Document loaders - Source Code - Add support for PL/SQL

[stelapo]

Checked

• I searched existing ideas and did not find a similar one
• I added a very descriptive title
• I've clearly described the feature request and motivation for it

Feature request

I propose to add the support of Oracle PLSQL language between code language parsers used to load source code files.

Referenced documentation: https://python.langchain.com/docs/integrations/document_loaders/source_code

Motivation

I'm working on a project where we'd like to use GPT AI model for the following use-cases:

• Q&A over the PLSQL code base to understand how it works
• using LLMs for documenting the code
• using LLM to help convert some stored procedures to other code languages, also generating unit tests.

Proposal (If applicable)

Following langchain documentation (https://python.langchain.com/docs/use_cases/code_understanding) it seems that first of all it's needed to split all source code files using a splitting strategy that does a few things:

• Keeps each top-level function in the code is loaded into separate documents.
• Puts remaining into a separate document.
• Retains metadata about where each split comes from

I don't know if anyone is already working on it (I couldn't find any open discussion about it), but I propose to add support for PLSQL, following the instructions here: https://python.langchain.com/docs/integrations/document_loaders/source_code#adding-languages-using-tree-sitter-template

I am not very knowledgeable about langchain and LLVM in general, so please let me know if my proposal does not make sense or if I'm not considering important points.

Thank you very much in advance