nasa-ads-citations

name	nasa-ads-citations
description	Guide for verifying and extracting BibTeX metadata from NASA Astrophysics Data System (ADS). Includes verified API endpoints, search strategies, and CrossRef fallback.

NASA ADS Citation Verification

Purpose

Verify, correct, and extract BibTeX metadata for astronomy/astrophysics papers from NASA ADS, with CrossRef as fallback.

Workflow

1. Know the ADS API Endpoints

Endpoint	Purpose	Status
https://ui.adsabs.harvard.edu/abs/{bibcode}/abstract	View paper details in browser	Often JS-rendered, may timeout
https://ui.adsabs.harvard.edu/api/search?	Search API	Returns HTTP 202 (async), not directly usable
https://ui.adsabs.harvard.edu/api/article?bibcode=	Best option for single bibcode lookup	Check if available
https://api.crossref.org/works/{doi}	Fallback — CrossRef DOI lookup	Most reliable, fast
http://export.arxiv.org/api/query?	arXiv search by query/author/title	Good for arXiv IDs

2. Primary Method: CrossRef DOI Lookup

CrossRef is the most reliable endpoint. Use it to verify:

• Author list
• Journal, volume, pages
• Year of publication
• Title
• arXiv ID (in arxiv_id field)

import requests
doi = "10.1051/0004-6361/202451427"
resp = requests.get(f"https://api.crossref.org/works/{doi}")
d = resp.json()['message']
print(f"Pages: {d.get('page','?')}")
print(f"Volume: {d.get('volume','?')}")
print(f"Authors: {[a.get('family') for a in d.get('author',[])]}")
print(f"Year: {d.get('published-print',{}).get('date-parts',[[None]])[0][0]}")
print(f"arXiv: {d.get('link',[])[0].get('URL','?')}")  # if present

3. Secondary Method: arXiv API

Use when you have the arXiv ID or need to find a paper by author/title:

# Search by arXiv ID
resp = requests.get("http://export.arxiv.org/api/query?id_list=2407.06963")
# Search by author+title
resp = requests.get("http://export.arxiv.org/api/query?search_query=au:Anderson+AND+ti:Gaia&max_results=5")

4. ADS Search Strategy

When CrossRef/DOI lookups are ambiguous or you need the ADS bibcode:

1. Known DOI → CrossRef lookup first (returns journal, volume, pages, authors)
2. Known author + title keywords → arXiv API search, then check arXiv page for ADS bibcode reference
3. Known volume + first page → Construct potential ADS bibcode (format: YYYYJournal...Volume...Page)

ADS bibcode format: YYYYJournal...Volume...Page

• 2024A&A...691A..98K = Year 2024, A&A, Volume 691, Article A98, first author with surname starting K

5. Common Pitfalls

• ADS API is flaky — direct API calls often time out or return 202 (async). Don't rely on it.
• CrossRef DOI matching can be wrong — always verify the title matches the intended paper, not just the DOI format.
• Author name variations — ADS and CrossRef may format names differently (e.g., "Maiz-Apellániz" vs "Maiz Apellániz" vs "Maiz-Apellaniz"). Always cross-check the title.
• Wrong paper in bib entry — a bib entry can have the correct title but wrong author (as happened with Zhang et al. misattributed to Maiz-Apellániz). Always verify author + title match.
• Page numbers in A&A — use A94, A215, etc. — not numeric pages.
• Gaia Collaboration papers — author field should be {Gaia Collaboration}, NOT individual names.
• Consortium/community papers can return garbage authors — e.g. Galaxy 2022 (10.1093/nar/gkac247) returns a broken leading author (", and Afgan, Enis and ...", hundreds of names). Use the collective form {The Galaxy Community} instead of pasting the raw list.
• "Website" citations often have a real proceedings paper — Binder is not \url{https://mybinder.org}: it is a SciPy 2018 proceedings paper (doi 10.25080/Majora-4af1f417-011, pages 113–120). Before citing a tool as @misc+URL, check CrossRef/the project docs for a citable paper.

5b. Batch verification (shell one-liners)

Fast pattern for verifying many DOIs at once with curl + jq (no Python needed):

# Author lists for a batch of DOIs
for doi in 10.1073/pnas.1708290115 10.1126/science.1213847; do
  echo "--- $doi"
  curl -s --max-time 15 "https://api.crossref.org/works/$doi" | \
    jq -r '.message.author | map(.family + ", " + (.given // "")) | join(" and ")'
done

# Full metadata line: DOI || title || journal || vol || issue || pages || year
curl -s "https://api.crossref.org/works/$doi" | jq -r '.message |
  [.DOI, .title[0], (.["container-title"][0] // "?"), (.volume // "?"),
   (.issue // "-"), (.page // "?"), (.issued["date-parts"][0][0]|tostring)] | join(" || ")'

6. BibTeX Entry Template

@article{citationkey,
  author  = {Author, A. and Coauthor, B.},
  title   = {{T}itle {W}ith {C}aps},
  journal = {Journal},
  year    = {YYYY},
  volume  = {NNN},
  pages   = {PPP},
  doi     = {10.xxxx/xxxxx},
  note    = {ADS: YYYYJournal...Vol...Page; arXiv:xxxx.xxxxx}
}

7. Verification Checklist

For each reference, verify:

• Title matches the paper being cited
• Author list is correct (not just surname)
• Journal, volume, and pages match ADS/CrossRef
• DOI resolves to the correct paper
• ADS bibcode is constructible and valid format
• arXiv ID is listed in the note

8. Author Name Conventions

• Individual authors: Last, F. or Last, F. and Coauthor, A.
• Gaia Collaboration: always use {{Gaia Collaboration}} (double braces for single-name author)
• Names with special characters: use LaTeX escapes (e.g., Ma{\\'i}z-Apell{\\'a}niz)
• For natbib with a ordering (A&A style), ensure bibpunct is set: \bibpunct{(}{)}{;}{a}{}{,}

9. When All Else Fails

If CrossRef + arXiv both fail:

1. Check the paper's arXiv abstract page (https://arxiv.org/abs/YYYY.NNNNN) — ADS bibcode is often listed in the metadata
2. Try the ADS abstract page directly (https://ui.adsabs.harvard.edu/abs/{potential_bibcode}/abstract)
3. Search the journal's publisher site for the specific volume/article
4. As a last resort, ask the user for the ADS bibcode directly