le-wechat-article v1.0：Leroy 专属公众号写作工坊（类型路由 / 分层审读 / 排版转换 / 草稿箱同步）
This commit is contained in:
Leroy
@@ -0,0 +1,224 @@
|
||||
#!/usr/bin/env python3
|
||||
# -*- coding: utf-8 -*-
|
||||
"""公众号文章自检：机器能量化的硬规则与产出配额，不靠眼睛。
|
||||
|
||||
用法：python3 check_article.py <article.md> [--type <类型>] [--strict]
|
||||
退出码：有 ERROR → 1；仅 WARN/INFO → 0（--strict 时 WARN 也算 1）。
|
||||
脚本量不出的部分（加粗挑得对不对、套话是否承担功能）仍需人工判断。
|
||||
"""
|
||||
import argparse
|
||||
import re
|
||||
import sys
|
||||
|
||||
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",
|
||||
"不难发现", "让我们来看看", "下面让我们", "众所周知"]
|
||||
HAN = r"\u4e00-\u9fff"
|
||||
|
||||
|
||||
def strip_inline_code(text):
|
||||
return re.sub(r"`[^`]*`", "", text)
|
||||
|
||||
|
||||
def split_blocks(md):
|
||||
"""按行分类，标记代码围栏内外的行。"""
|
||||
lines = md.split("\n")
|
||||
fenced, blocks = False, []
|
||||
for ln in lines:
|
||||
if ln.lstrip().startswith("```"):
|
||||
fenced = not fenced
|
||||
continue
|
||||
blocks.append((ln, fenced))
|
||||
return blocks
|
||||
|
||||
|
||||
def check(path, art_type, strict, allow_placeholder=False):
|
||||
errors, warns, infos = [], [], []
|
||||
md = open(path, encoding="utf-8").read()
|
||||
blocks = split_blocks(md)
|
||||
|
||||
# ---- 标题层级 ----
|
||||
h1 = [ln for ln, f in blocks if not f and re.match(r"^# ", ln)]
|
||||
bad_h = [ln for ln, f in blocks if not f and re.match(r"^(##|####+) ", ln)]
|
||||
if len(h1) != 1:
|
||||
errors.append(f"标题层级：全文应有且仅有一个 `# ` 主标题，当前 {len(h1)} 个")
|
||||
if bad_h:
|
||||
errors.append(f"标题层级：正文只允许 `###`，发现 {len(bad_h)} 处其他层级（如 {bad_h[0][:30]}…）")
|
||||
|
||||
# ---- 占位残留 ----
|
||||
for ln, f in blocks:
|
||||
if f:
|
||||
continue
|
||||
if re.search(r"placeholder|TODO[:：]?", ln):
|
||||
if allow_placeholder and re.match(r"^\s*!\[.*\]\(placeholder", ln):
|
||||
continue # 草稿期图位占位合法：alt 里是信息位文字规划
|
||||
errors.append(f"占位残留：{ln.strip()[:50]}")
|
||||
|
||||
# ---- 表格（按连续行分组，每张表报一次） ----
|
||||
groups, cur = [], []
|
||||
for ln, f in blocks:
|
||||
if not f and ln.lstrip().startswith("|"):
|
||||
cur.append(ln)
|
||||
elif cur:
|
||||
groups.append(cur)
|
||||
cur = []
|
||||
if cur:
|
||||
groups.append(cur)
|
||||
for g in groups:
|
||||
rows = [r for r in g if not re.match(r"^\s*\|[\s:|-]+\|\s*$", r)]
|
||||
if not rows:
|
||||
continue
|
||||
cols = max(r.count("|") - 1 for r in rows)
|
||||
if cols > 3:
|
||||
errors.append(f"表格超过 3 列（{cols} 列），移动端不可读，请改写为列表")
|
||||
elif max(len(c) for c in rows[0].split("|")[1:-1]) > 12:
|
||||
warns.append(f"表格单元格文字过长：{rows[0].strip()[:40]}…")
|
||||
|
||||
# ---- front matter digest 与导语重复 ----
|
||||
fm = re.match(r"^---\n(.*?)\n---\n", md, re.S)
|
||||
digest = ""
|
||||
if fm:
|
||||
m = re.search(r"^digest:\s*(.+)$", fm.group(1), re.M)
|
||||
if m:
|
