le-wechat-article v1.3:融合 human-writing 活人感写法(材料关 / 说话位置 / 活人感审读 + check_prose.py 双自检);附 check_article front matter 剥离修复

This commit is contained in:
leroy
2026-10-06 21:30:36 +08:00
parent 500d8224db
commit 3af717a76d
10 changed files with 788 additions and 19 deletions
@@ -14,7 +14,7 @@ TYPE_RANGES = {
"简报": (1500, 3000), "测评": (2500, 5000), "复盘": (2500, 6000),
"技术": (2500, 6000), "叙事": (2000, 5000),
}
FILLER_PATTERNS = ["在当今", "随着{}的发展", "综上所述", "总而言之", "值得注意的是",
FILLER_PATTERNS = ["在当今", "随着({})的发展", "综上所述", "总而言之", "值得注意的是",
"不难发现", "让我们来看看", "下面让我们", "众所周知"]
HAN = r"\u4e00-\u9fff"
@@ -38,7 +38,8 @@ def split_blocks(md):
def check(path, art_type, strict, allow_placeholder=False):
errors, warns, infos = [], [], []
md = open(path, encoding="utf-8").read()
blocks = split_blocks(md)
fm_pre = re.match(r"^---\n.*?\n---\n", md, re.S)
blocks = split_blocks(md[fm_pre.end():] if fm_pre else md)
# ---- 标题层级 ----
h1 = [ln for ln, f in blocks if not f and re.match(r"^# ", ln)]