Add support for loading files like CSV, PDF, DOC(X), etc. to use with ComfyUI RAG

[l-Aeden-l]

Hi!

I'd appreciate to be able to work with some files like the one I mentioned in the title.
I asked ChatGPT to give me an example some file types to give you an idea:

Text Documents and Office Formats: .txt, .doc, .docx, .pdf, .rtf ; Tabular Data and Spreadsheets: .csv, .tsv, .xls, .xlsx, .ods ; Structured Data and Configuration: .xml, .json, .yaml, .yml, .ini, .cfg, .conf ; Scripts and Source Code: .html, .css, .js, .py, .java, .c, .cpp, .rb, .sql, .bat, .sh, .ps1 ; Serialization and Models: .jsonl, .pickle, .pkl, .h5.

The most important is to retain only the more relevant ones. The first one, to me, is CSV and the second one, probably JSON.

Last thing: I'd appreciate that you realize a tutorial / give an explanation of what is exactly RAG, Embeddings and Store Vector for people who are not familiar with. Like a video tutorial, a written article and a ComfyUI workflow example.

Thanks.

[shhlife]

I've just pushed a new release that should handle loading of a bunch of different text files - will be showing some examples in the examples section of the repo shortly!

[l-Aeden-l]

Thanks a lot!