feat(api): Remove dependency on exiftool (#82)

2025-10-30 16:11:16 +05:30
parent fa60de79fd b9ec2b1b5a
commit cc9ac6093f
9 changed files with 1184 additions and 767 deletions
@@ -0,0 +1,10 @@
+__pycache__
+*ipynb
+build
+.idea
+.vscode
+.venv
+.env
+*log
+*.md
+uv.lock
@@ -13,23 +13,17 @@ FROM python:3.13-slim

 WORKDIR /app

-# Install apt-fast, git, exiftool
-
 COPY --from=nobodyxu/apt-fast:latest-debian-buster-slim /usr/local/ /usr/local/

 RUN apt update
 RUN apt install -y aria2
-RUN apt-fast install -y git build-essential libpango-1.0-0 libpangoft2-1.0-0 wget
+RUN apt-fast install -y git build-essential python3.13-dev libgobject-2.0 libpango-1.0 libpangoft2-1.0
+# aiohttp-client-cache depends on multipart, which requires python3.13-dev to build successfully on 3.13
+# weasyprint depends on libgoject, libpango, and libpangoft2

-ENV EXIFTOOL_VERSION="13.06"
-RUN wget "https://exiftool.org/Image-ExifTool-${EXIFTOOL_VERSION}.tar.gz"
-RUN gzip -dc "Image-ExifTool-${EXIFTOOL_VERSION}.tar.gz" | tar -xf -
-WORKDIR /app/Image-ExifTool-${EXIFTOOL_VERSION}
-RUN perl Makefile.PL
-RUN make test
-RUN make install
+RUN rm -rf /var/lib/apt/lists/*
+# https://github.com/TheOnlyWayUp/WattpadDownloader/pull/82#discussion_r2470358950

-RUN rm -rf /var/lib/apt/lists/* /app/Image-ExifTool-${EXIFTOOL_VERSION}

 WORKDIR /app

@@ -37,17 +31,15 @@ WORKDIR /app

 COPY --from=ghcr.io/astral-sh/uv:latest /uv /uvx /bin/

-COPY src/api/requirements.txt requirements.txt
-COPY src/api/src/create_book/generators/pdf/exiftool.config exiftool.config
-RUN uv pip install -r requirements.txt --system
+COPY src/api/pyproject.toml /app
+RUN uv sync
+COPY src/api/ /app
 COPY --from=0 /build/build /app/src/build
-COPY src/api/src src

-# Is this still needed?
 RUN ln -s /app/src/pdf/fonts /tmp/fonts

 WORKDIR /app/src

 EXPOSE 80

-CMD [ "python3", "main.py"]
+CMD [ "uv", "run", "main.py"]
@@ -17,7 +17,6 @@ dependencies = [
    "aiohttp-client-cache[all]",
    "bs4>=0.0.2",
    "uvicorn>=0.32.1",
-    "pyexiftool>=0.5.6",
    "weasyprint>=63.0",
    "jinja2>=3.1.6",
 ]
@@ -31,5 +30,6 @@ aiohttp-client-cache = { git = "https://github.com/TheOnlyWayUp/aiohttp-client-c
 [dependency-groups]
 dev = [
    "ipykernel>=6.29.5",
+    "ipynb>=0.5.1",
    "ruff>=0.11.12",
 ]
@@ -1,76 +0,0 @@
-aioboto3==13.2.0
-aiobotocore==2.15.2
-aiofiles==24.1.0
-aiohappyeyeballs==2.4.4
-aiohttp==3.11.9
-aiohttp-client-cache @ git+https://github.com/TheOnlyWayUp/aiohttp-client-cache.git@1f94f1d751e7320c0ea981d532ff02924782dae6
-aioitertools==0.12.0
-aiosignal==1.3.1
-aiosqlite==0.20.0
-annotated-types==0.7.0
-anyio==4.6.2.post1
-async-timeout==4.0.3
-attrs==23.1.0
-backoff==2.2.1
-beautifulsoup4==4.12.3
-boltons==24.1.0
-boto3==1.35.36
-botocore==1.35.36
-brotli==1.1.0
-bs4==0.0.2
-cffi==1.17.1
-click==8.1.7
-cssselect2==0.7.0
-dnspython==2.7.0
-ebooklib==0.18
-eliot==1.16.0
-exceptiongroup==1.2.2
-fastapi==0.115.5
-fonttools==4.55.2
-frozenlist==1.4.1
-h11==0.14.0
-idna==3.6
-itsdangerous==2.2.0
-jinja2==3.1.6
-jmespath==1.0.1
-lxml==5.3.0
-markdown-it-py==3.0.0
-mdurl==0.1.2
-motor==3.6.0
-multidict==6.0.4
-orjson==3.10.12
-pillow==10.4.0
-propcache==0.2.1
-pycparser==2.22
-pydantic==2.10.2
-pydantic-core==2.27.1
-pydantic-settings==2.6.1
-pydyf==0.11.0
-pyexiftool==0.5.6
-pygments==2.18.0
-pymongo==4.9.2
-pyphen==0.15.0
-pyrsistent==0.20.0
-python-dateutil==2.9.0.post0
-python-dotenv==1.0.1
-redis==5.2.0
-rich==13.9.4
-s3transfer==0.10.4
-setuptools==75.6.0
-six==1.16.0
-sniffio==1.3.1
-soupsieve==2.6
-starlette==0.41.3
-tinycss2==1.4.0
-tinyhtml5==2.0.0
-type-extensions==0.1.2
-typing-extensions==4.12.2
-url-normalize==1.4.3
-urllib3==2.2.3
-uvicorn==0.32.1
-weasyprint==63.0
-webencodings==0.5.1
-wrapt==1.17.0
-yarl==1.18.3
-zope-interface==7.2
-zopfli==0.2.3.post1
@@ -3,10 +3,10 @@ from io import BytesIO
 from pathlib import Path
 from tempfile import NamedTemporaryFile, _TemporaryFileWrapper

