fallback python-based .rtf extraction #88
Description
This is currently using the unrtf command line tool, but it would be nice to have a pure python extraction method as a fallback.
{{ message }}
This is currently using the unrtf command line tool, but it would be nice to have a pure python extraction method as a fallback.
Activity
pombredanne commentedon Aug 20, 2015
@deanmalmgren Have a look at https://github.com/brendonh/pyth
deanmalmgren commentedon Aug 29, 2015
Nice find, @pombredanne; thanks for pointing this out!
If anyone is interested in taking a crack at this, the pdf parser has a pretty good example for how to have multiple
methods for extracting text from documents.jpadilla commentedon Aug 29, 2015
@deanmalmgren I'll try to get this done in a few.
deanmalmgren commentedon Aug 29, 2015
Awesome, thanks @jpadilla!
jpadilla commentedon Aug 29, 2015
For one thing, pyth seems to have issues with charsets on the existing RTF files in textract which will make it harder to test. Might be related to brendonh/pyth#30
pombredanne commentedon Aug 29, 2015
@jpadilla indeed, and I ran pyth through the test files from unrtf and submitted a ticket there: brendonh/pyth#34
Which is an opportunity to help make @brendonh pyth better?
KyleKing commentedon Aug 6, 2026
Unfortunately pyth doesn't appear to be maintained and there doesn't appear to be a good unrtf Python alternative. I'm going to close this ticket while cleaning up the backlog, but I'm open to ideas on implementation and reopening this if there is!
KyleKing commentedon Aug 6, 2026
Actually, https://github.com/joshy/striprtf looks like a good option. If that sounds good to others, I can submit a PR or I would welcome any contributions