document-convert

Convert local documents into normalized markdown or plain text with lightweight metadata.

Why install it

Knowledge workflows need a common representation for mixed source files. This tool gives an agent a simple document normalization primitive for text, markdown, HTML, JSON, and CSV inputs.

Inputs

• path: path to the local document to convert
• to_format: markdown or text
• extract_metadata: whether to include file metadata such as size and line count

Outputs

• path: original file path
• media_type: detected media type for the input
• content: converted document body
• metadata: file and conversion metadata

Local development

The source code for this tool can be found here

Test:

python -m unittest discover -s tests -p 'test_*.py'

Example invocation

python -u document_convert/__main__.py < input.json

With input.json containing:

 {
    "path": "fixtures/sample.html",
    "to_format": "markdown"
  }