list', Ellipsis)), [],
-                         msg='if branched but object is empty return `[]`, not `default`')
-
-        # Testing expected_type behavior
-        _EXPECTED_TYPE_DATA = {'str': 'str', 'int': 0}
-        self.assertEqual(traverse_obj(_EXPECTED_TYPE_DATA, 'str', expected_type=compat_str), 'str',
-                         msg='accept matching `expected_type` type')
-        self.assertEqual(traverse_obj(_EXPECTED_TYPE_DATA, 'str', expected_type=int), None,
-                         msg='reject non matching `expected_type` type')
-        self.assertEqual(traverse_obj(_EXPECTED_TYPE_DATA, 'int', expected_type=lambda x: compat_str(x)), '0',
-                         msg='transform type using type function')
-        self.assertEqual(traverse_obj(_EXPECTED_TYPE_DATA, 'str',
-                                      expected_type=lambda _: 1 / 0), None,
-                         msg='wrap expected_type function in try_call')
-        self.assertEqual(traverse_obj(_EXPECTED_TYPE_DATA, Ellipsis, expected_type=compat_str), ['str'],
-                         msg='eliminate items that expected_type fails on')
-
-        # Test get_all behavior
-        _GET_ALL_DATA = {'key': [0, 1, 2]}
-        self.assertEqual(traverse_obj(_GET_ALL_DATA, ('key', Ellipsis), get_all=False), 0,
-                         msg='if not `get_all`, return only first matching value')
-        self.assertEqual(traverse_obj(_GET_ALL_DATA, Ellipsis, get_all=False), [0, 1, 2],
-                         msg='do not overflatten if not `get_all`')
-
-        # Test casesense behavior
-        _CASESENSE_DATA = {
-            'KeY': 'value0',
-            0: {
-                'KeY': 'value1',
-                0: {'KeY': 'value2'},
-            },
-            # FULLWIDTH LATIN CAPITAL LETTER K
-            '\uff2bey': 'value3',
-        }
-        self.assertEqual(traverse_obj(_CASESENSE_DATA, 'key'), None,
-                         msg='dict keys should be case sensitive unless `casesense`')
-        self.assertEqual(traverse_obj(_CASESENSE_DATA, 'keY',
-                                      casesense=False), 'value0',
-                         msg='allow non matching key case if `casesense`')
-        self.assertEqual(traverse_obj(_CASESENSE_DATA, '\uff4bey',  # FULLWIDTH LATIN SMALL LETTER K
-                                      casesense=False), 'value3',
-                         msg='allow non matching Unicode key case if `casesense`')
-        self.assertEqual(traverse_obj(_CASESENSE_DATA, (0, ('keY',)),
-                                      casesense=False), ['value1'],
-                         msg='allow non matching key case in branch if `casesense`')
-        self.assertEqual(traverse_obj(_CASESENSE_DATA, (0, ((0, 'keY'),)),
-                                      casesense=False), ['value2'],
-                         msg='allow non matching key case in branch path if `casesense`')
-
-        # Test traverse_string behavior
-        _TRAVERSE_STRING_DATA = {'str': 'str', 1.2: 1.2}
-        self.assertEqual(traverse_obj(_TRAVERSE_STRING_DATA, ('str', 0)), None,
-                         msg='do not traverse into string if not `traverse_string`')
-        self.assertEqual(traverse_obj(_TRAVERSE_STRING_DATA, ('str', 0),
-                                      _traverse_string=True), 's',
-                         msg='traverse into string if `traverse_string`')
-        self.assertEqual(traverse_obj(_TRAVERSE_STRING_DATA, (1.2, 1),
-                                      _traverse_string=True), '.',
-                         msg='traverse into converted data if `traverse_string`')
-        self.assertEqual(traverse_obj(_TRAVERSE_STRING_DATA, ('str', Ellipsis),
-                                      _traverse_string=True), list('str'),
-                         msg='`...` branching into string should result in list')
-        self.assertEqual(traverse_obj(_TRAVERSE_STRING_DATA, ('str', (0, 2)),
-                                      _traverse_string=True), ['s', 'r'],
-                         msg='branching into string should result in list')
-        self.assertEqual(traverse_obj(_TRAVERSE_STRING_DATA, ('str', lambda _, x: x),
-                                      _traverse_string=True), list('str'),
-                         msg='function branching into string should result in list')
-
-        # Test is_user_input behavior
-        _IS_USER_INPUT_DATA = {'range8': list(range(8))}
-        self.assertEqual(traverse_obj(_IS_USER_INPUT_DATA, ('range8', '3'),
-                                      _is_user_input=True), 3,
-                         msg='allow for string indexing if `is_user_input`')
-        self.assertCountEqual(traverse_obj(_IS_USER_INPUT_DATA, ('range8', '3:'),
-                                           _is_user_input=True), tuple(range(8))[3:],
-                              msg='allow for string slice if `is_user_input`')
-        self.assertCountEqual(traverse_obj(_IS_USER_INPUT_DATA, ('range8', ':4:2'),
-                                           _is_user_input=True), tuple(range(8))[:4:2],
-                              msg='allow step in string slice if `is_user_input`')
-        self.assertCountEqual(traverse_obj(_IS_USER_INPUT_DATA, ('range8', ':'),
-                                           _is_user_input=True), range(8),
-                              msg='`:` should be treated as `...` if `is_user_input`')
-        with self.assertRaises(TypeError, msg='too many params should result in error'):
-            traverse_obj(_IS_USER_INPUT_DATA, ('range8', ':::'), _is_user_input=True)
-
-        # Test re.Match as input obj
-        mobj = re.match(r'^0(12)(?P<group>3)(4)?$', '0123')
-        self.assertEqual(traverse_obj(mobj, Ellipsis), [x for x in mobj.groups() if x is not None],
-                         msg='`...` on a `re.Match` should give its `groups()`')
-        self.assertEqual(traverse_obj(mobj, lambda k, _: k in (0, 2)), ['0123', '3'],
-                         msg='function on a `re.Match` should give groupno, value starting at 0')
-        self.assertEqual(traverse_obj(mobj, 'group'), '3',
-                         msg='str key on a `re.Match` should give group with that name')
-        self.assertEqual(traverse_obj(mobj, 2), '3',
-                         msg='int key on a `re.Match` should give group with that name')
-        self.assertEqual(traverse_obj(mobj, 'gRoUp', casesense=False), '3',
-                         msg='str key on a `re.Match` should respect casesense')
-        self.assertEqual(traverse_obj(mobj, 'fail'), None,
-                         msg='failing str key on a `re.Match` should return `default`')
-        self.assertEqual(traverse_obj(mobj, 'gRoUpS', casesense=False), None,
-                         msg='failing str key on a `re.Match` should return `default`')
-        self.assertEqual(traverse_obj(mobj, 8), None,
-                         msg='failing int key on a `re.Match` should return `default`')
-
-    def test_get_first(self):
-        self.assertEqual(get_first([{'a': None}, {'a': 'spam'}], 'a'), 'spam')
-
     def test_join_nonempty(self):
         self.assertEqual(join_nonempty('a', 'b'), 'a-b')
         self.assertEqual(join_nonempty(

test/test_write_annotations.py

@@ -11,12 +11,11 @@ sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
 from test.helper import get_params, try_rm
 
 
-import io
-
 import xml.etree.ElementTree
 
 import youtube_dl.YoutubeDL
 import youtube_dl.extractor
+from youtube_dl.compat import compat_open as open
 
 
 class YoutubeDL(youtube_dl.YoutubeDL):
@@ -51,7 +50,7 @@ class TestAnnotations(unittest.TestCase):
         ydl.download([TEST_ID])
         self.assertTrue(os.path.exists(ANNOTATIONS_FILE))
         annoxml = None
-        with io.open(ANNOTATIONS_FILE, 'r', encoding='utf-8') as annof:
+        with open(ANNOTATIONS_FILE, 'r', encoding='utf-8') as annof:
             annoxml = xml.etree.ElementTree.parse(annof)
         self.assertTrue(annoxml is not None, 'Failed to parse annotations XML')
         root = annoxml.getroot()

test/test_youtube_signature.py

@@ -8,11 +8,14 @@ import sys
 import unittest
 sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
 
-import io
 import re
 import string
 
-from youtube_dl.compat import compat_str, compat_urlretrieve
+from youtube_dl.compat import (
+    compat_open as open,
+    compat_str,
+    compat_urlretrieve,
+)
 
 from test.helper import FakeYDL
 from youtube_dl.extractor import YoutubeIE
@@ -152,8 +155,12 @@ _NSIG_TESTS = [
         'qO0NiMtYQ7TeJnfFG2', 'k9cuJDHNS5O7kQ',
     ),
     (
-        'https://www.youtube.com/s/player/8c7583ff/player_ias.vflset/en_US/base.js',
-        'E2AQVN6y_zM7uN9w8z', '9A2dbY5GDZrt9A',
+        'https://www.youtube.com/s/player/b7910ca8/player_ias.vflset/en_US/base.js',
+        '_hXMCwMt9qE310D', 'LoZMgkkofRMCZQ',
+    ),
+    (
+        'https://www.youtube.com/s/player/590f65a6/player_ias.vflset/en_US/base.js',
+        '1tm7-g_A9zsI8_Lay_', 'xI4Vem4Put_rOg',
     ),
 ]
 
@@ -208,7 +215,7 @@ def t_factory(name, sig_func, url_pattern):
 
             if not os.path.exists(fn):
                 compat_urlretrieve(url, fn)
-            with io.open(fn, encoding='utf-8') as testf:
+            with open(fn, encoding='utf-8') as testf:
                 jscode = testf.read()
             self.assertEqual(sig_func(jscode, sig_input), expected_sig)
 

test/testdata/mpd/range_only.mpd

@@ -0,0 +1,35 @@
+<?xml version="1.0"?>
+<!-- MPD file Generated with GPAC version 1.0.1-revrelease at 2021-11-27T20:53:11.690Z -->
+<MPD xmlns="urn:mpeg:dash:schema:mpd:2011" minBufferTime="PT1.500S" type="static" mediaPresentationDuration="PT0H0M30.196S" maxSegmentDuration="PT0H0M10.027S" profiles="urn:mpeg:dash:profile:full:2011">
+ <ProgramInformation moreInformationURL="http://gpac.io">
+  <Title>manifest.mpd generated by GPAC</Title>
+ </ProgramInformation>
+
+ <Period duration="PT0H0M30.196S">
+  <AdaptationSet segmentAlignment="true" maxWidth="768" maxHeight="432" maxFrameRate="30000/1001" par="16:9" lang="und" startWithSAP="1">
+   <Representation id="1" mimeType="video/mp4" codecs="avc1.4D401E" width="768" height="432" frameRate="30000/1001" sar="1:1" bandwidth="526987">
+    <BaseURL>video_dashinit.mp4</BaseURL>
+    <SegmentList timescale="90000" duration="900000">
+     <Initialization range="0-881"/>
+     <SegmentURL mediaRange="882-876094" indexRange="882-925"/>
+     <SegmentURL mediaRange="876095-1466732" indexRange="876095-876138"/>
+     <SegmentURL mediaRange="1466733-1953615" indexRange="1466733-1466776"/>
+     <SegmentURL mediaRange="1953616-1994211" indexRange="1953616-1953659"/>
+    </SegmentList>
+   </Representation>
+  </AdaptationSet>
+  <AdaptationSet segmentAlignment="true" lang="und" startWithSAP="1">
+   <Representation id="2" mimeType="audio/mp4" codecs="mp4a.40.2" audioSamplingRate="48000" bandwidth="98096">
+    <AudioChannelConfiguration schemeIdUri="urn:mpeg:dash:23003:3:audio_channel_configuration:2011" value="2"/>
+    <BaseURL>audio_dashinit.mp4</BaseURL>
+    <SegmentList timescale="48000" duration="480000">
+     <Initialization range="0-752"/>
+     <SegmentURL mediaRange="753-124129" indexRange="753-796"/>
+     <SegmentURL mediaRange="124130-250544" indexRange="124130-124173"/>
+     <SegmentURL mediaRange="250545-374929" indexRange="250545-250588"/>
+    </SegmentList>
+   </Representation>
+  </AdaptationSet>
+ </Period>
+</MPD>
+

test/testdata/mpd/subtitles.mpd

@@ -0,0 +1,351 @@
+<?xml version="1.0" encoding="utf-8"?>
+<!-- Created with Unified Streaming Platform (version=1.10.18-20255) -->
+<MPD
+  xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
+  xmlns="urn:mpeg:dash:schema:mpd:2011"
+  xsi:schemaLocation="urn:mpeg:dash:schema:mpd:2011 http://standards.iso.org/ittf/PubliclyAvailableStandards/MPEG-DASH_schema_files/DASH-MPD.xsd"
+  type="static"
+  mediaPresentationDuration="PT14M48S"
+  maxSegmentDuration="PT1M"
+  minBufferTime="PT10S"
+  profiles="urn:mpeg:dash:profile:isoff-live:2011">
+  <Period
+    id="1"
+    duration="PT14M48S">
+    <BaseURL>dash/</BaseURL>
+    <AdaptationSet
+      id="1"
+      group="1"
+      contentType="audio"
+      segmentAlignment="true"
+      audioSamplingRate="48000"
+      mimeType="audio/mp4"
+      codecs="mp4a.40.2"
+      startWithSAP="1">
+      <AudioChannelConfiguration
+        schemeIdUri="urn:mpeg:dash:23003:3:audio_channel_configuration:2011"
+        value="2" />
+      <Role schemeIdUri="urn:mpeg:dash:role:2011" value="main" />
+      <SegmentTemplate
+        timescale="48000"
+        initialization="3144-kZT4LWMQw6Rh7Kpd-$RepresentationID$.dash"
+        media="3144-kZT4LWMQw6Rh7Kpd-$RepresentationID$-$Time$.dash">
+        <SegmentTimeline>
+          <S t="0" d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="3584" />
+        </SegmentTimeline>
+      </SegmentTemplate>
+      <Representation
+        id="audio=128001"
+        bandwidth="128001">
+      </Representation>
+    </AdaptationSet>
+    <AdaptationSet
+      id="2"
+      group="3"
+      contentType="text"
+      lang="en"
+      mimeType="application/mp4"
+      codecs="stpp"
+      startWithSAP="1">
+      <Role schemeIdUri="urn:mpeg:dash:role:2011" value="subtitle" />
+      <SegmentTemplate
+        timescale="1000"
+        initialization="3144-kZT4LWMQw6Rh7Kpd-$RepresentationID$.dash"
+        media="3144-kZT4LWMQw6Rh7Kpd-$RepresentationID$-$Time$.dash">
+        <SegmentTimeline>
+          <S t="0" d="60000" r="9" />
+          <S d="24000" />
+        </SegmentTimeline>
+      </SegmentTemplate>
+      <Representation
+        id="textstream_eng=1000"
+        bandwidth="1000">
+      </Representation>
+    </AdaptationSet>
+    <AdaptationSet
+      id="3"
+      group="2"
+      contentType="video"
+      par="960:409"
+      minBandwidth="100000"
+      maxBandwidth="4482000"
+      maxWidth="1689"
+      maxHeight="720"
+      segmentAlignment="true"
+      mimeType="video/mp4"
+      codecs="avc1.4D401F"
+      startWithSAP="1">
+      <Role schemeIdUri="urn:mpeg:dash:role:2011" value="main" />
+      <SegmentTemplate
+        timescale="12288"
+        initialization="3144-kZT4LWMQw6Rh7Kpd-$RepresentationID$.dash"
+        media="3144-kZT4LWMQw6Rh7Kpd-$RepresentationID$-$Time$.dash">
+        <SegmentTimeline>
+          <S t="0" d="24576" r="443" />
+        </SegmentTimeline>
+      </SegmentTemplate>
+      <Representation
+        id="video=100000"
+        bandwidth="100000"
+        width="336"
+        height="144"
+        sar="2880:2863"
+        scanType="progressive">
+      </Representation>
+      <Representation
+        id="video=326000"
+        bandwidth="326000"
+        width="562"
+        height="240"
+        sar="115200:114929"
+        scanType="progressive">
+      </Representation>
+      <Representation
+        id="video=698000"
+        bandwidth="698000"
+        width="844"
+        height="360"
+        sar="86400:86299"
+        scanType="progressive">
+      </Representation>
+      <Representation
+        id="video=1493000"
+        bandwidth="1493000"
+        width="1126"
+        height="480"
+        sar="230400:230267"
+        scanType="progressive">
+      </Representation>
+      <Representation
+        id="video=4482000"
+        bandwidth="4482000"
+        width="1688"
+        height="720"
+        sar="86400:86299"
+        scanType="progressive">
+      </Representation>
+    </AdaptationSet>
+  </Period>
+</MPD>

test/testdata/mpd/url_and_range.mpd

@@ -0,0 +1,32 @@
+<?xml version="1.0" ?>
+<MPD xmlns="urn:mpeg:dash:schema:mpd:2011" profiles="urn:mpeg:dash:profile:isoff-live:2011" minBufferTime="PT10.01S" mediaPresentationDuration="PT30.097S" type="static">
+  <!-- Created with Bento4 mp4-dash.py, VERSION=2.0.0-639 -->
+  <Period>
+    <!-- Video -->
+    <AdaptationSet mimeType="video/mp4" segmentAlignment="true" startWithSAP="1" maxWidth="768" maxHeight="432">
+      <Representation id="video-avc1" codecs="avc1.4D401E" width="768" height="432" scanType="progressive" frameRate="30000/1001" bandwidth="699597">
+        <SegmentList timescale="1000" duration="10010">
+          <Initialization sourceURL="video-frag.mp4" range="36-746"/>
+          <SegmentURL media="video-frag.mp4" mediaRange="747-876117"/>
+          <SegmentURL media="video-frag.mp4" mediaRange="876118-1466913"/>
+          <SegmentURL media="video-frag.mp4" mediaRange="1466914-1953954"/>
+          <SegmentURL media="video-frag.mp4" mediaRange="1953955-1994652"/>
+        </SegmentList>
+      </Representation>
+    </AdaptationSet>
+    <!-- Audio -->
+    <AdaptationSet mimeType="audio/mp4" startWithSAP="1" segmentAlignment="true">
+      <Representation id="audio-und-mp4a.40.2" codecs="mp4a.40.2" bandwidth="98808" audioSamplingRate="48000">
+        <AudioChannelConfiguration schemeIdUri="urn:mpeg:mpegB:cicp:ChannelConfiguration" value="2"/>
+        <SegmentList timescale="1000" duration="10010">
+          <Initialization sourceURL="audio-frag.mp4" range="32-623"/>
+          <SegmentURL media="audio-frag.mp4" mediaRange="624-124199"/>
+          <SegmentURL media="audio-frag.mp4" mediaRange="124200-250303"/>
+          <SegmentURL media="audio-frag.mp4" mediaRange="250304-374365"/>
+          <SegmentURL media="audio-frag.mp4" mediaRange="374366-374836"/>
+        </SegmentList>
+      </Representation>
+    </AdaptationSet>
+  </Period>
+</MPD>
+

youtube_dl/YoutubeDL.py

@@ -4,10 +4,9 @@
 from __future__ import absolute_import, unicode_literals
 
 import collections
-import contextlib
+import copy
 import datetime
 import errno
-import fileinput
 import io
 import itertools
 import json
@@ -25,19 +24,26 @@ import tokenize
 import traceback
 import random
 
-from ssl import OPENSSL_VERSION
+try:
+    from ssl import OPENSSL_VERSION
+except ImportError:
+    # Must be Python 2.6, should be built against 1.0.2
+    OPENSSL_VERSION = 'OpenSSL 1.0.2(?)'
 from string import ascii_letters
 
 from .compat import (
     compat_basestring,
-    compat_cookiejar,
+    compat_collections_chain_map as ChainMap,
     compat_filter as filter,
     compat_get_terminal_size,
     compat_http_client,
+    compat_http_cookiejar_Cookie,
+    compat_http_cookies_SimpleCookie,
     compat_integer_types,
     compat_kwargs,
     compat_map as map,
     compat_numeric_types,
+    compat_open as open,
     compat_os_name,
     compat_str,
     compat_tokenize_tokenize,
@@ -49,6 +55,7 @@ from .compat import (
 from .utils import (
     age_restricted,
     args_to_str,
+    bug_reports_message,
     ContentTooShortError,
     date_from_str,
     DateRange,
@@ -64,7 +71,6 @@ from .utils import (
     format_bytes,
     formatSeconds,
     GeoRestrictedError,
-    HEADRequest,
     int_or_none,
     ISO3166Utils,
     join_nonempty,
@@ -81,7 +87,6 @@ from .utils import (
     preferredencoding,
     prepend_extension,
     process_communicate_or_kill,
-    PUTRequest,
     register_socks_protocols,
     render_table,
     replace_extension,
@@ -93,6 +98,7 @@ from .utils import (
     std_headers,
     str_or_none,
     subtitles_filename,
+    traverse_obj,
     UnavailableVideoError,
     url_basename,
     version_tuple,
@@ -372,6 +378,9 @@ class YoutubeDL(object):
         self.params.update(params)
         self.cache = Cache(self)
 
+        self._header_cookies = []
+        self._load_cookies_from_headers(self.params.get('http_headers'))
+
         def check_deprecated(param, option, suggestion):
             if self.params.get(param) is not None:
                 self.report_warning(
@@ -578,7 +587,7 @@ class YoutubeDL(object):
         if self.params.get('cookiefile') is not None:
             self.cookiejar.save(ignore_discard=True, ignore_expires=True)
 
-    def trouble(self, message=None, tb=None):
+    def trouble(self, *args, **kwargs):
         """Determine action to take when a download problem appears.
 
         Depending on if the downloader has been configured to ignore
@@ -587,6 +596,11 @@ class YoutubeDL(object):
 
         tb, if given, is additional traceback information.
         """
+        # message=None, tb=None, is_error=True
+        message = args[0] if len(args) > 0 else kwargs.get('message', None)
+        tb = args[1] if len(args) > 1 else kwargs.get('tb', None)
+        is_error = args[2] if len(args) > 2 else kwargs.get('is_error', True)
+
         if message is not None:
             self.to_stderr(message)
         if self.params.get('verbose'):
@@ -599,7 +613,10 @@ class YoutubeDL(object):
                 else:
                     tb_data = traceback.format_list(traceback.extract_stack())
                     tb = ''.join(tb_data)
-            self.to_stderr(tb)
+            if tb:
+                self.to_stderr(tb)
+        if not is_error:
+            return
         if not self.params.get('ignoreerrors', False):
             if sys.exc_info()[0] and hasattr(sys.exc_info()[1], 'exc_info') and sys.exc_info()[1].exc_info[0]:
                 exc_info = sys.exc_info()[1].exc_info
@@ -608,11 +625,18 @@ class YoutubeDL(object):
             raise DownloadError(message, exc_info)
         self._download_retcode = 1
 
-    def report_warning(self, message):
+    def report_warning(self, message, only_once=False, _cache={}):
         '''
         Print the message to stderr, it will be prefixed with 'WARNING:'
         If stderr is a tty file the 'WARNING:' will be colored
         '''
+        if only_once:
+            m_hash = hash((self, message))
+            m_cnt = _cache.setdefault(m_hash, 0)
+            _cache[m_hash] = m_cnt + 1
+            if m_cnt > 0:
+                return
+
         if self.params.get('logger') is not None:
             self.params['logger'].warning(message)
         else:
@@ -625,7 +649,7 @@ class YoutubeDL(object):
             warning_message = '%s %s' % (_msg_header, message)
             self.to_stderr(warning_message)
 
-    def report_error(self, message, tb=None):
+    def report_error(self, message, *args, **kwargs):
         '''
         Do the same as trouble, but prefixes the message with 'ERROR:', colored
         in red if stderr is a tty file.
@@ -634,8 +658,18 @@ class YoutubeDL(object):
             _msg_header = '\033[0;31mERROR:\033[0m'
         else:
             _msg_header = 'ERROR:'
-        error_message = '%s %s' % (_msg_header, message)
-        self.trouble(error_message, tb)
+        kwargs['message'] = '%s %s' % (_msg_header, message)
+        self.trouble(*args, **kwargs)
+
+    def report_unscoped_cookies(self, *args, **kwargs):
+        # message=None, tb=False, is_error=False
+        if len(args) <= 2:
+            kwargs.setdefault('is_error', False)
+            if len(args) <= 0:
+                kwargs.setdefault(
+                    'message',
+                    'Unscoped cookies are not allowed: please specify some sort of scoping')
+        self.report_error(*args, **kwargs)
 
     def report_file_already_downloaded(self, file_name):
         """Report file has already been fully downloaded."""
@@ -831,7 +865,7 @@ class YoutubeDL(object):
                 msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.'
                 self.report_error(msg)
             except ExtractorError as e:  # An error we somewhat expected
-                self.report_error(compat_str(e), e.format_traceback())
+                self.report_error(compat_str(e), tb=e.format_traceback())
             except MaxDownloadsReached:
                 raise
             except Exception as e:
@@ -841,8 +875,83 @@ class YoutubeDL(object):
                     raise
         return wrapper
 
+    def _remove_cookie_header(self, http_headers):
+        """Filters out `Cookie` header from an `http_headers` dict
+        The `Cookie` header is removed to prevent leaks as a result of unscoped cookies.
+        See: https://github.com/yt-dlp/yt-dlp/security/advisories/GHSA-v8mc-9377-rwjj
+
+        @param http_headers     An `http_headers` dict from which any `Cookie` header
+                                should be removed, or None
+        """
+        return dict(filter(lambda pair: pair[0].lower() != 'cookie', (http_headers or {}).items()))
+
+    def _load_cookies(self, data, **kwargs):
+        """Loads cookies from a `Cookie` header
+
+        This tries to work around the security vulnerability of passing cookies to every domain.
+
+        @param data         The Cookie header as a string to load the cookies from
+        @param autoscope    If `False`, scope cookies using Set-Cookie syntax and error for cookie without domains
+                            If `True`, save cookies for later to be stored in the jar with a limited scope
+                            If a URL, save cookies in the jar with the domain of the URL
+        """
+        # autoscope=True (kw-only)
+        autoscope = kwargs.get('autoscope', True)
+
+        for cookie in compat_http_cookies_SimpleCookie(data).values() if data else []:
+            if autoscope and any(cookie.values()):
+                raise ValueError('Invalid syntax in Cookie Header')
+
+            domain = cookie.get('domain') or ''
+            expiry = cookie.get('expires')
+            if expiry == '':  # 0 is valid so we check for `''` explicitly
+                expiry = None
+            prepared_cookie = compat_http_cookiejar_Cookie(
+                cookie.get('version') or 0, cookie.key, cookie.value, None, False,
+                domain, True, True, cookie.get('path') or '', bool(cookie.get('path')),
+                bool(cookie.get('secure')), expiry, False, None, None, {})
+
+            if domain:
+                self.cookiejar.set_cookie(prepared_cookie)
+            elif autoscope is True:
+                self.report_warning(
+                    'Passing cookies as a header is a potential security risk; '
+                    'they will be scoped to the domain of the downloaded urls. '
+                    'Please consider loading cookies from a file or browser instead.',
+                    only_once=True)
+                self._header_cookies.append(prepared_cookie)
+            elif autoscope:
+                self.report_warning(
+                    'The extractor result contains an unscoped cookie as an HTTP header. '
+                    'If you are specifying an input URL, ' + bug_reports_message(),
+                    only_once=True)
+                self._apply_header_cookies(autoscope, [prepared_cookie])
+            else:
+                self.report_unscoped_cookies()
+
+    def _load_cookies_from_headers(self, headers):
+        self._load_cookies(traverse_obj(headers, 'cookie', casesense=False))
+
+    def _apply_header_cookies(self, url, cookies=None):
+        """This method applies stray header cookies to the provided url
+
+        This loads header cookies and scopes them to the domain provided in `url`.
+        While this is not ideal, it helps reduce the risk of them being sent to
+        an unintended destination.
+        """
+        parsed = compat_urllib_parse.urlparse(url)
+        if not parsed.hostname:
+            return
+
+        for cookie in map(copy.copy, cookies or self._header_cookies):
+            cookie.domain = '.' + parsed.hostname
+            self.cookiejar.set_cookie(cookie)
+
     @__handle_extraction_exceptions
     def __extract_info(self, url, ie, download, extra_info, process):
+        # Compat with passing cookies in http headers
+        self._apply_header_cookies(url)
+
         ie_result = ie.extract(url)
         if ie_result is None:  # Finished already (backwards compatibility; listformats and friends should be moved here)
             return
@@ -868,7 +977,7 @@ class YoutubeDL(object):
 
     def process_ie_result(self, ie_result, download=True, extra_info={}):
         """
-        Take the result of the ie(may be modified) and resolve all unresolved
+        Take the result of the ie (may be modified) and resolve all unresolved
         references (URLs, playlist items).
 
