Shell script that extracts all images from PDF, PPT/X, DOC/X and XLS/X documents

extract-images.sh

#!/bin/bash
# Extracts images from Microsoft Office or PDF files using the DocumentAlchemy API.

# USAGE: extract-images.sh <FILES>
#
# EXAMPLE: extract-images.sh MyDeck.pptx *.doc

# A ZIP archive containing the extracted images (if any) will be
# created for each document submitted.

# This is your DocumentAlchemy API key.  It may be set via the
# environment variable named `DA_API_KEY`.
# (See <https://documentalchemy.com/> to get one of your own.)
API_KEY=${DA_API_KEY:-"403l1zh3dkbakyb9"}

# This is a flag that can prevent this script from `echo`ing
# unnecessary information.  It can be set via the environment
# variable `QUIET`. A clever person could make this into a
# command line parameter like `-q`.
QUIET=${QUIET:-FALSE}

# Note that you can set environment variables on a per-invocation
# basis by prefixing NAME=VALUE before the command.  For example:
#   API_KEY=my-key QUIET=true doc2docx *.doc

# EXIT_CODE tracks the number of files we couldn't extract from.
EXIT_CODE=0


# Loop over the command line parameters....
for doc in "$@"; do
  # ...testing that is is an accessible file...
  if ! [ -s "$doc" ]; then
    $QUIET || echo "WARNING: File '$doc' was not found and will be ignored."
    EXIT_CODE=$((EXIT_CODE+1))
  else
    # ...if so, POST to DocumentAlchemy to extract the images...
    outfile="`dirname "$doc"`/images-from-`basename "$doc"`.zip";
    $QUIET || echo "Extracting images from '$doc' into '`basename "$outfile"`'...";
    response=$(curl --silent \
                    --write-out %{http_code} -H "Authorization: da.key=$API_KEY" \
                    -X POST --form "document=@$doc" \
                    https://documentalchemy.com/api/v1/document/-/rendition/images.zip \
                    -o "$outfile")
    # ...and report success or failure.
    if ! [ "$response" -eq "200" ]; then
      $QUIET || echo "WARNING: Expected a 200 response for file '$doc', found $response instead.";
      EXIT_CODE=$((EXIT_CODE+1))
    else
      $QUIET || echo "...OK. File '$outfile' created."
    fi
  fi
done

# Exit with the number of documents that could not be converted.
exit $EXIT_CODE