docs/how_to/markdown_header_metadata_splitter/

[giscus[bot]]

docs/how_to/markdown_header_metadata_splitter/

Motivation

https://python.langchain.com/docs/how_to/markdown_header_metadata_splitter/

[matthewyoung777]

What if I wanted to split documents that were split using the MarkdownHeaderTextSplitter, but still include every header in the page content of each chunk?

[scottmatthews7]

I have a similar issue. I want each chunk to retain metadata such as:

Section, Sub Section, Sub Sub Section, Figure Heading

So that in the response, it is clear where the information came from in my source material (a textbook in this case).

[asvishnyakov]

@matthewyoung777 @scottmatthews7 Isn't strip_headers = False is the thing you looking for?