searchpubmed

A Python helper‑library for programmatic PubMed work:

• Search – build and run complex boolean queries (E‑Search)
• Link – map PMIDs ↔ PMCIDs (E‑Link)
• Fetch – pull rich article metadata (E‑Fetch / E‑Summary)
• Full‑text – download JATS XML + HTML from PMC
• Analyse – get tidy results as pandas DataFrames

---

Query‑builder

The package now ships with a lightweight query‑builder and five pre‑tuned strategies for common real‑world‑data searches.

from searchpubmed import build_query, STRATEGY3_OPTS
print(build_query(STRATEGY3_OPTS)[:120] + "…")
# => (("observational" 5 "ehr"[tiab] OR …) AND english[lang] AND ("2010"[dp] : "3000"[dp]))

Create your own QueryOptions:

from searchpubmed import QueryOptions, build_query
opts = QueryOptions(
    data_sources=["ehr", "claims"],  # synonym keys
    design_terms=["observational", "secondary"],
    proximity_within=None,
)
print(build_query(opts))

---

Core features

• Perform complex boolean searches (AND/OR/NOT, phrase, proximity N)
• Batch retrieval of PMIDs and conversion to PMCIDs
• Fetch detailed metadata (title, abstract, authors, journal, date)
• Configurable rate‑limiting to stay within NCBI usage caps
• Results returned as pandas DataFrames for instant analysis

---

Installation

pip install searchpubmed

---

License

Apache 2.0 – see the LICENSE file for details.