Merge branch 'master' into readability-extractor

Dockerfile

@@ -1,90 +1,107 @@
-# This is the Dockerfile for ArchiveBox, it includes the following major pieces:
-#     git, curl, wget, python3, youtube-dl, google-chrome-stable, ArchiveBox
+# This is the Dockerfile for ArchiveBox, it bundles the following dependencies:
+#     python3, ArchiveBox, curl, wget, git, chromium, youtube-dl, single-file
 # Usage:
-#     docker build . -t archivebox
+#     docker build . -t archivebox --no-cache
 #     docker run -v "$PWD/data":/data archivebox init
 #     docker run -v "$PWD/data":/data archivebox add 'https://example.com'
-# Documentation:
-#     https://github.com/pirate/ArchiveBox/wiki/Docker#docker
+#     docker run -v "$PWD/data":/data -it archivebox manage createsuperuser
+#     docker run -v "$PWD/data":/data -p 8000:8000 archivebox server
 
 FROM python:3.8-slim-buster
 
 LABEL name="archivebox" \
-    maintainer="Nick Sweeting <archivebox-git@sweeting.me>" \
-    description="All-in-one personal internet archiving container"
+    maintainer="Nick Sweeting <archivebox-docker@sweeting.me>" \
+    description="All-in-one personal internet archiving container" \
+    homepage="https://github.com/pirate/ArchiveBox" \
+    documentation="https://github.com/pirate/ArchiveBox/wiki/Docker#docker"
 
+# System-level base config
 ENV TZ=UTC \
     LANGUAGE=en_US:en \
     LC_ALL=C.UTF-8 \
     LANG=C.UTF-8 \
     PYTHONIOENCODING=UTF-8 \
     PYTHONUNBUFFERED=1 \
-    APT_KEY_DONT_WARN_ON_DANGEROUS_USAGE=1 \
-    CODE_PATH=/app \
+    DEBIAN_FRONTEND=noninteractive \
+    APT_KEY_DONT_WARN_ON_DANGEROUS_USAGE=1
+
+# Application-level base config
+ENV CODE_DIR=/app \
     VENV_PATH=/venv \
-    DATA_PATH=/data \
-    EXTRA_PATH=/extra
+    DATA_DIR=/data \
+    NODE_DIR=/node \
+    ARCHIVEBOX_USER="archivebox"
 
-# First install CLI utils and base deps, then Chrome + Fons + nodejs
-RUN echo 'debconf debconf/frontend select Noninteractive' | debconf-set-selections \
+# Create non-privileged user for archivebox and chrome
+RUN groupadd --system $ARCHIVEBOX_USER \
+    && useradd --system --create-home --gid $ARCHIVEBOX_USER --groups audio,video $ARCHIVEBOX_USER
+
+# Install system dependencies
+RUN apt-get update -qq \
+    && apt-get install -qq -y --no-install-recommends \
+        apt-transport-https ca-certificates gnupg2 zlib1g-dev \
+        dumb-init gosu unzip curl \
+    && rm -rf /var/lib/apt/lists/*
+
+# Install apt dependencies
+RUN apt-get update -qq \
+    && apt-get install -qq -y --no-install-recommends \
+        wget curl chromium git ffmpeg youtube-dl \
+        fontconfig fonts-ipafont-gothic fonts-wqy-zenhei fonts-thai-tlwg fonts-kacst fonts-symbola fonts-noto fonts-freefont-ttf \
+    && rm -rf /var/lib/apt/lists/*
+
+# Install Node environment
+RUN curl -s https://deb.nodesource.com/gpgkey/nodesource.gpg.key | apt-key add - \
+    && echo 'deb https://deb.nodesource.com/node_14.x buster main' >> /etc/apt/sources.list \
     && apt-get update -qq \
     && apt-get install -qq -y --no-install-recommends \
-    apt-transport-https ca-certificates apt-utils gnupg gosu gnupg2 libgconf-2-4 zlib1g-dev \
-    dumb-init jq git wget curl youtube-dl ffmpeg \
-    && curl -sSL "https://dl.google.com/linux/linux_signing_key.pub" | apt-key add - \
-    && echo "deb https://dl.google.com/linux/chrome/deb/ stable main" > /etc/apt/sources.list.d/google-chrome.list \
-    && curl -sL https://deb.nodesource.com/setup_14.x | bash - \
-    && apt-get update -qq \
-    && apt-get install -qq -y --no-install-recommends \
-    google-chrome-stable \
-    fontconfig \
-    fonts-ipafont-gothic \
-    fonts-wqy-zenhei \
-    fonts-thai-tlwg \
-    fonts-kacst \
-    fonts-symbola \
-    fonts-noto \
-    fonts-freefont-ttf \
-    nodejs \
-    unzip \
-    && rm -rf /var/lib/apt/lists/* 
+        nodejs \
+    && rm -rf /var/lib/apt/lists/*
 