         It will also download the videos if 'download'.
@@ -930,8 +1039,8 @@ class YoutubeDL(object):
         elif result_type in ('playlist', 'multi_video'):
             # Protect from infinite recursion due to recursively nested playlists
             # (see https://github.com/ytdl-org/youtube-dl/issues/27833)
-            webpage_url = ie_result['webpage_url']
-            if webpage_url in self._playlist_urls:
+            webpage_url = ie_result.get('webpage_url')  # not all pl/mv have this
+            if webpage_url and webpage_url in self._playlist_urls:
                 self.to_screen(
                     '[download] Skipping already downloaded playlist: %s'
                     % ie_result.get('title') or ie_result.get('id'))
@@ -939,6 +1048,10 @@ class YoutubeDL(object):
 
             self._playlist_level += 1
             self._playlist_urls.add(webpage_url)
+            new_result = dict((k, v) for k, v in extra_info.items() if k not in ie_result)
+            if new_result:
+                new_result.update(ie_result)
+                ie_result = new_result
             try:
                 return self.__process_playlist(ie_result, download)
             finally:
@@ -1439,29 +1552,73 @@ class YoutubeDL(object):
         parsed_selector = _parse_format_selection(iter(TokenIterator(tokens)))
         return _build_selector_function(parsed_selector)
 
-    def _calc_headers(self, info_dict):
-        res = std_headers.copy()
+    def _calc_headers(self, info_dict, load_cookies=False):
+        if load_cookies:  # For --load-info-json
+            # load cookies from http_headers in legacy info.json
+            self._load_cookies(traverse_obj(info_dict, ('http_headers', 'Cookie'), casesense=False),
+                               autoscope=info_dict['url'])
+            # load scoped cookies from info.json
+            self._load_cookies(info_dict.get('cookies'), autoscope=False)
 
-        add_headers = info_dict.get('http_headers')
-        if add_headers:
-            res.update(add_headers)
-
-        cookies = self._calc_cookies(info_dict)
+        cookies = self.cookiejar.get_cookies_for_url(info_dict['url'])
         if cookies:
-            res['Cookie'] = cookies
+            # Make a string like name1=val1; attr1=a_val1; ...name2=val2; ...
+            # By convention a cookie name can't be a well-known attribute name
+            # so this syntax is unambiguous and can be parsed by (eg) SimpleCookie
+            encoder = compat_http_cookies_SimpleCookie()
+            values = []
+            attributes = (('Domain', '='), ('Path', '='), ('Secure',), ('Expires', '='), ('Version', '='))
+            attributes = tuple([x[0].lower()] + list(x) for x in attributes)
+            for cookie in cookies:
+                _, value = encoder.value_encode(cookie.value)
+                # Py 2 '' --> '', Py 3 '' --> '""'
+                if value == '':
+                    value = '""'
+                values.append('='.join((cookie.name, value)))
+                for attr in attributes:
+                    value = getattr(cookie, attr[0], None)
+                    if value:
+                        values.append('%s%s' % (''.join(attr[1:]), value if len(attr) == 3 else ''))
+            info_dict['cookies'] = '; '.join(values)
+
+        res = std_headers.copy()
+        res.update(info_dict.get('http_headers') or {})
+        res = self._remove_cookie_header(res)
 
         if 'X-Forwarded-For' not in res:
             x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
             if x_forwarded_for_ip:
                 res['X-Forwarded-For'] = x_forwarded_for_ip
 
-        return res
+        return res or None
 
     def _calc_cookies(self, info_dict):
         pr = sanitized_Request(info_dict['url'])
         self.cookiejar.add_cookie_header(pr)
         return pr.get_header('Cookie')
 
+    def _fill_common_fields(self, info_dict, final=True):
+
+        for ts_key, date_key in (
+                ('timestamp', 'upload_date'),
+                ('release_timestamp', 'release_date'),
+        ):
+            if info_dict.get(date_key) is None and info_dict.get(ts_key) is not None:
+                # Working around out-of-range timestamp values (e.g. negative ones on Windows,
+                # see http://bugs.python.org/issue1646728)
+                try:
+                    upload_date = datetime.datetime.utcfromtimestamp(info_dict[ts_key])
+                    info_dict[date_key] = compat_str(upload_date.strftime('%Y%m%d'))
+                except (ValueError, OverflowError, OSError):
+                    pass
+
+        # Auto generate title fields corresponding to the *_number fields when missing
+        # in order to always have clean titles. This is very common for TV series.
+        if final:
+            for field in ('chapter', 'season', 'episode'):
+                if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
+                    info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
+
     def process_video_result(self, info_dict, download=True):
         assert info_dict.get('_type', 'video') == 'video'
 
@@ -1529,24 +1686,7 @@ class YoutubeDL(object):
         if 'display_id' not in info_dict and 'id' in info_dict:
             info_dict['display_id'] = info_dict['id']
 
-        for ts_key, date_key in (
-                ('timestamp', 'upload_date'),
-                ('release_timestamp', 'release_date'),
-        ):
-            if info_dict.get(date_key) is None and info_dict.get(ts_key) is not None:
-                # Working around out-of-range timestamp values (e.g. negative ones on Windows,
-                # see http://bugs.python.org/issue1646728)
-                try:
-                    upload_date = datetime.datetime.utcfromtimestamp(info_dict[ts_key])
-                    info_dict[date_key] = compat_str(upload_date.strftime('%Y%m%d'))
-                except (ValueError, OverflowError, OSError):
-                    pass
-
-        # Auto generate title fields corresponding to the *_number fields when missing
-        # in order to always have clean titles. This is very common for TV series.
-        for field in ('chapter', 'season', 'episode'):
-            if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
-                info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
+        self._fill_common_fields(info_dict)
 
         for cc_kind in ('subtitles', 'automatic_captions'):
             cc = info_dict.get(cc_kind)
@@ -1634,10 +1774,13 @@ class YoutubeDL(object):
                 format['protocol'] = determine_protocol(format)
             # Add HTTP headers, so that external programs can use them from the
             # json output
-            full_format_info = info_dict.copy()
-            full_format_info.update(format)
-            format['http_headers'] = self._calc_headers(full_format_info)
-        # Remove private housekeeping stuff
+            format['http_headers'] = self._calc_headers(ChainMap(format, info_dict), load_cookies=True)
+
+        # Safeguard against old/insecure infojson when using --load-info-json
+        info_dict['http_headers'] = self._remove_cookie_header(
+            info_dict.get('http_headers') or {}) or None
+
+        # Remove private housekeeping stuff (copied to http_headers in _calc_headers())
         if '__x_forwarded_for_ip' in info_dict:
             del info_dict['__x_forwarded_for_ip']
 
@@ -1840,7 +1983,7 @@ class YoutubeDL(object):
             else:
                 try:
                     self.to_screen('[info] Writing video description to: ' + descfn)
-                    with io.open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
+                    with open(encodeFilename(descfn), 'w', encoding='utf-8') as descfile:
                         descfile.write(info_dict['description'])
                 except (OSError, IOError):
                     self.report_error('Cannot write description file ' + descfn)
@@ -1855,7 +1998,7 @@ class YoutubeDL(object):
             else:
                 try:
                     self.to_screen('[info] Writing video annotations to: ' + annofn)
-                    with io.open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
+                    with open(encodeFilename(annofn), 'w', encoding='utf-8') as annofile:
                         annofile.write(info_dict['annotations'])
                 except (KeyError, TypeError):
                     self.report_warning('There are no annotations to write.')
@@ -1882,7 +2025,7 @@ class YoutubeDL(object):
                         try:
                             # Use newline='' to prevent conversion of newline characters
                             # See https://github.com/ytdl-org/youtube-dl/issues/10268
-                            with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8', newline='') as subfile:
+                            with open(encodeFilename(sub_filename), 'w', encoding='utf-8', newline='') as subfile:
                                 subfile.write(sub_info['data'])
                         except (OSError, IOError):
                             self.report_error('Cannot write subtitles file ' + sub_filename)
@@ -1891,24 +2034,16 @@ class YoutubeDL(object):
                         try:
                             sub_data = ie._request_webpage(
                                 sub_info['url'], info_dict['id'], note=False).read()
-                            with io.open(encodeFilename(sub_filename), 'wb') as subfile:
+                            with open(encodeFilename(sub_filename), 'wb') as subfile:
                                 subfile.write(sub_data)
                         except (ExtractorError, IOError, OSError, ValueError) as err:
                             self.report_warning('Unable to download subtitle for "%s": %s' %
                                                 (sub_lang, error_to_compat_str(err)))
                             continue
 
-        if self.params.get('writeinfojson', False):
-            infofn = replace_extension(filename, 'info.json', info_dict.get('ext'))
-            if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(infofn)):
-                self.to_screen('[info] Video description metadata is already present')
-            else:
-                self.to_screen('[info] Writing video description metadata as JSON to: ' + infofn)
-                try:
-                    write_json_file(self.filter_requested_info(info_dict), infofn)
-                except (OSError, IOError):
-                    self.report_error('Cannot write metadata to JSON file ' + infofn)
-                    return
+        self._write_info_json(
+            'video description', info_dict,
+            replace_extension(filename, 'info.json', info_dict.get('ext')))
 
         self._write_thumbnails(info_dict, filename)
 
@@ -1929,7 +2064,11 @@ class YoutubeDL(object):
                         fd.add_progress_hook(ph)
                     if self.params.get('verbose'):
                         self.to_screen('[debug] Invoking downloader on %r' % info.get('url'))
-                    return fd.download(name, info)
+
+                    new_info = dict((k, v) for k, v in info.items() if not k.startswith('__p'))
+                    new_info['http_headers'] = self._calc_headers(new_info)
+
+                    return fd.download(name, new_info)
 
                 if info_dict.get('requested_formats') is not None:
                     downloaded = []
@@ -2099,12 +2238,8 @@ class YoutubeDL(object):
         return self._download_retcode
 
     def download_with_info_file(self, info_filename):
-        with contextlib.closing(fileinput.FileInput(
-                [info_filename], mode='r',
-                openhook=fileinput.hook_encoded('utf-8'))) as f:
-            # FileInput doesn't have a read method, we can't call json.load
-            # TODO: let's use io.open(), then
-            info = self.filter_requested_info(json.loads('\n'.join(f)))
+        with open(info_filename, encoding='utf-8') as f:
+            info = self.filter_requested_info(json.load(f))
         try:
             self.process_ie_result(info, download=True)
         except DownloadError:
@@ -2332,27 +2467,6 @@ class YoutubeDL(object):
         """ Start an HTTP download """
         if isinstance(req, compat_basestring):
             req = sanitized_Request(req)
-        # an embedded /../ sequence is not automatically handled by urllib2
-        # see https://github.com/yt-dlp/yt-dlp/issues/3355
-        url = req.get_full_url()
-        parts = url.partition('/../')
-        if parts[1]:
-            url = compat_urllib_parse.urljoin(parts[0] + parts[1][:1], parts[1][1:] + parts[2])
-        if url:
-            # worse, URL path may have initial /../ against RFCs: work-around
-            # by stripping such prefixes, like eg Firefox
-            parts = compat_urllib_parse.urlsplit(url)
-            path = parts.path
-            while path.startswith('/../'):
-                path = path[3:]
-            url = parts._replace(path=path).geturl()
-            # get a new Request with the munged URL
-            if url != req.get_full_url():
-                req_type = {'HEAD': HEADRequest, 'PUT': PUTRequest}.get(
-                    req.get_method(), compat_urllib_request.Request)
-                req = req_type(
-                    url, data=req.data, headers=dict(req.header_items()),
-                    origin_req_host=req.origin_req_host, unverifiable=req.unverifiable)
         return self._opener.open(req, timeout=self._socket_timeout)
 
     def print_debug_header(self):
@@ -2374,10 +2488,12 @@ class YoutubeDL(object):
                 self.get_encoding()))
         write_string(encoding_str, encoding=None)
 
-        self._write_string('[debug] youtube-dl version ' + __version__ + (' (single file build)\n' if ytdl_is_updateable() else '\n'))
+        writeln_debug = lambda *s: self._write_string('[debug] %s\n' % (''.join(s), ))
+        writeln_debug('youtube-dl version ', __version__)
         if _LAZY_LOADER:
-            self._write_string('[debug] Lazy loading extractors enabled\n')
-        writeln_debug = lambda *s: self._write_string('[debug] %s\n' % (''.join(s), ))  # moved down for easier merge
+            writeln_debug('Lazy loading extractors enabled')
+        if ytdl_is_updateable():
+            writeln_debug('Single file build')
         try:
             sp = subprocess.Popen(
                 ['git', 'rev-parse', '--short', 'HEAD'],
@@ -2453,7 +2569,7 @@ class YoutubeDL(object):
         opts_proxy = self.params.get('proxy')
 
         if opts_cookiefile is None:
-            self.cookiejar = compat_cookiejar.CookieJar()
+            self.cookiejar = YoutubeDLCookieJar()
         else:
             opts_cookiefile = expand_path(opts_cookiefile)
             self.cookiejar = YoutubeDLCookieJar(opts_cookiefile)
@@ -2514,6 +2630,28 @@ class YoutubeDL(object):
             encoding = preferredencoding()
         return encoding
 
+    def _write_info_json(self, label, info_dict, infofn, overwrite=None):
+        if not self.params.get('writeinfojson', False):
+            return False
+
+        def msg(fmt, lbl):
+            return fmt % (lbl + ' metadata',)
+
+        if overwrite is None:
+            overwrite = not self.params.get('nooverwrites', False)
+
+        if not overwrite and os.path.exists(encodeFilename(infofn)):
+            self.to_screen(msg('[info] %s is already present', label.title()))
+            return 'exists'
+        else:
+            self.to_screen(msg('[info] Writing %s as JSON to: ', label) + infofn)
+            try:
+                write_json_file(self.filter_requested_info(info_dict), infofn)
+                return True
+            except (OSError, IOError):
+                self.report_error(msg('Cannot write %s to JSON file ', label) + infofn)
+                return
+
     def _write_thumbnails(self, info_dict, filename):
         if self.params.get('writethumbnail', False):
             thumbnails = info_dict.get('thumbnails')

youtube_dl/__init__.py

@@ -5,7 +5,6 @@ from __future__ import unicode_literals
 
 __license__ = 'Public Domain'
 
-import codecs
 import io
 import os
 import random
@@ -17,6 +16,7 @@ from .options import (
 )
 from .compat import (
     compat_getpass,
+    compat_register_utf8,
     compat_shlex_split,
     workaround_optparse_bug9161,
 )
@@ -46,10 +46,8 @@ from .YoutubeDL import YoutubeDL
 
 
 def _real_main(argv=None):
-    # Compatibility fixes for Windows
-    if sys.platform == 'win32':
-        # https://github.com/ytdl-org/youtube-dl/issues/820
-        codecs.register(lambda name: codecs.lookup('utf-8') if name == 'cp65001' else None)
+    # Compatibility fix for Windows
+    compat_register_utf8()
 
     workaround_optparse_bug9161()
 

youtube_dl/cache.py

@@ -1,14 +1,16 @@
 from __future__ import unicode_literals
 
 import errno
-import io
 import json
 import os
 import re
 import shutil
 import traceback
 
-from .compat import compat_getenv
+from .compat import (
+    compat_getenv,
+    compat_open as open,
+)
 from .utils import (
     error_to_compat_str,
     expand_path,
@@ -83,7 +85,7 @@ class Cache(object):
         cache_fn = self._get_cache_fn(section, key, dtype)
         try:
             try:
-                with io.open(cache_fn, 'r', encoding='utf-8') as cachef:
+                with open(cache_fn, 'r', encoding='utf-8') as cachef:
                     return self._validate(json.load(cachef), min_ver)
             except ValueError:
                 try:

youtube_dl/compat.py

@@ -1,10 +1,12 @@
 # coding: utf-8
 from __future__ import unicode_literals
+from __future__ import division
 
 import base64
 import binascii
 import collections
 import ctypes
+import datetime
 import email
 import getpass
 import io
@@ -19,6 +21,7 @@ import socket
 import struct
 import subprocess
 import sys
+import types
 import xml.etree.ElementTree
 
 # naming convention
@@ -31,13 +34,17 @@ try:
     compat_str, compat_basestring, compat_chr = (
         unicode, basestring, unichr
     )
-    from .casefold import casefold as compat_casefold
-
 except NameError:
     compat_str, compat_basestring, compat_chr = (
-        str, str, chr
+        str, (str, bytes), chr
     )
+
+# casefold
+try:
+    compat_str.casefold
     compat_casefold = lambda s: s.casefold()
+except AttributeError:
+    from .casefold import casefold as compat_casefold
 
 try:
     import collections.abc as compat_collections_abc
@@ -49,6 +56,29 @@ try:
 except ImportError:  # Python 2
     import urllib2 as compat_urllib_request
 
+# Also fix up lack of method arg in old Pythons
+try:
+    type(compat_urllib_request.Request('http://127.0.0.1', method='GET'))
+except TypeError:
+    def _add_init_method_arg(cls):
+
+        init = cls.__init__
+
+        def wrapped_init(self, *args, **kwargs):
+            method = kwargs.pop('method', 'GET')
+            init(self, *args, **kwargs)
+            if any(callable(x.__dict__.get('get_method')) for x in (self.__class__, self) if x != cls):
+                # allow instance or its subclass to override get_method()
+                return
+            if self.has_data() and method == 'GET':
+                method = 'POST'
+            self.get_method = types.MethodType(lambda _: method, self)
+
+        cls.__init__ = wrapped_init
+
+    _add_init_method_arg(compat_urllib_request.Request)
+    del _add_init_method_arg
+
 try:
     import urllib.error as compat_urllib_error
 except ImportError:  # Python 2
@@ -73,6 +103,12 @@ try:
 except ImportError:  # Python 2
     import urllib as compat_urllib_response
 
+try:
+    compat_urllib_response.addinfourl.status
+except AttributeError:
+    # .getcode() is deprecated in Py 3.
+    compat_urllib_response.addinfourl.status = property(lambda self: self.getcode())
+
 try:
     import http.cookiejar as compat_cookiejar
 except ImportError:  # Python 2
@@ -97,12 +133,24 @@ except ImportError:  # Python 2
     import Cookie as compat_cookies
 compat_http_cookies = compat_cookies
 
-if sys.version_info[0] == 2:
+if sys.version_info[0] == 2 or sys.version_info < (3, 3):
     class compat_cookies_SimpleCookie(compat_cookies.SimpleCookie):
         def load(self, rawdata):
-            if isinstance(rawdata, compat_str):
-                rawdata = str(rawdata)
-            return super(compat_cookies_SimpleCookie, self).load(rawdata)
+            must_have_value = 0
+            if not isinstance(rawdata, dict):
+                if sys.version_info[:2] != (2, 7) or sys.platform.startswith('java'):
+                    # attribute must have value for parsing
+                    rawdata, must_have_value = re.subn(
+                        r'(?i)(;\s*)(secure|httponly)(\s*(?:;|$))', r'\1\2=\2\3', rawdata)
+                if sys.version_info[0] == 2:
+                    if isinstance(rawdata, compat_str):
+                        rawdata = str(rawdata)
+            super(compat_cookies_SimpleCookie, self).load(rawdata)
+            if must_have_value > 0:
+                for morsel in self.values():
+                    for attr in ('secure', 'httponly'):
+                        if morsel.get(attr):
+                            morsel[attr] = True
 else:
     compat_cookies_SimpleCookie = compat_cookies.SimpleCookie
 compat_http_cookies_SimpleCookie = compat_cookies_SimpleCookie
@@ -2354,6 +2402,11 @@ try:
     import http.client as compat_http_client
 except ImportError:  # Python 2
     import httplib as compat_http_client
+try:
+    compat_http_client.HTTPResponse.getcode
+except AttributeError:
+    # Py < 3.1
+    compat_http_client.HTTPResponse.getcode = lambda self: self.status
 
 try:
     from urllib.error import HTTPError as compat_HTTPError
@@ -2368,29 +2421,26 @@ except ImportError:  # Python 2
 compat_urllib_request_urlretrieve = compat_urlretrieve
 
 try:
+    from HTMLParser import (
+        HTMLParser as compat_HTMLParser,
+        HTMLParseError as compat_HTMLParseError)
+except ImportError:  # Python 3
     from html.parser import HTMLParser as compat_HTMLParser
-except ImportError:  # Python 2
-    from HTMLParser import HTMLParser as compat_HTMLParser
-compat_html_parser_HTMLParser = compat_HTMLParser
-
-try:  # Python 2
-    from HTMLParser import HTMLParseError as compat_HTMLParseError
-except ImportError:  # Python <3.4
     try:
         from html.parser import HTMLParseError as compat_HTMLParseError
     except ImportError:  # Python >3.4
-
-        # HTMLParseError has been deprecated in Python 3.3 and removed in
+        # HTMLParseError was deprecated in Python 3.3 and removed in
         # Python 3.5. Introducing dummy exception for Python >3.5 for compatible
         # and uniform cross-version exception handling
         class compat_HTMLParseError(Exception):
             pass
+compat_html_parser_HTMLParser = compat_HTMLParser
 compat_html_parser_HTMLParseError = compat_HTMLParseError
 
 try:
-    from subprocess import DEVNULL
-    compat_subprocess_get_DEVNULL = lambda: DEVNULL
-except ImportError:
+    _DEVNULL = subprocess.DEVNULL
+    compat_subprocess_get_DEVNULL = lambda: _DEVNULL
+except AttributeError:
     compat_subprocess_get_DEVNULL = lambda: open(os.path.devnull, 'w')
 
 try:
@@ -2669,8 +2719,222 @@ if sys.version_info < (2, 7):
         if isinstance(xpath, compat_str):
             xpath = xpath.encode('ascii')
         return xpath
+
+    # further code below based on CPython 2.7 source
+    import functools
+
+    _xpath_tokenizer_re = re.compile(r'''(?x)
+        (                                   # (1)
+            '[^']*'|"[^"]*"|                # quoted strings, or
+            ::|//?|\.\.|\(\)|[/.*:[\]()@=]  # navigation specials
+        )|                                  # or (2)
+        ((?:\{[^}]+\})?[^/[\]()@=\s]+)|     # token: optional {ns}, no specials
+        \s+                                 # or white space
+    ''')
+
+    def _xpath_tokenizer(pattern, namespaces=None):
+        for token in _xpath_tokenizer_re.findall(pattern):
+            tag = token[1]
+            if tag and tag[0] != "{" and ":" in tag:
+                try:
+                    if not namespaces:
+                        raise KeyError
+                    prefix, uri = tag.split(":", 1)
+                    yield token[0], "{%s}%s" % (namespaces[prefix], uri)
+                except KeyError:
+                    raise SyntaxError("prefix %r not found in prefix map" % prefix)
+            else:
+                yield token
+
+    def _get_parent_map(context):
+        parent_map = context.parent_map
+        if parent_map is None:
+            context.parent_map = parent_map = {}
+            for p in context.root.getiterator():
+                for e in p:
+                    parent_map[e] = p
+        return parent_map
+
+    def _select(context, result, filter_fn=lambda *_: True):
+        for elem in result:
+            for e in elem:
+                if filter_fn(e, elem):
+                    yield e
+
+    def _prepare_child(next_, token):
+        tag = token[1]
+        return functools.partial(_select, filter_fn=lambda e, _: e.tag == tag)
+
+    def _prepare_star(next_, token):
+        return _select
+
+    def _prepare_self(next_, token):
+        return lambda _, result: (e for e in result)
+
+    def _prepare_descendant(next_, token):
+        token = next(next_)
+        if token[0] == "*":
+            tag = "*"
+        elif not token[0]:
+            tag = token[1]
+        else:
+            raise SyntaxError("invalid descendant")
+
+        def select(context, result):
+            for elem in result:
+                for e in elem.getiterator(tag):
+                    if e is not elem:
+                        yield e
+        return select
+
+    def _prepare_parent(next_, token):
+        def select(context, result):
+            # FIXME: raise error if .. is applied at toplevel?
+            parent_map = _get_parent_map(context)
+            result_map = {}
+            for elem in result:
+                if elem in parent_map:
+                    parent = parent_map[elem]
+                    if parent not in result_map:
+                        result_map[parent] = None
+                        yield parent
+        return select
+
+    def _prepare_predicate(next_, token):
+        signature = []
+        predicate = []
+        for token in next_:
+            if token[0] == "]":
+                break
+            if token[0] and token[0][:1] in "'\"":
+                token = "'", token[0][1:-1]
+            signature.append(token[0] or "-")
+            predicate.append(token[1])
+
+        def select(context, result, filter_fn=lambda _: True):
+            for elem in result:
+                if filter_fn(elem):
+                    yield elem
+
+        signature = "".join(signature)
+        # use signature to determine predicate type
+        if signature == "@-":
+            # [@attribute] predicate
+            key = predicate[1]
+            return functools.partial(
+                select, filter_fn=lambda el: el.get(key) is not None)
+        if signature == "@-='":
+            # [@attribute='value']
+            key = predicate[1]
+            value = predicate[-1]
+            return functools.partial(
+                select, filter_fn=lambda el: el.get(key) == value)
+        if signature == "-" and not re.match(r"\d+$", predicate[0]):
+            # [tag]
+            tag = predicate[0]
+            return functools.partial(
+                select, filter_fn=lambda el: el.find(tag) is not None)
+        if signature == "-='" and not re.match(r"\d+$", predicate[0]):
+            # [tag='value']
+            tag = predicate[0]
+            value = predicate[-1]
+
+            def itertext(el):
+                for e in el.getiterator():
+                    e = e.text
+                    if e:
+                        yield e
+
+            def select(context, result):
+                for elem in result:
+                    for e in elem.findall(tag):
+                        if "".join(itertext(e)) == value:
+                            yield elem
+                            break
+            return select
+        if signature == "-" or signature == "-()" or signature == "-()-":
+            # [index] or [last()] or [last()-index]
+            if signature == "-":
+                index = int(predicate[0]) - 1
+            else:
+                if predicate[0] != "last":
+                    raise SyntaxError("unsupported function")
+                if signature == "-()-":
+                    try:
+                        index = int(predicate[2]) - 1
+                    except ValueError:
+                        raise SyntaxError("unsupported expression")
+                else:
+                    index = -1
+
+            def select(context, result):
+                parent_map = _get_parent_map(context)
+                for elem in result:
+                    try:
+                        parent = parent_map[elem]
+                        # FIXME: what if the selector is "*" ?
+                        elems = list(parent.findall(elem.tag))
+                        if elems[index] is elem:
+                            yield elem
+                    except (IndexError, KeyError):
+                        pass
+            return select
+        raise SyntaxError("invalid predicate")
+
+    ops = {
+        "": _prepare_child,
+        "*": _prepare_star,
+        ".": _prepare_self,
+        "..": _prepare_parent,
+        "//": _prepare_descendant,
+        "[": _prepare_predicate,
+    }
+
+    _cache = {}
+
+    class _SelectorContext:
+        parent_map = None
+
+        def __init__(self, root):
+            self.root = root
+
+    ##
+    # Generate all matching objects.
+
+    def compat_etree_iterfind(elem, path, namespaces=None):
+        # compile selector pattern
+        if path[-1:] == "/":
+            path = path + "*"  # implicit all (FIXME: keep this?)
+        try:
+            selector = _cache[path]
+        except KeyError:
+            if len(_cache) > 100:
+                _cache.clear()
+            if path[:1] == "/":
+                raise SyntaxError("cannot use absolute path on element")
+            tokens = _xpath_tokenizer(path, namespaces)
+            selector = []
+            for token in tokens:
+                if token[0] == "/":
+                    continue
+                try:
+                    selector.append(ops[token[0]](tokens, token))
+                except StopIteration:
+                    raise SyntaxError("invalid path")
+            _cache[path] = selector
+        # execute selector pattern
+        result = [elem]
+        context = _SelectorContext(elem)
+        for select in selector:
+            result = select(context, result)
+        return result
+
+    # end of code based on CPython 2.7 source
+
+
 else:
     compat_xpath = lambda xpath: xpath
+    compat_etree_iterfind = lambda element, match: element.iterfind(match)
 
 
 compat_os_name = os._name if os.name == 'java' else os.name
@@ -2706,7 +2970,7 @@ except (AssertionError, UnicodeEncodeError):
 
 
 def compat_ord(c):
-    if type(c) is int:
+    if isinstance(c, int):
         return c
     else:
         return ord(c)
@@ -2890,6 +3154,51 @@ else:
     compat_socket_create_connection = socket.create_connection
 
 
+try:
+    from contextlib import suppress as compat_contextlib_suppress
+except ImportError:
+    class compat_contextlib_suppress(object):
+        _exceptions = None
+
+        def __init__(self, *exceptions):
+            super(compat_contextlib_suppress, self).__init__()
+            # TODO: [Base]ExceptionGroup (3.12+)
+            self._exceptions = exceptions
+
+        def __enter__(self):
+            return self
+
+        def __exit__(self, exc_type, exc_val, exc_tb):
+            return exc_type is not None and issubclass(exc_type, self._exceptions or tuple())
+
+
+# subprocess.Popen context manager
+# avoids leaking handles if .communicate() is not called
+try:
+    _Popen = subprocess.Popen
+    # check for required context manager attributes
+    _Popen.__enter__ and _Popen.__exit__
+    compat_subprocess_Popen = _Popen
+except AttributeError:
+    # not a context manager - make one
+    from contextlib import contextmanager
+
+    @contextmanager
+    def compat_subprocess_Popen(*args, **kwargs):
+        popen = None
+        try:
+            popen = _Popen(*args, **kwargs)
+            yield popen
+        finally:
+            if popen:
+                for f in (popen.stdin, popen.stdout, popen.stderr):
+                    if f:
+                        # repeated .close() is OK, but just in case
+                        with compat_contextlib_suppress(EnvironmentError):
+                            f.close()
+                popen.wait()
+
+
 # Fix https://github.com/ytdl-org/youtube-dl/issues/4223
 # See http://bugs.python.org/issue9161 for what is broken
 def workaround_optparse_bug9161():
@@ -3137,6 +3446,36 @@ else:
     compat_open = open
 
 
+# compat_register_utf8
+def compat_register_utf8():
+    if sys.platform == 'win32':
+        # https://github.com/ytdl-org/youtube-dl/issues/820
+        from codecs import register, lookup
+        register(
+            lambda name: lookup('utf-8') if name == 'cp65001' else None)
+
+
+# compat_datetime_timedelta_total_seconds
+try:
+    compat_datetime_timedelta_total_seconds = datetime.timedelta.total_seconds
+except AttributeError:
+    # Py 2.6
+    def compat_datetime_timedelta_total_seconds(td):
+        return (td.microseconds + (td.seconds + td.days * 24 * 3600) * 10**6) / 10**6
+
+# optional decompression packages
+# PyPi brotli package implements 'br' Content-Encoding
+try:
+    import brotli as compat_brotli
+except ImportError:
+    compat_brotli = None
+# PyPi ncompress package implements 'compress' Content-Encoding
+try:
+    import ncompress as compat_ncompress
+except ImportError:
+    compat_ncompress = None
+
+
 legacy = [
     'compat_HTMLParseError',
     'compat_HTMLParser',
@@ -3170,16 +3509,20 @@ __all__ = [
     'compat_Struct',
     'compat_base64_b64decode',
     'compat_basestring',
+    'compat_brotli',
     'compat_casefold',
     'compat_chr',
     'compat_collections_abc',
     'compat_collections_chain_map',
+    'compat_datetime_timedelta_total_seconds',
     'compat_http_cookiejar',
     'compat_http_cookiejar_Cookie',
     'compat_http_cookies',
     'compat_http_cookies_SimpleCookie',
+    'compat_contextlib_suppress',
     'compat_ctypes_WINFUNCTYPE',
     'compat_etree_fromstring',
+    'compat_etree_iterfind',
     'compat_filter',
     'compat_get_terminal_size',
     'compat_getenv',
@@ -3194,6 +3537,7 @@ __all__ = [
     'compat_itertools_zip_longest',
     'compat_kwargs',
     'compat_map',
+    'compat_ncompress',
     'compat_numeric_types',
     'compat_open',
     'compat_ord',
@@ -3203,6 +3547,7 @@ __all__ = [
     'compat_print',
     'compat_re_Match',
     'compat_re_Pattern',
+    'compat_register_utf8',
     'compat_setenv',
     'compat_shlex_quote',
     'compat_shlex_split',
@@ -3211,6 +3556,7 @@ __all__ = [
     'compat_struct_pack',
     'compat_struct_unpack',
     'compat_subprocess_get_DEVNULL',
+    'compat_subprocess_Popen',
     'compat_tokenize_tokenize',
     'compat_urllib_error',
     'compat_urllib_parse',

youtube_dl/downloader/common.py

@@ -96,7 +96,7 @@ class FileDownloader(object):
                 return None
             return int(float(remaining) / rate)
         start, now = (start_or_rate, now_or_remaining)
-        total, current = args
+        total, current = args[:2]
         if total is None:
             return None
         if now is None:
@@ -339,6 +339,10 @@ class FileDownloader(object):
     def download(self, filename, info_dict):
         """Download to a filename using the info from info_dict
         Return True on success and False otherwise
+
+        This method filters the `Cookie` header from the info_dict to prevent leaks.
+        Downloaders have their own way of handling cookies.
+        See: https://github.com/yt-dlp/yt-dlp/security/advisories/GHSA-v8mc-9377-rwjj
         """
 
         nooverwrites_and_exists = (

youtube_dl/downloader/dash.py

@@ -35,6 +35,7 @@ class DashSegmentsFD(FragmentFD):
         for frag_index, fragment in enumerate(fragments, 1):
             if frag_index <= ctx['fragment_index']:
                 continue
+            success = False
             # In DASH, the first segment contains necessary headers to
             # generate a valid MP4 file, so always abort for the first segment
             fatal = frag_index == 1 or not skip_unavailable_fragments
@@ -42,10 +43,14 @@ class DashSegmentsFD(FragmentFD):
             if not fragment_url:
                 assert fragment_base_url
                 fragment_url = urljoin(fragment_base_url, fragment['path'])
-            success = False
+            headers = info_dict.get('http_headers')
+            fragment_range = fragment.get('range')
+            if fragment_range:
+                headers = headers.copy() if headers else {}
+                headers['Range'] = 'bytes=%s' % (fragment_range,)
             for count in itertools.count():
                 try:
-                    success, frag_content = self._download_fragment(ctx, fragment_url, info_dict)
+                    success, frag_content = self._download_fragment(ctx, fragment_url, info_dict, headers)
                     if not success:
                         return False
                     self._append_fragment(ctx, frag_content)

youtube_dl/downloader/external.py

@@ -1,17 +1,24 @@
 from __future__ import unicode_literals
 
-import os.path
+import os
 import re
 import subprocess
 import sys
+import tempfile
 import time
 
 from .common import FileDownloader
 from ..compat import (
     compat_setenv,
     compat_str,
+    compat_subprocess_Popen,
 )
-from ..postprocessor.ffmpeg import FFmpegPostProcessor, EXT_TO_OUT_FORMATS
+
+try:
+    from ..postprocessor.ffmpeg import FFmpegPostProcessor, EXT_TO_OUT_FORMATS
+except ImportError:
+    FFmpegPostProcessor = None
+
 from ..utils import (
     cli_option,
     cli_valueless_option,
@@ -23,6 +30,8 @@ from ..utils import (
     check_executable,
     is_outdated_version,
     process_communicate_or_kill,
+    T,
+    traverse_obj,
 )
 
 
@@ -30,6 +39,7 @@ class ExternalFD(FileDownloader):
     def real_download(self, filename, info_dict):
         self.report_destination(filename)
         tmpfilename = self.temp_name(filename)
+        self._cookies_tempfile = None
 
         try:
             started = time.time()
@@ -42,6 +52,13 @@ class ExternalFD(FileDownloader):
             # should take place
             retval = 0
             self.to_screen('[%s] Interrupted by user' % self.get_basename())
+        finally:
+            if self._cookies_tempfile and os.path.isfile(self._cookies_tempfile):
+                try:
+                    os.remove(self._cookies_tempfile)
+                except OSError:
+                    self.report_warning(
+                        'Unable to delete temporary cookies file "{0}"'.format(self._cookies_tempfile))
 
         if retval == 0:
             status = {
@@ -97,6 +114,16 @@ class ExternalFD(FileDownloader):
     def _configuration_args(self, default=[]):
         return cli_configuration_args(self.params, 'external_downloader_args', default)
 
+    def _write_cookies(self):
+        if not self.ydl.cookiejar.filename:
+            tmp_cookies = tempfile.NamedTemporaryFile(suffix='.cookies', delete=False)
+            tmp_cookies.close()
+            self._cookies_tempfile = tmp_cookies.name
+            self.to_screen('[download] Writing temporary cookies file to "{0}"'.format(self._cookies_tempfile))
+        # real_download resets _cookies_tempfile; if it's None, save() will write to cookiejar.filename
+        self.ydl.cookiejar.save(self._cookies_tempfile, ignore_discard=True, ignore_expires=True)
+        return self.ydl.cookiejar.filename or self._cookies_tempfile
+
     def _call_downloader(self, tmpfilename, info_dict):
         """ Either overwrite this or implement _make_cmd """
         cmd = [encodeArgument(a) for a in self._make_cmd(tmpfilename, info_dict)]
@@ -110,13 +137,21 @@ class ExternalFD(FileDownloader):
             self.to_stderr(stderr.decode('utf-8', 'replace'))
         return p.returncode
 
