"""
Labeler C: Frame Check detector labels.

Runs framing.py + frame_library.suggest_frames on each corpus document.
Produces binary per-frame labels: frame is exhibited iff it appears in
the detector's output list.

Per DESIGN.md §4: run AFTER A and B are frozen.
"""

from __future__ import annotations

import json
import sys
from pathlib import Path

REPO_ROOT = Path(__file__).resolve().parents[2]
STUDY_DIR = REPO_ROOT / "fvs_eval" / "validation_study"
CORPUS_DIR = STUDY_DIR / "corpus"
LABELS_DIR = STUDY_DIR / "labels"

sys.path.insert(0, str(REPO_ROOT))

from framing import (
    detect_coverage,
    detect_voice,
    temporal_orientation,
    detect_epistemic_basis,
)
from frame_library import suggest_frames


FRAME_IDS = [f"FVS-{i:03d}" for i in range(1, 21)]


def analyze_document(text: str) -> list:
    """Run full framing analysis and return list of suggested FVS IDs."""
    cov = detect_coverage(text)
    voice = detect_voice(text)
    temp = temporal_orientation(text)
    epist = detect_epistemic_basis(text)
    suggestions = suggest_frames(cov, voice, temp, epist)
    return suggestions, {
        "coverage": {k: (len(v) if isinstance(v, list) else v)
                     for k, v in cov.items() if k != "categories"},
        "coverage_density": {c: cov["categories"][c]["density_per_1kw"]
                             for c in cov.get("categories", {})},
        "voice": voice.get("voice"),
        "voice_total_sentences": voice.get("total_sentences"),
        "temporal": {
            "past": temp.get("past_pct"),
            "present": temp.get("present_pct"),
            "future": temp.get("future_pct"),
        },
        "epistemic_sourced_pct": epist.get("sourced_pct"),
    }


def main():
    with open(CORPUS_DIR / "manifest.json", "r", encoding="utf-8") as f:
        manifest = json.load(f)

    detector_labels = {
        "version": 1,
        "labeler": "detector",
        "labeler_note": (
            "frame_library.suggest_frames via framing.py signals. "
            "Deterministic regex-based detection. Labels are binary: "
            "frame listed in output = exhibits, else not."
        ),
        "frame_universe": FRAME_IDS,
        "labels": {},
        "signals": {},
    }

    for doc in manifest["documents"]:
        doc_id = doc["doc_id"]
        doc_path = CORPUS_DIR / f"{doc_id}.txt"
        if not doc_path.exists():
            print(f"  {doc_id}: SKIP (file missing)")
            continue
        text = doc_path.read_text(encoding="utf-8")
        try:
            suggestions, signals = analyze_document(text)
        except Exception as e:
            print(f"  {doc_id}: FAILED - {e}")
            continue
        suggested_ids = {s["fvs_id"] for s in suggestions}
        labels = {}
        for fid in FRAME_IDS:
            labels[fid] = {
                "exhibits": fid in suggested_ids,
                "signal": next(
                    (s["signal"] for s in suggestions if s["fvs_id"] == fid),
                    "",
                ) if fid in suggested_ids else "",
            }
        detector_labels["labels"][doc_id] = labels
        detector_labels["signals"][doc_id] = signals
        n_exhibits = sum(1 for v in labels.values() if v["exhibits"])
        print(f"  {doc_id}: {n_exhibits} frames suggested "
              f"({[fid for fid in FRAME_IDS if labels[fid]['exhibits']]})")

    out_path = LABELS_DIR / "labels_detector_v1_v3corpus.json"
    with open(out_path, "w", encoding="utf-8") as f:
        json.dump(detector_labels, f, indent=2)
    print(f"\nWritten: {out_path}")
    print(f"Total documents: {len(detector_labels['labels'])}")


if __name__ == "__main__":
    main()