genesis/api_clients/pubmed_api.py

# genesis/api_clients/pubmed_api.py
import requests
from xml.etree import ElementTree as ET

def search_pubmed(query: str, max_results: int = 10):
    """
    Search PubMed using the NCBI E-utilities API.
    Returns a list of dicts with 'title', 'authors', 'pub_date', 'link'.
    """
    base_url = "https://eutils.ncbi.nlm.nih.gov/entrez/eutils/esearch.fcgi"
    summary_url = "https://eutils.ncbi.nlm.nih.gov/entrez/eutils/esummary.fcgi"

    # Step 1: Search PubMed IDs
    search_params = {
        "db": "pubmed",
        "term": query,
        "retmax": max_results,
        "retmode": "json"
    }
    search_res = requests.get(base_url, params=search_params)
    search_res.raise_for_status()
    id_list = search_res.json().get("esearchresult", {}).get("idlist", [])

    if not id_list:
        return []

    # Step 2: Fetch summaries for IDs
    summary_params = {
        "db": "pubmed",
        "id": ",".join(id_list),
        "retmode": "xml"
    }
    summary_res = requests.get(summary_url, params=summary_params)
    summary_res.raise_for_status()

    root = ET.fromstring(summary_res.text)
    papers = []
    for docsum in root.findall(".//DocSum"):
        paper = {"title": None, "authors": [], "pub_date": None, "link": None}
        for item in docsum.findall("Item"):
            if item.attrib.get("Name") == "Title":
                paper["title"] = item.text
            elif item.attrib.get("Name") == "PubDate":
                paper["pub_date"] = item.text
            elif item.attrib.get("Name") == "AuthorList":
                paper["authors"] = [author.text for author in item.findall("Item")]
        uid_elem = docsum.find("Id")
        if uid_elem is not None:
            paper["link"] = f"https://pubmed.ncbi.nlm.nih.gov/{uid_elem.text}/"
        papers.append(paper)

    return papers