Reviewed to keep the size as low as possible.

Dockerfile

@@ -1,50 +1,66 @@
 # DigiPres Toolbox
-# A Docker image with some tools pre-installed
-FROM python:3.10-bullseye
+# ----------------
+#
+# A Docker image with some DigiPres tools pre-installed
+#
 
-# Core Jupyter support:
-RUN pip install --no-cache notebook jupyterlab bash_kernel
-RUN python -m bash_kernel.install
+# Start from as small an image as possible:
+FROM python:3.11-slim-bullseye
 
-# Some lightweight tools and support for installing more:
-RUN apt-get update && apt-get install -y sudo mediainfo cloc && \
+# Some (smallish) tools and support for installing more (69MB)
+RUN apt-get update && apt-get install -y curl mediainfo cloc unzip file && \
     apt-get clean && rm -rf /var/lib/apt/lists/*
 
-# Install Siegfried:
+# Install JRE for Java programs (headless 250MB, headful is 431MB)
+RUN apt-get update && apt-get install -y default-jre-headless && \
+    apt-get clean && rm -rf /var/lib/apt/lists/*
+
+# Install Siegfried (71MB)
 ENV SF_VERSION=1.11.1
 ENV SF_DEB=siegfried_${SF_VERSION}-1_amd64.deb
 RUN curl -s -L -O https://github.com/richardlehane/siegfried/releases/download/v${SF_VERSION}/${SF_DEB} && \
     dpkg -i ${SF_DEB} && \
     rm -f ${SF_DEB} && \
     sf -update
 
-# Install TRiD:
+# Install TRiD (9MB)
 RUN curl -s -L -O http://mark0.net/download/trid_linux_64.zip && \
-curl -s -L -O http://mark0.net/download/triddefs.zip && \
-unzip trid_linux_64.zip && unzip triddefs.zip && chmod +x ./trid && \
-mv ./trid /usr/local/bin/trid && mv triddefs.trd /usr/local/bin/ && \
-rm -f trid_linux_64.zip triddefs.zip
+    curl -s -L -O http://mark0.net/download/triddefs.zip && \
+    unzip trid_linux_64.zip && unzip triddefs.zip && chmod +x ./trid && \
+    mv ./trid /usr/local/bin/trid && mv triddefs.trd /usr/local/bin/ && \
+    rm -f trid_linux_64.zip triddefs.zip
 
-# Install Fido:
-RUN pip install --no-cache opf-fido
+# Install Fido (10MB)
+RUN pip install --no-cache-dir opf-fido
 
-# Install JRE for Java programs and ffmpeg for a/v formats (c. 0.6GB!):
-RUN apt-get update && apt-get install -y default-jre ffmpeg && \
-    apt-get clean && rm -rf /var/lib/apt/lists/*
+# Install RClone (62MB) as per https://rclone.org/install/#linux-precompiled
+RUN curl -O https://downloads.rclone.org/rclone-current-linux-amd64.zip && \
+    unzip rclone-current-linux-amd64.zip && \
+    cp rclone-*-linux-amd64/rclone /usr/bin/ && \
+    chown root:root /usr/bin/rclone && \
+    chmod 755 /usr/bin/rclone && \
+    rm -fr rclone-*-linux-*
 
-# Install GitHub Linguist and it's build dependencies (c. 0.2GB):
-RUN apt-get update && \
-    apt-get install -y cmake pkg-config libicu-dev zlib1g-dev libcurl4-openssl-dev libssl-dev ruby-dev && \ 
-    apt-get clean && rm -rf /var/lib/apt/lists/*
-RUN gem install github-linguist
+#
+# Skipping the following to try to keep the image size down...
+#
 