+    @staticmethod
+    def _header_items(info_dict):
+        return traverse_obj(
+            info_dict, ('http_headers', T(dict.items), Ellipsis))
+
 
 class CurlFD(ExternalFD):
     AVAILABLE_OPT = '-V'
 
     def _make_cmd(self, tmpfilename, info_dict):
-        cmd = [self.exe, '--location', '-o', tmpfilename]
-        for key, val in info_dict['http_headers'].items():
+        cmd = [self.exe, '--location', '-o', tmpfilename, '--compressed']
+        cookie_header = self.ydl.cookiejar.get_cookie_header(info_dict['url'])
+        if cookie_header:
+            cmd += ['--cookie', cookie_header]
+        for key, val in self._header_items(info_dict):
             cmd += ['--header', '%s: %s' % (key, val)]
         cmd += self._bool_option('--continue-at', 'continuedl', '-', '0')
         cmd += self._valueless_option('--silent', 'noprogress')
@@ -151,8 +186,11 @@ class AxelFD(ExternalFD):
 
     def _make_cmd(self, tmpfilename, info_dict):
         cmd = [self.exe, '-o', tmpfilename]
-        for key, val in info_dict['http_headers'].items():
+        for key, val in self._header_items(info_dict):
             cmd += ['-H', '%s: %s' % (key, val)]
+        cookie_header = self.ydl.cookiejar.get_cookie_header(info_dict['url'])
+        if cookie_header:
+            cmd += ['-H', 'Cookie: {0}'.format(cookie_header), '--max-redirect=0']
         cmd += self._configuration_args()
         cmd += ['--', info_dict['url']]
         return cmd
@@ -162,8 +200,10 @@ class WgetFD(ExternalFD):
     AVAILABLE_OPT = '--version'
 
     def _make_cmd(self, tmpfilename, info_dict):
-        cmd = [self.exe, '-O', tmpfilename, '-nv', '--no-cookies']
-        for key, val in info_dict['http_headers'].items():
+        cmd = [self.exe, '-O', tmpfilename, '-nv', '--compression=auto']
+        if self.ydl.cookiejar.get_cookie_header(info_dict['url']):
+            cmd += ['--load-cookies', self._write_cookies()]
+        for key, val in self._header_items(info_dict):
             cmd += ['--header', '%s: %s' % (key, val)]
         cmd += self._option('--limit-rate', 'ratelimit')
         retry = self._option('--tries', 'retries')
@@ -172,7 +212,10 @@ class WgetFD(ExternalFD):
                 retry[1] = '0'
             cmd += retry
         cmd += self._option('--bind-address', 'source_address')
-        cmd += self._option('--proxy', 'proxy')
+        proxy = self.params.get('proxy')
+        if proxy:
+            for var in ('http_proxy', 'https_proxy'):
+                cmd += ['--execute', '%s=%s' % (var, proxy)]
         cmd += self._valueless_option('--no-check-certificate', 'nocheckcertificate')
         cmd += self._configuration_args()
         cmd += ['--', info_dict['url']]
@@ -182,21 +225,58 @@ class WgetFD(ExternalFD):
 class Aria2cFD(ExternalFD):
     AVAILABLE_OPT = '-v'
 
+    @staticmethod
+    def _aria2c_filename(fn):
+        return fn if os.path.isabs(fn) else os.path.join('.', fn)
+
     def _make_cmd(self, tmpfilename, info_dict):
-        cmd = [self.exe, '-c']
-        cmd += self._configuration_args([
-            '--min-split-size', '1M', '--max-connection-per-server', '4'])
-        dn = os.path.dirname(tmpfilename)
-        if dn:
-            cmd += ['--dir', dn]
-        cmd += ['--out', os.path.basename(tmpfilename)]
-        for key, val in info_dict['http_headers'].items():
+        cmd = [self.exe, '-c',
+               '--console-log-level=warn', '--summary-interval=0', '--download-result=hide',
+               '--http-accept-gzip=true', '--file-allocation=none', '-x16', '-j16', '-s16']
+        if 'fragments' in info_dict:
+            cmd += ['--allow-overwrite=true', '--allow-piece-length-change=true']
+        else:
+            cmd += ['--min-split-size', '1M']
+
+        if self.ydl.cookiejar.get_cookie_header(info_dict['url']):
+            cmd += ['--load-cookies={0}'.format(self._write_cookies())]
+        for key, val in self._header_items(info_dict):
             cmd += ['--header', '%s: %s' % (key, val)]
+        cmd += self._configuration_args(['--max-connection-per-server', '4'])
+        cmd += ['--out', os.path.basename(tmpfilename)]
+        cmd += self._option('--max-overall-download-limit', 'ratelimit')
         cmd += self._option('--interface', 'source_address')
         cmd += self._option('--all-proxy', 'proxy')
         cmd += self._bool_option('--check-certificate', 'nocheckcertificate', 'false', 'true', '=')
         cmd += self._bool_option('--remote-time', 'updatetime', 'true', 'false', '=')
-        cmd += ['--', info_dict['url']]
+        cmd += self._bool_option('--show-console-readout', 'noprogress', 'false', 'true', '=')
+        cmd += self._configuration_args()
+
+        # aria2c strips out spaces from the beginning/end of filenames and paths.
+        # We work around this issue by adding a "./" to the beginning of the
+        # filename and relative path, and adding a "/" at the end of the path.
+        # See: https://github.com/yt-dlp/yt-dlp/issues/276
+        # https://github.com/ytdl-org/youtube-dl/issues/20312
+        # https://github.com/aria2/aria2/issues/1373
+        dn = os.path.dirname(tmpfilename)
+        if dn:
+            cmd += ['--dir', self._aria2c_filename(dn) + os.path.sep]
+        if 'fragments' not in info_dict:
+            cmd += ['--out', self._aria2c_filename(os.path.basename(tmpfilename))]
+        cmd += ['--auto-file-renaming=false']
+        if 'fragments' in info_dict:
+            cmd += ['--file-allocation=none', '--uri-selector=inorder']
+            url_list_file = '%s.frag.urls' % (tmpfilename, )
+            url_list = []
+            for frag_index, fragment in enumerate(info_dict['fragments']):
+                fragment_filename = '%s-Frag%d' % (os.path.basename(tmpfilename), frag_index)
+                url_list.append('%s\n\tout=%s' % (fragment['url'], self._aria2c_filename(fragment_filename)))
+            stream, _ = self.sanitize_open(url_list_file, 'wb')
+            stream.write('\n'.join(url_list).encode())
+            stream.close()
+            cmd += ['-i', self._aria2c_filename(url_list_file)]
+        else:
+            cmd += ['--', info_dict['url']]
         return cmd
 
 
@@ -235,8 +315,10 @@ class Aria2pFD(ExternalFD):
         }
         options['dir'] = os.path.dirname(tmpfilename) or os.path.abspath('.')
         options['out'] = os.path.basename(tmpfilename)
+        if self.ydl.cookiejar.get_cookie_header(info_dict['url']):
+            options['load-cookies'] = self._write_cookies()
         options['header'] = []
-        for key, val in info_dict['http_headers'].items():
+        for key, val in self._header_items(info_dict):
             options['header'].append('{0}: {1}'.format(key, val))
         download = aria2.add_uris([info_dict['url']], options)
         status = {
@@ -265,8 +347,16 @@ class HttpieFD(ExternalFD):
 
     def _make_cmd(self, tmpfilename, info_dict):
         cmd = ['http', '--download', '--output', tmpfilename, info_dict['url']]
-        for key, val in info_dict['http_headers'].items():
+        for key, val in self._header_items(info_dict):
             cmd += ['%s:%s' % (key, val)]
+
+        # httpie 3.1.0+ removes the Cookie header on redirect, so this should be safe for now. [1]
+        # If we ever need cookie handling for redirects, we can export the cookiejar into a session. [2]
+        # 1: https://github.com/httpie/httpie/security/advisories/GHSA-9w4w-cpc8-h2fq
+        # 2: https://httpie.io/docs/cli/sessions
+        cookie_header = self.ydl.cookiejar.get_cookie_header(info_dict['url'])
+        if cookie_header:
+            cmd += ['Cookie:%s' % cookie_header]
         return cmd
 
 
@@ -277,13 +367,14 @@ class FFmpegFD(ExternalFD):
 
     @classmethod
     def available(cls):
-        return FFmpegPostProcessor().available
+        # actual availability can only be confirmed for an instance
+        return bool(FFmpegPostProcessor)
 
     def _call_downloader(self, tmpfilename, info_dict):
-        url = info_dict['url']
-        ffpp = FFmpegPostProcessor(downloader=self)
+        # `downloader` means the parent `YoutubeDL`
+        ffpp = FFmpegPostProcessor(downloader=self.ydl)
         if not ffpp.available:
-            self.report_error('m3u8 download detected but ffmpeg or avconv could not be found. Please install one.')
+            self.report_error('ffmpeg required for download but no ffmpeg (nor avconv) executable could be found. Please install one.')
             return False
         ffpp.check_version()
 
@@ -312,7 +403,15 @@ class FFmpegFD(ExternalFD):
         # if end_time:
         #     args += ['-t', compat_str(end_time - start_time)]
 
-        if info_dict['http_headers'] and re.match(r'^https?://', url):
+        url = info_dict['url']
+        cookies = self.ydl.cookiejar.get_cookies_for_url(url)
+        if cookies:
+            args.extend(['-cookies', ''.join(
+                '{0}={1}; path={2}; domain={3};\r\n'.format(
+                    cookie.name, cookie.value, cookie.path, cookie.domain)
+                for cookie in cookies)])
+
+        if info_dict.get('http_headers') and re.match(r'^https?://', url):
             # Trailing \r\n after each HTTP header is important to prevent warning from ffmpeg/avconv:
             # [http @ 00000000003d2fa0] No trailing CRLF found in HTTP header.
             headers = handle_youtubedl_headers(info_dict['http_headers'])
@@ -392,21 +491,25 @@ class FFmpegFD(ExternalFD):
 
         self._debug_cmd(args)
 
-        proc = subprocess.Popen(args, stdin=subprocess.PIPE, env=env)
-        try:
-            retval = proc.wait()
-        except BaseException as e:
-            # subprocess.run would send the SIGKILL signal to ffmpeg and the
-            # mp4 file couldn't be played, but if we ask ffmpeg to quit it
-            # produces a file that is playable (this is mostly useful for live
-            # streams). Note that Windows is not affected and produces playable
-            # files (see https://github.com/ytdl-org/youtube-dl/issues/8300).
-            if isinstance(e, KeyboardInterrupt) and sys.platform != 'win32':
-                process_communicate_or_kill(proc, b'q')
-            else:
-                proc.kill()
-                proc.wait()
-            raise
+        # From [1], a PIPE opened in Popen() should be closed, unless
+        # .communicate() is called. Avoid leaking any PIPEs by using Popen
+        # as a context manager (newer Python 3.x and compat)
+        # Fixes "Resource Warning" in test/test_downloader_external.py
+        # [1] https://devpress.csdn.net/python/62fde12d7e66823466192e48.html
+        with compat_subprocess_Popen(args, stdin=subprocess.PIPE, env=env) as proc:
+            try:
+                retval = proc.wait()
+            except BaseException as e:
+                # subprocess.run would send the SIGKILL signal to ffmpeg and the
+                # mp4 file couldn't be played, but if we ask ffmpeg to quit it
+                # produces a file that is playable (this is mostly useful for live
+                # streams). Note that Windows is not affected and produces playable
+                # files (see https://github.com/ytdl-org/youtube-dl/issues/8300).
+                if isinstance(e, KeyboardInterrupt) and sys.platform != 'win32':
+                    process_communicate_or_kill(proc, b'q')
+                else:
+                    proc.kill()
+                raise
         return retval
 
 

youtube_dl/downloader/http.py

@@ -294,7 +294,7 @@ class HttpFD(FileDownloader):
 
                 # Progress message
                 speed = self.calc_speed(start, now, byte_counter - ctx.resume_len)
-                eta = self.calc_eta(speed, ctx.data_len and (ctx.data_len - ctx.resume_len))
+                eta = self.calc_eta(speed, ctx.data_len and (ctx.data_len - byte_counter))
 
                 self._hook_progress({
                     'status': 'downloading',

youtube_dl/extractor/caffeine.py

@@ -0,0 +1,79 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+from .common import InfoExtractor
+from ..utils import (
+    determine_ext,
+    int_or_none,
+    merge_dicts,
+    parse_iso8601,
+    T,
+    traverse_obj,
+    txt_or_none,
+    urljoin,
+)
+
+
+class CaffeineTVIE(InfoExtractor):
+    _VALID_URL = r'https?://(?:www\.)?caffeine\.tv/[^/]+/video/(?P<id>[0-9a-f-]+)'
+    _TESTS = [{
+        'url': 'https://www.caffeine.tv/TsuSurf/video/cffc0a00-e73f-11ec-8080-80017d29f26e',
+        'info_dict': {
+            'id': 'cffc0a00-e73f-11ec-8080-80017d29f26e',
+            'ext': 'mp4',
+            'title': 'GOOOOD MORNINNNNN #highlights',
+            'timestamp': 1654702180,
+            'upload_date': '20220608',
+            'uploader': 'TsuSurf',
+            'duration': 3145,
+            'age_limit': 17,
+        },
+        'params': {
+            'format': 'bestvideo',
+        },
+    }]
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+        json_data = self._download_json(
+            'https://api.caffeine.tv/social/public/activity/' + video_id,
+            video_id)
+        broadcast_info = traverse_obj(json_data, ('broadcast_info', T(dict))) or {}
+        title = broadcast_info['broadcast_title']
+        video_url = broadcast_info['video_url']
+
+        ext = determine_ext(video_url)
+        if ext == 'm3u8':
+            formats = self._extract_m3u8_formats(
+                video_url, video_id, 'mp4', entry_protocol='m3u8',
+                fatal=False)
+        else:
+            formats = [{'url': video_url}]
+        self._sort_formats(formats)
+
+        return merge_dicts({
+            'id': video_id,
+            'title': title,
+            'formats': formats,
+        }, traverse_obj(json_data, {
+            'uploader': ((None, 'user'), 'username'),
+        }, get_all=False), traverse_obj(json_data, {
+            'like_count': ('like_count', T(int_or_none)),
+            'view_count': ('view_count', T(int_or_none)),
+            'comment_count': ('comment_count', T(int_or_none)),
+            'tags': ('tags', Ellipsis, T(txt_or_none)),
+            'is_live': 'is_live',
+            'uploader': ('user', 'name'),
+        }), traverse_obj(broadcast_info, {
+            'duration': ('content_duration', T(int_or_none)),
+            'timestamp': ('broadcast_start_time', T(parse_iso8601)),
+            'thumbnail': ('preview_image_path', T(lambda u: urljoin(url, u))),
+            'age_limit': ('content_rating', T(lambda r: r and {
+                # assume Apple Store ratings [1]
+                # 1. https://en.wikipedia.org/wiki/Mobile_software_content_rating_system
+                'FOUR_PLUS': 0,
+                'NINE_PLUS': 9,
+                'TWELVE_PLUS': 12,
+                'SEVENTEEN_PLUS': 17,
+            }.get(r, 17))),
+        }))

youtube_dl/extractor/clipchamp.py

@@ -0,0 +1,69 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+from .common import InfoExtractor
+from ..compat import compat_str
+from ..utils import (
+    ExtractorError,
+    merge_dicts,
+    T,
+    traverse_obj,
+    unified_timestamp,
+    url_or_none,
+)
+
+
+class ClipchampIE(InfoExtractor):
+    _VALID_URL = r'https?://(?:www\.)?clipchamp\.com/watch/(?P<id>[\w-]+)'
+    _TESTS = [{
+        'url': 'https://clipchamp.com/watch/gRXZ4ZhdDaU',
+        'info_dict': {
+            'id': 'gRXZ4ZhdDaU',
+            'ext': 'mp4',
+            'title': 'Untitled video',
+            'uploader': 'Alexander Schwartz',
+            'timestamp': 1680805580,
+            'upload_date': '20230406',
+            'thumbnail': r're:^https?://.+\.jpg',
+        },
+        'params': {
+            'skip_download': 'm3u8',
+            'format': 'bestvideo',
+        },
+    }]
+
+    _STREAM_URL_TMPL = 'https://%s.cloudflarestream.com/%s/manifest/video.%s'
+    _STREAM_URL_QUERY = {'parentOrigin': 'https://clipchamp.com'}
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+        webpage = self._download_webpage(url, video_id)
+        data = self._search_nextjs_data(webpage, video_id)['props']['pageProps']['video']
+
+        storage_location = data.get('storage_location')
+        if storage_location != 'cf_stream':
+            raise ExtractorError('Unsupported clip storage location "%s"' % (storage_location,))
+
+        path = data['download_url']
+        iframe = self._download_webpage(
+            'https://iframe.cloudflarestream.com/' + path, video_id, 'Downloading player iframe')
+        subdomain = self._search_regex(
+            r'''\bcustomer-domain-prefix\s*=\s*("|')(?P<sd>[\w-]+)\1''', iframe,
+            'subdomain', group='sd', fatal=False) or 'customer-2ut9yn3y6fta1yxe'
+
+        formats = self._extract_mpd_formats(
+            self._STREAM_URL_TMPL % (subdomain, path, 'mpd'), video_id,
+            query=self._STREAM_URL_QUERY, fatal=False, mpd_id='dash')
+        formats.extend(self._extract_m3u8_formats(
+            self._STREAM_URL_TMPL % (subdomain, path, 'm3u8'), video_id, 'mp4',
+            query=self._STREAM_URL_QUERY, fatal=False, m3u8_id='hls'))
+
+        return merge_dicts({
+            'id': video_id,
+            'formats': formats,
+            'uploader': ' '.join(traverse_obj(data, ('creator', ('first_name', 'last_name'), T(compat_str)))) or None,
+        }, traverse_obj(data, {
+            'title': ('project', 'project_name', T(compat_str)),
+            'timestamp': ('created_at', T(unified_timestamp)),
+            'thumbnail': ('thumbnail_url', T(url_or_none)),
+        }), rev=True)

youtube_dl/extractor/dlf.py

@@ -0,0 +1,204 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+import re
+
+from .common import InfoExtractor
+from ..compat import (
+    compat_str,
+)
+from ..utils import (
+    determine_ext,
+    extract_attributes,
+    int_or_none,
+    merge_dicts,
+    traverse_obj,
+    url_or_none,
+    variadic,
+)
+
+
+class DLFBaseIE(InfoExtractor):
+    _VALID_URL_BASE = r'https?://(?:www\.)?deutschlandfunk\.de/'
+    _BUTTON_REGEX = r'(<button[^>]+alt="Anhören"[^>]+data-audio-diraid[^>]*>)'
+
+    def _parse_button_attrs(self, button, audio_id=None):
+        attrs = extract_attributes(button)
+        audio_id = audio_id or attrs['data-audio-diraid']
+
+        url = traverse_obj(
+            attrs, 'data-audio-download-src', 'data-audio', 'data-audioreference',
+            'data-audio-src', expected_type=url_or_none)
+        ext = determine_ext(url)
+        formats = (self._extract_m3u8_formats(url, audio_id, fatal=False)
+                   if ext == 'm3u8' else [{'url': url, 'ext': ext, 'vcodec': 'none'}])
+        self._sort_formats(formats)
+
+        def traverse_attrs(path):
+            path = list(variadic(path))
+            t = path.pop() if callable(path[-1]) else None
+            return traverse_obj(attrs, path, expected_type=t, get_all=False)
+
+        def txt_or_none(v, default=None):
+            return default if v is None else (compat_str(v).strip() or default)
+
+        return merge_dicts(*reversed([{
+            'id': audio_id,
+            # 'extractor_key': DLFIE.ie_key(),
+            # 'extractor': DLFIE.IE_NAME,
+            'formats': formats,
+        }, dict((k, traverse_attrs(v)) for k, v in {
+            'title': (('data-audiotitle', 'data-audio-title', 'data-audio-download-tracking-title'), txt_or_none),
+            'duration': (('data-audioduration', 'data-audio-duration'), int_or_none),
+            'thumbnail': ('data-audioimage', url_or_none),
+            'uploader': 'data-audio-producer',
+            'series': 'data-audio-series',
+            'channel': 'data-audio-origin-site-name',
+            'webpage_url': ('data-audio-download-tracking-path', url_or_none),
+        }.items())]))
+
+
+class DLFIE(DLFBaseIE):
+    IE_NAME = 'dlf'
+    _VALID_URL = DLFBaseIE._VALID_URL_BASE + r'[\w-]+-dlf-(?P<id>[\da-f]{8})-100\.html'
+    _TESTS = [
+        # Audio as an HLS stream
+        {
+            'url': 'https://www.deutschlandfunk.de/tanz-der-saiteninstrumente-das-wild-strings-trio-aus-slowenien-dlf-03a3eb19-100.html',
+            'info_dict': {
+                'id': '03a3eb19',
+                'title': r're:Tanz der Saiteninstrumente [-/] Das Wild Strings Trio aus Slowenien',
+                'ext': 'm4a',
+                'duration': 3298,
+                'thumbnail': 'https://assets.deutschlandfunk.de/FALLBACK-IMAGE-AUDIO/512x512.png?t=1603714364673',
+                'uploader': 'Deutschlandfunk',
+                'series': 'On Stage',
+                'channel': 'deutschlandfunk'
+            },
+            'params': {
+                'skip_download': 'm3u8'
+            },
+            'skip': 'This webpage no longer exists'
+        }, {
+            'url': 'https://www.deutschlandfunk.de/russische-athleten-kehren-zurueck-auf-die-sportbuehne-ein-gefaehrlicher-tueroeffner-dlf-d9cc1856-100.html',
+            'info_dict': {
+                'id': 'd9cc1856',
+                'title': 'Russische Athleten kehren zurück auf die Sportbühne: Ein gefährlicher Türöffner',
+                'ext': 'mp3',
+                'duration': 291,
+                'thumbnail': 'https://assets.deutschlandfunk.de/FALLBACK-IMAGE-AUDIO/512x512.png?t=1603714364673',
+                'uploader': 'Deutschlandfunk',
+                'series': 'Kommentare und Themen der Woche',
+                'channel': 'deutschlandfunk'
+            }
+        },
+    ]
+
+    def _real_extract(self, url):
+        audio_id = self._match_id(url)
+        webpage = self._download_webpage(url, audio_id)
+
+        return self._parse_button_attrs(
+            self._search_regex(self._BUTTON_REGEX, webpage, 'button'), audio_id)
+
+
+class DLFCorpusIE(DLFBaseIE):
+    IE_NAME = 'dlf:corpus'
+    IE_DESC = 'DLF Multi-feed Archives'
+    _VALID_URL = DLFBaseIE._VALID_URL_BASE + r'(?P<id>(?![\w-]+-dlf-[\da-f]{8})[\w-]+-\d+)\.html'
+    _TESTS = [
+        # Recorded news broadcast with referrals to related broadcasts
+        {
+            'url': 'https://www.deutschlandfunk.de/fechten-russland-belarus-ukraine-protest-100.html',
+            'info_dict': {
+                'id': 'fechten-russland-belarus-ukraine-protest-100',
+                'title': r're:Wiederzulassung als neutrale Athleten [-/] Was die Rückkehr russischer und belarussischer Sportler beim Fechten bedeutet',
+                'description': 'md5:91340aab29c71aa7518ad5be13d1e8ad'
+            },
+            'playlist_mincount': 5,
+            'playlist': [{
+                'info_dict': {
+                    'id': '1fc5d64a',
+                    'title': r're:Wiederzulassung als neutrale Athleten [-/] Was die Rückkehr russischer und belarussischer Sportler beim Fechten bedeutet',
+                    'ext': 'mp3',
+                    'duration': 252,
+                    'thumbnail': 'https://assets.deutschlandfunk.de/aad16241-6b76-4a09-958b-96d0ee1d6f57/512x512.jpg?t=1679480020313',
+                    'uploader': 'Deutschlandfunk',
+                    'series': 'Sport',
+                    'channel': 'deutschlandfunk'
+                }
+            }, {
+                'info_dict': {
+                    'id': '2ada145f',
+                    'title': r're:(?:Sportpolitik / )?Fechtverband votiert für Rückkehr russischer Athleten',
+                    'ext': 'mp3',
+                    'duration': 336,
+                    'thumbnail': 'https://assets.deutschlandfunk.de/FILE_93982766f7317df30409b8a184ac044a/512x512.jpg?t=1678547581005',
+                    'uploader': 'Deutschlandfunk',
+                    'series': 'Deutschlandfunk Nova',
+                    'channel': 'deutschlandfunk-nova'
+                }
+            }, {
+                'info_dict': {
+                    'id': '5e55e8c9',
+                    'title': r're:Wiederzulassung von Russland und Belarus [-/] "Herumlavieren" des Fechter-Bundes sorgt für Unverständnis',
+                    'ext': 'mp3',
+                    'duration': 187,
+                    'thumbnail': 'https://assets.deutschlandfunk.de/a595989d-1ed1-4a2e-8370-b64d7f11d757/512x512.jpg?t=1679173825412',
+                    'uploader': 'Deutschlandfunk',
+                    'series': 'Sport am Samstag',
+                    'channel': 'deutschlandfunk'
+                }
+            }, {
+                'info_dict': {
+                    'id': '47e1a096',
+                    'title': r're:Rückkehr Russlands im Fechten [-/] "Fassungslos, dass es einfach so passiert ist"',
+                    'ext': 'mp3',
+                    'duration': 602,
+                    'thumbnail': 'https://assets.deutschlandfunk.de/da4c494a-21cc-48b4-9cc7-40e09fd442c2/512x512.jpg?t=1678562155770',
+                    'uploader': 'Deutschlandfunk',
+                    'series': 'Sport am Samstag',
+                    'channel': 'deutschlandfunk'
+                }
+            }, {
+                'info_dict': {
+                    'id': '5e55e8c9',
+                    'title': r're:Wiederzulassung von Russland und Belarus [-/] "Herumlavieren" des Fechter-Bundes sorgt für Unverständnis',
+                    'ext': 'mp3',
+                    'duration': 187,
+                    'thumbnail': 'https://assets.deutschlandfunk.de/a595989d-1ed1-4a2e-8370-b64d7f11d757/512x512.jpg?t=1679173825412',
+                    'uploader': 'Deutschlandfunk',
+                    'series': 'Sport am Samstag',
+                    'channel': 'deutschlandfunk'
+                }
+            }]
+        },
+        # Podcast feed with tag buttons, playlist count fluctuates
+        {
+            'url': 'https://www.deutschlandfunk.de/kommentare-und-themen-der-woche-100.html',
+            'info_dict': {
+                'id': 'kommentare-und-themen-der-woche-100',
+                'title': 'Meinung - Kommentare und Themen der Woche',
+                'description': 'md5:2901bbd65cd2d45e116d399a099ce5d5',
+            },
+            'playlist_mincount': 10,
+        },
+        # Podcast feed with no description
+        {
+            'url': 'https://www.deutschlandfunk.de/podcast-tolle-idee-100.html',
+            'info_dict': {
+                'id': 'podcast-tolle-idee-100',
+                'title': 'Wissenschaftspodcast - Tolle Idee! - Was wurde daraus?',
+            },
+            'playlist_mincount': 11,
+        },
+    ]
+
+    def _real_extract(self, url):
+        playlist_id = self._match_id(url)
+        webpage = self._download_webpage(url, playlist_id)
+
+        return self.playlist_result(
+            map(self._parse_button_attrs, re.findall(self._BUTTON_REGEX, webpage)),
+            playlist_id, self._html_search_meta(['og:title', 'twitter:title'], webpage, default=None),
+            self._html_search_meta(['description', 'og:description', 'twitter:description'], webpage, default=None))

youtube_dl/extractor/epidemicsound.py

@@ -0,0 +1,101 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+from .common import InfoExtractor
+from ..utils import (
+    float_or_none,
+    T,
+    traverse_obj,
+    txt_or_none,
+    unified_timestamp,
+    url_or_none,
+)
+
+
+class EpidemicSoundIE(InfoExtractor):
+    _VALID_URL = r'https?://(?:www\.)?epidemicsound\.com/track/(?P<id>[0-9a-zA-Z]+)'
+    _TESTS = [{
+        'url': 'https://www.epidemicsound.com/track/yFfQVRpSPz/',
+        'md5': 'd98ff2ddb49e8acab9716541cbc9dfac',
+        'info_dict': {
+            'id': '45014',
+            'display_id': 'yFfQVRpSPz',
+            'ext': 'mp3',
+            'tags': ['foley', 'door', 'knock', 'glass', 'window', 'glass door knock'],
+            'title': 'Door Knock Door 1',
+            'duration': 1,
+            'thumbnail': 'https://cdn.epidemicsound.com/curation-assets/commercial-release-cover-images/default-sfx/3000x3000.jpg',
+            'timestamp': 1415320353,
+            'upload_date': '20141107',
+            'age_limit': None,
+            # check that the "best" format was found, since test file MD5 doesn't
+            # distinguish the formats
+            'format': 'full',
+        },
+    }, {
+        'url': 'https://www.epidemicsound.com/track/mj8GTTwsZd/',
+        'md5': 'c82b745890f9baf18dc2f8d568ee3830',
+        'info_dict': {
+            'id': '148700',
+            'display_id': 'mj8GTTwsZd',
+            'ext': 'mp3',
+            'tags': ['liquid drum n bass', 'energetic'],
+            'title': 'Noplace',
+            'duration': 237,
+            'thumbnail': 'https://cdn.epidemicsound.com/curation-assets/commercial-release-cover-images/11138/3000x3000.jpg',
+            'timestamp': 1694426482,
+            'release_timestamp': 1700535606,
+            'upload_date': '20230911',
+            'age_limit': None,
+            'format': 'full',
+        },
+    }]
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+        json_data = self._download_json('https://www.epidemicsound.com/json/track/' + video_id, video_id)
+
+        def fmt_or_none(f):
+            if not f.get('format'):
+                f['format'] = f.get('format_id')
+            elif not f.get('format_id'):
+                f['format_id'] = f['format']
+            if not (f['url'] and f['format']):
+                return
+            if f.get('format_note'):
+                f['format_note'] = 'track ID ' + f['format_note']
+            f['preference'] = -1 if f['format'] == 'full' else -2
+            return f
+
+        formats = traverse_obj(json_data, (
+            'stems', T(dict.items), Ellipsis, {
+                'format': (0, T(txt_or_none)),
+                'format_note': (1, 's3TrackId', T(txt_or_none)),
+                'format_id': (1, 'stemType', T(txt_or_none)),
+                'url': (1, 'lqMp3Url', T(url_or_none)),
+            }, T(fmt_or_none)))
+
+        self._sort_formats(formats)
+
+        info = traverse_obj(json_data, {
+            'id': ('id', T(txt_or_none)),
+            'tags': ('metadataTags', Ellipsis, T(txt_or_none)),
+            'title': ('title', T(txt_or_none)),
+            'duration': ('length', T(float_or_none)),
+            'timestamp': ('added', T(unified_timestamp)),
+            'thumbnail': (('imageUrl', 'cover'), T(url_or_none)),
+            'age_limit': ('isExplicit', T(lambda b: 18 if b else None)),
+            'release_timestamp': ('releaseDate', T(unified_timestamp)),
+        }, get_all=False)
+
+        info.update(traverse_obj(json_data, {
+            'categories': ('genres', Ellipsis, 'tag', T(txt_or_none)),
+            'tags': ('metadataTags', Ellipsis, T(txt_or_none)),
+        }))
+
+        info.update({
+            'display_id': video_id,
+            'formats': formats,
+        })
+
+        return info

youtube_dl/extractor/extractors.py

@@ -159,6 +159,7 @@ from .businessinsider import BusinessInsiderIE
 from .buzzfeed import BuzzFeedIE
 from .byutv import BYUtvIE
 from .c56 import C56IE
+from .caffeine import CaffeineTVIE
 from .callin import CallinIE
 from .camdemy import (
     CamdemyIE,
@@ -226,6 +227,7 @@ from .ciscolive import (
     CiscoLiveSearchIE,
 )
 from .cjsw import CJSWIE
+from .clipchamp import ClipchampIE
 from .cliphunter import CliphunterIE
 from .clippit import ClippitIE
 from .cliprs import ClipRsIE
@@ -295,6 +297,10 @@ from .dbtv import DBTVIE
 from .dctp import DctpTvIE
 from .deezer import DeezerPlaylistIE
 from .democracynow import DemocracynowIE
+from .dlf import (
+    DLFCorpusIE,
+    DLFIE,
+)
 from .dfb import DFBIE
 from .dhm import DHMIE
 from .digg import DiggIE
@@ -352,6 +358,7 @@ from .ellentube import (
 from .elpais import ElPaisIE
 from .embedly import EmbedlyIE
 from .engadget import EngadgetIE
+from .epidemicsound import EpidemicSoundIE
 from .eporner import EpornerIE
 from .eroprofile import EroProfileIE
 from .escapist import EscapistIE
@@ -376,7 +383,6 @@ from .fc2 import (
     FC2EmbedIE,
 )
 from .fczenit import FczenitIE
-from .filemoon import FileMoonIE
 from .fifa import FifaIE
 from .filmon import (
     FilmOnIE,
@@ -437,6 +443,7 @@ from .gamespot import GameSpotIE
 from .gamestar import GameStarIE
 from .gaskrank import GaskrankIE
 from .gazeta import GazetaIE
+from .gbnews import GBNewsIE
 from .gdcvault import GDCVaultIE
 from .gedidigital import GediDigitalIE
 from .generic import GenericIE
@@ -444,6 +451,13 @@ from .gfycat import GfycatIE
 from .giantbomb import GiantBombIE
 from .giga import GigaIE
 from .glide import GlideIE
+from .globalplayer import (
+    GlobalPlayerLiveIE,
+    GlobalPlayerLivePlaylistIE,
+    GlobalPlayerAudioIE,
+    GlobalPlayerAudioEpisodeIE,
+    GlobalPlayerVideoIE
+)
 from .globo import (
     GloboIE,
     GloboArticleIE,
@@ -884,21 +898,13 @@ from .ooyala import (
 )
 from .ora import OraTVIE
 from .orf import (
-    ORFTVthekIE,
-    ORFFM4IE,
+    ORFONIE,
+    ORFONLiveIE,
     ORFFM4StoryIE,
-    ORFOE1IE,
-    ORFOE3IE,
-    ORFNOEIE,
-    ORFWIEIE,
-    ORFBGLIE,
-    ORFOOEIE,
-    ORFSTMIE,
-    ORFKTNIE,
-    ORFSBGIE,
-    ORFTIRIE,
-    ORFVBGIE,
     ORFIPTVIE,
+    ORFPodcastIE,
+    ORFRadioIE,
+    ORFRadioCollectionIE,
 )
 from .outsidetv import OutsideTVIE
 from .packtpub import (
@@ -975,6 +981,10 @@ from .pornhub import (
 from .pornotube import PornotubeIE
 from .pornovoisines import PornoVoisinesIE
 from .pornoxo import PornoXOIE
+from .pr0gramm import (
+    Pr0grammIE,
+    Pr0grammStaticIE,
+)
 from .puhutv import (
     PuhuTVIE,
     PuhuTVSerieIE,
@@ -1071,6 +1081,10 @@ from .rutube import (
 from .rutv import RUTVIE
 from .ruutu import RuutuIE
 from .ruv import RuvIE
+from .s4c import (
+    S4CIE,
+    S4CSeriesIE,
+)
 from .safari import (
     SafariIE,
     SafariApiIE,
@@ -1565,6 +1579,7 @@ from .weibo import (
     WeiboMobileIE
 )
 from .weiqitv import WeiqiTVIE
+from .whyp import WhypIE
 from .wistia import (
     WistiaIE,
     WistiaPlaylistIE,
@@ -1630,7 +1645,15 @@ from .younow import (
     YouNowChannelIE,
     YouNowMomentIE,
 )
-from .youporn import YouPornIE
+from .youporn import (
+    YouPornIE,
+    YouPornCategoryIE,
+    YouPornChannelIE,
+    YouPornCollectionIE,
+    YouPornStarIE,
+    YouPornTagIE,
+    YouPornVideosIE,
+)
 from .yourporn import YourPornIE
 from .yourupload import YourUploadIE
 from .youtube import (
@@ -1678,7 +1701,3 @@ from .zingmp3 import (
 )
 from .zoom import ZoomIE
 from .zype import ZypeIE
-from .pr0gramm import (
-    Pr0grammIE,
-    Pr0grammStaticIE,
-)

