I like this project, it seems sensible to have an agreed format for this :)
The issue that jumps out to me is that, as far as I understand, the format would not support having files with the same basename but different extensions? e.g. example.txt and example.md would both have a metadata file name of example.meta.yml?
I'd therefore suggest it's preferable not to drop the final extension.
As a similar example, AWS Bedrock Knowledge Bases appends .metadata.json to the full filename including the extension (image from https://aws.amazon.com/blogs/machine-learning/multi-tenancy-in-rag-applications-in-a-single-amazon-bedrock-knowledge-base-with-metadata-filtering/ ).
I like this project, it seems sensible to have an agreed format for this :)
The issue that jumps out to me is that, as far as I understand, the format would not support having files with the same basename but different extensions? e.g.
example.txtandexample.mdwould both have a metadata file name ofexample.meta.yml?I'd therefore suggest it's preferable not to drop the final extension.
As a similar example, AWS Bedrock Knowledge Bases appends
.metadata.jsonto the full filename including the extension (image from https://aws.amazon.com/blogs/machine-learning/multi-tenancy-in-rag-applications-in-a-single-amazon-bedrock-knowledge-base-with-metadata-filtering/ ).