#!/usr/bin/env python3
"""Scan api/app for AI-related symbols (LLM, embedding, langchain JSON helpers).

Run from repo root:
  uv run python api/scripts/ai_touchpoints_scan.py
  uv run python api/scripts/ai_touchpoints_scan.py --markdown api/docs/ai-touchpoints.md

Default prints sorted unique file paths to stdout.
"""

from __future__ import annotations

import argparse
import re
import sys
from pathlib import Path

# Line must match at least one pattern to count as a touchpoint for that file.
PATTERNS: list[tuple[str, re.Pattern[str]]] = [
    ("llm_provider", re.compile(r"get_llm_provider(_fast)?\b")),
    ("embedding", re.compile(r"get_embedding_provider\b|EmbeddingProvider\b")),
    ("json_llm_helpers", re.compile(r"invoke_json_object|ainvoke_json_object|allm_json_call|llm_json_call\b")),
    ("llm_call_module", re.compile(r"from app\.core\.llm_call|import app\.core\.llm_call")),
    ("langchain", re.compile(r"\blangchain_|from langchain|import langchain")),
    ("ports_ai", re.compile(r"from app\.ports\.(llm|embedding|asr|tts)\b|LLMProvider\b")),
    ("agents_layer", re.compile(r"from app\.agents\.|import app\.agents\.")),
    ("memory_ai", re.compile(r"MemoryService\b|HybridRetriever\b|retrieve_evidence_bundle_|schedule_memory_enrichment\b")),
]


def app_root_from_script() -> Path:
    return Path(__file__).resolve().parent.parent


def iter_python_files(root: Path) -> list[Path]:
    base = root / "app"
    if not base.is_dir():
        raise SystemExit(f"expected package dir: {base}")
    return sorted(p for p in base.rglob("*.py") if p.is_file())


def scan_files(files: list[Path]) -> dict[Path, list[str]]:
    hits: dict[Path, list[str]] = {}
    for path in files:
        try:
            text = path.read_text(encoding="utf-8")
        except OSError:
            continue
        tags: list[str] = []
        for tag, rx in PATTERNS:
            if rx.search(text):
                tags.append(tag)
        if tags:
            hits[path] = sorted(set(tags))
    return hits


def main() -> int:
    parser = argparse.ArgumentParser(description=__doc__)
    parser.add_argument(
        "--root",
        type=Path,
        default=None,
        help="api/ directory (default: parent of this script)",
    )
    parser.add_argument(
        "--markdown",
        type=Path,
        default=None,
        help="if set, write markdown report to this path",
    )
    args = parser.parse_args()
    root = args.root or app_root_from_script()
    files = iter_python_files(root)
    hits = scan_files(files)

    lines = [f"{len(hits)} files with AI touchpoints under {root / 'app'}"]
    for path in sorted(hits.keys()):
        rel = path.relative_to(root.parent) if path.is_relative_to(root.parent) else path
        lines.append(f"{rel}\t{','.join(hits[path])}")

    report = "\n".join(lines) + "\n"
    sys.stdout.write(report)

    if args.markdown:
        md_lines = [
            "# AI touchpoints (generated)",
            "",
            "Regenerate: `uv run python api/scripts/ai_touchpoints_scan.py --markdown api/docs/ai-touchpoints.md`",
            "",
            "| File | Tags |",
            "|------|------|",
        ]
        for path in sorted(hits.keys()):
            rel = path.relative_to(root.parent)
            tags = ", ".join(f"`{t}`" for t in hits[path])
            md_lines.append(f"| `{rel}` | {tags} |")
        md_lines.append("")
        args.markdown.parent.mkdir(parents=True, exist_ok=True)
        args.markdown.write_text("\n".join(md_lines), encoding="utf-8")
        print(f"Wrote {args.markdown}", file=sys.stderr)

    return 0


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