youtube_dl/extractor/filemoon.py

@@ -1,43 +0,0 @@
-# coding: utf-8
-from __future__ import unicode_literals
-
-import re
-
-from .common import InfoExtractor
-from ..utils import (
-    decode_packed_codes,
-    js_to_json,
-)
-
-
-class FileMoonIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:www\.)?filemoon\.sx/./(?P<id>\w+)'
-    _TEST = {
-        'url': 'https://filemoon.sx/e/dw40rxrzruqz',
-        'md5': '5a713742f57ac4aef29b74733e8dda01',
-        'info_dict': {
-            'id': 'dw40rxrzruqz',
-            'title': 'dw40rxrzruqz',
-            'ext': 'mp4'
-        }
-    }
-
-    def _real_extract(self, url):
-        video_id = self._match_id(url)
-
-        webpage = self._download_webpage(url, video_id)
-        matches = re.findall(r'(?s)(eval.*?)</script>', webpage)
-        packed = matches[-1]
-        unpacked = decode_packed_codes(packed)
-        jwplayer_sources = self._parse_json(
-            self._search_regex(
-                r'(?s)player\s*\.\s*setup\s*\(\s*\{\s*sources\s*:\s*(.*?])', unpacked, 'jwplayer sources'),
-            video_id, transform_source=js_to_json)
-
-        formats = self._parse_jwplayer_formats(jwplayer_sources, video_id)
-
-        return {
-            'id': video_id,
-            'title': self._generic_title(url) or video_id,
-            'formats': formats
-        }

youtube_dl/extractor/gbnews.py

@@ -0,0 +1,139 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+from .common import InfoExtractor
+from ..utils import (
+    extract_attributes,
+    ExtractorError,
+    T,
+    traverse_obj,
+    txt_or_none,
+    url_or_none,
+)
+
+
+class GBNewsIE(InfoExtractor):
+    IE_DESC = 'GB News clips, features and live stream'
+
+    # \w+ is normally shows or news, but apparently any word redirects to the correct URL
+    _VALID_URL = r'https?://(?:www\.)?gbnews\.(?:uk|com)/(?:\w+/)?(?P<id>[^#?]+)'
+
+    _PLATFORM = 'safari'
+    _SSMP_URL = 'https://mm-v2.simplestream.com/ssmp/api.php'
+    _TESTS = [{
+        'url': 'https://www.gbnews.uk/shows/andrew-neils-message-to-companies-choosing-to-boycott-gb-news/106889',
+        'info_dict': {
+            'id': '106889',
+            'ext': 'mp4',
+            'title': "Andrew Neil's message to companies choosing to boycott GB News",
+            'description': 'md5:b281f5d22fd6d5eda64a4e3ba771b351',
+        },
+        'skip': '404 not found',
+    }, {
+        'url': 'https://www.gbnews.com/news/bbc-claudine-gay-harvard-university-antisemitism-row',
+        'info_dict': {
+            'id': '52264136',
+            'display_id': 'bbc-claudine-gay-harvard-university-antisemitism-row',
+            'ext': 'mp4',
+            'title': 'BBC deletes post after furious backlash over headline downplaying antisemitism',
+            'description': 'The post was criticised by former employers of the broadcaster',
+        },
+    }, {
+        'url': 'https://www.gbnews.uk/watchlive',
+        'info_dict': {
+            'id': '1069',
+            'display_id': 'watchlive',
+            'ext': 'mp4',
+            'title': 'GB News Live',
+            'is_live': True,
+        },
+        'params': {
+            'skip_download': 'm3u8',
+        },
+    }]
+
+    def _real_extract(self, url):
+        display_id = self._match_id(url).split('/')[-1]
+
+        webpage = self._download_webpage(url, display_id)
+        # extraction based on https://github.com/ytdl-org/youtube-dl/issues/29341
+        '''
+        <div id="video-106908"
+            class="simplestream"
+            data-id="GB001"
+            data-type="vod"
+            data-key="3Li3Nt2Qs8Ct3Xq9Fi5Uy0Mb2Bj0Qs"
+            data-token="f9c317c727dc07f515b20036c8ef14a6"
+            data-expiry="1624300052"
+            data-uvid="37900558"
+            data-poster="https://thumbnails.simplestreamcdn.com/gbnews/ondemand/37900558.jpg?width=700&"
+            data-npaw="false"
+            data-env="production">
+        '''
+        # exception if no match
+        video_data = self._search_regex(
+            r'(<div\s[^>]*\bclass\s*=\s*(\'|")(?!.*sidebar\b)simplestream(?:\s[\s\w$-]*)?\2[^>]*>)',
+            webpage, 'video data')
+
+        video_data = extract_attributes(video_data)
+        ss_id = video_data.get('data-id')
+        if not ss_id:
+            raise ExtractorError('Simplestream ID not found')
+
+        json_data = self._download_json(
+            self._SSMP_URL, display_id,
+            note='Downloading Simplestream JSON metadata',
+            errnote='Unable to download Simplestream JSON metadata',
+            query={
+                'id': ss_id,
+                'env': video_data.get('data-env', 'production'),
+            }, fatal=False)
+
+        meta_url = traverse_obj(json_data, ('response', 'api_hostname'))
+        if not meta_url:
+            raise ExtractorError('No API host found')
+
+        uvid = video_data['data-uvid']
+        dtype = video_data.get('data-type')
+        stream_data = self._download_json(
+            '%s/api/%s/stream/%s' % (meta_url, 'show' if dtype == 'vod' else dtype, uvid),
+            uvid,
+            query={
+                'key': video_data.get('data-key'),
+                'platform': self._PLATFORM,
+            },
+            headers={
+                'Token': video_data.get('data-token'),
+                'Token-Expiry': video_data.get('data-expiry'),
+                'Uvid': uvid,
+            }, fatal=False)
+
+        stream_url = traverse_obj(stream_data, (
+            'response', 'stream', T(url_or_none)))
+        if not stream_url:
+            raise ExtractorError('No stream data/URL')
+
+        # now known to be a dict
+        stream_data = stream_data['response']
+        drm = stream_data.get('drm')
+        if drm:
+            self.report_drm(uvid)
+
+        formats = self._extract_m3u8_formats(
+            stream_url, uvid, ext='mp4', entry_protocol='m3u8_native',
+            fatal=False)
+        # exception if no formats
+        self._sort_formats(formats)
+
+        return {
+            'id': uvid,
+            'display_id': display_id,
+            'title': (traverse_obj(stream_data, ('title', T(txt_or_none)))
+                      or self._og_search_title(webpage, default=None)
+                      or display_id.replace('-', ' ').capitalize()),
+            'description': self._og_search_description(webpage, default=None),
+            'thumbnail': (traverse_obj(video_data, ('data-poster', T(url_or_none)))
+                          or self._og_search_thumbnail(webpage)),
+            'formats': formats,
+            'is_live': (dtype == 'live') or None,
+        }

youtube_dl/extractor/globalplayer.py

@@ -0,0 +1,273 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+from .common import InfoExtractor
+from ..utils import (
+    clean_html,
+    join_nonempty,
+    merge_dicts,
+    parse_duration,
+    str_or_none,
+    T,
+    traverse_obj,
+    unified_strdate,
+    unified_timestamp,
+    urlhandle_detect_ext,
+)
+
+
+class GlobalPlayerBaseIE(InfoExtractor):
+
+    def _get_page_props(self, url, video_id):
+        webpage = self._download_webpage(url, video_id)
+        return self._search_nextjs_data(webpage, video_id)['props']['pageProps']
+
+    def _request_ext(self, url, video_id):
+        return urlhandle_detect_ext(self._request_webpage(  # Server rejects HEAD requests
+            url, video_id, note='Determining source extension'))
+
+    @staticmethod
+    def _clean_desc(x):
+        x = clean_html(x)
+        if x:
+            x = x.replace('\xa0', ' ')
+        return x
+
+    def _extract_audio(self, episode, series):
+
+        return merge_dicts({
+            'vcodec': 'none',
+        }, traverse_obj(series, {
+            'series': 'title',
+            'series_id': 'id',
+            'thumbnail': 'imageUrl',
+            'uploader': 'itunesAuthor',  # podcasts only
+        }), traverse_obj(episode, {
+            'id': 'id',
+            'description': ('description', T(self._clean_desc)),
+            'duration': ('duration', T(parse_duration)),
+            'thumbnail': 'imageUrl',
+            'url': 'streamUrl',
+            'timestamp': (('pubDate', 'startDate'), T(unified_timestamp)),
+            'title': 'title',
+        }, get_all=False), rev=True)
+
+
+class GlobalPlayerLiveIE(GlobalPlayerBaseIE):
+    _VALID_URL = r'https?://www\.globalplayer\.com/live/(?P<id>\w+)/\w+'
+    _TESTS = [{
+        'url': 'https://www.globalplayer.com/live/smoothchill/uk/',
+        'info_dict': {
+            'id': '2mx1E',
+            'ext': 'aac',
+            'display_id': 'smoothchill-uk',
+            'title': 're:^Smooth Chill.+$',
+            'thumbnail': 'https://herald.musicradio.com/media/f296ade8-50c9-4f60-911f-924e96873620.png',
+            'description': 'Music To Chill To',
+            # 'live_status': 'is_live',
+            'is_live': True,
+        },
+    }, {
+        # national station
+        'url': 'https://www.globalplayer.com/live/heart/uk/',
+        'info_dict': {
+            'id': '2mwx4',
+            'ext': 'aac',
+            'description': 'turn up the feel good!',
+            'thumbnail': 'https://herald.musicradio.com/media/49b9e8cb-15bf-4bf2-8c28-a4850cc6b0f3.png',
+            # 'live_status': 'is_live',
+            'is_live': True,
+            'title': 're:^Heart UK.+$',
+            'display_id': 'heart-uk',
+        },
+    }, {
+        # regional variation
+        'url': 'https://www.globalplayer.com/live/heart/london/',
+        'info_dict': {
+            'id': 'AMqg',
+            'ext': 'aac',
+            'thumbnail': 'https://herald.musicradio.com/media/49b9e8cb-15bf-4bf2-8c28-a4850cc6b0f3.png',
+            'title': 're:^Heart London.+$',
+            # 'live_status': 'is_live',
+            'is_live': True,
+            'display_id': 'heart-london',
+            'description': 'turn up the feel good!',
+        },
+    }]
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+        station = self._get_page_props(url, video_id)['station']
+        stream_url = station['streamUrl']
+
+        return merge_dicts({
+            'id': station['id'],
+            'display_id': (
+                join_nonempty('brandSlug', 'slug', from_dict=station)
+                or station.get('legacyStationPrefix')),
+            'url': stream_url,
+            'ext': self._request_ext(stream_url, video_id),
+            'vcodec': 'none',
+            'is_live': True,
+        }, {
+            'title': self._live_title(traverse_obj(
+                station, (('name', 'brandName'), T(str_or_none)),
+                get_all=False)),
+        }, traverse_obj(station, {
+            'description': 'tagline',
+            'thumbnail': 'brandLogo',
+        }), rev=True)
+
+
+class GlobalPlayerLivePlaylistIE(GlobalPlayerBaseIE):
+    _VALID_URL = r'https?://www\.globalplayer\.com/playlists/(?P<id>\w+)'
+    _TESTS = [{
+        # "live playlist"
+        'url': 'https://www.globalplayer.com/playlists/8bLk/',
+        'info_dict': {
+            'id': '8bLk',
+            'ext': 'aac',
+            # 'live_status': 'is_live',
+            'is_live': True,
+            'description': r're:(?s).+\bclassical\b.+\bClassic FM Hall [oO]f Fame\b',
+            'thumbnail': 'https://images.globalplayer.com/images/551379?width=450&signature=oMLPZIoi5_dBSHnTMREW0Xg76mA=',
+            'title': 're:Classic FM Hall of Fame.+$'
+        },
+    }]
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+        station = self._get_page_props(url, video_id)['playlistData']
+        stream_url = station['streamUrl']
+
+        return merge_dicts({
+            'id': video_id,
+            'url': stream_url,
+            'ext': self._request_ext(stream_url, video_id),
+            'vcodec': 'none',
+            'is_live': True,
+        }, traverse_obj(station, {
+            'title': 'title',
+            'description': ('description', T(self._clean_desc)),
+            'thumbnail': 'image',
+        }), rev=True)
+
+
+class GlobalPlayerAudioIE(GlobalPlayerBaseIE):
+    _VALID_URL = r'https?://www\.globalplayer\.com/(?:(?P<podcast>podcasts)/|catchup/\w+/\w+/)(?P<id>\w+)/?(?:$|[?#])'
+    _TESTS = [{
+        # podcast
+        'url': 'https://www.globalplayer.com/podcasts/42KuaM/',
+        'playlist_mincount': 5,
+        'info_dict': {
+            'id': '42KuaM',
+            'title': 'Filthy Ritual',
+            'thumbnail': 'md5:60286e7d12d795bd1bbc9efc6cee643e',
+            'categories': ['Society & Culture', 'True Crime'],
+            'uploader': 'Global',
+            'description': r're:(?s).+\bscam\b.+?\bseries available now\b',
+        },
+    }, {
+        # radio catchup
+        'url': 'https://www.globalplayer.com/catchup/lbc/uk/46vyD7z/',
+        'playlist_mincount': 2,
+        'info_dict': {
+            'id': '46vyD7z',
+            'description': 'Nick Ferrari At Breakfast is Leading Britain\'s Conversation.',
+            'title': 'Nick Ferrari',
+            'thumbnail': 'md5:4df24d8a226f5b2508efbcc6ae874ebf',
+        },
+    }]
+
+    def _real_extract(self, url):
+        video_id, podcast = self._match_valid_url(url).group('id', 'podcast')
+        props = self._get_page_props(url, video_id)
+        series = props['podcastInfo'] if podcast else props['catchupInfo']
+
+        return merge_dicts({
+            '_type': 'playlist',
+            'id': video_id,
+            'entries': [self._extract_audio(ep, series) for ep in traverse_obj(
+                        series, ('episodes', lambda _, v: v['id'] and v['streamUrl']))],
+            'categories': traverse_obj(series, ('categories', Ellipsis, 'name')) or None,
+        }, traverse_obj(series, {
+            'description': ('description', T(self._clean_desc)),
+            'thumbnail': 'imageUrl',
+            'title': 'title',
+            'uploader': 'itunesAuthor',  # podcasts only
+        }), rev=True)
+
+
+class GlobalPlayerAudioEpisodeIE(GlobalPlayerBaseIE):
+    _VALID_URL = r'https?://www\.globalplayer\.com/(?:(?P<podcast>podcasts)|catchup/\w+/\w+)/episodes/(?P<id>\w+)/?(?:$|[?#])'
+    _TESTS = [{
+        # podcast
+        'url': 'https://www.globalplayer.com/podcasts/episodes/7DrfNnE/',
+        'info_dict': {
+            'id': '7DrfNnE',
+            'ext': 'mp3',
+            'title': 'Filthy Ritual - Trailer',
+            'description': 'md5:1f1562fd0f01b4773b590984f94223e0',
+            'thumbnail': 'md5:60286e7d12d795bd1bbc9efc6cee643e',
+            'duration': 225.0,
+            'timestamp': 1681254900,
+            'series': 'Filthy Ritual',
+            'series_id': '42KuaM',
+            'upload_date': '20230411',
+            'uploader': 'Global',
+        },
+    }, {
+        # radio catchup
+        'url': 'https://www.globalplayer.com/catchup/lbc/uk/episodes/2zGq26Vcv1fCWhddC4JAwETXWe/',
+        'only_matching': True,
+        # expired: refresh the details with a current show for a full test
+        'info_dict': {
+            'id': '2zGq26Vcv1fCWhddC4JAwETXWe',
+            'ext': 'm4a',
+            'timestamp': 1682056800,
+            'series': 'Nick Ferrari',
+            'thumbnail': 'md5:4df24d8a226f5b2508efbcc6ae874ebf',
+            'upload_date': '20230421',
+            'series_id': '46vyD7z',
+            'description': 'Nick Ferrari At Breakfast is Leading Britain\'s Conversation.',
+            'title': 'Nick Ferrari',
+            'duration': 10800.0,
+        },
+    }]
+
+    def _real_extract(self, url):
+        video_id, podcast = self._match_valid_url(url).group('id', 'podcast')
+        props = self._get_page_props(url, video_id)
+        episode = props['podcastEpisode'] if podcast else props['catchupEpisode']
+
+        return self._extract_audio(
+            episode, traverse_obj(episode, 'podcast', 'show', expected_type=dict) or {})
+
+
+class GlobalPlayerVideoIE(GlobalPlayerBaseIE):
+    _VALID_URL = r'https?://www\.globalplayer\.com/videos/(?P<id>\w+)'
+    _TESTS = [{
+        'url': 'https://www.globalplayer.com/videos/2JsSZ7Gm2uP/',
+        'info_dict': {
+            'id': '2JsSZ7Gm2uP',
+            'ext': 'mp4',
+            'description': 'md5:6a9f063c67c42f218e42eee7d0298bfd',
+            'thumbnail': 'md5:d4498af48e15aae4839ce77b97d39550',
+            'upload_date': '20230420',
+            'title': 'Treble Malakai Bayoh sings a sublime Handel aria at Classic FM Live',
+        },
+    }]
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+        meta = self._get_page_props(url, video_id)['videoData']
+
+        return merge_dicts({
+            'id': video_id,
+        }, traverse_obj(meta, {
+            'url': 'url',
+            'thumbnail': ('image', 'url'),
+            'title': 'title',
+            'upload_date': ('publish_date', T(unified_strdate)),
+            'description': 'description',
+        }), rev=True)

youtube_dl/extractor/imgur.py

@@ -1,101 +1,267 @@
+# coding: utf-8
 from __future__ import unicode_literals
 
 import re
 
 from .common import InfoExtractor
 from ..utils import (
+    determine_ext,
+    ExtractorError,
+    float_or_none,
     int_or_none,
     js_to_json,
+    merge_dicts,
     mimetype2ext,
-    ExtractorError,
+    parse_iso8601,
+    T,
+    traverse_obj,
+    txt_or_none,
+    url_or_none,
 )
 
 
-class ImgurIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:i\.)?imgur\.com/(?!(?:a|gallery|(?:t(?:opic)?|r)/[^/]+)/)(?P<id>[a-zA-Z0-9]+)'
+class ImgurBaseIE(InfoExtractor):
+    # hard-coded value, as also used by ArchiveTeam
+    _CLIENT_ID = '546c25a59c58ad7'
+
+    @classmethod
+    def _imgur_result(cls, item_id):
+        return cls.url_result('imgur:%s' % item_id, ImgurIE.ie_key(), item_id)
+
+    def _call_api(self, endpoint, video_id, **kwargs):
+        return self._download_json(
+            'https://api.imgur.com/post/v1/%s/%s?client_id=%s&include=media,account' % (endpoint, video_id, self._CLIENT_ID),
+            video_id, **kwargs)
+
+    @staticmethod
+    def get_description(s):
+        if 'Discover the magic of the internet at Imgur' in s:
+            return None
+        return txt_or_none(s)
+
+
+class ImgurIE(ImgurBaseIE):
+    _VALID_URL = r'''(?x)
+        (?:
+            https?://(?:i\.)?imgur\.com/(?!(?:a|gallery|t|topic|r)/)|
+            imgur:
+        )(?P<id>[a-zA-Z0-9]+)
+    '''
 
     _TESTS = [{
-        'url': 'https://i.imgur.com/A61SaA1.gifv',
+        'url': 'https://imgur.com/A61SaA1',
         'info_dict': {
             'id': 'A61SaA1',
             'ext': 'mp4',
             'title': 're:Imgur GIF$|MRW gifv is up and running without any bugs$',
+            'timestamp': 1416446068,
+            'upload_date': '20141120',
         },
     }, {
-        'url': 'https://imgur.com/A61SaA1',
+        'url': 'https://i.imgur.com/A61SaA1.gifv',
         'only_matching': True,
     }, {
         'url': 'https://i.imgur.com/crGpqCV.mp4',
         'only_matching': True,
     }, {
-        # no title
+        # previously, no title
         'url': 'https://i.imgur.com/jxBXAMC.gifv',
-        'only_matching': True,
+        'info_dict': {
+            'id': 'jxBXAMC',
+            'ext': 'mp4',
+            'title': 'Fahaka puffer feeding',
+            'timestamp': 1533835503,
+            'upload_date': '20180809',
+        },
     }]
 
+    def _extract_twitter_formats(self, html, tw_id='twitter', **kwargs):
+        fatal = kwargs.pop('fatal', False)
+        tw_stream = self._html_search_meta('twitter:player:stream', html, fatal=fatal, **kwargs)
+        if not tw_stream:
+            return []
+        ext = mimetype2ext(self._html_search_meta(
+            'twitter:player:stream:content_type', html, default=None))
+        width, height = (int_or_none(self._html_search_meta('twitter:player:' + v, html, default=None))
+                         for v in ('width', 'height'))
+        return [{
+            'format_id': tw_id,
+            'url': tw_stream,
+            'ext': ext or determine_ext(tw_stream),
+            'width': width,
+            'height': height,
+        }]
+
     def _real_extract(self, url):
         video_id = self._match_id(url)
+        data = self._call_api('media', video_id, fatal=False, expected_status=404)
         webpage = self._download_webpage(
-            'https://i.imgur.com/{id}.gifv'.format(id=video_id), video_id)
+            'https://i.imgur.com/{id}.gifv'.format(id=video_id), video_id, fatal=not data) or ''
 
-        width = int_or_none(self._og_search_property(
-            'video:width', webpage, default=None))
-        height = int_or_none(self._og_search_property(
-            'video:height', webpage, default=None))
+        if not traverse_obj(data, ('media', 0, (
+                ('type', T(lambda t: t == 'video' or None)),
+                ('metadata', 'is_animated'))), get_all=False):
+            raise ExtractorError(
+                '%s is not a video or animated image' % video_id,
+                expected=True)
+
+        media_fmt = traverse_obj(data, ('media', 0, {
+            'url': ('url', T(url_or_none)),
+            'ext': 'ext',
+            'width': ('width', T(int_or_none)),
+            'height': ('height', T(int_or_none)),
+            'filesize': ('size', T(int_or_none)),
+            'acodec': ('metadata', 'has_sound', T(lambda b: None if b else 'none')),
+        }))
+
+        media_url = traverse_obj(media_fmt, 'url')
+        if media_url:
+            if not media_fmt.get('ext'):
+                media_fmt['ext'] = mimetype2ext(traverse_obj(
+                    data, ('media', 0, 'mime_type'))) or determine_ext(media_url)
+            if traverse_obj(data, ('media', 0, 'type')) == 'image':
+                media_fmt['acodec'] = 'none'
+                media_fmt.setdefault('preference', -10)
+
+        tw_formats = self._extract_twitter_formats(webpage)
+        if traverse_obj(tw_formats, (0, 'url')) == media_url:
+            tw_formats = []
+        else:
+            # maybe this isn't an animated image/video?
+            self._check_formats(tw_formats, video_id)
 
         video_elements = self._search_regex(
             r'(?s)<div class="video-elements">(.*?)</div>',
             webpage, 'video elements', default=None)
-        if not video_elements:
+        if not (video_elements or tw_formats or media_url):
             raise ExtractorError(
-                'No sources found for video %s. Maybe an image?' % video_id,
+                'No sources found for video %s. Maybe a plain image?' % video_id,
                 expected=True)
 
-        formats = []
-        for m in re.finditer(r'<source\s+src="(?P<src>[^"]+)"\s+type="(?P<type>[^"]+)"', video_elements):
-            formats.append({
-                'format_id': m.group('type').partition('/')[2],
-                'url': self._proto_relative_url(m.group('src')),
-                'ext': mimetype2ext(m.group('type')),
-                'width': width,
-                'height': height,
+        def mung_format(fmt, *extra):
+            fmt.update({
                 'http_headers': {
                     'User-Agent': 'youtube-dl (like wget)',
                 },
             })
+            for d in extra:
+                fmt.update(d)
+            return fmt
 
-        gif_json = self._search_regex(
-            r'(?s)var\s+videoItem\s*=\s*(\{.*?\})',
-            webpage, 'GIF code', fatal=False)
-        if gif_json:
-            gifd = self._parse_json(
-                gif_json, video_id, transform_source=js_to_json)
-            formats.append({
-                'format_id': 'gif',
-                'preference': -10,
-                'width': width,
-                'height': height,
-                'ext': 'gif',
-                'acodec': 'none',
-                'vcodec': 'gif',
-                'container': 'gif',
-                'url': self._proto_relative_url(gifd['gifUrl']),
-                'filesize': gifd.get('size'),
-                'http_headers': {
-                    'User-Agent': 'youtube-dl (like wget)',
-                },
-            })
+        if video_elements:
+            def og_get_size(media_type):
+                return dict((p, int_or_none(self._og_search_property(
+                    ':'.join((media_type, p)), webpage, default=None)))
+                    for p in ('width', 'height'))
+
+            size = og_get_size('video')
+            if all(v is None for v in size.values()):
+                size = og_get_size('image')
+
+            formats = traverse_obj(
+                re.finditer(r'<source\s+src="(?P<src>[^"]+)"\s+type="(?P<type>[^"]+)"', video_elements),
+                (Ellipsis, {
+                    'format_id': ('type', T(lambda s: s.partition('/')[2])),
+                    'url': ('src', T(self._proto_relative_url)),
+                    'ext': ('type', T(mimetype2ext)),
+                }, T(lambda f: mung_format(f, size))))
+
+            gif_json = self._search_regex(
+                r'(?s)var\s+videoItem\s*=\s*(\{.*?\})',
+                webpage, 'GIF code', fatal=False)
+            MUST_BRANCH = (None, T(lambda _: None))
+            formats.extend(traverse_obj(gif_json, (
+                T(lambda j: self._parse_json(
+                    j, video_id, transform_source=js_to_json, fatal=False)), {
+                        'url': ('gifUrl', T(self._proto_relative_url)),
+                        'filesize': ('size', T(int_or_none)),
+                }, T(lambda f: mung_format(f, size, {
+                    'format_id': 'gif',
+                    'preference': -10,  # gifs are worse than videos
+                    'ext': 'gif',
+                    'acodec': 'none',
+                    'vcodec': 'gif',
+                    'container': 'gif',
+                })), MUST_BRANCH)))
+        else:
+            formats = []
+
+        # maybe add formats from JSON or page Twitter metadata
+        if not any((u == media_url) for u in traverse_obj(formats, (Ellipsis, 'url'))):
+            formats.append(mung_format(media_fmt))
+        tw_url = traverse_obj(tw_formats, (0, 'url'))
+        if not any((u == tw_url) for u in traverse_obj(formats, (Ellipsis, 'url'))):
+            formats.extend(mung_format(f) for f in tw_formats)
 
         self._sort_formats(formats)
 
-        return {
+        return merge_dicts(traverse_obj(data, {
+            'uploader_id': ('account_id', T(txt_or_none),
+                            T(lambda a: a if int_or_none(a) != 0 else None)),
+            'uploader': ('account', 'username', T(txt_or_none)),
+            'uploader_url': ('account', 'avatar_url', T(url_or_none)),
+            'like_count': ('upvote_count', T(int_or_none)),
+            'dislike_count': ('downvote_count', T(int_or_none)),
+            'comment_count': ('comment_count', T(int_or_none)),
+            'age_limit': ('is_mature', T(lambda x: 18 if x else None)),
+            'timestamp': (('updated_at', 'created_at'), T(parse_iso8601)),
+            'release_timestamp': ('created_at', T(parse_iso8601)),
+        }, get_all=False), traverse_obj(data, ('media', 0, 'metadata', {
+            'title': ('title', T(txt_or_none)),
+            'description': ('description', T(self.get_description)),
+            'duration': ('duration', T(float_or_none)),
+            'timestamp': (('updated_at', 'created_at'), T(parse_iso8601)),
+            'release_timestamp': ('created_at', T(parse_iso8601)),
+        })), {
             'id': video_id,
             'formats': formats,
-            'title': self._og_search_title(webpage, default=video_id),
-        }
+            'title': self._og_search_title(webpage, default='Imgur video ' + video_id),
+            'description': self.get_description(self._og_search_description(webpage)),
+            'thumbnail': url_or_none(self._html_search_meta('thumbnailUrl', webpage, default=None)),
+        })
 
 
-class ImgurGalleryIE(InfoExtractor):
+class ImgurGalleryBaseIE(ImgurBaseIE):
+    _GALLERY = True
+
+    def _real_extract(self, url):
+        gallery_id = self._match_id(url)
+
+        data = self._call_api('albums', gallery_id, fatal=False, expected_status=404)
+
+        info = traverse_obj(data, {
+            'title': ('title', T(txt_or_none)),
+            'description': ('description', T(self.get_description)),
+        })
+
+        if traverse_obj(data, 'is_album'):
+
+            def yield_media_ids():
+                for m_id in traverse_obj(data, (
+                        'media', lambda _, v: v.get('type') == 'video' or v['metadata']['is_animated'],
+                        'id', T(txt_or_none))):
+                    yield m_id
+
+            # if a gallery with exactly one video, apply album metadata to video
+            media_id = (
+                self._GALLERY
+                and traverse_obj(data, ('image_count', T(lambda c: c == 1)))
+                and next(yield_media_ids(), None))
+
+            if not media_id:
+                result = self.playlist_result(
+                    map(self._imgur_result, yield_media_ids()), gallery_id)
+                result.update(info)
+                return result
+            gallery_id = media_id
+
+        result = self._imgur_result(gallery_id)
+        info['_type'] = 'url_transparent'
+        result.update(info)
+        return result
+
+
+class ImgurGalleryIE(ImgurGalleryBaseIE):
     IE_NAME = 'imgur:gallery'
     _VALID_URL = r'https?://(?:i\.)?imgur\.com/(?:gallery|(?:t(?:opic)?|r)/[^/]+)/(?P<id>[a-zA-Z0-9]+)'
 
@@ -106,49 +272,93 @@ class ImgurGalleryIE(InfoExtractor):
             'title': 'Adding faces make every GIF better',
         },
         'playlist_count': 25,
+        'skip': 'Zoinks! You\'ve taken a wrong turn.',
     }, {
+        # TODO: static images - replace with animated/video gallery
         'url': 'http://imgur.com/topic/Aww/ll5Vk',
         'only_matching': True,
     }, {
         'url': 'https://imgur.com/gallery/YcAQlkx',
+        'add_ies': ['Imgur'],
         'info_dict': {
             'id': 'YcAQlkx',
             'ext': 'mp4',
             'title': 'Classic Steve Carell gif...cracks me up everytime....damn the repost downvotes....',
-        }
+            'timestamp': 1358554297,
+            'upload_date': '20130119',
+            'uploader_id': '1648642',
+            'uploader': 'wittyusernamehere',
+        },
     }, {
+        # TODO: static image - replace with animated/video gallery
         'url': 'http://imgur.com/topic/Funny/N8rOudd',
         'only_matching': True,
     }, {
         'url': 'http://imgur.com/r/aww/VQcQPhM',
-        'only_matching': True,
+        'add_ies': ['Imgur'],
+        'info_dict': {
+            'id': 'VQcQPhM',
+            'ext': 'mp4',
+            'title': 'The boss is here',
+            'timestamp': 1476494751,
+            'upload_date': '20161015',
+            'uploader_id': '19138530',
+            'uploader': 'thematrixcam',
+        },
+    },
+        # from PR #16674
+        {
+        'url': 'https://imgur.com/t/unmuted/6lAn9VQ',
+        'info_dict': {
+            'id': '6lAn9VQ',
+            'title': 'Penguins !',
+        },
+        'playlist_count': 3,
+    }, {
+        'url': 'https://imgur.com/t/unmuted/kx2uD3C',
+        'add_ies': ['Imgur'],
+        'info_dict': {
+            'id': 'ZVMv45i',
+            'ext': 'mp4',
+            'title': 'Intruder',
+            'timestamp': 1528129683,
+            'upload_date': '20180604',
+        },
+    }, {
+        'url': 'https://imgur.com/t/unmuted/wXSK0YH',
+        'add_ies': ['Imgur'],
+        'info_dict': {
+            'id': 'JCAP4io',
+            'ext': 'mp4',
+            'title': 're:I got the blues$',
+            'description': 'Luka’s vocal stylings.\n\nFP edit: don’t encourage me. I’ll never stop posting Luka and friends.',
+            'timestamp': 1527809525,
+            'upload_date': '20180531',
+        },
     }]
 
