Closed smallyear closed 3 months ago
I sincerely apologize for the confusion in my previous comments. Upon further investigation, I've discovered that there is indeed a method to modify the schema in the codebase. This oversight was entirely my mistake.
Given this new information, I believe this issue can be closed. The functionality I was requesting already exists within the project.
Feature Request: Custom Schema Modification
Current Behavior
Currently, the
extractor.infer_schema
method generates a schema that may not fully meet specific user requirements.Proposed Solution
I would like to propose adding functionality that allows users to modify the generated schema before using it with the
extractor.extract
method. This would enable users to tailor the schema to better fit their specific use cases.Use Case
In my particular scenario, I need to adjust the inferred schema to more accurately reflect the structure of the data I'm working with. Having the ability to modify the schema would greatly enhance the flexibility and usability of llama-extract for my application.
Potential Implementation
Perhaps a new method could be introduced, such as
extractor.modify_schema(custom_schema)
, which would allow users to pass in their modified version of the ExtractionSchema.Benefits
This feature would:
Thank you for considering this feature request. I appreciate the work you've done on llama-extract and look forward to your thoughts on this proposal.