Merge branch 'dev'

docker/compose/docker-compose.ci-test.yml

@@ -11,9 +11,12 @@ services:
     container_name: gotenberg
     network_mode: host
     restart: unless-stopped
+    # The gotenberg chromium route is used to convert .eml files. We do not
+    # want to allow external content like tracking pixels or even javascript.
     command:
       - "gotenberg"
-      - "--chromium-disable-routes=true"
+      - "--chromium-disable-javascript=true"
+      - "--chromium-allow-list=file:///tmp/.*"
   tika:
     image: ghcr.io/paperless-ngx/tika:latest
     hostname: tika

docker/compose/docker-compose.mariadb-tika.yml

@@ -85,9 +85,12 @@ services:
   gotenberg:
     image: docker.io/gotenberg/gotenberg:7.6
     restart: unless-stopped
+    # The gotenberg chromium route is used to convert .eml files. We do not
+    # want to allow external content like tracking pixels or even javascript.
     command:
       - "gotenberg"
-      - "--chromium-disable-routes=true"
+      - "--chromium-disable-javascript=true"
+      - "--chromium-allow-list=file:///tmp/.*"
 
   tika:
     image: ghcr.io/paperless-ngx/tika:latest

docker/compose/docker-compose.postgres-tika.yml

@@ -79,9 +79,13 @@ services:
   gotenberg:
     image: docker.io/gotenberg/gotenberg:7.6
     restart: unless-stopped
+
+    # The gotenberg chromium route is used to convert .eml files. We do not
+    # want to allow external content like tracking pixels or even javascript.
     command:
       - "gotenberg"
-      - "--chromium-disable-routes=true"
+      - "--chromium-disable-javascript=true"
+      - "--chromium-allow-list=file:///tmp/.*"
 
   tika:
     image: ghcr.io/paperless-ngx/tika:latest

docker/compose/docker-compose.sqlite-tika.yml

@@ -67,9 +67,13 @@ services:
   gotenberg:
     image: docker.io/gotenberg/gotenberg:7.6
     restart: unless-stopped
+
+    # The gotenberg chromium route is used to convert .eml files. We do not
+    # want to allow external content like tracking pixels or even javascript.
     command:
       - "gotenberg"
-      - "--chromium-disable-routes=true"
+      - "--chromium-disable-javascript=true"
+      - "--chromium-allow-list=file:///tmp/.*"
 
   tika:
     image: ghcr.io/paperless-ngx/tika:latest

docker/docker-entrypoint.sh

@@ -53,30 +53,6 @@ map_folders() {
 	export CONSUME_DIR="${PAPERLESS_CONSUMPTION_DIR:-/usr/src/paperless/consume}"
 }
 
-nltk_data () {
-	# Store the NLTK data outside the Docker container
-	local -r nltk_data_dir="${DATA_DIR}/nltk"
-	local -r truthy_things=("yes y 1 t true")
-
-	# If not set, or it looks truthy
-	if [[ -z "${PAPERLESS_ENABLE_NLTK}" ]] || [[ "${truthy_things[*]}" =~ ${PAPERLESS_ENABLE_NLTK,} ]]; then
-
-		# Download or update the snowball stemmer data
-		python3 -W ignore::RuntimeWarning -m nltk.downloader -d "${nltk_data_dir}" snowball_data
-
-		# Download or update the stopwords corpus
-		python3 -W ignore::RuntimeWarning -m nltk.downloader -d "${nltk_data_dir}" stopwords
-
-		# Download or update the punkt tokenizer data
-		python3 -W ignore::RuntimeWarning -m nltk.downloader -d "${nltk_data_dir}" punkt
-
-	else
-		echo "Skipping NLTK data download"
-
-	fi
-
-}
-
 custom_container_init() {
 	# Mostly borrowed from the LinuxServer.io base image
 	# https://github.com/linuxserver/docker-baseimage-ubuntu/tree/bionic/root/etc/cont-init.d
@@ -157,8 +133,6 @@ initialize() {
 	echo "Creating directory ${tmp_dir}"
 	mkdir -p "${tmp_dir}"
 
-	nltk_data
-
 	set +e
 	echo "Adjusting permissions of paperless files. This may take a while."
 	chown -R paperless:paperless ${tmp_dir}
@@ -191,10 +165,6 @@ install_languages() {
 
 	for lang in "${langs[@]}"; do
 		pkg="tesseract-ocr-$lang"
-		# English is installed by default
-		#if [[ "$lang" ==  "eng" ]]; then
-		#    continue
-		#fi
 
 		if dpkg -s "$pkg" &>/dev/null; then
 			echo "Package $pkg already installed!"

docker/docker-prepare.sh

@@ -20,7 +20,6 @@ wait_for_postgres() {
 			exit 1
 		else
 			echo "Attempt $attempt_num failed! Trying again in 5 seconds..."
-
 		fi
 
 		attempt_num=$(("$attempt_num" + 1))
@@ -37,6 +36,8 @@ wait_for_mariadb() {
 	local attempt_num=1
 	local -r max_attempts=5
 
+	# Disable warning, host and port can't have spaces
+	# shellcheck disable=SC2086
 	while ! true > /dev/tcp/$host/$port; do
 
 		if [ $attempt_num -eq $max_attempts ]; then
@@ -67,10 +68,16 @@ migrations() {
 		# of the current container starts.
 		flock 200
 		echo "Apply database migrations..."
-		python3 manage.py migrate
+		python3 manage.py migrate --skip-checks --no-input
 	) 200>"${DATA_DIR}/migration_lock"
 }
 
+django_checks() {
+	# Explicitly run the Django system checks
+	echo "Running Django checks"
+	python3 manage.py check
+}
+
 search_index() {
 
 	local -r index_version=1
@@ -100,6 +107,8 @@ do_work() {
 
 	migrations
 
+	django_checks
+
 	search_index
 
 	superuser