Skip to main content

PptxToMarkdownPipeline

What it is​

A pipeline that converts a .pptx (PowerPoint) file into Markdown by extracting text from each slide’s XML and formatting it as slide headers with bullet points.

Public API​

  • PptxToMarkdownPipelineConfiguration
    • mime_type: str — MIME type this pipeline targets. Default:
      • application/vnd.openxmlformats-officedocument.presentationml.presentation
  • PptxToMarkdownPipelineParameters
    • processor_iri: str — IRI identifying the processor. Default:
      • http://ontology.naas.ai/abi/document/PptxToMarkdownProcessor
  • PptxToMarkdownPipeline
    • convert_to_markdown(file: File) -> str — Reads PPTX content from file, extracts slide text, and returns Markdown.
    • Internal helpers (not intended as public API):
      • _normalize_whitespace(value: str) -> str
      • _list_slide_paths(pptx_file: zipfile.ZipFile) -> list[str]
      • _extract_text_from_slide(slide_xml: bytes) -> list[str]

Configuration/Dependencies​

  • Depends on:
    • naas_abi_marketplace...File providing read() -> bytes
    • Base classes:
      • ConvertToMarkdownBasePipeline
      • ConvertToMarkdownBasePipelineConfiguration
      • ConvertToMarkdownBasePipelineParameters
    • Standard library: zipfile, io, re, xml.etree.ElementTree
    • pydantic.Field for parameter metadata

Usage​

from naas_abi_marketplace.domains.document.pipelines.PptxToMarkdownPipeline import (
    PptxToMarkdownPipeline,
)
from naas_abi_marketplace.domains.document.ontologies.classes.ontology_naas_ai.abi.document.File import File

pipeline = PptxToMarkdownPipeline()

pptx_file: File = ...  # must implement .read() returning PPTX bytes
md = pipeline.convert_to_markdown(pptx_file)
print(md)

Markdown output shape:

  • For each non-empty slide:
    • ## Slide N
    • - <line 1>
    • - <line 2>
    • ...

Caveats​

  • Only extracts text present in slide XML text runs (a:t) and treats line breaks (a:br) as newline markers before whitespace normalization.
  • Images, tables, speaker notes, and most formatting are not converted.
  • If the input is not a valid PPTX zip (or slide XML cannot be read), convert_to_markdown raises ValueError("Invalid PPTX content: unable to read slide XML").
  • Slide numbering is derived from ppt/slides/slide*.xml filenames and sorted by their numeric index.