Home

Format parsers

parse_pdf, parse_docx, parse_pptx, and parse_html are format-specific variants of parse_document. They accept the same parameters and return the same ParseResult.

Use them when you want parsing to fail immediately if the file doesn't match the expected format. In all other cases, parse_document is simpler.

Signatures

docslicer.parse_pdf(source, **kwargs)                    -> ParseResult
docslicer.parse_docx(source, **kwargs)                   -> ParseResult
docslicer.parse_pptx(source, **kwargs)                   -> ParseResult
docslicer.parse_html(source, source_url=None, **kwargs)  -> ParseResult

All **kwargs are identical to the parameters on parse_document.

What source accepts

FunctionAccepted sources
parse_pdfFile path, raw PDF bytes, file-like object
parse_docxFile path, raw DOCX bytes, file-like object
parse_pptxFile path, raw PPTX bytes, file-like object
parse_htmlURL, file path, raw HTML string, file-like object

parse_html also accepts a source_url keyword argument (original URL of the page, used for link normalisation when source is a file path or raw HTML string).

Example

# Raises immediately if the bytes aren't a valid PDF
with open(upload_path, "rb") as f:
    result = docslicer.parse_pdf(f)