190 lines
11 KiB
Python
190 lines
11 KiB
Python
#!/usr/bin/env python3
|
||
r"""doc_term_audit.py — GDD 退役术语扫描（半自动）
|
||
|
||
动机：机制换名/退役批（如 2026-09-20 伏击→警觉）之后，正文里容易残留旧措辞的
|
||
"现行语义"引用——修订记录与退役注记里的旧词是合法历史，混在一起肉眼难分。
|
||
本工具维护一份退役词表，全文档扫描命中并按上下文分两类：
|
||
[疑似现行] 该行不含日期/退役/换装等白名单信号，也没落在版本历史节——人工判读
|
||
[留档类] 行内带日期、退役/作废/换装/继任/留档/历史等信号，或位于修订记录节——默认忽略
|
||
|
||
用法:
|
||
python tools/doc_term_audit.py # 扫默认范围，报告模式（exit 0）
|
||
python tools/doc_term_audit.py --strict # 有疑似现行命中时 exit 1（供门禁/CI）
|
||
python tools/doc_term_audit.py <file.md> ... # 只扫指定文件
|
||
python tools/doc_term_audit.py --json # 机器可读
|
||
|
||
口径：只扫 design/gdd/ design/registry/——design/research
|
||
（历史提案）、docs/（样例）与 production/ 不在扫描域；词表扩充 = 改 RETIRED_TERMS。
|
||
"""
|
||
import argparse
|
||
import json
|
||
import re
|
||
import sys
|
||
from pathlib import Path
|
||
|
||
# 退役词表：(正则词, 退役日期, 继任词/口径, 备注)。新增退役批时在此追加一行。
|
||
RETIRED_TERMS = [
|
||
("伏击", "2026-09-20", "警觉（blast 负向）", "伏击机制整体退役"),
|
||
("ambush_double|ambush_layers|flat_attach", "2026-09-20", "enemy_alert / effects 投影", "伏击数据键退役"),
|
||
("advantage_per_round_cap", "2026-09-20", "（已删字段——行动点天然封顶）", "单轮授予上限删除批"),
|
||
("溢出未生效", "2026-09-20", "（行类型退役；满层拒收走奖励未生效行）", "行协议 10→9 类"),
|
||
("m_amb|M_AMB", "2026-09-20", "（伏击暴露当量口径作废）", "挖掘风险不等式退役"),
|
||
("七型|五型×正负|五型 × 正负", "2026-09-20", "八型（十六格）", "格面资源八型批"),
|
||
("10 类行|行型封闭为 10 类|10 类封闭协议|行型 10 类", "2026-09-20", "9 类", "溢出未生效行下线"),
|
||
("拔旗|撤旗", "2026-09-19", "（旗无移除手段——插旗不可逆）", "插旗不可逆批"),
|
||
("三选一反噬|标错反噬|反噬三选一", "2026-09-15", "极性矩阵负向兑付", "格面资源批"),
|
||
("defense_up", "2026-09-16", "ac_up（护甲强化）", "护甲等级掷点批换装"),
|
||
("蓄力行|charge 行", "2026-09-16", "施放行（cast）", "敌人技能池批改型"),
|
||
("行为环", "2026-09-16", "技能池（池模板+摇号）", "固定出招剧本退役"),
|
||
("heavy_multiplier|重击倍率", "2026-09-16", "技能行 damage_mult", "行为环退役随批"),
|
||
("多段命中", "2026-09-19", "（退役，无防御磨损体系）", "技能池新效应词汇批"),
|
||
("充能", "2026-09-19", "（挖掘期/战斗段充能退役；蓄力 empower 仍在）", "充能退役批——注意误报"),
|
||
("英雄卡|敌人卡", "2026-09-20", "站桩立绘", "敌人卡形态裁决批"),
|
||
("道具窗口|可选类道具|已消耗未生效|使用被拒|窗口打开锚行|窗口跳过锚行", "2026-09-15", "（随可选类道具整体删除）", "格面资源批"),
|
||
("雾格", "2026-09-14", "（雾格移除）", "不做清单"),
|
||
("兄弟精", "2026-09-26", "L2 熊地精（bugbear 复用）", "变体名未经批准退役批——用户质问查证后全仓清除"),
|
||
# —— 用语规范（2026-10-15 用户令）：权威=design/gdd/术语表.md「用语规范」A 表，机器闸吸收 ——
|
||
# 注意：「随批」「台账」两词本身是上方行级白名单信号（行内出现即归留档类），对其扫描会自豁免，
|
||
# 故不列入；「收口」「在飞」等照常扫描。术语表「用语规范」节已加入节级豁免（词表自身不命中）。
|
||
("对标", "2026-10-15", "参照", "用语规范 A 表（STC 1.7）"),
|
||
("若干", "2026-10-15", "写具体数值或写明数值出处", "用语规范 A 表（STC 5.4）"),
|
||
("不得", "2026-10-15", "禁止", "用语规范 A 表（STC 词表）"),
|
||
("报错", "2026-10-15", "名词用错误；动词用返回错误/显示错误", "用语规范 A 表（STC 词表）"),