-    def _real_extract(self, url):
-        gallery_id = self._match_id(url)
 
-        data = self._download_json(
-            'https://imgur.com/gallery/%s.json' % gallery_id,
-            gallery_id)['data']['image']
-
-        if data.get('is_album'):
-            entries = [
-                self.url_result('http://imgur.com/%s' % image['hash'], ImgurIE.ie_key(), image['hash'])
-                for image in data['album_images']['images'] if image.get('hash')]
-            return self.playlist_result(entries, gallery_id, data.get('title'), data.get('description'))
-
-        return self.url_result('http://imgur.com/%s' % gallery_id, ImgurIE.ie_key(), gallery_id)
-
-
-class ImgurAlbumIE(ImgurGalleryIE):
+class ImgurAlbumIE(ImgurGalleryBaseIE):
     IE_NAME = 'imgur:album'
     _VALID_URL = r'https?://(?:i\.)?imgur\.com/a/(?P<id>[a-zA-Z0-9]+)'
-
+    _GALLERY = False
     _TESTS = [{
+        # TODO: only static images - replace with animated/video gallery
         'url': 'http://imgur.com/a/j6Orj',
+        'only_matching': True,
+    },
+        # from PR #21693
+        {
+        'url': 'https://imgur.com/a/iX265HX',
         'info_dict': {
-            'id': 'j6Orj',
-            'title': 'A Literary Analysis of "Star Wars: The Force Awakens"',
+            'id': 'iX265HX',
+            'title': 'enen-no-shouboutai'
         },
-        'playlist_count': 12,
+        'playlist_count': 2,
+    }, {
+        'url': 'https://imgur.com/a/8pih2Ed',
+        'info_dict': {
+            'id': '8pih2Ed'
+        },
+        'playlist_mincount': 1,
     }]

youtube_dl/extractor/mixcloud.py

@@ -1,3 +1,4 @@
+# coding: utf-8
 from __future__ import unicode_literals
 
 import itertools
@@ -10,7 +11,7 @@ from ..compat import (
     compat_ord,
     compat_str,
     compat_urllib_parse_unquote,
-    compat_zip
+    compat_zip as zip,
 )
 from ..utils import (
     int_or_none,
@@ -24,7 +25,7 @@ class MixcloudBaseIE(InfoExtractor):
     def _call_api(self, object_type, object_fields, display_id, username, slug=None):
         lookup_key = object_type + 'Lookup'
         return self._download_json(
-            'https://www.mixcloud.com/graphql', display_id, query={
+            'https://app.mixcloud.com/graphql', display_id, query={
                 'query': '''{
   %s(lookup: {username: "%s"%s}) {
     %s
@@ -44,7 +45,7 @@ class MixcloudIE(MixcloudBaseIE):
             'ext': 'm4a',
             'title': 'Cryptkeeper',
             'description': 'After quite a long silence from myself, finally another Drum\'n\'Bass mix with my favourite current dance floor bangers.',
-            'uploader': 'Daniel Holbach',
+            'uploader': 'dholbach',  # was: 'Daniel Holbach',
             'uploader_id': 'dholbach',
             'thumbnail': r're:https?://.*\.jpg',
             'view_count': int,
@@ -57,7 +58,7 @@ class MixcloudIE(MixcloudBaseIE):
             'id': 'gillespeterson_caribou-7-inch-vinyl-mix-chat',
             'ext': 'mp3',
             'title': 'Caribou 7 inch Vinyl Mix & Chat',
-            'description': 'md5:2b8aec6adce69f9d41724647c65875e8',
+            'description': r're:Last week Dan Snaith aka Caribou swung by the Brownswood.{136}',
             'uploader': 'Gilles Peterson Worldwide',
             'uploader_id': 'gillespeterson',
             'thumbnail': 're:https?://.*',
@@ -65,6 +66,23 @@ class MixcloudIE(MixcloudBaseIE):
             'timestamp': 1422987057,
             'upload_date': '20150203',
         },
+        'params': {
+            'skip_download': '404 not found',
+        },
+    }, {
+        'url': 'https://www.mixcloud.com/gillespeterson/carnival-m%C3%BAsica-popular-brasileira-mix/',
+        'info_dict': {
+            'id': 'gillespeterson_carnival-música-popular-brasileira-mix',
+            'ext': 'm4a',
+            'title': 'Carnival Música Popular Brasileira Mix',
+            'description': r're:Gilles was recently in Brazil to play at Boiler Room.{208}',
+            'timestamp': 1454347174,
+            'upload_date': '20160201',
+            'uploader': 'Gilles Peterson Worldwide',
+            'uploader_id': 'gillespeterson',
+            'thumbnail': 're:https?://.*',
+            'view_count': int,
+        },
     }, {
         'url': 'https://beta.mixcloud.com/RedLightRadio/nosedrip-15-red-light-radio-01-18-2016/',
         'only_matching': True,
@@ -76,10 +94,10 @@ class MixcloudIE(MixcloudBaseIE):
         """Encrypt/Decrypt XOR cipher. Both ways are possible because it's XOR."""
         return ''.join([
             compat_chr(compat_ord(ch) ^ compat_ord(k))
-            for ch, k in compat_zip(ciphertext, itertools.cycle(key))])
+            for ch, k in zip(ciphertext, itertools.cycle(key))])
 
     def _real_extract(self, url):
-        username, slug = re.match(self._VALID_URL, url).groups()
+        username, slug = self._match_valid_url(url).groups()
         username, slug = compat_urllib_parse_unquote(username), compat_urllib_parse_unquote(slug)
         track_id = '%s_%s' % (username, slug)
 

youtube_dl/extractor/openload.py

@@ -7,6 +7,7 @@ import subprocess
 import tempfile
 
 from ..compat import (
+    compat_open as open,
     compat_urlparse,
     compat_kwargs,
 )

youtube_dl/extractor/palcomp3.py

@@ -8,7 +8,7 @@ from ..compat import compat_str
 from ..utils import (
     int_or_none,
     str_or_none,
-    try_get,
+    traverse_obj,
 )
 
 
@@ -109,7 +109,7 @@ class PalcoMP3ArtistIE(PalcoMP3BaseIE):
     }
     name'''
 
-    @ classmethod
+    @classmethod
     def suitable(cls, url):
         return False if re.match(PalcoMP3IE._VALID_URL, url) else super(PalcoMP3ArtistIE, cls).suitable(url)
 
@@ -118,7 +118,8 @@ class PalcoMP3ArtistIE(PalcoMP3BaseIE):
         artist = self._call_api(artist_slug, self._ARTIST_FIELDS_TMPL)['artist']
 
         def entries():
-            for music in (try_get(artist, lambda x: x['musics']['nodes'], list) or []):
+            for music in traverse_obj(artist, (
+                    'musics', 'nodes', lambda _, m: m['musicID'])):
                 yield self._parse_music(music)
 
         return self.playlist_result(
@@ -137,7 +138,7 @@ class PalcoMP3VideoIE(PalcoMP3BaseIE):
             'title': 'Maiara e Maraisa - Você Faz Falta Aqui - DVD Ao Vivo Em Campo Grande',
             'description': 'md5:7043342c09a224598e93546e98e49282',
             'upload_date': '20161107',
-            'uploader_id': 'maiaramaraisaoficial',
+            'uploader_id': '@maiaramaraisaoficial',
             'uploader': 'Maiara e Maraisa',
         }
     }]

youtube_dl/extractor/s4c.py

@@ -0,0 +1,124 @@
+# coding: utf-8
+
+from __future__ import unicode_literals
+
+from functools import partial as partial_f
+
+from .common import InfoExtractor
+from ..utils import (
+    float_or_none,
+    merge_dicts,
+    T,
+    traverse_obj,
+    txt_or_none,
+    url_or_none,
+)
+
+
+class S4CIE(InfoExtractor):
+    _VALID_URL = r'https?://(?:www\.)?s4c\.cymru/clic/programme/(?P<id>\d+)'
+    _TESTS = [{
+        'url': 'https://www.s4c.cymru/clic/programme/861362209',
+        'info_dict': {
+            'id': '861362209',
+            'ext': 'mp4',
+            'title': 'Y Swn',
+            'description': 'md5:f7681a30e4955b250b3224aa9fe70cf0',
+            'duration': 5340,
+            'thumbnail': 'https://www.s4c.cymru/amg/1920x1080/Y_Swn_2023S4C_099_ii.jpg',
+        },
+    }, {
+        'url': 'https://www.s4c.cymru/clic/programme/856636948',
+        'info_dict': {
+            'id': '856636948',
+            'ext': 'mp4',
+            'title': 'Am Dro',
+            'duration': 2880,
+            'description': 'md5:100d8686fc9a632a0cb2db52a3433ffe',
+            'thumbnail': 'https://www.s4c.cymru/amg/1920x1080/Am_Dro_2022-23S4C_P6_4005.jpg',
+        },
+    }]
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+        details = self._download_json(
+            'https://www.s4c.cymru/df/full_prog_details',
+            video_id, query={
+                'lang': 'e',
+                'programme_id': video_id,
+            }, fatal=False)
+
+        player_config = self._download_json(
+            'https://player-api.s4c-cdn.co.uk/player-configuration/prod', video_id, query={
+                'programme_id': video_id,
+                'signed': '0',
+                'lang': 'en',
+                'mode': 'od',
+                'appId': 'clic',
+                'streamName': '',
+            }, note='Downloading player config JSON')
+
+        m3u8_url = self._download_json(
+            'https://player-api.s4c-cdn.co.uk/streaming-urls/prod', video_id, query={
+                'mode': 'od',
+                'application': 'clic',
+                'region': 'WW',
+                'extra': 'false',
+                'thirdParty': 'false',
+                'filename': player_config['filename'],
+            }, note='Downloading streaming urls JSON')['hls']
+        formats = self._extract_m3u8_formats(m3u8_url, video_id, 'mp4', m3u8_id='hls', entry_protocol='m3u8_native')
+        self._sort_formats(formats)
+
+        subtitles = {}
+        for sub in traverse_obj(player_config, ('subtitles', lambda _, v: url_or_none(v['0']))):
+            subtitles.setdefault(sub.get('3', 'en'), []).append({
+                'url': sub['0'],
+                'name': sub.get('1'),
+            })
+
+        return merge_dicts({
+            'id': video_id,
+            'formats': formats,
+            'subtitles': subtitles,
+            'thumbnail': url_or_none(player_config.get('poster')),
+        }, traverse_obj(details, ('full_prog_details', 0, {
+            'title': (('programme_title', 'series_title'), T(txt_or_none)),
+            'description': ('full_billing', T(txt_or_none)),
+            'duration': ('duration', T(partial_f(float_or_none, invscale=60))),
+        }), get_all=False),
+            rev=True)
+
+
+class S4CSeriesIE(InfoExtractor):
+    _VALID_URL = r'https?://(?:www\.)?s4c\.cymru/clic/series/(?P<id>\d+)'
+    _TESTS = [{
+        'url': 'https://www.s4c.cymru/clic/series/864982911',
+        'playlist_mincount': 6,
+        'info_dict': {
+            'id': '864982911',
+            'title': 'Iaith ar Daith',
+        },
+    }, {
+        'url': 'https://www.s4c.cymru/clic/series/866852587',
+        'playlist_mincount': 8,
+        'info_dict': {
+            'id': '866852587',
+            'title': 'FFIT Cymru',
+        },
+    }]
+
+    def _real_extract(self, url):
+        series_id = self._match_id(url)
+        series_details = self._download_json(
+            'https://www.s4c.cymru/df/series_details', series_id, query={
+                'lang': 'e',
+                'series_id': series_id,
+                'show_prog_in_series': 'Y'
+            }, note='Downloading series details JSON')
+
+        return self.playlist_result(
+            (self.url_result('https://www.s4c.cymru/clic/programme/' + episode_id, S4CIE, episode_id)
+             for episode_id in traverse_obj(series_details, ('other_progs_in_series', Ellipsis, 'id'))),
+            playlist_id=series_id, playlist_title=traverse_obj(
+                series_details, ('full_prog_details', 0, 'series_title', T(txt_or_none))))

youtube_dl/extractor/soundcloud.py

@@ -233,8 +233,8 @@ class SoundcloudIE(InfoExtractor):
                 'id': '583011102',
                 'ext': 'mp3',
                 'title': 'Mezzo Valzer',
-                'description': 'md5:4138d582f81866a530317bae316e8b61',
-                'uploader': 'Micronie',
+                'description': 'md5:8de664f12895716c8ac6f1da6348eb8e',
+                'uploader': 'Giovanni Sarani',
                 'uploader_id': '3352531',
                 'timestamp': 1551394171,
                 'upload_date': '20190228',
@@ -293,6 +293,11 @@ class SoundcloudIE(InfoExtractor):
         if non_fatal:
             del kwargs['fatal']
         query = kwargs.get('query', {}).copy()
+        oauth_token = self._get_cookies(self._BASE_URL).get('oauth_token')
+        if oauth_token:
+            if not kwargs.get('headers'):
+                kwargs['headers'] = {}
+            kwargs['headers']['authorization'] = 'OAuth ' + oauth_token.value
         for _ in range(2):
             query['client_id'] = self._CLIENT_ID
             kwargs['query'] = query
@@ -319,6 +324,7 @@ class SoundcloudIE(InfoExtractor):
         track_id = compat_str(info['id'])
         title = info['title']
 
+        duration = info.get('duration')
         format_urls = set()
         formats = []
         query = {'client_id': self._CLIENT_ID}
@@ -335,12 +341,14 @@ class SoundcloudIE(InfoExtractor):
                 if urlh:
                     format_url = urlh.geturl()
                     format_urls.add(format_url)
+                    filesize = int_or_none(urlh.headers.get('Content-Length'))
                     formats.append({
                         'format_id': 'download',
                         'ext': urlhandle_detect_ext(urlh) or 'mp3',
-                        'filesize': int_or_none(urlh.headers.get('Content-Length')),
+                        'filesize': filesize,
                         'url': format_url,
                         'preference': 10,
+                        'abr': (filesize and duration) and (filesize / duration * 8),
                     })
 
         def invalid_url(url):

youtube_dl/extractor/telewebion.py

@@ -3,17 +3,23 @@ from __future__ import unicode_literals
 
 from .common import InfoExtractor
 
+from ..utils import (
+    float_or_none,
+    int_or_none,
+    url_or_none,
+)
+
 
 class TelewebionIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:www\.)?telewebion\.com/#!/episode/(?P<id>\d+)'
+    _VALID_URL = r'https?://(?:www\.)?telewebion\.com/(episode|clip)/(?P<id>[a-zA-Z0-9]+)'
 
     _TEST = {
-        'url': 'http://www.telewebion.com/#!/episode/1263668/',
+        'url': 'http://www.telewebion.com/episode/0x1b3139c/',
         'info_dict': {
-            'id': '1263668',
+            'id': '0x1b3139c',
             'ext': 'mp4',
             'title': 'قرعه\u200cکشی لیگ قهرمانان اروپا',
-            'thumbnail': r're:^https?://.*\.jpg',
+            'thumbnail': r're:^https?://static\.telewebion\.com/episodeImages/.*/default',
             'view_count': int,
         },
         'params': {
@@ -25,31 +31,24 @@ class TelewebionIE(InfoExtractor):
     def _real_extract(self, url):
         video_id = self._match_id(url)
 
-        secure_token = self._download_webpage(
-            'http://m.s2.telewebion.com/op/op?action=getSecurityToken', video_id)
-        episode_details = self._download_json(
-            'http://m.s2.telewebion.com/op/op', video_id,
-            query={'action': 'getEpisodeDetails', 'episode_id': video_id})
+        episode_details = self._download_json('https://gateway.telewebion.ir/kandoo/episode/getEpisodeDetail/?EpisodeId={0}'.format(video_id), video_id)
+        episode_details = episode_details['body']['queryEpisode'][0]
 
-        m3u8_url = 'http://m.s1.telewebion.com/smil/%s.m3u8?filepath=%s&m3u8=1&secure_token=%s' % (
-            video_id, episode_details['file_path'], secure_token)
+        channel_id = episode_details['channel']['descriptor']
+        episode_image_id = episode_details.get('image')
+        episode_image = 'https://static.telewebion.com/episodeImages/{0}/default'.format(episode_image_id) if episode_image_id else None
+
+        m3u8_url = 'https://cdna.telewebion.com/{0}/episode/{1}/playlist.m3u8'.format(channel_id, video_id)
         formats = self._extract_m3u8_formats(
-            m3u8_url, video_id, ext='mp4', m3u8_id='hls')
-
-        picture_paths = [
-            episode_details.get('picture_path'),
-            episode_details.get('large_picture_path'),
-        ]
-
-        thumbnails = [{
-            'url': picture_path,
-            'preference': idx,
-        } for idx, picture_path in enumerate(picture_paths) if picture_path is not None]
+            m3u8_url, video_id, ext='mp4', m3u8_id='hls',
+            entry_protocol='m3u8_native')
+        self._sort_formats(formats)
 
         return {
             'id': video_id,
             'title': episode_details['title'],
             'formats': formats,
-            'thumbnails': thumbnails,
-            'view_count': episode_details.get('view_count'),
+            'thumbnail': url_or_none(episode_image),
+            'view_count': int_or_none(episode_details.get('view_count')),
+            'duration': float_or_none(episode_details.get('duration')),
         }

youtube_dl/extractor/vbox7.py

@@ -2,9 +2,22 @@
 from __future__ import unicode_literals
 
 import re
+import time
 
 from .common import InfoExtractor
-from ..utils import ExtractorError
+from ..compat import compat_kwargs
+from ..utils import (
+    base_url,
+    determine_ext,
+    ExtractorError,
+    float_or_none,
+    merge_dicts,
+    T,
+    traverse_obj,
+    txt_or_none,
+    url_basename,
+    url_or_none,
+)
 
 
 class Vbox7IE(InfoExtractor):
@@ -20,23 +33,27 @@ class Vbox7IE(InfoExtractor):
                         )
                         (?P<id>[\da-fA-F]+)
                     '''
+    _EMBED_REGEX = [r'<iframe[^>]+src=(?P<q>["\'])(?P<url>(?:https?:)?//vbox7\.com/emb/external\.php.+?)(?P=q)']
     _GEO_COUNTRIES = ['BG']
     _TESTS = [{
-        'url': 'http://vbox7.com/play:0946fff23c',
-        'md5': 'a60f9ab3a3a2f013ef9a967d5f7be5bf',
+        # the http: URL just redirects here
+        'url': 'https://vbox7.com/play:0946fff23c',
+        'md5': '50ca1f78345a9c15391af47d8062d074',
         'info_dict': {
             'id': '0946fff23c',
             'ext': 'mp4',
             'title': 'Борисов: Притеснен съм за бъдещето на България',
             'description': 'По думите му е опасно страната ни да бъде обявена за "сигурна"',
-            'thumbnail': r're:^https?://.*\.jpg$',
             'timestamp': 1470982814,
             'upload_date': '20160812',
             'uploader': 'zdraveibulgaria',
+            'thumbnail': r're:^https?://.*\.jpg$',
+            'view_count': int,
+            'duration': 2640,
         },
-        'params': {
-            'proxy': '127.0.0.1:8118',
-        },
+        'expected_warnings': [
+            'Unable to download webpage',
+        ],
     }, {
         'url': 'http://vbox7.com/play:249bb972c2',
         'md5': '99f65c0c9ef9b682b97313e052734c3f',
@@ -44,8 +61,15 @@ class Vbox7IE(InfoExtractor):
             'id': '249bb972c2',
             'ext': 'mp4',
             'title': 'Смях! Чудо - чист за секунди - Скрита камера',
+            'description': 'Смях! Чудо - чист за секунди - Скрита камера',
+            'timestamp': 1360215023,
+            'upload_date': '20130207',
+            'uploader': 'svideteliat_ot_varshava',
+            'thumbnail': 'https://i49.vbox7.com/o/249/249bb972c20.jpg',
+            'view_count': int,
+            'duration': 83,
         },
-        'skip': 'georestricted',
+        'expected_warnings': ['Failed to download m3u8 information'],
     }, {
         'url': 'http://vbox7.com/emb/external.php?vid=a240d20f9c&autoplay=1',
         'only_matching': True,
@@ -54,52 +78,127 @@ class Vbox7IE(InfoExtractor):
         'only_matching': True,
     }]
 
-    @staticmethod
-    def _extract_url(webpage):
-        mobj = re.search(
-            r'<iframe[^>]+src=(?P<q>["\'])(?P<url>(?:https?:)?//vbox7\.com/emb/external\.php.+?)(?P=q)',
-            webpage)
+    @classmethod
+    def _extract_url(cls, webpage):
+        mobj = re.search(cls._EMBED_REGEX[0], webpage)
         if mobj:
             return mobj.group('url')
 
+    # specialisation to transform what looks like ld+json that
+    # may contain invalid character combinations
+
+    # transform_source=None, fatal=True
+    def _parse_json(self, json_string, video_id, *args, **kwargs):
+        if '"@context"' in json_string[:30]:
+            # this is ld+json, or that's the way to bet
+            transform_source = args[0] if len(args) > 0 else kwargs.get('transform_source')
+            if not transform_source:
+
+                def fix_chars(src):
+                    # fix malformed ld+json: replace raw CRLFs with escaped LFs
+                    return re.sub(
+                        r'"[^"]+"', lambda m: re.sub(r'\r?\n', r'\\n', m.group(0)), src)
+
+                if len(args) > 0:
+                    args = (fix_chars,) + args[1:]
+                else:
+                    kwargs['transform_source'] = fix_chars
+                    kwargs = compat_kwargs(kwargs)
+
+        return super(Vbox7IE, self)._parse_json(
+            json_string, video_id, *args, **kwargs)
+
     def _real_extract(self, url):
         video_id = self._match_id(url)
+        url = 'https://vbox7.com/play:%s' % (video_id,)
 
+        now = time.time()
         response = self._download_json(
-            'https://www.vbox7.com/ajax/video/nextvideo.php?vid=%s' % video_id,
-            video_id)
+            'https://www.vbox7.com/aj/player/item/options', video_id,
+            query={'vid': video_id}, headers={'Referer': url})
+        # estimate time to which possible `ago` member is relative
+        now = now + 0.5 * (time.time() - now)
 
-        if 'error' in response:
+        if traverse_obj(response, 'error'):
             raise ExtractorError(
                 '%s said: %s' % (self.IE_NAME, response['error']), expected=True)
 
-        video = response['options']
+        src_url = traverse_obj(response, ('options', 'src', T(url_or_none))) or ''
 
-        title = video['title']
-        video_url = video['src']
-
-        if '/na.mp4' in video_url:
+        fmt_base = url_basename(src_url).rsplit('.', 1)[0].rsplit('_', 1)[0]
+        if fmt_base in ('na', 'vn'):
             self.raise_geo_restricted(countries=self._GEO_COUNTRIES)
 
-        uploader = video.get('uploader')
+        ext = determine_ext(src_url)
+        if ext == 'mpd':
+            # extract MPD
+            try:
+                formats, subtitles = self._extract_mpd_formats_and_subtitles(
+                    src_url, video_id, 'dash', fatal=False)
+            except KeyError:  # fatal doesn't catch this
+                self.report_warning('Failed to parse MPD manifest')
+                formats, subtitles = [], {}
+        elif ext != 'm3u8':
+            formats = [{
+                'url': src_url,
+            }] if src_url else []
+            subtitles = {}
 
-        webpage = self._download_webpage(
-            'http://vbox7.com/play:%s' % video_id, video_id, fatal=None)
+        if src_url:
+            # possibly extract HLS, based on https://github.com/yt-dlp/yt-dlp/pull/9100
+            fmt_base = base_url(src_url) + fmt_base
+            # prepare for _extract_m3u8_formats_and_subtitles()
+            # hls_formats, hls_subs = self._extract_m3u8_formats_and_subtitles(
+            hls_formats = self._extract_m3u8_formats(
+                '{0}.m3u8'.format(fmt_base), video_id, m3u8_id='hls', fatal=False)
+            formats.extend(hls_formats)
+            # self._merge_subtitles(hls_subs, target=subtitles)
 
-        info = {}
+            # In case MPD/HLS cannot be parsed, or anyway, get mp4 combined
+            # formats usually provided to Safari, iOS, and old Windows
+            video = response['options']
+            resolutions = (1080, 720, 480, 240, 144)
+            highest_res = traverse_obj(video, (
+                'highestRes', T(int))) or resolutions[0]
+            resolutions = traverse_obj(video, (
+                'resolutions', lambda _, r: highest_res >= int(r) > 0)) or resolutions
+            mp4_formats = traverse_obj(resolutions, (
+                Ellipsis, T(lambda res: {
+                    'url': '{0}_{1}.mp4'.format(fmt_base, res),
+                    'format_id': 'http-{0}'.format(res),
+                    'height': res,
+                })))
+            # if above formats are flaky, enable the line below
+            # self._check_formats(mp4_formats, video_id)
+            formats.extend(mp4_formats)
 
-        if webpage:
-            info = self._search_json_ld(
-                webpage.replace('"/*@context"', '"@context"'), video_id,
-                fatal=False)
+        self._sort_formats(formats)
 
-        info.update({
+        webpage = self._download_webpage(url, video_id, fatal=False) or ''
+
+        info = self._search_json_ld(
+            webpage.replace('"/*@context"', '"@context"'), video_id,
+            fatal=False) if webpage else {}
+
+        if not info.get('title'):
+            info['title'] = traverse_obj(response, (
+                'options', 'title', T(txt_or_none))) or self._og_search_title(webpage)
+
+        def if_missing(k):
+            return lambda x: None if k in info else x
+
+        info = merge_dicts(info, {
             'id': video_id,
-            'title': title,
-            'url': video_url,
-            'uploader': uploader,
-            'thumbnail': self._proto_relative_url(
+            'formats': formats,
+            'subtitles': subtitles or None,
+        }, info, traverse_obj(response, ('options', {
+            'uploader': ('uploader', T(txt_or_none)),
+            'timestamp': ('ago', T(if_missing('timestamp')), T(lambda t: int(round((now - t) / 60.0)) * 60)),
+            'duration': ('duration', T(if_missing('duration')), T(float_or_none)),
+        })))
+        if 'thumbnail' not in info:
+            info['thumbnail'] = self._proto_relative_url(
                 info.get('thumbnail') or self._og_search_thumbnail(webpage),
-                'http:'),
-        })
+                'https:'),
+
         return info

youtube_dl/extractor/videa.py

@@ -6,22 +6,31 @@ import re
 import string
 
 from .common import InfoExtractor
+from ..compat import (
+    compat_b64decode,
+    compat_ord,
+    compat_struct_pack,
+)
 from ..utils import (
     ExtractorError,
     int_or_none,
     mimetype2ext,
     parse_codecs,
+    parse_qs,
     update_url_query,
     urljoin,
     xpath_element,
     xpath_text,
 )
-from ..compat import (
-    compat_b64decode,
-    compat_ord,
-    compat_struct_pack,
-    compat_urlparse,
-)
+
+
+def compat_random_choices(population, *args, **kwargs):
+    # weights=None, *, cum_weights=None, k=1
+    # limited implementation needed here
+    weights = args[0] if args else kwargs.get('weights')
+    assert all(w is None for w in (weights, kwargs.get('cum_weights')))
+    k = kwargs.get('k', 1)
+    return ''.join(random.choice(population) for _ in range(k))
 
 
 class VideaIE(InfoExtractor):
@@ -35,6 +44,7 @@ class VideaIE(InfoExtractor):
                         )
                         (?P<id>[^?#&]+)
                     '''
+    _EMBED_REGEX = [r'<iframe[^>]+src=(["\'])(?P<url>(?:https?:)?//videa\.hu/player\?.*?\bv=.+?)\1']
     _TESTS = [{
         'url': 'http://videa.hu/videok/allatok/az-orult-kigyasz-285-kigyot-kigyo-8YfIAjxwWGwT8HVQ',
         'md5': '97a7af41faeaffd9f1fc864a7c7e7603',
@@ -44,6 +54,7 @@ class VideaIE(InfoExtractor):
             'title': 'Az őrült kígyász 285 kígyót enged szabadon',
             'thumbnail': r're:^https?://.*',
             'duration': 21,
+            'age_limit': 0,
         },
     }, {
         'url': 'http://videa.hu/videok/origo/jarmuvek/supercars-elozes-jAHDWfWSJH5XuFhH',
@@ -54,6 +65,7 @@ class VideaIE(InfoExtractor):
             'title': 'Supercars előzés',
             'thumbnail': r're:^https?://.*',
             'duration': 64,
+            'age_limit': 0,
         },
     }, {
         'url': 'http://videa.hu/player?v=8YfIAjxwWGwT8HVQ',
@@ -64,6 +76,7 @@ class VideaIE(InfoExtractor):
             'title': 'Az őrült kígyász 285 kígyót enged szabadon',
             'thumbnail': r're:^https?://.*',
             'duration': 21,
+            'age_limit': 0,
         },
     }, {
         'url': 'http://videa.hu/player/v/8YfIAjxwWGwT8HVQ?autoplay=1',
@@ -80,11 +93,14 @@ class VideaIE(InfoExtractor):
     }]
     _STATIC_SECRET = 'xHb0ZvME5q8CBcoQi6AngerDu3FGO9fkUlwPmLVY_RTzj2hJIS4NasXWKy1td7p'
 
-    @staticmethod
-    def _extract_urls(webpage):
-        return [url for _, url in re.findall(
-            r'<iframe[^>]+src=(["\'])(?P<url>(?:https?:)?//videa\.hu/player\?.*?\bv=.+?)\1',
-            webpage)]
+    @classmethod
+    def _extract_urls(cls, webpage):
+        def yield_urls():
+            for pattern in cls._EMBED_REGEX:
+                for m in re.finditer(pattern, webpage):
+                    yield m.group('url')
+
+        return list(yield_urls())
 
     @staticmethod
     def rc4(cipher_text, key):
@@ -130,13 +146,13 @@ class VideaIE(InfoExtractor):
         for i in range(0, 32):
             result += s[i - (self._STATIC_SECRET.index(l[i]) - 31)]
 
-        query = compat_urlparse.parse_qs(compat_urlparse.urlparse(player_url).query)
-        random_seed = ''.join(random.choice(string.ascii_letters + string.digits) for _ in range(8))
+        query = parse_qs(player_url)
+        random_seed = ''.join(compat_random_choices(string.ascii_letters + string.digits, k=8))
         query['_s'] = random_seed
         query['_t'] = result[:16]
 
         b64_info, handle = self._download_webpage_handle(
-            'http://videa.hu/videaplayer_get_xml.php', video_id, query=query)
+            'http://videa.hu/player/xml', video_id, query=query)
         if b64_info.startswith('<?xml'):
             info = self._parse_xml(b64_info, video_id)
         else:

youtube_dl/extractor/vidlii.py

@@ -4,6 +4,7 @@ from __future__ import unicode_literals
 import re
 
 from .common import InfoExtractor
+
 from ..utils import (
     float_or_none,
     get_element_by_id,
@@ -11,6 +12,7 @@ from ..utils import (
     strip_or_none,
     unified_strdate,
     urljoin,
+    str_to_int,
 )
 
 
@@ -35,6 +37,26 @@ class VidLiiIE(InfoExtractor):
             'categories': ['News & Politics'],
             'tags': ['Vidlii', 'Jan', 'Videogames'],
         }
+    }, {
+        # HD
+        'url': 'https://www.vidlii.com/watch?v=2Ng8Abj2Fkl',
+        'md5': '450e7da379c884788c3a4fa02a3ce1a4',
+        'info_dict': {
+            'id': '2Ng8Abj2Fkl',
+            'ext': 'mp4',
+            'title': 'test',
+            'description': 'md5:cc55a86032a7b6b3cbfd0f6b155b52e9',
+            'thumbnail': 'https://www.vidlii.com/usfi/thmp/2Ng8Abj2Fkl.jpg',
+            'uploader': 'VidLii',
+            'uploader_url': 'https://www.vidlii.com/user/VidLii',
+            'upload_date': '20200927',
+            'duration': 5,
+            'view_count': int,
+            'comment_count': int,
+            'average_rating': float,
+            'categories': ['Film & Animation'],
+            'tags': list,
+        },
     }, {
         'url': 'https://www.vidlii.com/embed?v=tJluaH4BJ3v&a=0',
         'only_matching': True,
@@ -46,11 +68,32 @@ class VidLiiIE(InfoExtractor):
         webpage = self._download_webpage(
             'https://www.vidlii.com/watch?v=%s' % video_id, video_id)
 
