Add the new paperless_tika parser

This parser will use an external Tika and Gotenberg server to parse
"Office" documents (.doc, .xls, .odt, etc.)

Signed-off-by: Jo Vandeginste <Jo.Vandeginste@kuleuven.be>

Pipfile

@@ -42,6 +42,7 @@ whoosh="~=2.7.4"
 inotifyrecursive = "~=0.3.4"
 ocrmypdf = "*"
 tqdm = "*"
+tika = "*"
 
 [dev-packages]
 coveralls = "*"

docker/hub/docker-compose.tika.yml

@@ -0,0 +1,43 @@
+version: "3.4"
+services:
+  broker:
+    image: redis:6.0
+    restart: always
+
+  webserver:
+    image: jonaswinkler/paperless-ng:0.9.9
+    restart: always
+    depends_on:
+      - broker
+    ports:
+      - 8000:8000
+    healthcheck:
+      test: ["CMD", "curl", "-f", "http://localhost:8000"]
+      interval: 30s
+      timeout: 10s
+      retries: 5
+    volumes:
+      - data:/usr/src/paperless/data
+      - media:/usr/src/paperless/media
+      - ./export:/usr/src/paperless/export
+      - ./consume:/usr/src/paperless/consume
+    env_file: docker-compose.env
+    environment:
+      PAPERLESS_REDIS: redis://broker:6379
+      PAPERLESS_TIKA: 1
+      GOTENBERG_SERVER_ENDPOINT: http://gotenberg:3000
+      TIKA_SERVER_ENDPOINT: http://tika:9998
+
+  gotenberg:
+    image: thecodingmachine/gotenberg
+    restart: unless-stopped
+    environment:
+      DISABLE_GOOGLE_CHROME: 1
+
+  tika:
+    image: apache/tika
+    restart: unless-stopped
+
+volumes:
+  data:
+  media:

docker/local/docker-compose.tika.yml

@@ -0,0 +1,43 @@
+version: "3.4"
+services:
+  broker:
+    image: redis:6.0
+    restart: always
+
+  webserver:
+    build: .
+    restart: always
+    depends_on:
+      - broker
+    ports:
+      - 8000:8000
+    healthcheck:
+      test: ["CMD", "curl", "-f", "http://localhost:8000"]
+      interval: 30s
+      timeout: 10s
+      retries: 5
+    volumes:
+      - data:/usr/src/paperless/data
+      - media:/usr/src/paperless/media
+      - ./export:/usr/src/paperless/export
+      - ./consume:/usr/src/paperless/consume
+    env_file: docker-compose.env
+    environment:
+      PAPERLESS_REDIS: redis://broker:6379
+      PAPERLESS_TIKA: 1
+      GOTENBERG_SERVER_ENDPOINT: http://gotenberg:3000
+      TIKA_SERVER_ENDPOINT: http://tika:9998
+
+  gotenberg:
+    image: thecodingmachine/gotenberg
+    restart: unless-stopped
+    environment:
+      DISABLE_GOOGLE_CHROME: 1
+
+  tika:
+    image: apache/tika
+    restart: unless-stopped
+
+volumes:
+  data:
+  media:

docs/configuration.rst

@@ -277,6 +277,35 @@ PAPERLESS_OCR_USER_ARG=<json>
 
         {"deskew": true, "optimize": 3, "unpaper_args": "--pre-rotate 90"}    
     
+.. _configuration-tika:
+
+Tika settings
+#############
+
+Paperless can make use of `Tika <https://tika.apache.org/>`_ and 
+`Gotenberg <https://thecodingmachine.github.io/gotenberg/>`_ for parsing and
+converting "Office" documents (such as ".doc", ".xlsx" and ".odt"). If you
+wish to use this, you must provide a Tika server and a Gotenberg server,
+configure their endpoints, and enable the feature.
+
+If you run paperless on docker, you can add those services to the docker-compose
+file (see the examples provided).
+
+PAPERLESS_TIKA=<bool>
+    Enable (or disable) the Tika parser.
+
+    Defaults to false.
+
+TIKA_SERVER_ENDPOINT=<url>
+    Set the endpoint URL were Paperless can reach your Tika server.
+
+    Defaults to "http://localhost:9998".
+
+GOTENBERG_SERVER_ENDPOINT=<url>
+    Set the endpoint URL were Paperless can reach your Gotenberg server.
+
+    Defaults to "http://localhost:3000".
+
     
 Software tweaks
 ###############

src/paperless/settings.py

@@ -87,6 +87,7 @@ INSTALLED_APPS = [
     "documents.apps.DocumentsConfig",
     "paperless_tesseract.apps.PaperlessTesseractConfig",
     "paperless_text.apps.PaperlessTextConfig",
+    "paperless_tika.apps.PaperlessTikaConfig",
     "paperless_mail.apps.PaperlessMailConfig",
 
     "django.contrib.admin",
@@ -424,3 +425,7 @@ for t in json.loads(os.getenv("PAPERLESS_FILENAME_PARSE_TRANSFORMS", "[]")):
 PAPERLESS_FILENAME_FORMAT = os.getenv("PAPERLESS_FILENAME_FORMAT")
 
