scripts/backfill_narrative.py

"""Attach a press-style narrative to a brief that was generated without one.

The narrative is composed from the brief itself -- verified facts, computed
quarter deltas, the reported trajectory -- and needs no filings, no vector
store and no agent loop. So a stored brief can gain its article for the cost
of one short model call instead of a full regeneration.

Usage::

    python -m scripts.backfill_narrative AAPL
    python -m scripts.backfill_narrative AAPL --dry-run
    python -m scripts.backfill_narrative --all

``--dry-run`` prints the article and the validation report without touching
the database.
"""
from __future__ import annotations

import argparse
import json
import sqlite3
import sys

from dotenv import load_dotenv

from agent.llm import default_config
from agent.narrative import compose_narrative
from storage import briefs_db

load_dotenv()


def _update_stored_brief(ticker: str, brief: dict) -> int:
    """Write *brief* back over the rows it came from. Returns rows touched.

    save_brief() cannot be used here: its snapshot insert is keyed on
    generated_at and would be ignored, leaving get_brief() -- which reads
    brief_snapshots first -- still serving the article-less payload.
    """
    payload = json.dumps(brief)
    generated_at = brief.get("generated_at") or ""
    with sqlite3.connect(briefs_db.DB_PATH) as conn:
        touched = conn.execute(
            "UPDATE brief_snapshots SET brief_json = ? "
            "WHERE ticker = ? AND generated_at = ?",
            (payload, ticker.upper(), generated_at),
        ).rowcount
        touched += conn.execute(
            "UPDATE briefs SET brief_json = ? WHERE ticker = ?",
            (payload, ticker.upper()),
        ).rowcount
        # A stored translation predates the article and would render the
        # brief without it; drop it so it is rebuilt on next request.
        conn.execute("DELETE FROM brief_translations WHERE ticker = ?", (ticker.upper(),))
    return touched


def backfill(ticker: str, *, dry_run: bool = False, force: bool = False) -> int:
    brief = briefs_db.get_brief(ticker)
    if brief is None:
        print(f"{ticker}: no stored brief", file=sys.stderr)
        return 1
    if brief.get("narrative") and not force:
        print(f"{ticker}: already has a narrative (use --force to replace)")
        return 0

    try:
        from analytics.segments import digest_lines
        brief["segment_lines"] = digest_lines(ticker)
    except Exception as exc:
        print(f"  · segment lines unavailable: {exc}")

    article, report, usage = compose_narrative(
        brief, default_config(), brief.get("language") or "English"
    )
    for line in report:
        print(f"  · {line}")
    if usage:
        print(f"  · usage: {usage}")

    if article is None:
        print(f"{ticker}: no article survived validation")
        return 1

    print(f"\n  {article.get('headline') or '(headline dropped)'}")
    print(f"  {article.get('dek') or ''}\n")
    for para in article["paragraphs"]:
        print(f"  [{para['role']}] ({', '.join(para['fact_ids'])})")
        print(f"  {para['text']}\n")

    if dry_run:
        print(f"{ticker}: dry run, nothing written")
        return 0

    brief["narrative"] = article
    brief["narrative_usage"] = usage
    rows = _update_stored_brief(ticker, brief)
    print(f"{ticker}: narrative attached ({rows} row(s) updated)")
    return 0


def main() -> int:
    parser = argparse.ArgumentParser(description=__doc__)
    parser.add_argument("ticker", nargs="?", help="Ticker to backfill.")
    parser.add_argument("--all", action="store_true", help="Every ticker with a stored brief.")
    parser.add_argument("--dry-run", action="store_true", help="Print without writing.")
    parser.add_argument("--force", action="store_true", help="Replace an existing narrative.")
    args = parser.parse_args()

    if args.all:
        tickers = briefs_db.list_tickers()
    elif args.ticker:
        tickers = [args.ticker.upper()]
    else:
        parser.error("give a ticker or --all")

    failures = 0
    for ticker in tickers:
        print(f"\n== {ticker} ==")
        failures += backfill(ticker, dry_run=args.dry_run, force=args.force)
    return 1 if failures else 0


if __name__ == "__main__":
    raise SystemExit(main())