-        video_url = self._search_regex(
-            r'src\s*:\s*(["\'])(?P<url>(?:https?://)?(?:(?!\1).)+)\1', webpage,
-            'video url', group='url')
+        formats = []
 
-        title = self._search_regex(
+        def add_format(format_url, height=None):
+            height = int(self._search_regex(r'(\d+)\.mp4',
+                         format_url, 'height', default=360))
+
+            formats.append({
+                'url': format_url,
+                'format_id': '%dp' % height if height else None,
+                'height': height,
+            })
+
+        sources = re.findall(
+            r'src\s*:\s*(["\'])(?P<url>(?:https?://)?(?:(?!\1).)+)\1',
+            webpage)
+
+        formats = []
+        if len(sources) > 1:
+            add_format(sources[1][1])
+            self._check_formats(formats, video_id)
+        if len(sources) > 0:
+            add_format(sources[0][1])
+
+        self._sort_formats(formats)
+
+        title = self._html_search_regex(
             (r'<h1>([^<]+)</h1>', r'<title>([^<]+) - VidLii<'), webpage,
             'title')
 
@@ -82,9 +125,9 @@ class VidLiiIE(InfoExtractor):
             default=None) or self._search_regex(
             r'duration\s*:\s*(\d+)', webpage, 'duration', fatal=False))
 
-        view_count = int_or_none(self._search_regex(
-            (r'<strong>(\d+)</strong> views',
-             r'Views\s*:\s*<strong>(\d+)</strong>'),
+        view_count = str_to_int(self._html_search_regex(
+            (r'<strong>([\d,.]+)</strong> views',
+             r'Views\s*:\s*<strong>([\d,.]+)</strong>'),
             webpage, 'view count', fatal=False))
 
         comment_count = int_or_none(self._search_regex(
@@ -109,7 +152,7 @@ class VidLiiIE(InfoExtractor):
 
         return {
             'id': video_id,
-            'url': video_url,
+            'formats': formats,
             'title': title,
             'description': description,
             'thumbnail': thumbnail,

youtube_dl/extractor/vimeo.py

@@ -673,8 +673,8 @@ class VimeoIE(VimeoBaseInfoExtractor):
             raise
 
         if '//player.vimeo.com/video/' in url:
-            config = self._parse_json(self._search_regex(
-                r'(?s)\b(?:playerC|c)onfig\s*=\s*({.+?})\s*[;\n]', webpage, 'info section'), video_id)
+            config = self._search_json(
+                r'\b(?:playerC|c)onfig\s*=', webpage, 'info section', video_id)
             if config.get('view') == 4:
                 config = self._verify_player_video_password(
                     redirect_url, video_id, headers)

youtube_dl/extractor/whyp.py

@@ -0,0 +1,55 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+from .common import InfoExtractor
+from ..utils import (
+    float_or_none,
+    merge_dicts,
+    str_or_none,
+    T,
+    traverse_obj,
+    url_or_none,
+)
+
+
+class WhypIE(InfoExtractor):
+    _VALID_URL = r'https?://(?:www\.)?whyp\.it/tracks/(?P<id>\d+)'
+    _TESTS = [{
+        'url': 'https://www.whyp.it/tracks/18337/home-page-example-track-b4kq7',
+        'md5': 'c1187b42ebf8605284e3dc92aeb33d16',
+        'info_dict': {
+            'url': 'https://cdn.whyp.it/50eb17cc-e9ff-4e18-b89b-dc9206a95cb1.mp3',
+            'id': '18337',
+            'title': 'Home Page Example Track',
+            'description': r're:(?s).+\bexample track\b',
+            'ext': 'mp3',
+            'duration': 52.82,
+            'uploader': 'Brad',
+            'uploader_id': '1',
+            'thumbnail': 'https://cdn.whyp.it/a537bb36-3373-4c61-96c8-27fc1b2f427a.jpg',
+        },
+    }, {
+        'url': 'https://www.whyp.it/tracks/18337',
+        'only_matching': True,
+    }]
+
+    def _real_extract(self, url):
+        unique_id = self._match_id(url)
+        webpage = self._download_webpage(url, unique_id)
+        data = self._search_nuxt_data(webpage, unique_id)['rawTrack']
+
+        return merge_dicts({
+            'url': data['audio_url'],
+            'id': unique_id,
+        }, traverse_obj(data, {
+            'title': 'title',
+            'description': 'description',
+            'duration': ('duration', T(float_or_none)),
+            'uploader': ('user', 'username'),
+            'uploader_id': ('user', 'id', T(str_or_none)),
+            'thumbnail': ('artwork_url', T(url_or_none)),
+        }), {
+            'ext': 'mp3',
+            'vcodec': 'none',
+            'http_headers': {'Referer': 'https://whyp.it/'},
+        }, rev=True)

youtube_dl/extractor/xfileshare.py

@@ -4,20 +4,28 @@ from __future__ import unicode_literals
 import re
 
 from .common import InfoExtractor
-from ..compat import compat_chr
+from ..compat import (
+    compat_chr,
+    compat_zip as zip,
+)
 from ..utils import (
+    clean_html,
     decode_packed_codes,
     determine_ext,
     ExtractorError,
+    get_element_by_id,
     int_or_none,
-    js_to_json,
+    merge_dicts,
+    T,
+    traverse_obj,
+    url_or_none,
     urlencode_postdata,
 )
 
 
 # based on openload_decode from 2bfeee69b976fe049761dd3012e30b637ee05a58
 def aa_decode(aa_code):
-    symbol_table = [
+    symbol_table = (
         ('7', '((ﾟｰﾟ) + (o^_^o))'),
         ('6', '((o^_^o) +(o^_^o))'),
         ('5', '((ﾟｰﾟ) + (ﾟΘﾟ))'),
@@ -26,84 +34,180 @@ def aa_decode(aa_code):
         ('3', '(o^_^o)'),
         ('1', '(ﾟΘﾟ)'),
         ('0', '(c^_^o)'),
-    ]
+        ('+', ''),
+    )
     delim = '(ﾟДﾟ)[ﾟεﾟ]+'
-    ret = ''
-    for aa_char in aa_code.split(delim):
+
+    def chr_from_code(c):
         for val, pat in symbol_table:
-            aa_char = aa_char.replace(pat, val)
-        aa_char = aa_char.replace('+ ', '')
-        m = re.match(r'^\d+', aa_char)
-        if m:
-            ret += compat_chr(int(m.group(0), 8))
+            c = c.replace(pat, val)
+        if c.startswith(('u', 'U')):
+            base = 16
+            c = c[1:]
         else:
-            m = re.match(r'^u([\da-f]+)', aa_char)
-            if m:
-                ret += compat_chr(int(m.group(1), 16))
-    return ret
+            base = 10
+        c = int_or_none(c, base=base)
+        return '' if c is None else compat_chr(c)
+
+    return ''.join(
+        chr_from_code(aa_char)
+        for aa_char in aa_code.split(delim))
 
 
 class XFileShareIE(InfoExtractor):
     _SITES = (
-        (r'aparat\.cam', 'Aparat'),
-        (r'clipwatching\.com', 'ClipWatching'),
-        (r'gounlimited\.to', 'GoUnlimited'),
-        (r'govid\.me', 'GoVid'),
-        (r'holavid\.com', 'HolaVid'),
-        (r'streamty\.com', 'Streamty'),
-        (r'thevideobee\.to', 'TheVideoBee'),
-        (r'uqload\.com', 'Uqload'),
-        (r'vidbom\.com', 'VidBom'),
-        (r'vidlo\.us', 'vidlo'),
-        (r'vidlocker\.xyz', 'VidLocker'),
-        (r'vidshare\.tv', 'VidShare'),
-        (r'vup\.to', 'VUp'),
+        # status check 2024-02: site availability, G site: search
+        (r'aparat\.cam', 'Aparat'),  # Cloudflare says host error 522, apparently changed to wolfstreeam.tv
+        (r'filemoon\.sx/.', 'FileMoon'),
+        (r'gounlimited\.to', 'GoUnlimited'),  # no media pages listed
+        (r'govid\.me', 'GoVid'),  # no media pages listed
+        (r'highstream\.tv', 'HighStream'),  # clipwatching.com redirects here
+        (r'holavid\.com', 'HolaVid'),  # Cloudflare says host error 522
+        # (r'streamty\.com', 'Streamty'),  # no media pages listed, connection timeout
+        # (r'thevideobee\.to', 'TheVideoBee'),  # no pages listed, refuses connection
+        (r'uqload\.to', 'Uqload'),  # .com, .co redirect here
+        (r'(?:vedbam\.xyz|vadbam.net)', 'V?dB?m'),  # vidbom.com redirects here, but no valid media pages listed
+        (r'vidlo\.us', 'vidlo'),  # no valid media pages listed
+        (r'vidlocker\.xyz', 'VidLocker'),  # no media pages listed
+        (r'(?:w\d\.)?viidshar\.com', 'VidShare'),  # vidshare.tv redirects here
+        # (r'vup\.to', 'VUp'),  # domain not found
         (r'wolfstream\.tv', 'WolfStream'),
-        (r'xvideosharing\.com', 'XVideoSharing'),
+        (r'xvideosharing\.com', 'XVideoSharing'),  # just started showing 'maintenance mode'
     )
 
-    IE_DESC = 'XFileShare based sites: %s' % ', '.join(list(zip(*_SITES))[1])
+    IE_DESC = 'XFileShare-based sites: %s' % ', '.join(list(zip(*_SITES))[1])
     _VALID_URL = (r'https?://(?:www\.)?(?P<host>%s)/(?:embed-)?(?P<id>[0-9a-zA-Z]+)'
                   % '|'.join(site for site in list(zip(*_SITES))[0]))
+    _EMBED_REGEX = [r'<iframe\b[^>]+\bsrc=(["\'])(?P<url>(?:https?:)?//(?:%s)/embed-[0-9a-zA-Z]+.*?)\1' % '|'.join(site for site in list(zip(*_SITES))[0])]
 
     _FILE_NOT_FOUND_REGEXES = (
         r'>(?:404 - )?File Not Found<',
         r'>The file was removed by administrator<',
     )
+    _TITLE_REGEXES = (
+        r'style="z-index: [0-9]+;">([^<]+)</span>',
+        r'<td nowrap>([^<]+)</td>',
+        r'h4-fine[^>]*>([^<]+)<',
+        r'>Watch (.+)[ <]',
+        r'<h2 class="video-page-head">([^<]+)</h2>',
+        r'<h2 style="[^"]*color:#403f3d[^"]*"[^>]*>([^<]+)<',  # streamin.to (dead)
+        r'title\s*:\s*"([^"]+)"',  # govid.me
+    )
+    _SOURCE_URL_REGEXES = (
+        r'(?:file|src)\s*:\s*(["\'])(?P<url>http(?:(?!\1).)+\.(?:m3u8|mp4|flv)(?:(?!\1).)*)\1',
+        r'file_link\s*=\s*(["\'])(?P<url>http(?:(?!\1).)+)\1',
+        r'addVariable\((\\?["\'])file\1\s*,\s*(\\?["\'])(?P<url>http(?:(?!\2).)+)\2\)',
+        r'<embed[^>]+src=(["\'])(?P<url>http(?:(?!\1).)+\.(?:m3u8|mp4|flv)(?:(?!\1).)*)\1',
+    )
+    _THUMBNAIL_REGEXES = (
+        r'<video[^>]+poster="([^"]+)"',
+        r'(?:image|poster)\s*:\s*["\'](http[^"\']+)["\'],',
+    )
 
     _TESTS = [{
-        'url': 'http://xvideosharing.com/fq65f94nd2ve',
-        'md5': '4181f63957e8fe90ac836fa58dc3c8a6',
+        'note': 'link in `sources`',
+        'url': 'https://uqload.to/dcsu06gdb45o',
+        'md5': '7f8db187b254379440bf4fcad094ae86',
         'info_dict': {
-            'id': 'fq65f94nd2ve',
+            'id': 'dcsu06gdb45o',
             'ext': 'mp4',
-            'title': 'sample',
-            'thumbnail': r're:http://.*\.jpg',
+            'title': 'f2e31015957e74c8c8427982e161c3fc mp4',
+            'thumbnail': r're:https://.*\.jpg'
+        },
+        'params': {
+            'nocheckcertificate': True,
+        },
+        'expected_warnings': ['Unable to extract JWPlayer data'],
+    }, {
+        'note': 'link in decoded `sources`',
+        'url': 'https://xvideosharing.com/1tlg6agrrdgc',
+        'md5': '2608ce41932c1657ae56258a64e647d9',
+        'info_dict': {
+            'id': '1tlg6agrrdgc',
+            'ext': 'mp4',
+            'title': '0121',
+            'thumbnail': r're:https?://.*\.jpg',
+        },
+        'skip': 'This server is in maintenance mode.',
+    }, {
+        'note': 'JWPlayer link in un-p,a,c,k,e,d JS',
+        'url': 'https://filemoon.sx/e/dw40rxrzruqz',
+        'md5': '5a713742f57ac4aef29b74733e8dda01',
+        'info_dict': {
+            'id': 'dw40rxrzruqz',
+            'title': 'dw40rxrzruqz',
+            'ext': 'mp4'
+        },
+    }, {
+        'note': 'JWPlayer link in un-p,a,c,k,e,d JS',
+        'url': 'https://vadbam.net/6lnbkci96wly.html',
+        'md5': 'a1616800076177e2ac769203957c54bc',
+        'info_dict': {
+            'id': '6lnbkci96wly',
+            'title': 'Heart Crime S01 E03 weciima autos',
+            'ext': 'mp4'
+        },
+    }, {
+        'note': 'JWPlayer link in clear',
+        'url': 'https://w1.viidshar.com/nnibe0xf0h79.html',
+        'md5': 'f0a580ce9df06cc61b4a5c979d672367',
+        'info_dict': {
+            'id': 'nnibe0xf0h79',
+            'title': 'JaGa 68ar',
+            'ext': 'mp4'
+        },
+        'params': {
+            'skip_download': 'ffmpeg',
+        },
+        'expected_warnings': ['hlsnative has detected features it does not support'],
+    }, {
+        'note': 'JWPlayer link in clear',
+        'url': 'https://wolfstream.tv/a3drtehyrg52.html',
+        'md5': '1901d86a79c5e0c6a51bdc9a4cfd3769',
+        'info_dict': {
+            'id': 'a3drtehyrg52',
+            'title': 'NFL 2023 W04 DET@GB',
+            'ext': 'mp4'
         },
     }, {
         'url': 'https://aparat.cam/n4d6dh0wvlpr',
         'only_matching': True,
     }, {
-        'url': 'https://wolfstream.tv/nthme29v9u2x',
+        'url': 'https://uqload.to/ug5somm0ctnk.html',
+        'only_matching': True,
+    }, {
+        'url': 'https://highstream.tv/2owiyz3sjoux',
+        'only_matching': True,
+    }, {
+        'url': 'https://vedbam.xyz/6lnbkci96wly.html',
         'only_matching': True,
     }]
 
-    @staticmethod
-    def _extract_urls(webpage):
-        return [
-            mobj.group('url')
-            for mobj in re.finditer(
-                r'<iframe\b[^>]+\bsrc=(["\'])(?P<url>(?:https?:)?//(?:%s)/embed-[0-9a-zA-Z]+.*?)\1'
-                % '|'.join(site for site in list(zip(*XFileShareIE._SITES))[0]),
-                webpage)]
+    @classmethod
+    def _extract_urls(cls, webpage):
+
+        def yield_urls():
+            for regex in cls._EMBED_REGEX:
+                for mobj in re.finditer(regex, webpage):
+                    yield mobj.group('url')
+
+        return list(yield_urls())
 
     def _real_extract(self, url):
-        host, video_id = re.match(self._VALID_URL, url).groups()
+        host, video_id = self._match_valid_url(url).group('host', 'id')
 
-        url = 'https://%s/' % host + ('embed-%s.html' % video_id if host in ('govid.me', 'vidlo.us') else video_id)
+        url = 'https://%s/%s' % (
+            host,
+            'embed-%s.html' % video_id if host in ('govid.me', 'vidlo.us') else video_id)
         webpage = self._download_webpage(url, video_id)
-
-        if any(re.search(p, webpage) for p in self._FILE_NOT_FOUND_REGEXES):
+        container_div = get_element_by_id('container', webpage) or webpage
+        if self._search_regex(
+                r'>This server is in maintenance mode\.', container_div,
+                'maint error', group=0, default=None):
+            raise ExtractorError(clean_html(container_div), expected=True)
+        if self._search_regex(
+                self._FILE_NOT_FOUND_REGEXES, container_div,
+                'missing video error', group=0, default=None):
             raise ExtractorError('Video %s does not exist' % video_id, expected=True)
 
         fields = self._hidden_inputs(webpage)
@@ -122,59 +226,43 @@ class XFileShareIE(InfoExtractor):
                     'Content-type': 'application/x-www-form-urlencoded',
                 })
 
-        title = (self._search_regex(
-            (r'style="z-index: [0-9]+;">([^<]+)</span>',
-             r'<td nowrap>([^<]+)</td>',
-             r'h4-fine[^>]*>([^<]+)<',
-             r'>Watch (.+)[ <]',
-             r'<h2 class="video-page-head">([^<]+)</h2>',
-             r'<h2 style="[^"]*color:#403f3d[^"]*"[^>]*>([^<]+)<',  # streamin.to
-             r'title\s*:\s*"([^"]+)"'),  # govid.me
-            webpage, 'title', default=None) or self._og_search_title(
-            webpage, default=None) or video_id).strip()
+        title = (
+            self._search_regex(self._TITLE_REGEXES, webpage, 'title', default=None)
+            or self._og_search_title(webpage, default=None)
+            or video_id).strip()
 
-        for regex, func in (
-                (r'(eval\(function\(p,a,c,k,e,d\){.+)', decode_packed_codes),
-                (r'(ﾟ.+)', aa_decode)):
-            obf_code = self._search_regex(regex, webpage, 'obfuscated code', default=None)
-            if obf_code:
-                webpage = webpage.replace(obf_code, func(obf_code))
+        obf_code = True
+        while obf_code:
+            for regex, func in (
+                    (r'(?s)(?<!-)\b(eval\(function\(p,a,c,k,e,d\)\{(?:(?!</script>).)+\)\))',
+                     decode_packed_codes),
+                    (r'(ﾟ.+)', aa_decode)):
+                obf_code = self._search_regex(regex, webpage, 'obfuscated code', default=None)
+                if obf_code:
+                    webpage = webpage.replace(obf_code, func(obf_code))
+                    break
 
-        formats = []
+        jwplayer_data = self._find_jwplayer_data(
+            webpage.replace(r'\'', '\''), video_id)
+        result = self._parse_jwplayer_data(
+            jwplayer_data, video_id, require_title=False,
+            m3u8_id='hls', mpd_id='dash')
 
-        jwplayer_data = self._search_regex(
-            [
-                r'jwplayer\("[^"]+"\)\.load\(\[({.+?})\]\);',
-                r'jwplayer\("[^"]+"\)\.setup\(({.+?})\);',
-            ], webpage,
-            'jwplayer data', default=None)
-        if jwplayer_data:
-            jwplayer_data = self._parse_json(
-                jwplayer_data.replace(r"\'", "'"), video_id, js_to_json)
+        if not traverse_obj(result, 'formats'):
             if jwplayer_data:
-                formats = self._parse_jwplayer_data(
-                    jwplayer_data, video_id, False,
-                    m3u8_id='hls', mpd_id='dash')['formats']
-
-        if not formats:
-            urls = []
-            for regex in (
-                    r'(?:file|src)\s*:\s*(["\'])(?P<url>http(?:(?!\1).)+\.(?:m3u8|mp4|flv)(?:(?!\1).)*)\1',
-                    r'file_link\s*=\s*(["\'])(?P<url>http(?:(?!\1).)+)\1',
-                    r'addVariable\((\\?["\'])file\1\s*,\s*(\\?["\'])(?P<url>http(?:(?!\2).)+)\2\)',
-                    r'<embed[^>]+src=(["\'])(?P<url>http(?:(?!\1).)+\.(?:m3u8|mp4|flv)(?:(?!\1).)*)\1'):
+                self.report_warning(
+                    'Failed to extract JWPlayer formats', video_id=video_id)
+            urls = set()
+            for regex in self._SOURCE_URL_REGEXES:
                 for mobj in re.finditer(regex, webpage):
-                    video_url = mobj.group('url')
-                    if video_url not in urls:
-                        urls.append(video_url)
+                    urls.add(mobj.group('url'))
 
             sources = self._search_regex(
                 r'sources\s*:\s*(\[(?!{)[^\]]+\])', webpage, 'sources', default=None)
-            if sources:
-                urls.extend(self._parse_json(sources, video_id))
+            urls.update(traverse_obj(sources, (T(lambda s: self._parse_json(s, video_id)), Ellipsis)))
 
             formats = []
-            for video_url in urls:
+            for video_url in traverse_obj(urls, (Ellipsis, T(url_or_none))):
                 if determine_ext(video_url) == 'm3u8':
                     formats.extend(self._extract_m3u8_formats(
                         video_url, video_id, 'mp4',
@@ -185,17 +273,19 @@ class XFileShareIE(InfoExtractor):
                         'url': video_url,
                         'format_id': 'sd',
                     })
-        self._sort_formats(formats)
+            result = {'formats': formats}
+
+        self._sort_formats(result['formats'])
 
         thumbnail = self._search_regex(
-            [
-                r'<video[^>]+poster="([^"]+)"',
-                r'(?:image|poster)\s*:\s*["\'](http[^"\']+)["\'],',
-            ], webpage, 'thumbnail', default=None)
+            self._THUMBNAIL_REGEXES, webpage, 'thumbnail', default=None)
 
-        return {
+        if not (title or result.get('title')):
+            title = self._generic_title(url) or video_id
+
+        return merge_dicts(result, {
             'id': video_id,
-            'title': title,
+            'title': title or None,
             'thumbnail': thumbnail,
-            'formats': formats,
-        }
+            'http_headers': {'Referer': url}
+        })

youtube_dl/extractor/youporn.py

@@ -1,20 +1,38 @@
 # coding: utf-8
 from __future__ import unicode_literals
 
+import itertools
 import re
+from time import sleep
 
 from .common import InfoExtractor
 from ..utils import (
+    clean_html,
     extract_attributes,
+    ExtractorError,
+    get_element_by_class,
+    get_element_by_id,
     int_or_none,
-    str_to_int,
+    merge_dicts,
+    parse_count,
+    parse_qs,
+    T,
+    traverse_obj,
     unified_strdate,
     url_or_none,
+    urljoin,
 )
 
 
 class YouPornIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:www\.)?youporn\.com/(?:watch|embed)/(?P<id>\d+)(?:/(?P<display_id>[^/?#&]+))?'
+    _VALID_URL = (
+        r'youporn:(?P<id>\d+)',
+        r'''(?x)
+            https?://(?:www\.)?youporn\.com/(?:watch|embed)/(?P<id>\d+)
+            (?:/(?:(?P<display_id>[^/?#&]+)/?)?)?(?:[#?]|$)
+    '''
+    )
+    _EMBED_REGEX = [r'<iframe[^>]+\bsrc=["\'](?P<url>(?:https?:)?//(?:www\.)?youporn\.com/embed/\d+)']
     _TESTS = [{
         'url': 'http://www.youporn.com/watch/505835/sex-ed-is-it-safe-to-masturbate-daily/',
         'md5': '3744d24c50438cf5b6f6d59feb5055c2',
@@ -34,7 +52,7 @@ class YouPornIE(InfoExtractor):
             'tags': list,
             'age_limit': 18,
         },
-        'skip': 'This video has been disabled',
+        'skip': 'This video has been deactivated',
     }, {
         # Unknown uploader
         'url': 'http://www.youporn.com/watch/561726/big-tits-awesome-brunette-on-amazing-webcam-show/?from=related3&al=2&from_id=561726&pos=4',
@@ -66,57 +84,104 @@ class YouPornIE(InfoExtractor):
     }, {
         'url': 'https://www.youporn.com/watch/13922959/femdom-principal/',
         'only_matching': True,
+    }, {
+        'url': 'https://www.youporn.com/watch/16290308/tinderspecial-trailer1/',
+        'info_dict': {
+            'id': '16290308',
+            'age_limit': 18,
+            'categories': [],
+            'description': None,  # SEO spam using title removed
+            'display_id': 'tinderspecial-trailer1',
+            'duration': 298.0,
+            'ext': 'mp4',
+            'upload_date': '20201123',
+            'uploader': 'Ersties',
+            'tags': [],
+            'thumbnail': 'https://fi1.ypncdn.com/m=eaSaaTbWx/202011/23/16290308/original/3.jpg',
+            'timestamp': 1606147564,
+            'title': 'Tinder In Real Life',
+            'view_count': int,
+        }
     }]
 
-    @staticmethod
-    def _extract_urls(webpage):
-        return re.findall(
-            r'<iframe[^>]+\bsrc=["\']((?:https?:)?//(?:www\.)?youporn\.com/embed/\d+)',
-            webpage)
+    @classmethod
+    def _extract_urls(cls, webpage):
+        def yield_urls():
+            for p in cls._EMBED_REGEX:
+                for m in re.finditer(p, webpage):
+                    yield m.group('url')
+
+        return list(yield_urls())
 
     def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
-        video_id = mobj.group('id')
-        display_id = mobj.group('display_id') or video_id
+        # A different video ID (data-video-id) is hidden in the page but
+        # never seems to be used
+        video_id, display_id = self._match_valid_url(url).group('id', 'display_id')
+        url = 'http://www.youporn.com/watch/%s' % (video_id,)
+        webpage = self._download_webpage(
+            url, video_id, headers={'Cookie': 'age_verified=1'})
 
-        definitions = self._download_json(
-            'https://www.youporn.com/api/video/media_definitions/%s/' % video_id,
-            display_id)
+        watchable = self._search_regex(
+            r'''(<div\s[^>]*\bid\s*=\s*('|")?watch-container(?(2)\2|(?!-)\b)[^>]*>)''',
+            webpage, 'watchability', default=None)
+        if not watchable:
+            msg = re.split(r'\s{4}', clean_html(get_element_by_id(
+                'mainContent', webpage)) or '')[0]
+            raise ExtractorError(
+                ('%s says: %s' % (self.IE_NAME, msg))
+                if msg else 'Video unavailable: no reason found',
+                expected=True)
+        # internal ID ?
+        # video_id = extract_attributes(watchable).get('data-video-id')
+
+        playervars = self._search_json(
+            r'\bplayervars\s*:', webpage, 'playervars', video_id)
+
+        def get_fmt(x):
+            v_url = url_or_none(x.get('videoUrl'))
+            if v_url:
+                x['videoUrl'] = v_url
+                return (x['format'], x)
+
+        defs_by_format = dict(traverse_obj(playervars, (
+            'mediaDefinitions', lambda _, v: v.get('format'), T(get_fmt))))
+
+        def get_format_data(f):
+            if f not in defs_by_format:
+                return []
+            return self._download_json(
+                defs_by_format[f]['videoUrl'], video_id, '{0}-formats'.format(f))
 
         formats = []
-        for definition in definitions:
-            if not isinstance(definition, dict):
-                continue
-            video_url = url_or_none(definition.get('videoUrl'))
-            if not video_url:
-                continue
-            f = {
-                'url': video_url,
-                'filesize': int_or_none(definition.get('videoSize')),
-            }
-            height = int_or_none(definition.get('quality'))
+        # Try to extract only the actual master m3u8 first, avoiding the duplicate single resolution "master" m3u8s
+        for hls_url in traverse_obj(
+                get_format_data('hls'),
+                (lambda _, v: not isinstance(v['defaultQuality'], bool), 'videoUrl'),
+                (Ellipsis, 'videoUrl')):
+            formats.extend(self._extract_m3u8_formats(
+                hls_url, video_id, 'mp4', fatal=False, m3u8_id='hls',
+                entry_protocol='m3u8_native'))
+
+        for f in traverse_obj(get_format_data('mp4'), (
+                lambda _, v: v.get('videoUrl'), {
+                    'url': ('videoUrl', T(url_or_none)),
+                    'filesize': ('videoSize', T(int_or_none)),
+                    'height': ('quality', T(int_or_none)),
+                }, T(lambda x: x.get('videoUrl') and x))):
             # Video URL's path looks like this:
             #  /201012/17/505835/720p_1500k_505835/YouPorn%20-%20Sex%20Ed%20Is%20It%20Safe%20To%20Masturbate%20Daily.mp4
             #  /201012/17/505835/vl_240p_240k_505835/YouPorn%20-%20Sex%20Ed%20Is%20It%20Safe%20To%20Masturbate%20Daily.mp4
             #  /videos/201703/11/109285532/1080P_4000K_109285532.mp4
             # We will benefit from it by extracting some metadata
-            mobj = re.search(r'(?P<height>\d{3,4})[pP]_(?P<bitrate>\d+)[kK]_\d+', video_url)
+            mobj = re.search(r'(?P<height>\d{3,4})[pP]_(?P<bitrate>\d+)[kK]_\d+', f['videoUrl'])
             if mobj:
-                if not height:
-                    height = int(mobj.group('height'))
-                bitrate = int(mobj.group('bitrate'))
-                f.update({
-                    'format_id': '%dp-%dk' % (height, bitrate),
-                    'tbr': bitrate,
-                })
-            f['height'] = height
+                if not f.get('height'):
+                    f['height'] = int(mobj.group('height'))
+                f['tbr'] = int(mobj.group('bitrate'))
+                f['format_id'] = '%dp-%dk' % (f['height'], f['tbr'])
             formats.append(f)
         self._sort_formats(formats)
 
-        webpage = self._download_webpage(
-            'http://www.youporn.com/watch/%s' % video_id, display_id,
-            headers={'Cookie': 'age_verified=1'})
-
         title = self._html_search_regex(
             r'(?s)<div[^>]+class=["\']watchVideoTitle[^>]+>(.+?)</div>',
             webpage, 'title', default=None) or self._og_search_title(
@@ -131,8 +196,10 @@ class YouPornIE(InfoExtractor):
         thumbnail = self._search_regex(
             r'(?:imageurl\s*=|poster\s*:)\s*(["\'])(?P<thumbnail>.+?)\1',
             webpage, 'thumbnail', fatal=False, group='thumbnail')
-        duration = int_or_none(self._html_search_meta(
-            'video:duration', webpage, 'duration', fatal=False))
+        duration = traverse_obj(playervars, ('duration', T(int_or_none)))
+        if duration is None:
+            duration = int_or_none(self._html_search_meta(
+                'video:duration', webpage, 'duration', fatal=False))
 
         uploader = self._html_search_regex(
             r'(?s)<div[^>]+class=["\']submitByLink["\'][^>]*>(.+?)</div>',
@@ -148,11 +215,11 @@ class YouPornIE(InfoExtractor):
 
         view_count = None
         views = self._search_regex(
-            r'(<div[^>]+\bclass=["\']js_videoInfoViews["\']>)', webpage,
-            'views', default=None)
+            r'(<div\s[^>]*\bdata-value\s*=[^>]+>)\s*<label>Views:</label>',
+            webpage, 'views', default=None)
         if views:
-            view_count = str_to_int(extract_attributes(views).get('data-value'))
-        comment_count = str_to_int(self._search_regex(
+            view_count = parse_count(extract_attributes(views).get('data-value'))
+        comment_count = parse_count(self._search_regex(
             r'>All [Cc]omments? \(([\d,.]+)\)',
             webpage, 'comment count', default=None))
 
@@ -168,7 +235,10 @@ class YouPornIE(InfoExtractor):
             r'(?s)Tags:.*?</div>\s*<div[^>]+class=["\']tagBoxContent["\'][^>]*>(.+?)</div>',
             'tags')
 
