convert-file2md — File to Markdown Converter

Overview

Batch convert PDF, DOCX, PPTX, and XLSX files to Markdown format using the mineru CLI engine.

Requirements

• Python 3.7+
• mineru CLI (pip install mineru)
• curl (for image download)

The script automatically locates mineru in your PATH or common installation directories.

Quick Start

# Single file conversion
python3 convert2md.py -i document.pdf -o ./output/

# Batch conversion (all supported files in a directory)
python3 convert2md.py -i ./input/ -o ./output/

# Chinese document
python3 convert2md.py -i document.pdf -o ./output/ -l ch

# Scanned document (OCR mode)
python3 convert2md.py -i scanned.pdf -o ./output/ -m ocr

CLI Parameters

Parameter	Short	Description	Default
--input	-i	Input file or directory	./input/
--output	-o	Output directory	./output/
--backend	-b	mineru backend engine	pipeline
--method	-m	Parse method	auto
--lang	-l	Document language	en
--keep-intermediates	—	Keep intermediate files	false

Backend Options

Value	Description
pipeline	(Default) Local Pipeline engine
vlm-http-client	VLM HTTP client mode
hybrid-http-client	Hybrid HTTP client mode
vlm-auto-engine	VLM auto engine
hybrid-auto-engine	Hybrid auto engine

Parse Methods

Value	Description
auto	(Default) Auto-detect
txt	Plain text mode (fast)
ocr	OCR mode for scanned documents

Language Options

Value	Use Case
en	(Default) English
ch	Chinese
ch_server	Chinese (server model)
ch_lite	Chinese (lightweight model)
japan	Japanese
korean	Korean
chinese_cht	Traditional Chinese
latin	Latin scripts
arabic	Arabic
cyrillic	Cyrillic
devanagari	Devanagari

Output Structure

output_dir/
└── <filename>/
    ├── <filename>.md
    └── images/
        ├── image1.png
        └── image2.jpg

Supported Formats

Only .pdf, .docx, .pptx, .xlsx are supported. Legacy .doc files require conversion first.

Troubleshooting

1. mineru not found: Install via pip install mineru or ensure it's in PATH.

   pip install mineru
   which mineru

2. Poor Chinese recognition: Use -l ch instead of the default -l en.
   Traditional Chinese: -l chinese_cht.

3. Conversion timeout: Single file timeout is 600 seconds (10 minutes). For large files, try -m txt to skip OCR.

4. Unsupported formats: Only .pdf, .docx, .pptx, .xlsx are supported.