||||
digest = m.group(1).strip().strip("\"'「」")
|
||||
body = md[fm.end():] if fm else md
|
||||
first_quote = ""
|
||||
for ln, f in split_blocks(body):
|
||||
if f or not ln.strip() or ln.startswith("# "):
|
||||
continue # 跳过空行与主标题，找正文首个引用
|
||||
if ln.startswith(">"):
|
||||
first_quote = ln.lstrip("> ").strip()
|
||||
break
|
||||
if digest and first_quote and digest[:20] == first_quote[:20]:
|
||||
warns.append(f"摘要 ≠ 导语：digest 与正文首个引用开头 20 字相同，读者会同一句读两遍")
|
||||
|
||||
# ---- 加粗配额 ----
|
||||
plain_texts, bolds = [], [] # plain_texts: (去加粗去代码文本, 本行是否有加粗)
|
||||
section_has_bold, cur_section_bold, section_names, cur_section = [], False, [], None
|
||||
for ln, f in blocks:
|
||||
if f or ln.lstrip().startswith("|"):
|
||||
continue
|
||||
if re.match(r"^### ", ln):
|
||||
if cur_section is not None:
|
||||
section_names.append(cur_section)
|
||||
section_has_bold.append(cur_section_bold)
|
||||
cur_section, cur_section_bold = ln, False
|
||||
plain_texts.append(("\n", True)) # 小节标题本身是落点，重置累计
|
||||
continue
|
||||
line_bolds = re.findall(r"\*\*([^*]+)\*\*", ln)
|
||||
bolds.extend(line_bolds)
|
||||
if line_bolds:
|
||||
cur_section_bold = True
|
||||
plain = strip_inline_code(re.sub(r"\*\*[^*]+\*\*", "", ln))
|
||||
plain_texts.append((plain + "\n", bool(line_bolds)))
|
||||
if cur_section is not None:
|
||||
section_names.append(cur_section)
|
||||
section_has_bold.append(cur_section_bold)
|
||||
|
||||
for name, has in zip(section_names, section_has_bold):
|
||||
if not has:
|
||||
warns.append(f"加粗配额：小节「{name.lstrip('# ')}」全节无加粗落点")
|
||||
|
||||
# 连续无加粗字符数（跨段累计；含加粗/小节标题/金句卡的行重置计数）
|
||||
gap_runs, run = [], 0
|
||||
for text, has_bold in plain_texts:
|
||||
if has_bold or re.match(r"^>\s*.{4,}\s*——", text.strip()):
|
||||
run = 0
|
||||
continue
|
||||
run += len(re.sub(r"\s+", "", text))
|
||||
if run > 200:
|
||||
gap_runs.append(run)
|
||||
run = 0
|
||||
if gap_runs:
|
||||
warns.append(f"加粗配额：存在 {len(gap_runs)} 处连续 >200 字无加粗（最长 {max(gap_runs)} 字），读者划过整屏找不到重点")
|
||||
|
||||
bold_chain = " / ".join(bolds)
|
||||
if not bolds:
|
||||
errors.append("加粗配额：全文无加粗——读者只看标题和加粗应能理解全篇，当前做不到")
|
||||
|
||||
# ---- 金句卡 ----
|
||||
golden = [ln for ln, f in blocks if not f and re.match(r"^> .{4,} ——", ln)]
|
||||
if len(golden) > 1:
|
||||
warns.append(f"金句卡：`> 金句 —— 出处` 出现 {len(golden)} 处，至多 1 处，多了退化成装饰条")
|
||||
|
||||
# ---- 中英文间距 / 引号 / 省略号（排除代码、链接、图片） ----
|
||||
scan_lines = []
|
||||
for ln, f in blocks:
|
||||
if f or ln.lstrip().startswith("|"):
|
||||
continue
|
||||
clean = re.sub(r"!\[[^\]]*\]\([^)]*\)", "", ln) # 图片
|
||||
clean = re.sub(r"\[([^\]]*)\]\([^)]*\)", r"\1", clean) # 链接留文字
|
||||
clean = strip_inline_code(clean)
|
||||
scan_lines.append(clean)
|
||||
scan = "\n".join(scan_lines)
|
||||
