findar: Search arXiv papers with GitHub integration

Search arXiv papers and automatically find associated GitHub repositories with a single command.

Features

• 🔍 Search arXiv - Query academic papers from arXiv.org
• 🐙 GitHub Integration - Automatically find associated GitHub repositories
• ⭐ Smart Comparison - Compare multiple repos by star count
• 📊 Export Results - Save results as Stata dataset
• 🔗 Clickable Links - Direct links to papers, PDFs, DOIs, and repos
• 🎯 Flexible Syntax - Simple keywords or detailed query options

Installation

net install findar, from("https://raw.githubusercontent.com/BlueDayDreeaming/findar/main/")

Or download and place in your Stata ado directory.

Requirements

• Stata 16.0 or later
• Windows or macOS (Linux not supported)
• Internet connection

Quick Start

Basic Search

findar deep learning

Search with Options

findar machine learning, maxresults(10)

Display Abstracts

findar transformer attention, abstract

Detailed Information

findar neural networks, detail

Save Results

findar causal inference, saving(results) replace

Syntax

Simplified Syntax

findar keywords [, options]

Standard Syntax

findar, query(string) [options]

Options

Option	Description
query(string)	Search query (required for standard syntax)
maxresults(#)	Maximum number of results (default: 10)
detail	Display detailed information
abstract	Show paper abstracts
nogithub	Disable GitHub search
save	Keep results in memory (v1.1.5+)
saving(filename)	Save results to file
replace	Overwrite existing dataset

Examples

See findar_example.do for comprehensive examples.

* Basic search
findar deep learning

* More results
findar machine learning, maxresults(20)

* With abstracts
findar transformer attention, abstract

* Detailed mode
findar reinforcement learning, detail

* Save to memory (new in v1.1.5)
findar computer vision, maxresults(10) save

* Save to file
findar computer vision, saving(cv_papers) replace

* Standard syntax
findar, query("natural language processing") maxresults(5)

Output Variables

When using saving(), the dataset includes:

Variable	Description
arxiv_id	arXiv paper ID
title	Paper title
authors	Author list
published	Publication date
summary	Abstract
doi	Digital Object Identifier
gh_repo	GitHub repository (owner/repo)
gh_url	GitHub URL
gh_stars	Star count
gh_lang	Programming language

Stored Results

Result	Description
r(count)	Number of papers found
r(arxiv_found)	Papers with GitHub repos
r(not_found)	Papers without repos

How It Works

1. Query arXiv API - Searches arXiv for matching papers
2. Extract GitHub URLs - Scans paper metadata for GitHub links
3. Fetch Repo Info - Queries GitHub API for repository details
4. Compare & Rank - When multiple repos found, compares by stars
5. Display Results - Shows formatted output with clickable links

Screenshots

Basic Search

--------------------------------------------------------------------------------
arXiv Search + GitHub Integration
--------------------------------------------------------------------------------
Query: deep learning
Max results: 5
GitHub search: ENABLED (default)
--------------------------------------------------------------------------------
Searching arXiv...
Found 5 papers.

[1] Opening the black box of deep learning
Links: arXiv PDF DOI GitHub(★1234)

Detailed Mode

[1] Neural Network Architectures
Authors: John Doe, Jane Smith
Published: 2024-01-15

Links: arXiv  PDF  GitHub(repo_name_★5678_Python)

Known Limitations

• Linux/Unix not supported (Windows and macOS only)
• GitHub API rate limits apply (60 requests/hour unauthenticated)
• Requires internet connection
• Corporate firewalls may block API access

Help

help findar

Citation

If you use findar in your research, please cite:

Chucheng Wan, Yile Zhang, Xinyi Huang, Qin Qin, and Xinyi Yi (2025). findar: Search arXiv papers with GitHub integration.
Stata package. https://github.com/BlueDayDreeaming/findar

Support

• 📝 Documentation
• 🐛 Issue Tracker
• 💬 Discussions

License

MIT License

Authors

Chucheng Wan

• Affiliation: Sun Yat-sen University, Guangzhou, China
• Email: chucheng.wan@outlook.com

Yile Zhang

• Affiliation: Sun Yat-sen University, Guangzhou, China
• Email: zhangyle5@mail2.sysu.edu.cn

Xinyi Huang

• Affiliation: Sun Yat-sen University, Guangzhou, China
• Email: huangxy577@mail2.sysu.edu.cn

Qin Qin

• Affiliation: Sun Yat-sen University, Guangzhou, China
• Email: qinq25@mail2.sysu.edu.cn

Xinyi Yi

• Affiliation: Sun Yat-sen University, Guangzhou, China
• Email: 3031727931@qq.com

Changelog

Version 1.1.5 (2025-10-18)

• Added save option to keep results in memory
• Fixed macOS connectivity issues (HTTPS support)
• Added XML declaration skip for better parsing
• Improved data handling (auto-clear when not saving)
• Removed diagnostic tool (findar_test)

Version 1.1.4 (2025-10-18)

• HTTPS support for macOS compatibility
• XML parsing improvements

Version 1.1.1 (2025-10-18)

• Added system compatibility check (Windows/macOS only)
• Code optimization and cleanup
• Improved documentation
• Added comprehensive examples

Version 1.0.0 (Initial Release)

• arXiv search functionality
• GitHub integration
• Multiple syntax support
• Dataset export capability