"""
Test efetch date parsing for YYYY-01-01 records — CORRECTED parse
"""
import asyncio, httpx, lxml.etree as ET
from datetime import date
from sqlalchemy import text
from sqlalchemy.ext.asyncio import create_async_engine

DATABASE_URL = 'postgresql+asyncpg://scilit:scilit_prod_2026@postgres:5432/scilit'

MONTH_MAP = {
    'jan':1,'january':1,'1':1,'feb':2,'february':2,'2':2,'mar':3,'march':3,'3':3,
    'apr':4,'april':4,'4':4,'may':5,'5':5,'jun':6,'june':6,'6':6,
    'jul':7,'july':7,'7':7,'aug':8,'august':8,'8':8,'sep':9,'september':9,'9':9,
    'oct':10,'october':10,'10':10,'nov':11,'november':11,'11':11,'dec':12,'december':12,'12':12,
}
DAYS = [0,31,29,31,30,31,30,31,31,30,31,30,31]

def month_num(m): return MONTH_MAP.get(m.strip().lower(), 1)
def clamp_day(m, d, y):
    if d < 1: return 1
    md = DAYS[m]
    if m == 2 and d == 29 and not (y%4==0 and (y%100!=0 or y%400==0)): md = 28
    return min(d, md)

def parse_pub_date(elem):
    """Same logic as fix_dates_via_efetch.py"""
    pde = elem.find('.//PubDate')
    if pde is None: return None
    ye = pde.find('Year'); me = pde.find('Month'); de = pde.find('Day')
    if ye is not None and ye.text:
        y = int(ye.text.strip())
        hm = me is not None and me.text; hd = de is not None and de.text
        if not hm and not hd: return None
        m = month_num(me.text.strip()) if hm else 1
        d = int(de.text.strip()) if hd else 1
        d = clamp_day(m, d, y)  # clamp, not swap
        return date(y, m, d)
    return None

async def main():
    engine = create_async_engine(DATABASE_URL)
    async with engine.connect() as conn:
        # Sample across all years with YYYY-01-01
        r = await conn.execute(text("""
            SELECT pub_year, COUNT(*) as cnt FROM global_literature
            WHERE pub_date IS NOT NULL AND EXTRACT(MONTH FROM pub_date)=1 AND EXTRACT(DAY FROM pub_date)=1
            AND pmid IS NOT NULL
            GROUP BY pub_year ORDER BY pub_year DESC
        """))
        print('=== YYYY-01-01 records by year ===')
        for row in r:
            print(f'  {row.pub_year}: {row.cnt}')

        # Get 2 PMIDs from each year that has >0 records
        r = await conn.execute(text("""
            SELECT id, pmid, pub_date, pub_year FROM global_literature
            WHERE pub_date IS NOT NULL AND EXTRACT(MONTH FROM pub_date)=1 AND EXTRACT(DAY FROM pub_date)=1
            AND pmid IS NOT NULL
            ORDER BY pub_year DESC, pmid DESC LIMIT 40
        """))
        samples = r.all()
    await engine.dispose()

    # Efetch and report which PMIDs have better dates
    test_pmids = [row.pmid for row in samples]
    print(f'\n=== Testing {len(test_pmids)} PMIDs via efetch ===')
    async with httpx.AsyncClient(timeout=30) as c:
        resp = await c.get(
            'https://eutils.ncbi.nlm.nih.gov/entrez/eutils/efetch.fcgi',
            params={'db': 'pubmed', 'id': ','.join(str(p) for p in test_pmids), 'retmode': 'xml'}
        )
        print(f'HTTP status={resp.status_code} content_len={len(resp.content)}')
        root = ET.fromstring(resp.content)
        arts = root.findall('PubmedArticle')
        print(f'Articles returned: {len(arts)}')

        fixed = 0; year_only = 0; still_jan01 = 0
        efetch_map = {}
        for art in arts:
            pmid_e = art.find('MedlineCitation/PMID')
            pmid = int(pmid_e.text) if pmid_e is not None else None
            pd = parse_pub_date(art)
            if pmid: efetch_map[pmid] = pd

        for row in samples:
            pd = efetch_map.get(row.pmid)
            if pd is None:
                year_only += 1
            elif pd.month == 1 and pd.day == 1:
                still_jan01 += 1
            else:
                fixed += 1
                print(f'  CAN FIX: PMID={row.pmid} {row.pub_date} -> {pd}')

        print(f'\n=== Summary ===')
        print(f'  Can fix (better date found): {fixed}')
        print(f'  Year-only (no month/day):   {year_only}')
        print(f'  Still Jan-01:               {still_jan01}')

asyncio.run(main())