arxivparse 0.1.3

Convert arXiv papers to plain text using LTeXML

arxivparse

Convert arXiv papers to clean plain text using LaTeXML. Downloads LaTeX source from arXiv, converts to XML, and extracts body text (title, abstract, sections, math, captions) — no HTML intermediary, no bibliography, no footnotes.

Prerequisites

• Python >= 3.13
• uv (or pip)
• LaTeXML (v0.8.x) — install via Homebrew:

brew install latexml

Verify it's on your PATH:

latexml --VERSION

Install

# From PyPI
pip install arxivparse

# Or from source
git clone <repo-url> arxivparse
cd arxivparse
uv sync

Quick Start

from arxivparse import arxiv_to_text

text = arxiv_to_text("1706.03762")
print(text[:200])

CLI Usage

# Single paper
arxivparse 1706.03762

# Multiple papers (sequential)
arxivparse 1706.03762 2301.07041

# Custom output path
arxivparse -o output.txt 1706.03762

# Custom output directory
arxivparse -d ./papers 1706.03762 2301.07041

# Verbose output (download, convert, extract steps)
arxivparse -v 1706.03762

# Keep temp files for debugging
arxivparse --keep-temp 1706.03762

Each paper produces a <arxiv_id>.txt file.

Library Usage

Simple: get text as a string

from arxivparse import arxiv_to_text

text = arxiv_to_text("1706.03762")

Full pipeline control

from arxivparse.pipeline import convert_arxiv_to_text
from arxivparse.errors import Arxiv2TextError

try:
    output_path = convert_arxiv_to_text("1706.03762")
    text = output_path.read_text(encoding="utf-8")
except Arxiv2TextError as e:
    print(f"Failed: {e}")

Call the CLI from code

from main import main

main(["1706.03762", "2301.07041"])
main(["-o", "output.txt", "1706.03762"])

Arguments

Param	Type	Default	Description
arxiv_id	str	—	arXiv ID (e.g. "1706.03762")
output_path	Path	<arxiv_id>.txt	Where to write the output
keep_temp	bool	False	Keep temp files after conversion

Error Handling

from arxivparse.errors import (
    DownloadError,         # network/HTTP failure
    NoLatexSourceError,    # paper is PDF-only
    ConversionError,       # latexml failed
    MainTexNotFoundError,  # no .tex file found in bundle
)

Build / Publish

# Build a wheel
uv build

# Publish to PyPI
uv publish

# Or with twine
twine upload dist/*

Output Format

QASPER-style plain text: title, abstract, section headings, body paragraphs, inline math (LaTeX notation like h_{t}), and figure/table captions. No bibliography entries, footnotes, author affiliations, or citation numbers.