-        return {
+        data = self._search_json_ld(webpage, video_id, expected_type='VideoObject', fatal=False) or {}
+        data.pop('url', None)
+
+        result = merge_dicts(data, {
             'id': video_id,
             'display_id': display_id,
             'title': title,
@@ -183,4 +253,442 @@ class YouPornIE(InfoExtractor):
             'tags': tags,
             'age_limit': age_limit,
             'formats': formats,
-        }
+        })
+        # Remove promotional non-description
+        if result.get('description', '').startswith(
+                'Watch %s online' % (result['title'],)):
+            del result['description']
+        return result
+
+
+class YouPornListBase(InfoExtractor):
+    # pattern in '.title-text' element of page section containing videos
+    _PLAYLIST_TITLEBAR_RE = r'\s+[Vv]ideos\s*$'
+    _PAGE_RETRY_COUNT = 0  # ie, no retry
+    _PAGE_RETRY_DELAY = 2  # seconds
+
+    def _get_next_url(self, url, pl_id, html):
+        return urljoin(url, self._search_regex(
+            r'''<a\s[^>]*?\bhref\s*=\s*("|')(?P<url>(?:(?!\1)[^>])+)\1''',
+            get_element_by_id('next', html) or '', 'next page',
+            group='url', default=None))
+
+    @classmethod
+    def _get_title_from_slug(cls, title_slug):
+        return re.sub(r'[_-]', ' ', title_slug)
+
+    def _entries(self, url, pl_id, html=None, page_num=None):
+
+        # separates page sections
+        PLAYLIST_SECTION_RE = (
+            r'''<div\s[^>]*\bclass\s*=\s*('|")(?:[\w$-]+\s+|\s)*?title-bar(?:\s+[\w$-]+|\s)*\1[^>]*>'''
+        )
+        # contains video link
+        VIDEO_URL_RE = r'''(?x)
+            <div\s[^>]*\bdata-video-id\s*=\s*('|")\d+\1[^>]*>\s*
+            (?:<div\b[\s\S]+?</div>\s*)*
+            <a\s[^>]*\bhref\s*=\s*('|")(?P<url>(?:(?!\2)[^>])+)\2
+        '''
+
+        def yield_pages(url, html=html, page_num=page_num):
+            fatal = not html
+            for pnum in itertools.count(start=page_num or 1):
+                if not html:
+                    html = self._download_webpage(
+                        url, pl_id, note='Downloading page %d' % pnum,
+                        fatal=fatal)
+                if not html:
+                    break
+                fatal = False
+                yield (url, html, pnum)
+                # explicit page: extract just that page
+                if page_num is not None:
+                    break
+                next_url = self._get_next_url(url, pl_id, html)
+                if not next_url or next_url == url:
+                    break
+                url, html = next_url, None
+
+        def retry_page(msg, tries_left, page_data):
+            if tries_left <= 0:
+                return
+            self.report_warning(msg, pl_id)
+            sleep(self._PAGE_RETRY_DELAY)
+            return next(
+                yield_pages(page_data[0], page_num=page_data[2]), None)
+
+        def yield_entries(html):
+            for frag in re.split(PLAYLIST_SECTION_RE, html):
+                if not frag:
+                    continue
+                t_text = get_element_by_class('title-text', frag or '')
+                if not (t_text and re.search(self._PLAYLIST_TITLEBAR_RE, t_text)):
+                    continue
+                for m in re.finditer(VIDEO_URL_RE, frag):
+                    video_url = urljoin(url, m.group('url'))
+                    if video_url:
+                        yield self.url_result(video_url)
+
+        last_first_url = None
+        for page_data in yield_pages(url, html=html, page_num=page_num):
+            # page_data: url, html, page_num
+            first_url = None
+            tries_left = self._PAGE_RETRY_COUNT + 1
+            while tries_left > 0:
+                tries_left -= 1
+                for from_ in yield_entries(page_data[1]):
+                    # may get the same page twice instead of empty page
+                    # or (site bug) intead of actual next page
+                    if not first_url:
+                        first_url = from_['url']
+                        if first_url == last_first_url:
+                            # sometimes (/porntags/) the site serves the previous page
+                            # instead but may provide the correct page after a delay
+                            page_data = retry_page(
+                                'Retrying duplicate page...', tries_left, page_data)
+                            if page_data:
+                                first_url = None
+                                break
+                            continue
+                    yield from_
+                else:
+                    if not first_url and 'no-result-paragarph1' in page_data[1]:
+                        page_data = retry_page(
+                            'Retrying empty page...', tries_left, page_data)
+                        if page_data:
+                            continue
+                    else:
+                        # success/failure
+                        break
+            # may get an infinite (?) sequence of empty pages
+            if not first_url:
+                break
+            last_first_url = first_url
+
+    def _real_extract(self, url, html=None):
+        # exceptionally, id may be None
+        m_dict = self._match_valid_url(url).groupdict()
+        pl_id, page_type, sort = (m_dict.get(k) for k in ('id', 'type', 'sort'))
+
+        qs = parse_qs(url)
+        for q, v in qs.items():
+            if v:
+                qs[q] = v[-1]
+            else:
+                del qs[q]
+
+        base_id = pl_id or 'YouPorn'
+        title = self._get_title_from_slug(base_id)
+        if page_type:
+            title = '%s %s' % (page_type.capitalize(), title)
+        base_id = [base_id.lower()]
+        if sort is None:
+            title += ' videos'
+        else:
+            title = '%s videos by %s' % (title, re.sub(r'[_-]', ' ', sort))
+            base_id.append(sort)
+        if qs:
+            ps = ['%s=%s' % item for item in sorted(qs.items())]
+            title += ' (%s)' % ','.join(ps)
+            base_id.extend(ps)
+        pl_id = '/'.join(base_id)
+
+        return self.playlist_result(
+            self._entries(url, pl_id, html=html,
+                          page_num=int_or_none(qs.get('page'))),
+            playlist_id=pl_id, playlist_title=title)
+
+
+class YouPornCategoryIE(YouPornListBase):
+    IE_DESC = 'YouPorn category, with sorting, filtering and pagination'
+    _VALID_URL = r'''(?x)
+        https?://(?:www\.)?youporn\.com/
+        (?P<type>category)/(?P<id>[^/?#&]+)
+        (?:/(?P<sort>popular|views|rating|time|duration))?/?(?:[#?]|$)
+    '''
+    _TESTS = [{
+        'note': 'Full list with pagination',
+        'url': 'https://www.youporn.com/category/lingerie/popular/',
+        'info_dict': {
+            'id': 'lingerie/popular',
+            'title': 'Category lingerie videos by popular',
+        },
+        'playlist_mincount': 39,
+    }, {
+        'note': 'Filtered paginated list with single page result',
+        'url': 'https://www.youporn.com/category/lingerie/duration/?min_minutes=10',
+        'info_dict': {
+            'id': 'lingerie/duration/min_minutes=10',
+            'title': 'Category lingerie videos by duration (min_minutes=10)',
+        },
+        'playlist_maxcount': 30,
+    }, {
+        'note': 'Single page of full list',
+        'url': 'https://www.youporn.com/category/lingerie/popular?page=1',
+        'info_dict': {
+            'id': 'lingerie/popular/page=1',
+            'title': 'Category lingerie videos by popular (page=1)',
+        },
+        'playlist_count': 30,
+    }]
+
+
+class YouPornChannelIE(YouPornListBase):
+    IE_DESC = 'YouPorn channel, with sorting and pagination'
+    _VALID_URL = r'''(?x)
+        https?://(?:www\.)?youporn\.com/
+        (?P<type>channel)/(?P<id>[^/?#&]+)
+        (?:/(?P<sort>rating|views|duration))?/?(?:[#?]|$)
+    '''
+    _TESTS = [{
+        'note': 'Full list with pagination',
+        'url': 'https://www.youporn.com/channel/x-feeds/',
+        'info_dict': {
+            'id': 'x-feeds',
+            'title': 'Channel X-Feeds videos',
+        },
+        'playlist_mincount': 37,
+    }, {
+        'note': 'Single page of full list (no filters here)',
+        'url': 'https://www.youporn.com/channel/x-feeds/duration?page=1',
+        'info_dict': {
+            'id': 'x-feeds/duration/page=1',
+            'title': 'Channel X-Feeds videos by duration (page=1)',
+        },
+        'playlist_count': 24,
+    }]
+
+    @staticmethod
+    def _get_title_from_slug(title_slug):
+        return re.sub(r'_', ' ', title_slug).title()
+
+
+class YouPornCollectionIE(YouPornListBase):
+    IE_DESC = 'YouPorn collection (user playlist), with sorting and pagination'
+    _VALID_URL = r'''(?x)
+        https?://(?:www\.)?youporn\.com/
+        (?P<type>collection)s/videos/(?P<id>\d+)
+        (?:/(?P<sort>rating|views|time|duration))?/?(?:[#?]|$)
+    '''
+    _PLAYLIST_TITLEBAR_RE = r'^\s*Videos\s+in\s'
+    _TESTS = [{
+        'note': 'Full list with pagination',
+        'url': 'https://www.youporn.com/collections/videos/33044251/',
+        'info_dict': {
+            'id': '33044251',
+            'title': 'Collection Sexy Lips videos',
+            'uploader': 'ph-littlewillyb',
+        },
+        'playlist_mincount': 50,
+    }, {
+        'note': 'Single page of full list (no filters here)',
+        'url': 'https://www.youporn.com/collections/videos/33044251/time?page=1',
+        'info_dict': {
+            'id': '33044251/time/page=1',
+            'title': 'Collection Sexy Lips videos by time (page=1)',
+            'uploader': 'ph-littlewillyb',
+        },
+        'playlist_count': 20,
+    }]
+
+    def _real_extract(self, url):
+        pl_id = self._match_id(url)
+        html = self._download_webpage(url, pl_id)
+        playlist = super(YouPornCollectionIE, self)._real_extract(url, html=html)
+        infos = re.sub(r'\s+', ' ', clean_html(get_element_by_class(
+            'collection-infos', html)) or '')
+        title, uploader = self._search_regex(
+            r'^\s*Collection: (?P<title>.+?) \d+ VIDEOS \d+ VIEWS \d+ days LAST UPDATED From: (?P<uploader>[\w_-]+)',
+            infos, 'title/uploader', group=('title', 'uploader'), default=(None, None))
+
+        return merge_dicts({
+            'title': playlist['title'].replace(playlist['id'].split('/')[0], title),
+            'uploader': uploader,
+        }, playlist) if title else playlist
+
+
+class YouPornTagIE(YouPornListBase):
+    IE_DESC = 'YouPorn tag (porntags), with sorting, filtering and pagination'
+    _VALID_URL = r'''(?x)
+        https?://(?:www\.)?youporn\.com/
+        porn(?P<type>tag)s/(?P<id>[^/?#&]+)
+        (?:/(?P<sort>views|rating|time|duration))?/?(?:[#?]|$)
+    '''
+    _PLAYLIST_TITLEBAR_RE = r'^\s*Videos\s+tagged\s'
+    _PAGE_RETRY_COUNT = 1
+    _TESTS = [{
+        'note': 'Full list with pagination',
+        'url': 'https://www.youporn.com/porntags/austrian',
+        'info_dict': {
+            'id': 'austrian',
+            'title': 'Tag austrian videos',
+        },
+        'playlist_mincount': 35,
+        'expected_warnings': ['Retrying duplicate page'],
+    }, {
+        'note': 'Filtered paginated list with single page result',
+        'url': 'https://www.youporn.com/porntags/austrian/duration/?min_minutes=10',
+        'info_dict': {
+            'id': 'austrian/duration/min_minutes=10',
+            'title': 'Tag austrian videos by duration (min_minutes=10)',
+        },
+        # number of videos per page is (row x col) 2x3 + 6x4 + 2, or + 3,
+        # or more, varying with number of ads; let's set max as 9x4
+        # NB col 1 may not be shown in non-JS page with site CSS and zoom 100%
+        'playlist_maxcount': 32,
+        'expected_warnings': ['Retrying duplicate page', 'Retrying empty page'],
+    }, {
+        'note': 'Single page of full list',
+        'url': 'https://www.youporn.com/porntags/austrian/?page=1',
+        'info_dict': {
+            'id': 'austrian/page=1',
+            'title': 'Tag austrian videos (page=1)',
+        },
+        'playlist_mincount': 32,
+        'playlist_maxcount': 34,
+        'expected_warnings': ['Retrying duplicate page', 'Retrying empty page'],
+    }]
+
+    # YP tag navigation is broken, loses sort
+    def _get_next_url(self, url, pl_id, html):
+        next_url = super(YouPornTagIE, self)._get_next_url(url, pl_id, html)
+        if next_url:
+            n = self._match_valid_url(next_url)
+            if n:
+                s = n.groupdict().get('sort')
+            if s:
+                u = self._match_valid_url(url)
+                if u:
+                    u = u.groupdict().get('sort')
+                    if s and not u:
+                        n = n.end('sort')
+                        next_url = next_url[:n] + '/' + u + next_url[n:]
+        return next_url
+
+
+class YouPornStarIE(YouPornListBase):
+    IE_DESC = 'YouPorn Pornstar, with description, sorting and pagination'
+    _VALID_URL = r'''(?x)
+        https?://(?:www\.)?youporn\.com/
+        (?P<type>pornstar)/(?P<id>[^/?#&]+)
+        (?:/(?P<sort>rating|views|duration))?/?(?:[#?]|$)
+    '''
+    _PLAYLIST_TITLEBAR_RE = r'^\s*Videos\s+[fF]eaturing\s'
+    _TESTS = [{
+        'note': 'Full list with pagination',
+        'url': 'https://www.youporn.com/pornstar/daynia/',
+        'info_dict': {
+            'id': 'daynia',
+            'title': 'Pornstar Daynia videos',
+            'description': r're:Daynia Rank \d+ Videos \d+ Views [\d,.]+ .+ Subscribers \d+',
+        },
+        'playlist_mincount': 45,
+    }, {
+        'note': 'Single page of full list (no filters here)',
+        'url': 'https://www.youporn.com/pornstar/daynia/?page=1',
+        'info_dict': {
+            'id': 'daynia/page=1',
+            'title': 'Pornstar Daynia videos (page=1)',
+            'description': 're:.{180,}',
+        },
+        'playlist_count': 26,
+    }]
+
+    @staticmethod
+    def _get_title_from_slug(title_slug):
+        return re.sub(r'_', ' ', title_slug).title()
+
+    def _real_extract(self, url):
+        pl_id = self._match_id(url)
+        html = self._download_webpage(url, pl_id)
+        playlist = super(YouPornStarIE, self)._real_extract(url, html=html)
+        INFO_ELEMENT_RE = r'''(?x)
+            <div\s[^>]*\bclass\s*=\s*('|")(?:[\w$-]+\s+|\s)*?pornstar-info-wrapper(?:\s+[\w$-]+|\s)*\1[^>]*>
+            (?P<info>[\s\S]+?)(?:</div>\s*){6,}
+        '''
+
+        infos = self._search_regex(INFO_ELEMENT_RE, html, 'infos', group='info', default='')
+        if infos:
+            infos = re.sub(
+                r'(?:\s*nl=nl)+\s*', ' ',
+                re.sub(r'(?u)\s+', ' ', clean_html(
+                    re.sub('\n', 'nl=nl', infos)))).replace('ribe Subsc', '')
+
+        return merge_dicts({
+            'description': infos.strip() or None,
+        }, playlist)
+
+
+class YouPornVideosIE(YouPornListBase):
+    IE_DESC = 'YouPorn video (browse) playlists, with sorting, filtering and pagination'
+    _VALID_URL = r'''(?x)
+        https?://(?:www\.)?youporn\.com/
+            (?:(?P<id>browse)/)?
+            (?P<sort>(?(id)
+                (?:duration|rating|time|views)|
+                (?:most_(?:favou?rit|view)ed|recommended|top_rated)?))
+            (?:[/#?]|$)
+    '''
+    _PLAYLIST_TITLEBAR_RE = r'\s+(?:[Vv]ideos|VIDEOS)\s*$'
+    _TESTS = [{
+        'note': 'Full list with pagination (too long for test)',
+        'url': 'https://www.youporn.com/',
+        'info_dict': {
+            'id': 'youporn',
+            'title': 'YouPorn videos',
+        },
+        'only_matching': True,
+    }, {
+        'note': 'Full list with pagination (too long for test)',
+        'url': 'https://www.youporn.com/recommended',
+        'info_dict': {
+            'id': 'youporn/recommended',
+            'title': 'YouPorn videos by recommended',
+        },
+        'only_matching': True,
+    }, {
+        'note': 'Full list with pagination (too long for test)',
+        'url': 'https://www.youporn.com/top_rated',
+        'info_dict': {
+            'id': 'youporn/top_rated',
+            'title': 'YouPorn videos by top rated',
+        },
+        'only_matching': True,
+    }, {
+        'note': 'Full list with pagination (too long for test)',
+        'url': 'https://www.youporn.com/browse/time',
+        'info_dict': {
+            'id': 'browse/time',
+            'title': 'YouPorn videos by time',
+        },
+        'only_matching': True,
+    }, {
+        'note': 'Filtered paginated list with single page result',
+        'url': 'https://www.youporn.com/most_favorited/?res=VR&max_minutes=2',
+        'info_dict': {
+            'id': 'youporn/most_favorited/max_minutes=2/res=VR',
+            'title': 'YouPorn videos by most favorited (max_minutes=2,res=VR)',
+        },
+        'playlist_mincount': 10,
+        'playlist_maxcount': 28,
+    }, {
+        'note': 'Filtered paginated list with several pages',
+        'url': 'https://www.youporn.com/most_favorited/?res=VR&max_minutes=5',
+        'info_dict': {
+            'id': 'youporn/most_favorited/max_minutes=5/res=VR',
+            'title': 'YouPorn videos by most favorited (max_minutes=5,res=VR)',
+        },
+        'playlist_mincount': 45,
+    }, {
+        'note': 'Single page of full list',
+        'url': 'https://www.youporn.com/browse/time?page=1',
+        'info_dict': {
+            'id': 'browse/time/page=1',
+            'title': 'YouPorn videos by time (page=1)',
+        },
+        'playlist_count': 36,
+    }]
+
+    @staticmethod
+    def _get_title_from_slug(title_slug):
+        return 'YouPorn' if title_slug == 'browse' else title_slug

youtube_dl/extractor/youtube.py

@@ -2,6 +2,7 @@
 
 from __future__ import unicode_literals
 
+import collections
 import itertools
 import json
 import os.path
@@ -23,10 +24,10 @@ from ..compat import (
 )
 from ..jsinterp import JSInterpreter
 from ..utils import (
-    ExtractorError,
     clean_html,
     dict_get,
     error_to_compat_str,
+    ExtractorError,
     float_or_none,
     extract_attributes,
     get_element_by_attribute,
@@ -36,7 +37,9 @@ from ..utils import (
     LazyList,
     merge_dicts,
     mimetype2ext,
+    NO_DEFAULT,
     parse_codecs,
+    parse_count,
     parse_duration,
     parse_qs,
     qualities,
@@ -44,7 +47,9 @@ from ..utils import (
     smuggle_url,
     str_or_none,
     str_to_int,
+    T,
     traverse_obj,
+    try_call,
     try_get,
     txt_or_none,
     unescapeHTML,
@@ -260,16 +265,10 @@ class YoutubeBaseInfoExtractor(InfoExtractor):
         cookies = self._get_cookies('https://www.youtube.com/')
         if cookies.get('__Secure-3PSID'):
             return
-        consent_id = None
-        consent = cookies.get('CONSENT')
-        if consent:
-            if 'YES' in consent.value:
-                return
-            consent_id = self._search_regex(
-                r'PENDING\+(\d+)', consent.value, 'consent', default=None)
-        if not consent_id:
-            consent_id = random.randint(100, 999)
-        self._set_cookie('.youtube.com', 'CONSENT', 'YES+cb.20210328-17-p0.en+FX+%s' % consent_id)
+        socs = cookies.get('SOCS')
+        if socs and not socs.value.startswith('CAA'):  # not consented
+            return
+        self._set_cookie('.youtube.com', 'SOCS', 'CAI', secure=True)  # accept all (required for mixes)
 
     def _real_initialize(self):
         self._initialize_consent()
@@ -448,7 +447,7 @@ class YoutubeBaseInfoExtractor(InfoExtractor):
             extract_attributes(self._search_regex(
                 r'''(?s)(<link\b[^>]+\bitemprop\s*=\s*("|')%s\2[^>]*>)'''
                 % re.escape(var_name),
-                get_element_by_attribute('itemprop', 'author', webpage) or '',
+                get_element_by_attribute('itemprop', 'author', webpage or '') or '',
                 'author link', default='')),
             paths[var_name][0])
 
@@ -1253,7 +1252,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                 'title': 'IMG 3456',
                 'description': '',
                 'upload_date': '20170613',
-                'uploader': 'ElevageOrVert',
+                'uploader': "l'Or Vert asbl",
                 'uploader_id': '@ElevageOrVert',
             },
             'params': {
@@ -1466,6 +1465,30 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
         self._code_cache = {}
         self._player_cache = {}
 
+    # *ytcfgs, webpage=None
+    def _extract_player_url(self, *ytcfgs, **kw_webpage):
+        if ytcfgs and not isinstance(ytcfgs[0], dict):
+            webpage = kw_webpage.get('webpage') or ytcfgs[0]
+        if webpage:
+            player_url = self._search_regex(
+                r'"(?:PLAYER_JS_URL|jsUrl)"\s*:\s*"([^"]+)"',
+                webpage or '', 'player URL', fatal=False)
+            if player_url:
+                ytcfgs = ytcfgs + ({'PLAYER_JS_URL': player_url},)
+        return traverse_obj(
+            ytcfgs, (Ellipsis, 'PLAYER_JS_URL'), (Ellipsis, 'WEB_PLAYER_CONTEXT_CONFIGS', Ellipsis, 'jsUrl'),
+            get_all=False, expected_type=lambda u: urljoin('https://www.youtube.com', u))
+
+    def _download_player_url(self, video_id, fatal=False):
+        res = self._download_webpage(
+            'https://www.youtube.com/iframe_api',
+            note='Downloading iframe API JS', video_id=video_id, fatal=fatal)
+        player_version = self._search_regex(
+            r'player\\?/([0-9a-fA-F]{8})\\?/', res or '', 'player version', fatal=fatal,
+            default=NO_DEFAULT if res else None)
+        if player_version:
+            return 'https://www.youtube.com/s/player/{0}/player_ias.vflset/en_US/base.js'.format(player_version)
+
     def _signature_cache_id(self, example_sig):
         """ Return a string representation of a signature """
         return '.'.join(compat_str(len(part)) for part in example_sig.split('.'))
@@ -1480,46 +1503,49 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
             raise ExtractorError('Cannot identify player %r' % player_url)
         return id_m.group('id')
 
-    def _get_player_code(self, video_id, player_url, player_id=None):
+    def _load_player(self, video_id, player_url, fatal=True, player_id=None):
         if not player_id:
             player_id = self._extract_player_info(player_url)
-
         if player_id not in self._code_cache:
-            self._code_cache[player_id] = self._download_webpage(
-                player_url, video_id,
+            code = self._download_webpage(
+                player_url, video_id, fatal=fatal,
                 note='Downloading player ' + player_id,
                 errnote='Download of %s failed' % player_url)
-        return self._code_cache[player_id]
+            if code:
+                self._code_cache[player_id] = code
+        return self._code_cache[player_id] if fatal else self._code_cache.get(player_id)
 
     def _extract_signature_function(self, video_id, player_url, example_sig):
         player_id = self._extract_player_info(player_url)
 
         # Read from filesystem cache
-        func_id = 'js_%s_%s' % (
+        func_id = 'js_{0}_{1}'.format(
             player_id, self._signature_cache_id(example_sig))
         assert os.path.basename(func_id) == func_id
 
-        cache_spec = self._downloader.cache.load('youtube-sigfuncs', func_id)
-        if cache_spec is not None:
-            return lambda s: ''.join(s[i] for i in cache_spec)
+        self.write_debug('Extracting signature function {0}'.format(func_id))
+        cache_spec, code = self.cache.load('youtube-sigfuncs', func_id), None
 
-        code = self._get_player_code(video_id, player_url, player_id)
-        res = self._parse_sig_js(code)
+        if not cache_spec:
+            code = self._load_player(video_id, player_url, player_id)
+        if code:
+            res = self._parse_sig_js(code)
+            test_string = ''.join(map(compat_chr, range(len(example_sig))))
+            cache_spec = [ord(c) for c in res(test_string)]
+            self.cache.store('youtube-sigfuncs', func_id, cache_spec)
 
-        test_string = ''.join(map(compat_chr, range(len(example_sig))))
-        cache_res = res(test_string)
-        cache_spec = [ord(c) for c in cache_res]
-
-        self._downloader.cache.store('youtube-sigfuncs', func_id, cache_spec)
-        return res
+        return lambda s: ''.join(s[i] for i in cache_spec)
 
     def _print_sig_code(self, func, example_sig):
+        if not self.get_param('youtube_print_sig_code'):
+            return
+
         def gen_sig_code(idxs):
             def _genslice(start, end, step):
                 starts = '' if start == 0 else str(start)
                 ends = (':%d' % (end + step)) if end + step >= 0 else ':'
                 steps = '' if step == 1 else (':%d' % step)
-                return 's[%s%s%s]' % (starts, ends, steps)
+                return 's[{0}{1}{2}]'.format(starts, ends, steps)
 
             step = None
             # Quelch pyflakes warnings - start will be set when step is set
@@ -1573,131 +1599,134 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
         initial_function = jsi.extract_function(funcname)
         return lambda s: initial_function([s])
 
+    def _cached(self, func, *cache_id):
+        def inner(*args, **kwargs):
+            if cache_id not in self._player_cache:
+                try:
+                    self._player_cache[cache_id] = func(*args, **kwargs)
+                except ExtractorError as e:
+                    self._player_cache[cache_id] = e
+                except Exception as e:
+                    self._player_cache[cache_id] = ExtractorError(traceback.format_exc(), cause=e)
+
+            ret = self._player_cache[cache_id]
+            if isinstance(ret, Exception):
+                raise ret
+            return ret
+        return inner
+
     def _decrypt_signature(self, s, video_id, player_url):
         """Turn the encrypted s field into a working signature"""
-
-        if player_url is None:
-            raise ExtractorError('Cannot decrypt signature without player_url')
-
-        try:
-            player_id = (player_url, self._signature_cache_id(s))
-            if player_id not in self._player_cache:
-                func = self._extract_signature_function(
-                    video_id, player_url, s
-                )
-                self._player_cache[player_id] = func
-            func = self._player_cache[player_id]
-            if self._downloader.params.get('youtube_print_sig_code'):
-                self._print_sig_code(func, s)
-            return func(s)
-        except Exception as e:
-            tb = traceback.format_exc()
-            raise ExtractorError(
-                'Signature extraction failed: ' + tb, cause=e)
-
-    def _extract_player_url(self, webpage):
-        player_url = self._search_regex(
-            r'"(?:PLAYER_JS_URL|jsUrl)"\s*:\s*"([^"]+)"',
-            webpage or '', 'player URL', fatal=False)
-        if not player_url:
-            return
-        if player_url.startswith('//'):
-            player_url = 'https:' + player_url
-        elif not re.match(r'https?://', player_url):
-            player_url = compat_urllib_parse.urljoin(
-                'https://www.youtube.com', player_url)
-        return player_url
+        extract_sig = self._cached(
+            self._extract_signature_function, 'sig', player_url, self._signature_cache_id(s))
+        func = extract_sig(video_id, player_url, s)
+        self._print_sig_code(func, s)
+        return func(s)
 
     # from yt-dlp
     # See also:
     # 1. https://github.com/ytdl-org/youtube-dl/issues/29326#issuecomment-894619419
     # 2. https://code.videolan.org/videolan/vlc/-/blob/4fb284e5af69aa9ac2100ccbdd3b88debec9987f/share/lua/playlist/youtube.lua#L116
     # 3. https://github.com/ytdl-org/youtube-dl/issues/30097#issuecomment-950157377
-    def _extract_n_function_name(self, jscode):
-        target = r'(?P<nfunc>[a-zA-Z_$][\w$]*)(?:\[(?P<idx>\d+)\])?'
-        nfunc_and_idx = self._search_regex(
-            r'\.get\("n"\)\)&&\(b=(%s)\([\w$]+\)' % (target, ),
-            jscode, 'Initial JS player n function name')
-        nfunc, idx = re.match(target, nfunc_and_idx).group('nfunc', 'idx')
-        if not idx:
-            return nfunc
-        if int_or_none(idx) == 0:
-            real_nfunc = self._search_regex(
-                r'var %s\s*=\s*\[([a-zA-Z_$][\w$]*)\];' % (re.escape(nfunc), ), jscode,
-                'Initial JS player n function alias ({nfunc}[{idx}])'.format(**locals()))
-            if real_nfunc:
-                return real_nfunc
-        return self._parse_json(self._search_regex(
-            r'var %s\s*=\s*(\[.+?\]);' % (re.escape(nfunc), ), jscode,
-            'Initial JS player n function name ({nfunc}[{idx}])'.format(**locals())), nfunc, transform_source=js_to_json)[int(idx)]
-
-    def _extract_n_function(self, video_id, player_url):
-        player_id = self._extract_player_info(player_url)
-        func_code = self._downloader.cache.load('youtube-nsig', player_id)
-
-        if func_code:
-            jsi = JSInterpreter(func_code)
-        else:
-            jscode = self._get_player_code(video_id, player_url, player_id)
-            funcname = self._extract_n_function_name(jscode)
-            jsi = JSInterpreter(jscode)
-            func_code = jsi.extract_function_code(funcname)
-            self._downloader.cache.store('youtube-nsig', player_id, func_code)
-
-        if self._downloader.params.get('youtube_print_sig_code'):
-            self.to_screen('Extracted nsig function from {0}:\n{1}\n'.format(player_id, func_code[1]))
-
-        return lambda s: jsi.extract_function_from_code(*func_code)([s])
-
-    def _n_descramble(self, n_param, player_url, video_id):
-        """Compute the response to YT's "n" parameter challenge,
-           or None
-
-        Args:
-        n_param     -- challenge string that is the value of the
-                       URL's "n" query parameter
-        player_url  -- URL of YT player JS
-        video_id
-        """
-
-        sig_id = ('nsig_value', n_param)
-        if sig_id in self._player_cache:
-            return self._player_cache[sig_id]
+    def _decrypt_nsig(self, n, video_id, player_url):
+        """Turn the encrypted n field into a working signature"""
+        if player_url is None:
+            raise ExtractorError('Cannot decrypt nsig without player_url')
 
         try:
-            player_id = ('nsig', player_url)
-            if player_id not in self._player_cache:
-                self._player_cache[player_id] = self._extract_n_function(video_id, player_url)
-            func = self._player_cache[player_id]
-            ret = func(n_param)
-            if ret.startswith('enhanced_except_'):
-                raise ExtractorError('Unhandled exception in decode')
-            self._player_cache[sig_id] = ret
-            if self._downloader.params.get('verbose', False):
-                self._downloader.to_screen('[debug] [%s] %s' % (self.IE_NAME, 'Decrypted nsig {0} => {1}'.format(n_param, self._player_cache[sig_id])))
-            return self._player_cache[sig_id]
-        except Exception as e:
-            self._downloader.report_warning(
-                '[%s] %s (%s %s)' % (
-                    self.IE_NAME,
+            jsi, player_id, func_code = self._extract_n_function_code(video_id, player_url)
+        except ExtractorError as e:
+            raise ExtractorError('Unable to extract nsig jsi, player_id, func_codefunction code', cause=e)
+        if self.get_param('youtube_print_sig_code'):
+            self.to_screen('Extracted nsig function from {0}:\n{1}\n'.format(
+                player_id, func_code[1]))
+
+        try:
+            extract_nsig = self._cached(self._extract_n_function_from_code, 'nsig func', player_url)
+            ret = extract_nsig(jsi, func_code)(n)
+        except JSInterpreter.Exception as e:
+            self.report_warning(
+                '%s (%s %s)' % (
                     'Unable to decode n-parameter: download likely to be throttled',
                     error_to_compat_str(e),
-                    traceback.format_exc()))
+                    traceback.format_exc()),
+                video_id=video_id)
+            return
+
+        self.write_debug('Decrypted nsig {0} => {1}'.format(n, ret))
+        return ret
+
+    def _extract_n_function_name(self, jscode):
+        func_name, idx = self._search_regex(
+            r'\.get\("n"\)\)&&\(b=(?P<nfunc>[a-zA-Z_$][\w$]*)(?:\[(?P<idx>\d+)\])?\([\w$]+\)',
+            jscode, 'Initial JS player n function name', group=('nfunc', 'idx'))
+        if not idx:
+            return func_name
+
+        return self._parse_json(self._search_regex(
+            r'var {0}\s*=\s*(\[.+?\])\s*[,;]'.format(re.escape(func_name)), jscode,
+            'Initial JS player n function list ({0}.{1})'.format(func_name, idx)),
+            func_name, transform_source=js_to_json)[int(idx)]
+
+    def _extract_n_function_code(self, video_id, player_url):
+        player_id = self._extract_player_info(player_url)
+        func_code = self.cache.load('youtube-nsig', player_id)
+        jscode = func_code or self._load_player(video_id, player_url)
+        jsi = JSInterpreter(jscode)
+
+        if func_code:
+            return jsi, player_id, func_code
+
+        func_name = self._extract_n_function_name(jscode)
+
+        # For redundancy
+        func_code = self._search_regex(
+            r'''(?xs)%s\s*=\s*function\s*\((?P<var>[\w$]+)\)\s*
+                     # NB: The end of the regex is intentionally kept strict
+                     {(?P<code>.+?}\s*return\ [\w$]+.join\(""\))};''' % func_name,
+            jscode, 'nsig function', group=('var', 'code'), default=None)
+        if func_code:
+            func_code = ([func_code[0]], func_code[1])
+        else:
+            self.write_debug('Extracting nsig function with jsinterp')
+            func_code = jsi.extract_function_code(func_name)
+
+        self.cache.store('youtube-nsig', player_id, func_code)
+        return jsi, player_id, func_code
+
+    def _extract_n_function_from_code(self, jsi, func_code):
+        func = jsi.extract_function_from_code(*func_code)
+
+        def extract_nsig(s):
+            try:
+                ret = func([s])
+            except JSInterpreter.Exception:
+                raise
+            except Exception as e:
+                raise JSInterpreter.Exception(traceback.format_exc(), cause=e)
+
+            if ret.startswith('enhanced_except_'):
+                raise JSInterpreter.Exception('Signature function returned an exception')
+            return ret
+
+        return extract_nsig
+
+    def _unthrottle_format_urls(self, video_id, player_url, *formats):
+
+        def decrypt_nsig(n):
+            return self._cached(self._decrypt_nsig, 'nsig', n, player_url)
 
