llama_index icon indicating copy to clipboard operation
llama_index copied to clipboard

[WIP] Add pandas-based tabular parser

Open ephe-meral opened this issue 2 years ago • 1 comments

As per the discussion in discord, this is a draft of a CSV parser that uses pandas for loading the CSV data.

If this goes in the right direction, we could also add an Excel / OpenDocument Table format in the same style. (And notebooks to showcase it, if that makes sense?)

I also fixed the documentation of the existing CSV parser.

ephe-meral avatar Jan 30 '23 09:01 ephe-meral

@ephe-meral sorry just checking, is this ready to review?

jerryjliu avatar Feb 01 '23 07:02 jerryjliu

@jerryjliu Yes please - my intention was to provide a conversation starting point, i.e. to potentially improve/iterate this to make it fit our needs. But functionality wise it could also be merged IMO, should already be useful (and certainly enough for my use case). I'll remove the WIP for now.

ephe-meral avatar Feb 01 '23 21:02 ephe-meral