Script resilience to version updates

[alix-dumoulin]

TLDR: some scripts found on GitHub / StackOverflow use deprecated methods/functions/libraries

Most common cases so far:

• Mismatch between script and version installed using pip, can lead to methods not existing, old set of parameters because the function changed, requirements for the package not being installed
• Script using libraries that are not maintained anymore
• Scripts importing a module that is not part of the package anymore (or may have a new name?)

Example:
This script which converts pdf to html/XML

• imports process_pdf() which after some research turns out to have been replaced with PDFPage.get_pages()
• and has a parameter codec for encoding which is not part of the set of parameters in the new version
• imports a module plywhich is not in the requirements.txt (therefore not installed via pip)

Why it's a problem

• Now: makes it a bit painful to get scripts running
• Later: if we integrate scripts into a product, we need to make sure they are resilient to updates of the packages they use (which is very high-maintenance)