-    def _unthrottle_format_urls(self, video_id, player_url, formats):
         for fmt in formats:
             parsed_fmt_url = compat_urllib_parse.urlparse(fmt['url'])
             n_param = compat_parse_qs(parsed_fmt_url.query).get('n')
             if not n_param:
                 continue
             n_param = n_param[-1]
-            n_response = self._n_descramble(n_param, player_url, video_id)
+            n_response = decrypt_nsig(n_param)(n_param, video_id, player_url)
             if n_response is None:
                 # give up if descrambling failed
                 break
-            for fmt_dct in traverse_obj(fmt, (None, (None, ('fragments', Ellipsis))), expected_type=dict):
-                fmt_dct['url'] = update_url(
-                    fmt_dct['url'], query_update={'n': [n_response]})
+            fmt['url'] = update_url_query(fmt['url'], {'n': n_response})
 
     # from yt-dlp, with tweaks
     def _extract_signature_timestamp(self, video_id, player_url, ytcfg=None, fatal=False):
@@ -1705,16 +1734,16 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
         Extract signatureTimestamp (sts)
         Required to tell API what sig/player version is in use.
         """
-        sts = int_or_none(ytcfg.get('STS')) if isinstance(ytcfg, dict) else None
+        sts = traverse_obj(ytcfg, 'STS', expected_type=int)
         if not sts:
             # Attempt to extract from player
             if player_url is None:
                 error_msg = 'Cannot extract signature timestamp without player_url.'
                 if fatal:
                     raise ExtractorError(error_msg)
-                self._downloader.report_warning(error_msg)
+                self.report_warning(error_msg)
                 return
-            code = self._get_player_code(video_id, player_url)
+            code = self._load_player(video_id, player_url, fatal=fatal)
             sts = int_or_none(self._search_regex(
                 r'(?:signatureTimestamp|sts)\s*:\s*(?P<sts>[0-9]{5})', code or '',
                 'JS player signature timestamp', group='sts', fatal=fatal))
@@ -1730,12 +1759,18 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
         # cpn generation algorithm is reverse engineered from base.js.
         # In fact it works even with dummy cpn.
         CPN_ALPHABET = 'abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789-_'
-        cpn = ''.join((CPN_ALPHABET[random.randint(0, 256) & 63] for _ in range(0, 16)))
+        cpn = ''.join(CPN_ALPHABET[random.randint(0, 256) & 63] for _ in range(0, 16))
 
-        playback_url = update_url(
-            playback_url, query_update={
-                'ver': ['2'],
-                'cpn': [cpn],
+        # more consistent results setting it to right before the end
+        qs = parse_qs(playback_url)
+        video_length = '{0}'.format(float((qs.get('len') or ['1.5'])[0]) - 1)
+
+        playback_url = update_url_query(
+            playback_url, {
+                'ver': '2',
+                'cpn': cpn,
+                'cmt': video_length,
+                'el': 'detailpage',  # otherwise defaults to "shorts"
             })
 
         self._download_webpage(
@@ -1983,8 +2018,11 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
             else:
                 self.to_screen('Downloading just video %s because of --no-playlist' % video_id)
 
+        if not player_url:
+            player_url = self._extract_player_url(webpage)
+
         formats = []
-        itags = []
+        itags = collections.defaultdict(set)
         itag_qualities = {}
         q = qualities(['tiny', 'small', 'medium', 'large', 'hd720', 'hd1080', 'hd1440', 'hd2160', 'hd2880', 'highres'])
         CHUNK_SIZE = 10 << 20
@@ -2000,58 +2038,92 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                 })
             } for range_start in range(0, f['filesize'], CHUNK_SIZE))
 
+        lower = lambda s: s.lower()
+
         for fmt in streaming_formats:
-            if fmt.get('targetDurationSec') or fmt.get('drmFamilies'):
+            if fmt.get('targetDurationSec'):
                 continue
 
             itag = str_or_none(fmt.get('itag'))
-            quality = fmt.get('quality')
-            if itag and quality:
+            audio_track = traverse_obj(fmt, ('audioTrack', T(dict))) or {}
+
+            quality = traverse_obj(fmt, ((
+                # The 3gp format (17) in android client has a quality of "small",
+                # but is actually worse than other formats
+                T(lambda _: 'tiny' if itag == 17 else None),
+                ('quality', T(lambda q: q if q and q != 'tiny' else None)),
+                ('audioQuality', T(lower)),
+                'quality'), T(txt_or_none)), get_all=False)
+            if quality and itag:
                 itag_qualities[itag] = quality
             # FORMAT_STREAM_TYPE_OTF(otf=1) requires downloading the init fragment
             # (adding `&sq=0` to the URL) and parsing emsg box to determine the
-            # number of fragment that would subsequently requested with (`&sq=N`)
+            # number of fragments that would subsequently be requested with (`&sq=N`)
             if fmt.get('type') == 'FORMAT_STREAM_TYPE_OTF':
                 continue
 
             fmt_url = fmt.get('url')
             if not fmt_url:
                 sc = compat_parse_qs(fmt.get('signatureCipher'))
-                fmt_url = url_or_none(try_get(sc, lambda x: x['url'][0]))
-                encrypted_sig = try_get(sc, lambda x: x['s'][0])
-                if not (sc and fmt_url and encrypted_sig):
+                fmt_url = traverse_obj(sc, ('url', -1, T(url_or_none)))
+                encrypted_sig = traverse_obj(sc, ('s', -1))
+                if not (fmt_url and encrypted_sig):
                     continue
-                if not player_url:
-                    player_url = self._extract_player_url(webpage)
+                player_url = player_url or self._extract_player_url(webpage)
                 if not player_url:
                     continue
-                signature = self._decrypt_signature(sc['s'][0], video_id, player_url)
-                sp = try_get(sc, lambda x: x['sp'][0]) or 'signature'
-                fmt_url += '&' + sp + '=' + signature
+                try:
+                    fmt_url = update_url_query(fmt_url, {
+                        traverse_obj(sc, ('sp', -1)) or 'signature':
+                            [self._decrypt_signature(encrypted_sig, video_id, player_url)],
+                    })
+                except ExtractorError as e:
+                    self.report_warning('Signature extraction failed: Some formats may be missing',
+                                        video_id=video_id, only_once=True)
+                    self.write_debug(error_to_compat_str(e), only_once=True)
+                    continue
 
-            if itag:
-                itags.append(itag)
-            tbr = float_or_none(
-                fmt.get('averageBitrate') or fmt.get('bitrate'), 1000)
+            language_preference = (
+                10 if audio_track.get('audioIsDefault')
+                else -10 if 'descriptive' in (traverse_obj(audio_track, ('displayName', T(lower))) or '')
+                else -1)
+            name = (
+                traverse_obj(fmt, ('qualityLabel', T(txt_or_none)))
+                or quality.replace('audio_quality_', ''))
             dct = {
-                'asr': int_or_none(fmt.get('audioSampleRate')),
-                'filesize': int_or_none(fmt.get('contentLength')),
-                'format_id': itag,
-                'format_note': fmt.get('qualityLabel') or quality,
-                'fps': int_or_none(fmt.get('fps')),
-                'height': int_or_none(fmt.get('height')),
-                'quality': q(quality),
-                'tbr': tbr,
+                'format_id': join_nonempty(itag, fmt.get('isDrc') and 'drc'),
                 'url': fmt_url,
-                'width': fmt.get('width'),
+                # Format 22 is likely to be damaged: see https://github.com/yt-dlp/yt-dlp/issues/3372
+                'source_preference': ((-5 if itag == '22' else -1)
+                                      + (100 if 'Premium' in name else 0)),
+                'quality': q(quality),
+                'language': join_nonempty(audio_track.get('id', '').split('.')[0],
+                                          'desc' if language_preference < -1 else '') or None,
+                'language_preference': language_preference,
+                # Strictly de-prioritize 3gp formats
+                'preference': -2 if itag == '17' else None,
             }
-            mimetype = fmt.get('mimeType')
-            if mimetype:
-                mobj = re.match(
-                    r'((?:[^/]+)/(?:[^;]+))(?:;\s*codecs="([^"]+)")?', mimetype)
-                if mobj:
-                    dct['ext'] = mimetype2ext(mobj.group(1))
-                    dct.update(parse_codecs(mobj.group(2)))
+            if itag:
+                itags[itag].add(('https', dct.get('language')))
+            self._unthrottle_format_urls(video_id, player_url, dct)
+            dct.update(traverse_obj(fmt, {
+                'asr': ('audioSampleRate', T(int_or_none)),
+                'filesize': ('contentLength', T(int_or_none)),
+                'format_note': ('qualityLabel', T(lambda x: x or quality)),
+                # for some formats, fps is wrongly returned as 1
+                'fps': ('fps', T(int_or_none), T(lambda f: f if f > 1 else None)),
+                'audio_channels': ('audioChannels', T(int_or_none)),
+                'height': ('height', T(int_or_none)),
+                'has_drm': ('drmFamilies', T(bool)),
+                'tbr': (('averageBitrate', 'bitrate'), T(lambda t: float_or_none(t, 1000))),
+                'width': ('width', T(int_or_none)),
+                '_duration_ms': ('approxDurationMs', T(int_or_none)),
+            }, get_all=False))
+            mime_mobj = re.match(
+                r'((?:[^/]+)/(?:[^;]+))(?:;\s*codecs="([^"]+)")?', fmt.get('mimeType') or '')
+            if mime_mobj:
+                dct['ext'] = mimetype2ext(mime_mobj.group(1))
+                dct.update(parse_codecs(mime_mobj.group(2)))
             single_stream = 'none' in (dct.get(c) for c in ('acodec', 'vcodec'))
             if single_stream and dct.get('ext'):
                 dct['container'] = dct['ext'] + '_dash'
@@ -2066,32 +2138,62 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
 
             formats.append(dct)
 
+        def process_manifest_format(f, proto, client_name, itag, all_formats=False):
+            key = (proto, f.get('language'))
+            if not all_formats and key in itags[itag]:
+                return False
+            itags[itag].add(key)
+
+            if itag:
+                f['format_id'] = (
+                    '{0}-{1}'.format(itag, proto)
+                    if all_formats or any(p != proto for p, _ in itags[itag])
+                    else itag)
+
+            if f.get('source_preference') is None:
+                f['source_preference'] = -1
+
+            if itag in ('616', '235'):
+                f['format_note'] = join_nonempty(f.get('format_note'), 'Premium', delim=' ')
+                f['source_preference'] += 100
+
+            f['quality'] = q(traverse_obj(f, (
+                'format_id', T(lambda s: itag_qualities[s.split('-')[0]])), default=-1))
+            if try_call(lambda: f['fps'] <= 1):
+                del f['fps']
+
+            if proto == 'hls' and f.get('has_drm'):
+                f['has_drm'] = 'maybe'
+                f['source_preference'] -= 5
+            return True
+
         hls_manifest_url = streaming_data.get('hlsManifestUrl')
         if hls_manifest_url:
             for f in self._extract_m3u8_formats(
                     hls_manifest_url, video_id, 'mp4', fatal=False):
-                itag = self._search_regex(
-                    r'/itag/(\d+)', f['url'], 'itag', default=None)
-                if itag:
-                    f['format_id'] = itag
-                formats.append(f)
+                if process_manifest_format(
+                        f, 'hls', None, self._search_regex(
+                            r'/itag/(\d+)', f['url'], 'itag', default=None)):
+                    formats.append(f)
 
         if self._downloader.params.get('youtube_include_dash_manifest', True):
             dash_manifest_url = streaming_data.get('dashManifestUrl')
             if dash_manifest_url:
                 for f in self._extract_mpd_formats(
                         dash_manifest_url, video_id, fatal=False):
-                    itag = f['format_id']
-                    if itag in itags:
-                        continue
-                    if itag in itag_qualities:
-                        f['quality'] = q(itag_qualities[itag])
-                    filesize = int_or_none(self._search_regex(
-                        r'/clen/(\d+)', f.get('fragment_base_url')
-                        or f['url'], 'file size', default=None))
-                    if filesize:
-                        f['filesize'] = filesize
-                    formats.append(f)
+                    if process_manifest_format(
+                            f, 'dash', None, f['format_id']):
+                        f['filesize'] = traverse_obj(f, (
+                            ('fragment_base_url', 'url'), T(lambda u: self._search_regex(
+                                r'/clen/(\d+)', u, 'file size', default=None)),
+                            T(int_or_none)), get_all=False)
+                        formats.append(f)
+
+        playable_formats = [f for f in formats if not f.get('has_drm')]
+        if formats and not playable_formats:
+            # If there are no formats that definitely don't have DRM, all have DRM
+            self.report_drm(video_id)
+        formats[:] = playable_formats
 
         if not formats:
             if streaming_data.get('licenseInfos'):
@@ -2163,6 +2265,17 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
             video_details.get('lengthSeconds')
             or microformat.get('lengthSeconds')) \
             or parse_duration(search_meta('duration'))
+
+        for f in formats:
+            # Some formats may have much smaller duration than others (possibly damaged during encoding)
+            # but avoid false positives with small duration differences.
+            # Ref: https://github.com/yt-dlp/yt-dlp/issues/2823
+            if try_call(lambda x: float(x.pop('_duration_ms')) / duration < 500, args=(f,)):
+                self.report_warning(
+                    '{0}: Some possibly damaged formats will be deprioritized'.format(video_id), only_once=True)
+                # Strictly de-prioritize damaged formats
+                f['preference'] = -10
+
         is_live = video_details.get('isLive')
 
         owner_profile_url = self._yt_urljoin(self._extract_author_var(
@@ -2171,10 +2284,6 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
         uploader = self._extract_author_var(
             webpage, 'name', videodetails=video_details, metadata=microformat)
 
-        if not player_url:
-            player_url = self._extract_player_url(webpage)
-        self._unthrottle_format_urls(video_id, player_url, formats)
-
         info = {
             'id': video_id,
             'title': self._live_title(video_title) if is_live else video_title,
@@ -2367,6 +2476,14 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                             'like_count': str_to_int(like_count),
                             'dislike_count': str_to_int(dislike_count),
                         })
+                    else:
+                        info['like_count'] = traverse_obj(vpir, (
+                            'videoActions', 'menuRenderer', 'topLevelButtons', Ellipsis,
+                            'segmentedLikeDislikeButtonViewModel', 'likeButtonViewModel', 'likeButtonViewModel',
+                            'toggleButtonViewModel', 'toggleButtonViewModel', 'defaultButtonViewModel',
+                            'buttonViewModel', (('title', ('accessibilityText', T(lambda s: s.split()), Ellipsis))), T(parse_count)),
+                            get_all=False)
+
                 vsir = content.get('videoSecondaryInfoRenderer')
                 if vsir:
                     rows = try_get(
@@ -2481,7 +2598,7 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor):
         'playlist_mincount': 94,
         'info_dict': {
             'id': 'UCqj7Cz7revf5maW9g5pgNcg',
-            'title': 'Igor Kleiner - Playlists',
+            'title': r're:Igor Kleiner(?: Ph\.D\.)? - Playlists',
             'description': 'md5:be97ee0f14ee314f1f002cf187166ee2',
             'uploader': 'Igor Kleiner',
             'uploader_id': '@IgorDataScience',
@@ -2492,7 +2609,7 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor):
         'playlist_mincount': 94,
         'info_dict': {
             'id': 'UCqj7Cz7revf5maW9g5pgNcg',
-            'title': 'Igor Kleiner - Playlists',
+            'title': r're:Igor Kleiner(?: Ph\.D\.)? - Playlists',
             'description': 'md5:be97ee0f14ee314f1f002cf187166ee2',
             'uploader': 'Igor Kleiner',
             'uploader_id': '@IgorDataScience',
@@ -2604,7 +2721,7 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor):
         'url': 'https://www.youtube.com/channel/UCKfVa3S1e4PHvxWcwyMMg8w/channels',
         'info_dict': {
             'id': 'UCKfVa3S1e4PHvxWcwyMMg8w',
-            'title': 'lex will - Channels',
+            'title': r're:lex will - (?:Home|Channels)',
             'description': 'md5:2163c5d0ff54ed5f598d6a7e6211e488',
             'uploader': 'lex will',
             'uploader_id': '@lexwill718',

youtube_dl/jsinterp.py

@@ -14,6 +14,7 @@ from .utils import (
     remove_quotes,
     unified_timestamp,
     variadic,
+    write_string,
 )
 from .compat import (
     compat_basestring,
@@ -53,15 +54,16 @@ def wraps_op(op):
 
 # NB In principle NaN cannot be checked by membership.
 # Here all NaN values are actually this one, so _NaN is _NaN,
-# although _NaN != _NaN.
+# although _NaN != _NaN. Ditto Infinity.
 
 _NaN = float('nan')
+_Infinity = float('inf')
 
 
 def _js_bit_op(op):
 
     def zeroise(x):
-        return 0 if x in (None, JS_Undefined, _NaN) else x
+        return 0 if x in (None, JS_Undefined, _NaN, _Infinity) else x
 
     @wraps_op(op)
     def wrapped(a, b):
@@ -84,7 +86,7 @@ def _js_arith_op(op):
 def _js_div(a, b):
     if JS_Undefined in (a, b) or not (a or b):
         return _NaN
-    return operator.truediv(a or 0, b) if b else float('inf')
+    return operator.truediv(a or 0, b) if b else _Infinity
 
 
 def _js_mod(a, b):
@@ -220,6 +222,42 @@ class LocalNameSpace(ChainMap):
         return 'LocalNameSpace%s' % (self.maps, )
 
 
+class Debugger(object):
+    ENABLED = False
+
+    @staticmethod
+    def write(*args, **kwargs):
+        level = kwargs.get('level', 100)
+
+        def truncate_string(s, left, right=0):
+            if s is None or len(s) <= left + right:
+                return s
+            return '...'.join((s[:left - 3], s[-right:] if right else ''))
+
+        write_string('[debug] JS: {0}{1}\n'.format(
+            '  ' * (100 - level),
+            ' '.join(truncate_string(compat_str(x), 50, 50) for x in args)))
+
+    @classmethod
+    def wrap_interpreter(cls, f):
+        def interpret_statement(self, stmt, local_vars, allow_recursion, *args, **kwargs):
+            if cls.ENABLED and stmt.strip():
+                cls.write(stmt, level=allow_recursion)
+            try:
+                ret, should_ret = f(self, stmt, local_vars, allow_recursion, *args, **kwargs)
+            except Exception as e:
+                if cls.ENABLED:
+                    if isinstance(e, ExtractorError):
+                        e = e.orig_msg
+                    cls.write('=> Raises:', e, '<-|', stmt, level=allow_recursion)
+                raise
+            if cls.ENABLED and stmt.strip():
+                if should_ret or not repr(ret) == stmt:
+                    cls.write(['->', '=>'][should_ret], repr(ret), '<-|', stmt, level=allow_recursion)
+            return ret, should_ret
+        return interpret_statement
+
+
 class JSInterpreter(object):
     __named_object_counter = 0
 
@@ -277,9 +315,20 @@ class JSInterpreter(object):
 
         def __getattr__(self, name):
             self.__instantiate()
-            if hasattr(self, name):
-                return getattr(self, name)
-            return super(JSInterpreter.JS_RegExp, self).__getattr__(name)
+            # make Py 2.6 conform to its lying documentation
+            if name == 'flags':
+                self.flags = self.__flags
+                return self.flags
+            elif name == 'pattern':
+                self.pattern = self.__pattern_txt
+                return self.pattern
+            elif hasattr(self.__self, name):
+                v = getattr(self.__self, name)
+                setattr(self, name, v)
+                return v
+            elif name in ('groupindex', 'groups'):
+                return 0 if name == 'groupindex' else {}
+            raise AttributeError('{0} has no attribute named {1}'.format(self, name))
 
         @classmethod
         def regex_flags(cls, expr):
@@ -296,8 +345,7 @@ class JSInterpreter(object):
     def __op_chars(cls):
         op_chars = set(';,[')
         for op in cls._all_operators():
-            for c in op[0]:
-                op_chars.add(c)
+            op_chars.update(op[0])
         return op_chars
 
     def _named_object(self, namespace, obj):
@@ -315,9 +363,8 @@ class JSInterpreter(object):
         # collections.Counter() is ~10% slower in both 2.7 and 3.9
         counters = dict((k, 0) for k in _MATCHING_PARENS.values())
         start, splits, pos, delim_len = 0, 0, 0, len(delim) - 1
-        in_quote, escaping, skipping = None, False, 0
-        after_op, in_regex_char_group = True, False
-
+        in_quote, escaping, after_op, in_regex_char_group = None, False, True, False
+        skipping = 0
         for idx, char in enumerate(expr):
             paren_delta = 0
             if not in_quote:
@@ -371,10 +418,12 @@ class JSInterpreter(object):
         return separated[0][1:].strip(), separated[1].strip()
 
     @staticmethod
-    def _all_operators():
-        return itertools.chain(
-            # Ref: https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Operators/Operator_Precedence
-            _SC_OPERATORS, _LOG_OPERATORS, _COMP_OPERATORS, _OPERATORS)
+    def _all_operators(_cached=[]):
+        if not _cached:
+            _cached.extend(itertools.chain(
+                # Ref: https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Operators/Operator_Precedence
+                _SC_OPERATORS, _LOG_OPERATORS, _COMP_OPERATORS, _OPERATORS))
+        return _cached
 
     def _operator(self, op, left_val, right_expr, expr, local_vars, allow_recursion):
         if op in ('||', '&&'):
@@ -405,7 +454,7 @@ class JSInterpreter(object):
         except Exception as e:
             if allow_undefined:
                 return JS_Undefined
-            raise self.Exception('Cannot get index {idx:.100}'.format(**locals()), expr=repr(obj), cause=e)
+            raise self.Exception('Cannot get index {idx!r:.100}'.format(**locals()), expr=repr(obj), cause=e)
 
     def _dump(self, obj, namespace):
         try:
@@ -427,6 +476,7 @@ class JSInterpreter(object):
     _FINALLY_RE = re.compile(r'finally\s*\{')
     _SWITCH_RE = re.compile(r'switch\s*\(')
 
+    @Debugger.wrap_interpreter
     def interpret_statement(self, stmt, local_vars, allow_recursion=100):
         if allow_recursion < 0:
             raise self.Exception('Recursion limit reached')
@@ -500,7 +550,6 @@ class JSInterpreter(object):
                 expr = self._dump(inner, local_vars) + outer
 
         if expr.startswith('('):
-
             m = re.match(r'\((?P<d>[a-z])%(?P<e>[a-z])\.length\+(?P=e)\.length\)%(?P=e)\.length', expr)
             if m:
                 # short-cut eval of frequently used `(d%e.length+e.length)%e.length`, worth ~6% on `pytest -k test_nsig`
@@ -682,7 +731,7 @@ class JSInterpreter(object):
                 (?P<op>{_OPERATOR_RE})?
                 =(?!=)(?P<expr>.*)$
             )|(?P<return>
-                (?!if|return|true|false|null|undefined)(?P<name>{_NAME_RE})$
+                (?!if|return|true|false|null|undefined|NaN|Infinity)(?P<name>{_NAME_RE})$
             )|(?P<indexing>
                 (?P<in>{_NAME_RE})\[(?P<idx>.+)\]$
             )|(?P<attribute>
@@ -716,11 +765,12 @@ class JSInterpreter(object):
             raise JS_Break()
         elif expr == 'continue':
             raise JS_Continue()
-
         elif expr == 'undefined':
             return JS_Undefined, should_return
         elif expr == 'NaN':
             return _NaN, should_return
+        elif expr == 'Infinity':
+            return _Infinity, should_return
 
         elif md.get('return'):
             return local_vars[m.group('name')], should_return
@@ -749,18 +799,28 @@ class JSInterpreter(object):
             right_expr = separated.pop()
             # handle operators that are both unary and binary, minimal BODMAS
             if op in ('+', '-'):
+                # simplify/adjust consecutive instances of these operators
                 undone = 0
                 while len(separated) > 1 and not separated[-1].strip():
                     undone += 1
                     separated.pop()
                 if op == '-' and undone % 2 != 0:
                     right_expr = op + right_expr
+                elif op == '+':
+                    while len(separated) > 1 and separated[-1].strip() in self.OP_CHARS:
+                        right_expr = separated.pop() + right_expr
+                # hanging op at end of left => unary + (strip) or - (push right)
                 left_val = separated[-1]
                 for dm_op in ('*', '%', '/', '**'):
                     bodmas = tuple(self._separate(left_val, dm_op, skip_delims=skip_delim))
                     if len(bodmas) > 1 and not bodmas[-1].strip():
                         expr = op.join(separated) + op + right_expr
-                        right_expr = None
+                        if len(separated) > 1:
+                            separated.pop()
+                            right_expr = op.join((left_val, right_expr))
+                        else:
+                            separated = [op.join((left_val, right_expr))]
+                            right_expr = None
                         break
                 if right_expr is None:
                     continue
@@ -786,6 +846,8 @@ class JSInterpreter(object):
 
             def eval_method():
                 if (variable, member) == ('console', 'debug'):
+                    if Debugger.ENABLED:
+                        Debugger.write(self.interpret_expression('[{}]'.format(arg_str), local_vars, allow_recursion))
                     return
                 types = {
                     'String': compat_str,
@@ -940,16 +1002,19 @@ class JSInterpreter(object):
     def extract_object(self, objname):
         _FUNC_NAME_RE = r'''(?:[a-zA-Z$0-9]+|"[a-zA-Z$0-9]+"|'[a-zA-Z$0-9]+')'''
         obj = {}
-        obj_m = re.search(
-            r'''(?x)
-                (?<!this\.)%s\s*=\s*{\s*
-                    (?P<fields>(%s\s*:\s*function\s*\(.*?\)\s*{.*?}(?:,\s*)?)*)
-                }\s*;
-            ''' % (re.escape(objname), _FUNC_NAME_RE),
-            self.code)
-        if not obj_m:
+        fields = None
+        for obj_m in re.finditer(
+                r'''(?xs)
+                    {0}\s*\.\s*{1}|{1}\s*=\s*\{{\s*
+                        (?P<fields>({2}\s*:\s*function\s*\(.*?\)\s*\{{.*?}}(?:,\s*)?)*)
+                    }}\s*;
+                '''.format(_NAME_RE, re.escape(objname), _FUNC_NAME_RE),
+                self.code):
+            fields = obj_m.group('fields')
+            if fields:
+                break
+        else:
             raise self.Exception('Could not find object ' + objname)
-        fields = obj_m.group('fields')
         # Currently, it only supports function definitions
         fields_m = re.finditer(
             r'''(?x)
@@ -985,9 +1050,9 @@ class JSInterpreter(object):
                 \((?P<args>[^)]*)\)\s*
                 (?P<code>{.+})''' % {'name': re.escape(funcname)},
             self.code)
-        code, _ = self._separate_at_paren(func_m.group('code'))  # refine the match
         if func_m is None:
             raise self.Exception('Could not find JS function "{funcname}"'.format(**locals()))
+        code, _ = self._separate_at_paren(func_m.group('code'))  # refine the match
         return self.build_arglist(func_m.group('args')), code
 
     def extract_function(self, funcname):

youtube_dl/options.py

@@ -544,12 +544,14 @@ def parseOpts(overrideArguments=None):
     workarounds.add_option(
         '--referer',
         metavar='URL', dest='referer', default=None,
-        help='Specify a custom referer, use if the video access is restricted to one domain',
+        help='Specify a custom Referer: use if the video access is restricted to one domain',
     )
     workarounds.add_option(
         '--add-header',
         metavar='FIELD:VALUE', dest='headers', action='append',
-        help='Specify a custom HTTP header and its value, separated by a colon \':\'. You can use this option multiple times',
+        help=('Specify a custom HTTP header and its value, separated by a colon \':\'. You can use this option multiple times. '
+              'NB Use --cookies rather than adding a Cookie header if its contents may be sensitive; '
+              'data from a Cookie header will be sent to all domains, not just the one intended')
     )
     workarounds.add_option(
         '--bidi-workaround',

youtube_dl/postprocessor/embedthumbnail.py

@@ -18,6 +18,8 @@ from ..utils import (
     shell_quote,
 )
 
+from ..compat import compat_open as open
+
 
 class EmbedThumbnailPPError(PostProcessingError):
     pass

youtube_dl/postprocessor/ffmpeg.py

@@ -1,6 +1,5 @@
 from __future__ import unicode_literals
 
-import io
 import os
 import subprocess
 import time
@@ -9,6 +8,7 @@ import re
 
 from .common import AudioConversionError, PostProcessor
 
+from ..compat import compat_open as open
 from ..utils import (
     encodeArgument,
     encodeFilename,
@@ -74,8 +74,11 @@ class FFmpegPostProcessor(PostProcessor):
         return FFmpegPostProcessor(downloader)._versions
 
     def _determine_executables(self):
-        programs = ['avprobe', 'avconv', 'ffmpeg', 'ffprobe']
+        # ordered to match prefer_ffmpeg!
+        convs = ['ffmpeg', 'avconv']
+        probes = ['ffprobe', 'avprobe']
         prefer_ffmpeg = True
+        programs = convs + probes
 
         def get_ffmpeg_version(path):
             ver = get_exe_version(path, args=['-version'])
@@ -96,6 +99,7 @@ class FFmpegPostProcessor(PostProcessor):
 
         self._paths = None
         self._versions = None
+        location = None
         if self._downloader:
             prefer_ffmpeg = self._downloader.params.get('prefer_ffmpeg', True)
             location = self._downloader.params.get('ffmpeg_location')
@@ -118,33 +122,21 @@ class FFmpegPostProcessor(PostProcessor):
                     location = os.path.dirname(os.path.abspath(location))
                     if basename in ('ffmpeg', 'ffprobe'):
                         prefer_ffmpeg = True
+        self._paths = dict(
+            (p, p if location is None else os.path.join(location, p))
+            for p in programs)
+        self._versions = dict(
+            x for x in (
+                (p, get_ffmpeg_version(self._paths[p])) for p in programs)
+            if x[1] is not None)
 
-                self._paths = dict(
-                    (p, os.path.join(location, p)) for p in programs)
-                self._versions = dict(
-                    (p, get_ffmpeg_version(self._paths[p])) for p in programs)
-        if self._versions is None:
-            self._versions = dict(
-                (p, get_ffmpeg_version(p)) for p in programs)
-            self._paths = dict((p, p) for p in programs)
-
-        if prefer_ffmpeg is False:
-            prefs = ('avconv', 'ffmpeg')
-        else:
-            prefs = ('ffmpeg', 'avconv')
-        for p in prefs:
-            if self._versions[p]:
-                self.basename = p
-                break
-
-        if prefer_ffmpeg is False:
-            prefs = ('avprobe', 'ffprobe')
-        else:
-            prefs = ('ffprobe', 'avprobe')
-        for p in prefs:
-            if self._versions[p]:
-                self.probe_basename = p
-                break
+        basenames = [None, None]
+        for i, progs in enumerate((convs, probes)):
+            for p in progs[::-1 if prefer_ffmpeg is False else 1]:
+                if self._versions.get(p):
+                    basenames[i] = p
+                    break
+        self.basename, self.probe_basename = basenames
 
     @property
     def available(self):
@@ -493,7 +485,7 @@ class FFmpegMetadataPP(FFmpegPostProcessor):
         chapters = info.get('chapters', [])
         if chapters:
             metadata_filename = replace_extension(filename, 'meta')
-            with io.open(metadata_filename, 'wt', encoding='utf-8') as f:
+            with open(metadata_filename, 'w', encoding='utf-8') as f:
                 def ffmpeg_escape(text):
                     return re.sub(r'(=|;|#|\\|\n)', r'\\\1', text)
 
@@ -636,7 +628,7 @@ class FFmpegSubtitlesConvertorPP(FFmpegPostProcessor):
                 with open(dfxp_file, 'rb') as f:
                     srt_data = dfxp2srt(f.read())
 
-                with io.open(srt_file, 'wt', encoding='utf-8') as f:
+                with open(srt_file, 'w', encoding='utf-8') as f:
                     f.write(srt_data)
                 old_file = srt_file
 
@@ -652,7 +644,7 @@ class FFmpegSubtitlesConvertorPP(FFmpegPostProcessor):
 
             self.run_ffmpeg(old_file, new_file, ['-f', new_format])
 
-            with io.open(new_file, 'rt', encoding='utf-8') as f:
+            with open(new_file, 'r', encoding='utf-8') as f:
                 subs[lang] = {
                     'ext': new_ext,
                     'data': f.read(),

youtube_dl/swfinterp.py

@@ -727,7 +727,7 @@ class SWFInterpreter(object):
                             stack.append(res)
                             continue
 
-                        assert isinstance(obj, (dict, _ScopeDict)),\
+                        assert isinstance(obj, (dict, _ScopeDict)), \
                             'Accessing member %r on %r' % (pname, obj)
                         res = obj.get(pname, undefined)
                         stack.append(res)

youtube_dl/traversal.py

@@ -0,0 +1,10 @@
+# coding: utf-8
+
+# TODO: move these utils.fns here and move import to utils
+# flake8: noqa
+from .utils import (
+    dict_get,
+    get_first,
+    T,
+    traverse_obj,
+)

youtube_dl/update.py

@@ -1,6 +1,5 @@
 from __future__ import unicode_literals
 
-import io
 import json
 import traceback
 import hashlib
@@ -9,7 +8,10 @@ import subprocess
 import sys
 from zipimport import zipimporter
 
-from .compat import compat_realpath
+from .compat import (
+    compat_open as open,
+    compat_realpath,
+)
 from .utils import encode_compat_str
 
 from .version import __version__
@@ -127,7 +129,7 @@ def update_self(to_screen, verbose, opener):
 
         try:
             bat = os.path.join(directory, 'youtube-dl-updater.bat')
-            with io.open(bat, 'w') as batfile:
+            with open(bat, 'w') as batfile:
                 batfile.write('''
 @echo off
 echo Waiting for file handle to be closed ...