 THUMBNAIL_FONT_NAME = os.getenv("PAPERLESS_THUMBNAIL_FONT_NAME", "/usr/share/fonts/liberation/LiberationSerif-Regular.ttf")
+
+# Tika settings
+PAPERLESS_TIKA = __get_boolean("PAPERLESS_TIKA", "NO")
+GOTENBERG_SERVER_ENDPOINT = os.getenv("GOTENBERG_SERVER_ENDPOINT", "http://localhost:3000")

src/paperless_tika/apps.py

@@ -0,0 +1,14 @@
+from django.apps import AppConfig
+from django.conf import settings
+from paperless_tika.signals import tika_consumer_declaration
+
+
+class PaperlessTikaConfig(AppConfig):
+    name = "paperless_tika"
+
+    def ready(self):
+        from documents.signals import document_consumer_declaration
+
+        if settings.PAPERLESS_TIKA:
+            document_consumer_declaration.connect(tika_consumer_declaration)
+        AppConfig.ready(self)

src/paperless_tika/parsers.py

@@ -0,0 +1,118 @@
+import os
+import subprocess
+import tika
+import requests
+import dateutil.parser
+
+from PIL import ImageDraw, ImageFont, Image
+from django.conf import settings
+
+from documents.parsers import DocumentParser, ParseError, run_convert
+from paperless_tesseract.parsers import RasterisedDocumentParser
+from tika import parser
+
+
+class TikaDocumentParser(DocumentParser):
+    """
+    This parser sends documents to a local tika server
+    """
+
+    def get_thumbnail(self, document_path, mime_type):
+        self.log("info", f"[TIKA_THUMB] Generating thumbnail for{document_path}")
+        archive_path = self.archive_path
+
+        out_path = os.path.join(self.tempdir, "convert.png")
+
+        # Run convert to get a decent thumbnail
+        try:
+            run_convert(
+                density=300,
+                scale="500x5000>",
+                alpha="remove",
+                strip=True,
+                trim=False,
+                input_file="{}[0]".format(archive_path),
+                output_file=out_path,
+                logging_group=self.logging_group,
+            )
+        except ParseError:
+            # if convert fails, fall back to extracting
+            # the first PDF page as a PNG using Ghostscript
+            self.log(
+                "warning",
+                "Thumbnail generation with ImageMagick failed, falling back "
+                "to ghostscript. Check your /etc/ImageMagick-x/policy.xml!",
+            )
+            gs_out_path = os.path.join(self.tempdir, "gs_out.png")
+            cmd = [
+                settings.GS_BINARY,
+                "-q",
+                "-sDEVICE=pngalpha",
+                "-o",
+                gs_out_path,
+                archive_path,
+            ]
+            if not subprocess.Popen(cmd).wait() == 0:
+                raise ParseError("Thumbnail (gs) failed at {}".format(cmd))
+            # then run convert on the output from gs
+            run_convert(
+                density=300,
+                scale="500x5000>",
+                alpha="remove",
+                strip=True,
+                trim=False,
+                input_file=gs_out_path,
+                output_file=out_path,
+                logging_group=self.logging_group,
+            )
+
+        return out_path
+
+    def parse(self, document_path, mime_type):
+        self.log("info", f"[TIKA_PARSE] Sending {document_path} to Tika server")
+
+        try:
+            parsed = parser.from_file(document_path)
+        except requests.exceptions.HTTPError as err:
+            raise ParseError(f"Could not parse {document_path} with tika server: {err}")
+
+        try:
+            content = parsed["content"].strip()
+        except:
+            content = ""
+
+        try:
+            creation_date = dateutil.parser.isoparse(
+                parsed["metadata"]["Creation-Date"]
+            )
+        except:
+            creation_date = None
+
+        archive_path = os.path.join(self.tempdir, "convert.pdf")
+        convert_to_pdf(self, document_path, archive_path)
+
+        self.archive_path = archive_path
+        self.date = creation_date
+        self.text = content
+
+
+def convert_to_pdf(self, document_path, pdf_path):
+    pdf_path = os.path.join(self.tempdir, "convert.pdf")
+    gotenberg_server = settings.GOTENBERG_SERVER_ENDPOINT
+    url = gotenberg_server + "/convert/office"
+
+    self.log("info", f"[TIKA] Converting {document_path} to PDF as {pdf_path}")
+    files = {"files": open(document_path, "rb")}
+    headers = {}
+
+    try:
+        response = requests.post(url, files=files, headers=headers)
+        response.raise_for_status()  # ensure we notice bad responses
+    except requests.exceptions.HTTPError as err:
+        raise ParseError(
+            f"Could not contact gotenberg server at {gotenberg_server}: {err}"
+        )
+
+    file = open(pdf_path, "wb")
+    file.write(response.content)
+    file.close()

src/paperless_tika/signals.py

@@ -0,0 +1,20 @@
+from .parsers import TikaDocumentParser
+
+
+def tika_consumer_declaration(sender, **kwargs):
+    return {
+        "parser": TikaDocumentParser,
+        "weight": 10,
+        "mime_types": {
+            "application/msword": ".doc",
+            "application/vnd.openxmlformats-officedocument.wordprocessingml.document": ".docx",
+            "application/vnd.ms-excel": ".xls",
+            "application/vnd.openxmlformats-officedocument.spreadsheetml.sheet": ".xlsx",
+            "application/vnd.ms-powerpoint": ".ppt",
+            "application/vnd.openxmlformats-officedocument.presentationml.presentation": ".pptx",
+            "application/vnd.openxmlformats-officedocument.presentationml.slideshow": ".ppsx",
+            "application/vnd.oasis.opendocument.presentation": ".odp",
+            "application/vnd.oasis.opendocument.spreadsheet": ".ods",
+            "application/vnd.oasis.opendocument.text": ".odt",
+        },
+    }

src/paperless_tika/test.py

@@ -0,0 +1,3 @@
+import magic
+m = magic.from_file("/nfsstorage/jo/syncthing/Documenten/20R-309.153.052.pdf")
+print(m)