Customizing PDF parsing

When a PDF file is added to a document, it is converted to Markdown before chunking (see Parsing documents). The pdf_parser advanced configuration parameter, set in the document's Advanced tab, customizes how this conversion is done:

Parameter	Type	Description
detect_tables	bool	Render detected tables as Markdown tables. Default true.
strip_running_headers	bool	Remove repeated text at the top and bottom of each page — running headers, footers and page numbers. Default true.

For example, to disable table detection while keeping running headers:

{
    "pdf_parser": {
        "detect_tables": false,
        "strip_running_headers": false
    }
}