|
||
(r"\bbuff\b", "2026-10-15", "增益", "用语规范 A 表（STC 1.1）"),
|
||
("收口", "2026-10-15", "完成/合并进 master", "用语规范 A 表（doc-conventions 第 4 节）"),
|
||
("在飞", "2026-10-15", "正在执行中", "用语规范 A 表（doc-conventions 第 4 节）"),
|
||
("对拍", "2026-10-15", "与基线数据逐项比对", "用语规范 A 表（doc-conventions 第 4 节）"),
|
||
("销账|销号", "2026-10-15", "注销", "用语规范 A 表（doc-conventions 第 4 节）"),
|
||
("落文", "2026-10-15", "写入文档/已写入", "用语规范 A 表（doc-conventions 第 4 节同族）"),
|
||
]
|
||
|
||
DEFAULT_DIRS = ["design/gdd", "design/registry"]
|
||
|
||
# 用语规范 A 表补充词专项补扫：随批/台账 两词本身是上方行级白名单信号（行内出现即归
|
||
# 留档类），主扫描对它们自豁免；本通道只做版本历史节外的专项报告，供后续触碰批清偿。
|
||
SUPPLEMENT_TERMS = [
|
||
("随批", "2026-10-15", "顺带", "用语规范 A 表（doc-conventions 第 4 节）"),
|
||
("台账", "2026-10-15", "状态记录", "用语规范 A 表（doc-conventions 第 4 节）"),
|
||
]
|
||
|
||
|
||
def scan_supplement(paths):
|
||
terms = [(re.compile(t), date, repl, note) for t, date, repl, note in SUPPLEMENT_TERMS]
|
||
findings = []
|
||
for p in paths:
|
||
text = Path(p).read_text(encoding="utf-8").replace("\r\n", "\n")
|
||
in_version_section = False
|
||
for lineno, line in enumerate(text.split("\n"), 1):
|
||
if SECTION_RE.match(line):
|
||
in_version_section = True
|
||
hits = [(m.group(0), t) for t in terms for m in [t[0].search(line)] if m]
|
||
if not hits or in_version_section or COMMENT_RE.match(line):
|
||
continue
|
||
findings.append({
|
||
"term": "、".join(sorted({h[0] for h in hits})),
|
||
"file": str(p).replace("\\", "/"),
|
||
"line": lineno,
|
||
"klass": "疑似现行",
|
||
"text": line.strip()[:160],
|
||
"retired": hits[0][1][1],
|
||
"replacement": hits[0][1][2],
|
||
})
|
||
return findings
|
||
|
||
# 词级语境豁免：某词退役的只是特定机制语境，另一现行机制合法沿用同一词时，
|
||
# 命中行含豁免语境信号即不算退役词命中（不进疑似、不进留档计数）。
|
||
# 充能：挖掘期/战斗段充能退役（2026-09-19），但秘宝主动件「每冒险充能」（同调件
|
||
# 使用次数账，迭代 11 起在役）系现行机制共用该词——秘宝/同调/主动件等语境不算命中。
|
||
TERM_CONTEXT_EXEMPT = {
|
||
"充能": re.compile(r"秘宝|同调|主动件|每冒险|冒险开始|充能账|充能维护|attunement"),
|
||
# buff：反引号行内代码（字段/键名）原样保留（STC 6.2），只查散文用词
|
||
r"\bbuff\b": re.compile(r"`[^`]*buff[^`]*`"),
|
||
}
|
||
|
||
# 行级白名单信号：命中任一即归"留档类"
|
||
WHITELIST_RE = re.compile(
|
||
r"2026-\d{2}-\d{2}"
|
||
r"|退役|作废|换装|继任|留档|历史口径|已下线|随批|随伏击|不再|已删除"
|
||
r"|\bv0\.\d+"
|
||
r"|修订记录|版本历史|履行回执|台账"
|
||
)
|
||
# 节级白名单：版本历史/修订记录/用语规范节内整段豁免（用语规范节的不用列是内容本身）
|
||
SECTION_RE = re.compile(r"^#{1,6}\s*(版本历史|修订记录|用语规范)")
|
||
COMMENT_RE = re.compile(r"^\s*[;#]")
|
||
|
||
|
||
def scan(paths):
|
||
terms = [(re.compile(t), date, repl, note) for t, date, repl, note in RETIRED_TERMS]
|
||
findings = [] # {term, file, line, text, klass, retired, replacement}
|
||
for p in paths:
|
