extracting html tables from 10k

[mmistroni]

Hi @dgunning without revealing too much,.,,could you pseudo code how are you extracting html tables which might be embedded in a 10k? i'll uupdate with examples as soon as i have one
thanks again, you rock!
Marco

[dgunning]

Hi Marco,

There are two table extractions from 10-K

• xbrl tables from the XBRL attachments
• direct from the HTML filing document itself.

Which one are you interested in?

BTW they both use AI generated code so I have a general idea of how it's done but not details

[mmistroni]

the one from the txt direct.

[mmistroni]

ai generrated... so u used LLM as pair programmer and the llm generated code foryou? brilliant... but will it cover all possible usecases for embeddd tables? ty

[dgunning]

Table parsing is generally good. There are some weaknesses which will be addressed in later releases. The limitation isn't the development of the code - it is in testing. I can't test that it covers all possible use cases unless I develop a test harness that can throw all possible use cases at it, which is an enormous task