unix 'grep' a word inside pdf or image based on OCR
./imagegrep foo.pdf invoice eng && echo "grab your wallet!"
wget https://raw.githubusercontent.com/coderofsalvation/imagegrep-bash/master/imagegrep chmod 755 imagegrep ./imagegrep foo.pdf invoice eng
these packages can be installed using apt-get or yum
To automate, categorize files and their destination folder. OCR usually fails in many cases, but sometimes knowing one word (and its length) is enough. Imagegrep can be used this to scrape gmail and copy invoice-attachments to a preferred folder on my harddrive.
# not covered here: gmail to local maildir using 'offlineimap' # not covered here: use mu ('maildir-utils' package) to extract pdf attachments find mailbox/latest/*.pdf | while read file; do ./imagegrep "$file" invoice eng &&\ echo "grab your wallet!" &&\ mv foo.pdf ~/admin/invoices/. done