axis/axis_core/validator.py

from __future__ import annotations

import re
from typing import Tuple


class OutputValidator:
    def validate(self, query: str, qtype: str, output: str) -> Tuple[bool, str]:
        s = (output or "").strip()
        if not s:
            return False, "empty"

        # “秒でそれっぽい”を弾く
        if len(s) < 260:
            return False, "too_short"

        # ガード文禁止
        bad = [
            "もう少し具体化", "精度が上がります", "情報が不足", "受理しました",
            "章タイトル", "Available routes"
        ]
        if any(x in s for x in bad):
            return False, "guard_phrase"

        # 章タイトル列挙っぽい（名詞だけ箇条書き）を弾く
        # 例: "- Finite Bond Dimension Bottleneck" みたいなのが連続
        lines = [ln.strip() for ln in s.splitlines() if ln.strip().startswith("-")]
        if lines:
            title_like = 0
            for ln in lines[:10]:
                # 動詞/因果記号が無い
                if not re.search(r"(=>|because|therefore|implies|bound|<|>|ため|したがって|ゆえに|上限|制約)", ln, re.IGNORECASE):
                    title_like += 1
            if title_like >= 6:
                return False, "headings_only"

        # advanced_physicsならキーワード最低限
        if qtype == "advanced_physics":
            need = ["χ", "N", "Page", "QES", "RT", "complexity"]
            hit = sum(1 for w in need if w.lower() in s.lower())
            if hit < 2:
                return False, "missing_phys_terms"

        return True, "ok"