-# Clone singlefile and move it to the /bin folder so archivebox can find it
-
-WORKDIR "$EXTRA_PATH"
-RUN wget -qO - https://github.com/gildas-lormeau/SingleFile/archive/master.zip > SingleFile.zip \
-    && unzip -q SingleFile.zip \
-    && npm install --prefix SingleFile-master/cli --production > /dev/null 2>&1 \
-    && chmod +x SingleFile-master/cli/single-file 
-
-RUN wget -qO - https://github.com/pirate/readability-extractor/archive/master.zip > readability.zip \
-    && unzip -q readability.zip \
-    && npm install --prefix readability-extractor-master --production > /dev/null 2>&1 \
-    && chmod +x readability-extractor-master/readability-extractor
-
-# Run everything from here on out as non-privileged user
-RUN groupadd --system archivebox \
-    && useradd --system --create-home --gid archivebox --groups audio,video archivebox
-
-ADD . "$CODE_PATH"
-WORKDIR "$CODE_PATH"
+# Install Python dependencies
+WORKDIR "$CODE_DIR"
 ENV PATH="${PATH}:$VENV_PATH/bin"
 RUN python -m venv --clear --symlinks "$VENV_PATH" \
-    && pip install --upgrade pip setuptools \
-    && pip install -e .
+    && pip install --upgrade --quiet pip setuptools
+ADD ./archivebox.egg-info/requires.txt "$CODE_DIR/archivebox.egg-info/requires.txt"
+RUN apt-get update -qq \
+    && apt-get install -qq -y --no-install-recommends \
+        build-essential python-dev python3-dev \
+    && grep -B 1000 -E '^$' "$CODE_DIR/archivebox.egg-info/requires.txt" | pip install --quiet -r /dev/stdin \
+    && apt-get purge -y build-essential python-dev python3-dev \
+    && apt-get autoremove -y \
+    && rm -rf /var/lib/apt/lists/*
 
-VOLUME "$DATA_PATH"
-WORKDIR "$DATA_PATH"
-EXPOSE 8000
+# Install Node dependencies
+WORKDIR "$NODE_DIR"
+ENV PATH="${PATH}:$NODE_DIR/node_modules/.bin" \
+    npm_config_loglevel=error
+RUN npm install -g npm
+ADD ./package.json ./package.json
+RUN npm install
+
+# Install ArchiveBox Python package
+WORKDIR "$CODE_DIR"
+ADD . "$CODE_DIR"
+RUN pip install -e .
+
+# Setup ArchiveBox runtime config
+WORKDIR "$DATA_DIR"
 ENV IN_DOCKER=True \
-    CHROME_BINARY=google-chrome \
     CHROME_SANDBOX=False \
-    USE_SINGLEFILE="true" \
-    SINGLEFILE_BINARY="$EXTRA_PATH/SingleFile-master/cli/single-file" \
-    USE_READABILITY="true" \
-    READABILITY_BINARY="$EXTRA_PATH/readability-extractor-master/readability-extractor"
+    CHROME_BINARY="chromium" \
+    USE_SINGLEFILE=True \
+    SINGLEFILE_BINARY="$NODE_DIR/node_modules/.bin/single-file" \
+    USE_READABILITY=True \
+    READABILITY_BINARY="$NODE_DIR/node_modules/.bin/readability-extractor"
 
-RUN env ALLOW_ROOT=True archivebox version
+# Print version for nice docker finish summary
+RUN archivebox version
+
+# Open up the interfaces to the outside world
+VOLUME "$DATA_DIR"
+VOLUME "$CODE_DIR"
+EXPOSE 8000
 
 ENTRYPOINT ["dumb-init", "--", "/app/bin/docker_entrypoint.sh"]
 CMD ["archivebox", "server", "0.0.0.0:8000"]