+import pydyf
 from bs4 import BeautifulSoup
-from exiftool import ExifTool
 from jinja2 import Template
-from weasyprint import CSS, HTML
+from weasyprint import CSS, HTML, Document
 from weasyprint.text.fonts import FontConfiguration

 from ..models import Story
@@ -97,7 +97,7 @@ class PDFGenerator(AbstractGenerator):
        self.images = images
        self.author = author_image

-        self.book: _TemporaryFileWrapper = NamedTemporaryFile(suffix=".pdf")
+        self.book: _TemporaryFileWrapper = NamedTemporaryFile(suffix=".pdf")  # type: ignore
        self.content = TEMPLATE

    def generate_chapters(self) -> dict[int, str]:
@@ -134,6 +134,12 @@ class PDFGenerator(AbstractGenerator):
            "book_title": self.story["title"],
            "cover": f"data:image/jpg;base64,{b64encode(self.cover).decode()}",
            "username": self.story["user"]["username"],
+            "author_bio": self.story["user"]["description"],
+            "clean_tags": ", ".join(self.story["tags"]),
+            "created": self.story["createDate"],
+            "modified": self.story["modifyDate"],
+            "is_completed": self.story["completed"],
+            "is_mature": self.story["mature"],
            "description": self.story["description"],
            "avatar": b64encode(self.author).decode(),
            "copyright": {
@@ -149,6 +155,11 @@ class PDFGenerator(AbstractGenerator):

        self.content: str = Template(self.content).render(data)

+    def write_custom_metadata(self, document: Document, pdf: pydyf.PDF):
+        """Write non-standard metadata fields to the PDF."""
+        pdf.info["completed"] = pydyf.String(str(self.story["completed"]))
+        pdf.info["mature"] = pydyf.String(str(self.story["mature"]))
+
    def generate_pdf(self):
        """Generate and write the PDF to a temporary file (self.book)."""
        font_config = FontConfiguration()
@@ -157,47 +168,17 @@ class PDFGenerator(AbstractGenerator):

        html_obj = HTML(string=self.content)
        html_obj.write_pdf(
-            self.book.name, stylesheets=[stylesheet_obj], font_config=font_config
-        )
-
-    def add_metadata(self):
-        """Write metadata to generated PDF file at self.book, using ExifTool."""
-
-        clean_description = (
-            self.story["description"].strip().replace("\n", "$/")
-        )  # exiftool doesn't parse \ns correctly, they support $/ for the same instead. `&#xa;` is another option.
-
-        metadata = {
-            "Author": self.story["user"]["username"],
-            "Title": self.story["title"],
-            "Subject": clean_description,
-            "CreationDate": self.story["createDate"],
-            "ModDate": self.story["modifyDate"],
-            "Keywords": ",".join(self.story["tags"]),
-            "Language": self.story["language"]["name"],
-            "Completed": self.story["completed"],
-            "MatureContent": self.story["mature"],
-            "Producer": "Dhanush Rambhatla (TheOnlyWayUp - https://rambhat.la) and WattpadDownloader",
-        }  # As per https://exiftool.org/TagNames/PDF.html
-
-        with ExifTool(config_file=DATA_PATH / "exiftool.config") as et:
-            # Custom configuration adds Completed and MatureContent tags.
-            # exiftool logger logs executed command
-            et.execute(
-                *(
-                    [f"-{key}={value}" for key, value in metadata.items()]
-                    + [
-                        "-overwrite_original",
-                        self.book.file.name,
-                    ]
-                )
+            self.book.name,
+            stylesheets=[stylesheet_obj],
+            font_config=font_config,
+            finisher=self.write_custom_metadata,
+            options={"custom_metadata": True},
        )

    def compile(self):
        parts = self.generate_chapters()
        self.populate_template(parts)
        self.generate_pdf()
-        self.add_metadata()
        return True

    def dump(self) -> BytesIO:
@@ -1,16 +1,30 @@
 <!DOCTYPE html>
 <html lang="{{ langcode }}">