||
text = Path(p).read_text(encoding="utf-8").replace("\r\n", "\n")
|
||
in_version_section = False
|
||
prev_lines = [] # 已处理行（供"上一 2 行留档标记继承"判读）
|
||
for lineno, line in enumerate(text.split("\n"), 1):
|
||
if SECTION_RE.match(line):
|
||
in_version_section = True
|
||
hits = [(m.group(0), t) for t in terms for m in [t[0].search(line)] if m]
|
||
hits = [
|
||
(word, t) for (word, t) in hits
|
||
if not (t[0].pattern in TERM_CONTEXT_EXEMPT and TERM_CONTEXT_EXEMPT[t[0].pattern].search(line))
|
||
]
|
||
# 上一 2 行带留档/作废/历史标记（如"下表留档作废"头注）→ 本行视为留档
|
||
prev = " ".join(prev_lines[-2:]) if prev_lines else ""
|
||
prev_lines.append(line)
|
||
if not hits:
|
||
continue
|
||
if in_version_section or WHITELIST_RE.search(line) or WHITELIST_RE.search(prev) or COMMENT_RE.match(line):
|
||
klass = "留档类"
|
||
else:
|
||
klass = "疑似现行"
|
||
findings.append({
|
||
"term": "、".join(sorted({h[0] for h in hits})),
|
||
"file": str(p).replace("\\", "/"),
|
||
"line": lineno,
|
||
"klass": klass,
|
||
"text": line.strip()[:160],
|
||
"retired": hits[0][1][1],
|
||
"replacement": hits[0][1][2],
|
||
})
|
||
return findings
|
||
|
||
|
||
def main():
|
||
ap = argparse.ArgumentParser(description="GDD 退役术语扫描（半自动）")
|
||
ap.add_argument("files", nargs="*", help="指定文件；缺省扫默认目录")
|
||
ap.add_argument("--strict", action="store_true", help="疑似现行命中时 exit 1")
|
||
ap.add_argument("--json", action="store_true")
|
||
args = ap.parse_args()
|
||
|
||
if args.files:
|
||
paths = args.files
|
||
else:
|
||
paths = []
|
||
for d in DEFAULT_DIRS:
|
||
paths += sorted(Path(d).rglob("*.md"))
|
||
paths += sorted(Path(d).rglob("*.yaml"))
|
||
paths = [p for p in paths if Path(p).exists()]
|
||
|
||
findings = scan(paths) + scan_supplement(paths)
|
||
suspect = [f for f in findings if f["klass"] == "疑似现行"]
|
||
kept = [f for f in findings if f["klass"] != "疑似现行"]
|
||
|
||
if args.json:
|
||
print(json.dumps({"suspect": suspect, "archived_count": len(kept)}, ensure_ascii=False, indent=1))
|
||
else:
|
||
print(f"扫描 {len(paths)} 个文件 · 退役词命中 {len(findings)} 处 · 疑似现行 {len(suspect)} 处（留档类 {len(kept)} 处默认忽略）")
|
||
by_term = {}
|
||
for f in suspect:
|
||
by_term.setdefault(f["term"], []).append(f)
|
||
for term, items in sorted(by_term.items()):
|
||
print(f"\n【{term}】×{len(items)}（退役 {items[0]['retired']} → {items[0]['replacement']}）")
|
||
for it in items[:8]:
|
||
print(f" {it['file']}:{it['line']} {it['text']}")
|
||
if len(items) > 8:
|
||
print(f" …另有 {len(items) - 8} 处")
|
||
return 1 if (args.strict and suspect) else 0
|
||
|
||
|
||
if __name__ == "__main__":
|
||
sys.exit(main())
|