#!/usr/bin/env python3
"""从 tools/gen_voice_dual.py 台词表提取 id -> 台词全文 映射，输出 JSON。

供试玩转写 harness（web/harness/）把语音播放事件转写成人类可读文本。
台词表是唯一事实源：新台词先改进 gen_voice_dual.py，再跑本脚本刷新 JSON。

用法：.venv-tts/bin/python tools/extract_voice_text.py
输出：web/harness/voice_text.json  {"game": {"id": "全文", ...}}
"""
import ast
import json
import sys
from pathlib import Path

ROOT = Path(__file__).resolve().parent.parent
sys.path.insert(0, str(ROOT / "tools"))

import gen_voice_dual as gvd  # noqa: E402

out = {}
for game, lines in gvd.GAMES.items():
    table = {}
    for fid, _ins, text in lines:
        if fid in table:
            raise SystemExit(f"重复 id：{game}/{fid}")
        table[fid] = text
    # NPC 单音色平铺（gen_voice_dual.NPCS：id, speaker, instruct, 台词）
    for fid, _sp, _ins, text in gvd.NPCS.get(game, []):
        if fid in table:
            raise SystemExit(f"重复 id：{game}/{fid}")
        table[fid] = text
    out[game] = table

# 键盘变体（P1 路线 a）：独立顶层表 "__key__"，不与主表混放——现有消费者
# （mock.mjs / transcribe.mjs / live/control.mjs 的 VOICE_TEXT[game][id] 查找）
# 行为零改动；键盘措辞会话（input=key）按 VOICE_TEXT.__key__[game][id] 取期望文本。
out["__key__"] = {
    game: {fid: text for fid, text in gvd.KEY_LINES[game]}
    for game in gvd.GAMES
}


def parse_legacy(py_path: Path) -> dict:
    """AST 解析旧版 gen_voice.py：提取 ('id', ("Speaker", "instruct"), "text") 元组。
    不能 import——这些脚本顶层直接执行生成。"""
    tree = ast.parse(py_path.read_text(encoding="utf-8"))
    found = {}
    for node in ast.walk(tree):
        if not isinstance(node, ast.Tuple) or len(node.elts) != 3:
            continue
        a, _b, c = node.elts
        if (isinstance(a, ast.Constant) and isinstance(a.value, str)
                and isinstance(c, ast.Constant) and isinstance(c.value, str)
                and len(c.value) >= 2):
            found[a.value] = c.value
    return found


# 旧 flavor 脚本补齐角色台词（storyline 林晚等、escape 角色音）；主表优先，不覆盖
for game, legacy in [("storyline", ROOT / "app/src/storyline/gen/gen_voice.py"),
                     ("escape", ROOT / "app/src/escape/gen/gen_voice.py")]:
    for fid, text in parse_legacy(legacy).items():
        out[game].setdefault(fid, text)

dest = ROOT / "web" / "harness" / "voice_text.json"
dest.parent.mkdir(parents=True, exist_ok=True)
dest.write_text(json.dumps(out, ensure_ascii=False, indent=1), encoding="utf-8")
total = sum(len(v) for k, v in out.items() if k != "__key__")
key_total = sum(len(v) for v in out["__key__"].values())
print(f"OK {dest.relative_to(ROOT)}  games={len(out) - 1}  lines={total}  key_variants={key_total}")