"""

GovInfo API poller  -  full bill text, Congressional Record, public laws.

Endpoint: GET https://api.govinfo.gov/published/{start}/{end}

Auth:      same api.data.gov key as Congress.gov (DOCKET_CONGRESS_KEY)

Rate:      36,000 req/hr  -  most generous of the four primary sources.

Lag:       1-2 business days behind the legislative action.

If DOCKET_CONGRESS_KEY is not set, this poller skips gracefully.

"""

import json

import logging

import os

import sqlite3

from datetime import datetime, timedelta

from typing import Optional

import requests

from ingest.change_detection import is_new_event, mark_seen

from ingest.db import get_connection, init_schema

from ingest.source_validator import validate_source_url

logger = logging.getLogger(__name__)

SOURCE = "govinfo"

BASE_URL = "https://api.govinfo.gov"

COLLECTIONS = ["BILLS", "PLAW", "CREC"]

def poll(days_back: int = 2, db_path: Optional[str] = None) -> int:

    """

    Fetch recently published GovInfo packages (bill text, public laws, CR).

    If DOCKET_CONGRESS_KEY is not set, skips gracefully and returns 0.

    """

    api_key = os.environ.get("DOCKET_CONGRESS_KEY")

    if not api_key:

        logger.warning(

            "GovInfo poller: DOCKET_CONGRESS_KEY is not set. "

            "Skipping GovInfo source (same key as Congress.gov - see congress.py)."

        )

        return 0

    end_dt   = datetime.utcnow()

    start_dt = end_dt - timedelta(days=days_back)

    start_str = start_dt.strftime("%Y-%m-%dT%H:%M:%SZ")

    end_str   = end_dt.strftime("%Y-%m-%dT%H:%M:%SZ")

    conn = get_connection(db_path) if db_path else get_connection()

    init_schema(conn)

    stored = 0

    for collection in COLLECTIONS:

        url = f"{BASE_URL}/published/{start_str}/{end_str}"

        params = {

            "api_key": api_key,

            "collection": collection,

            "pageSize": 100,

        }

        try:

            resp = requests.get(url, params=params, timeout=30)

            resp.raise_for_status()

        except requests.RequestException as exc:

            logger.error("GovInfo %s request failed: %s", collection, exc)

            continue

        data = resp.json()

        packages = data.get("packages", [])

        logger.info("GovInfo %s: received %d package(s)", collection, len(packages))

        for pkg in packages:

            pkg_id      = pkg.get("packageId", "")

            update_date = pkg.get("dateIssued", pkg.get("lastModified", ""))

            if not pkg_id or not is_new_event(conn, SOURCE, pkg_id, update_date):

                continue

            source_url = pkg.get("packageLink", f"https://www.govinfo.gov/content/pkg/{pkg_id}/pdf/{pkg_id}.pdf")

            try:

                validate_source_url(source_url)

            except Exception as exc:

                logger.error("GovInfo: invalid source_url %s: %s", source_url, exc)

                continue

            conn.execute(

                """

                INSERT INTO raw_events (source, external_id, payload, fetched_at)

                VALUES (?, ?, ?, ?)

                ON CONFLICT(source, external_id) DO UPDATE SET

                    payload    = excluded.payload,

                    fetched_at = excluded.fetched_at

                """,

                (SOURCE, pkg_id, json.dumps(pkg), datetime.utcnow().isoformat()),

            )

            mark_seen(conn, SOURCE, pkg_id, update_date)

            stored += 1

    conn.commit()

    conn.close()

    logger.info("GovInfo: stored %d new package(s)", stored)

    return stored