+
+<head>
+
    <meta charset="UTF-8">
    <meta name="viewport" content="width=device-width, initial-scale=1.0">

+    <!-- https://doc.courtbouillon.org/weasyprint/stable/api_reference.html#weasyprint.document.DocumentMetadata -->
    <title>{{ book_title }}</title>
+    <meta name=description content="{{description}}">
+    <meta name=author content="{{author}}">
+    <meta name=keywords content="{{clean_tags}}">
+    <meta name=language content="{{langcode}}">
+    <meta name=dcterms.created content="{{created}}">
+    <meta name=dcterms.modified content="{{modified}}">
+    <meta name=generator content="Dhanush Rambhatla (TheOnlyWayUp - https://rambhat.la) and WattpadDownloader">

-    <section class="fullpage">
+
+</head>
+
+<section class="fullpage">
    <img src="{{ cover }}" alt="Cover">
-    </section>
+</section>

-    <div id="copyright-container">
+<div id="copyright-container">
    <h1 id="copyright-notice">Copyright Notice</h1>

    <h2 id="copyright-title">{{ book_title }}</h2>
@@ -23,11 +37,8 @@
    <div id="copyright-separator"></div>

    {% if copyright.data %}
-        <img src="data:image/jpg;base64,{{copyright.data}}" 
-alt="{{copyright.name}}" 
-width="88" 
-height="31" 
-id="copyright-license-image">
+    <img src="data:image/jpg;base64,{{copyright.data}}" alt="{{copyright.name}}" width="88" height="31"
+        id="copyright-license-image">
    {% endif %}

    <p id="copyright-copyright">{{ statement }}</p>
@@ -40,9 +51,9 @@ id="copyright-license-image">
        ID: {{ book_id }}.
        <a href="https://wattpad.com/story/{{ book_id }}" target="_blank" id="copyright-link">View this Book Online</a>
    </p>
-    </div>
+</div>

-    <div id="book">
+<div id="book">
    <section id="contents" class="toc">
        <h1>Table of Contents</h1>
        <ul>
@@ -55,10 +66,10 @@ id="copyright-license-image">

    {{parts[part_id] | safe}}
    {% endfor %}
-    </div>
+</div>

-    <h1>About the Author</h1>
-    <div id="author-container">
+<h1>About the Author</h1>
+<div id="author-container">
    <div id="author-about">
        <img src="data:image/jpg;base64,{{avatar}}" alt="{{author}}'s profile picture" id="author-profile-picture">
        <h2 id="author-name">
@@ -66,8 +77,9 @@ id="copyright-license-image">
        </h2>
        <hr id="author-divider">
        <p id="author-bio">
-                {{ description }}
+            {{ author_bio }}
        </p>
    </div>
-    </div>
+</div>
+
 </html>
@@ -1,26 +0,0 @@
-
-%Image::ExifTool::UserDefined = (
-    'Image::ExifTool::XMP::xmp' => {
-        Completed => {
-            Writable => 'boolean',  # Can be a boolean (True/False)
-            Groups => { 2 => 'Content' },
-        },
-        MatureContent => {
-            Writable => 'boolean',  # Can be a boolean (True/False)
-            Groups => { 2 => 'Content' },
-        },
-    },
-    
-    'Image::ExifTool::IPTC::ApplicationRecord' => {
-        161 => {
-            Name => 'Completed',
-            Format => 'string[0,16]',  # Store as a string (e.g., "Yes"/"No")
-        },
-        162 => {
-            Name => 'MatureContent',
-            Format => 'string[0,16]',  # Store as a string (e.g., "Yes"/"No")
-        },
-    },
-);
-
-1;  # End
@@ -0,0 +1,5 @@
+The fonts need to be symlinked to /tmp/fonts, this allows the fonts to be loaded during development and during build-time.
+It's assumed fonts will be present at `/tmp/fonts`, during development they're at `/src/api/src/create_book/generators/pdf`, and during deployment they're at `/app/src/api/src/create_book/generators/pdf`. This seems like a clean solution.
+
+`Fontconfig error: Cannot load default config file: No such file: (null)`
+If the fonts aren't found, this warning pops up in console. It won't cause downloads to fail, though.