-# Install Apache Tika (needs Java):
-ENV TIKA_VERSION=2.9.2
-RUN curl -s -L -o /usr/share/java/tika-app-${TIKA_VERSION}.jar https://dlcdn.apache.org/tika/${TIKA_VERSION}/tika-app-${TIKA_VERSION}.jar && \
-    ln -s /usr/share/java/tika-app-${TIKA_VERSION}.jar /usr/share/java/tika-app.jar
-COPY tika.sh /usr/local/bin/tika.sh
+# Install GitHub Linguist (~200MB at least, depending on shared deps)
+#RUN apt-get update && \
+#    apt-get install -y build-essential cmake pkg-config libicu-dev zlib1g-dev libcurl4-openssl-dev libssl-dev ruby-dev && \ 
+#    gem install github-linguist && \
+#    apt-get clean && rm -rf /var/lib/apt/lists/*
 
-# Install DROID (needs Java)
-COPY droid /usr/share/java/droid
-RUN ln -s /usr/share/java/droid/droid.sh /usr/local/bin/droid.sh
+# Install ffmpeg for a/v formats (484MB!)
+# RUN apt-get update && apt-get install -y ffmpeg && \
+#    apt-get clean && rm -rf /var/lib/apt/lists/*
+
+# Install Apache Tika (needs Java)
+#ENV TIKA_VERSION=2.9.2
+#RUN curl -s -L -o /usr/share/java/tika-app-${TIKA_VERSION}.jar https://dlcdn.apache.org/tika/${TIKA_VERSION}/tika-app-${TIKA_VERSION}.jar && \
+#    ln -s /usr/share/java/tika-app-${TIKA_VERSION}.jar /usr/share/java/tika-app.jar
+#COPY tika.sh /usr/local/bin/tika.sh
 
+# Install DROID (needs Java)
+#COPY droid /usr/share/java/droid
+#RUN ln -s /usr/share/java/droid/droid.sh /usr/local/bin/droid.sh

README.md

@@ -3,27 +3,52 @@ DigiPres Toolbox
 
 A Docker image designed to make it easy to experiment with tools for Digital Preservation.  Designed to be used via the [DigiPres Sandbox](https://github.com/digipres/sandbox) and the [DigiPres Workbench](https://github.com/digipres/workbench).
 
+Build locally with e.g.
+
+```
+docker build . -t toolbox
+```
+
+Then run with
+
+```
+docker run -it toolbox bash
+```
+
 ## Supported Tools
 
+Large (>>1GB) images don't seem to run well on Binder, so we can't install everything we'd like to. e.g. `ffmpeg` takes up 0.5GB!
+
+These sizes can be determined by using separate installation lines in the `Dockerfile` and then using commands like this to see what happened and what size the additional layer is:
+
+```
+docker history --no-trunc toolbox | grep ffmpeg
+```
+
 ### Pre-installed
 
- - [Apache Tika](https://tika.apache.org/)
  - [CLOC](https://github.com/AlDanial/cloc)
- - [DROID](http://digital-preservation.github.io/droid/)
- - [ffmpeg](https://ffmpeg.org) including [ffprobe](https://ffmpeg.org/ffprobe.html)
  - [Fido](https://github.com/openpreserve/fido)
  - [File](https://www.darwinsys.com/file/)
- - [GitHub Linguist](https://github.com/github/linguist)
  - [MediaInfo](https://github.com/MediaArea/MediaInfo)
  - [Siegfried](https://www.itforarchivists.com/siegfried)
  - [TrID](http://mark0.net/soft-trid-e.html)
 
 ### Verified Installable
 
-These aren't installed by default, but the [Sandbox](https://github.com/digipres/sandbox) shows how to install them.
+These aren't installed by default because of their size, but the [Sandbox](https://github.com/digipres/sandbox) indicates how to download and install them.
 
+ - [Apache Tika](https://tika.apache.org/)
+ - [DROID](http://digital-preservation.github.io/droid/)
+ - [ffmpeg](https://ffmpeg.org) including [ffprobe](https://ffmpeg.org/ffprobe.html)
  - [pdfcpu](https://pdfcpu.io)
 
+### Cannot Be Installed
+
+These require root access to install but take up too much space
+
+ - [GitHub Linguist](https://github.com/github/linguist) (200-400MB in size depending on base image, mostly down to requiring a full build environment)
+
 ### To Consider
 
 - RClone