spacing_hits = re.findall(f"[{HAN}][A-Za-z0-9]|[A-Za-z0-9][{HAN}]", scan)
|
||||
if spacing_hits:
|
||||
warns.append(f"中英文间距：{len(spacing_hits)} 处汉字与英数字之间缺一个半角空格，如：「{spacing_hits[0]}」")
|
||||
curly = re.findall(r"[\u201c\u201d\u2018\u2019]", scan)
|
||||
if curly:
|
||||
warns.append(f"引号：发现 {len(curly)} 个中文弯引号，请统一改用直角引号「」『』")
|
||||
ascii_q = re.findall(f'"[^"]*[{HAN}][^"]*"', scan)
|
||||
if ascii_q:
|
||||
warns.append(f"引号：{len(ascii_q)} 处用 ASCII 直引号包裹中文（如「{ascii_q[0]}」），请改用直角引号「」")
|
||||
if re.search(r"\.\.\.|。{3,}", scan):
|
||||
infos.append("省略号：请使用……（两个 U+2026），不要用三个句点")
|
||||
|
||||
# ---- 高频套话 ----
|
||||
for pat in FILLER_PATTERNS:
|
||||
p = pat.format("AI|科技|技术") if "{}" in pat else pat
|
||||
hits = re.findall(p, scan)
|
||||
if hits:
|
||||
warns.append(f"套话检查：「{pat}」出现 {len(hits)} 次——保留与否取决于是否承担真实功能，请人工判断")
|
||||
|
||||
# ---- 字数 ----
|
||||
text = re.sub(r"[#>*`|\-\[\]()!\s]", "", scan)
|
||||
n = len(text)
|
||||
if art_type in TYPE_RANGES:
|
||||
lo, hi = TYPE_RANGES[art_type]
|
||||
if n < lo:
|
||||
infos.append(f"字数：约 {n} 字，低于「{art_type}」参考下限 {lo}——素材不足时允许，勿注水")
|
||||
elif n > hi * 1.3:
|
||||
warns.append(f"字数：约 {n} 字，明显超出「{art_type}」参考上限 {hi}，检查是否有注水")
|
||||
else:
|
||||
infos.append(f"字数：约 {n} 字（{art_type} 参考 {lo}–{hi}）")
|
||||
else:
|
||||
infos.append(f"字数：约 {n} 字（未指定类型，无参考区间）")
|
||||
|
||||
# ---- 输出 ----
|
||||
print("=" * 56)
|
||||
print(f"公众号自检 · {path}")
|
||||
print("=" * 56)
|
||||
for tag, items in (("❌ ERROR", errors), ("⚠️ WARN", warns), ("ℹ️ INFO", infos)):
|
||||
for it in items:
|
||||
print(f"{tag} {it}")
|
||||
if bold_chain:
|
||||
print("-" * 56)
|
||||
print("加粗串读提要（连读应是一篇能独立看懂的缩写版）：")
|
||||
print(" " + (bold_chain[:600] + ("…" if len(bold_chain) > 600 else "")))
|
||||
print("-" * 56)
|
||||
total = len(errors) + len(warns)
|
||||
print(f"结果：{len(errors)} 错误 / {len(warns)} 提醒 / {len(infos)} 信息")
|
||||
if errors:
|
||||
return 1
|
||||
if strict and warns:
|
||||
print("--strict 模式：WARN 视为不通过")
|
||||
return 1
|
||||
return 0
|
||||
|
||||
|
||||
def main():
|
||||
ap = argparse.ArgumentParser(description="公众号文章硬规则与产出配额自检")
|
||||
ap.add_argument("article")
|
||||
ap.add_argument("--type", choices=list(TYPE_RANGES), help="文章类型（提供字数参考区间）")
|
||||
ap.add_argument("--strict", action="store_true", help="WARN 也视为不通过")
|
||||
ap.add_argument("--allow-placeholder", action="store_true",
|
||||
help="草稿模式：图位  不算错误（交付版必须去掉此开关）")
|
||||
args = ap.parse_args()
|
||||
sys.exit(check(args.article, args.type, args.strict, args.allow_placeholder))
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
main()
|
||||
@@ -0,0 +1,311 @@
|
||||
#!/usr/bin/env python3
|
||||
# -*- coding: utf-8 -*-
|
||||
"""Markdown → 微信公众号可粘贴 HTML（全内联样式，零依赖）。
|
||||
|
||||
排版 = 骨架（结构装饰）× 色板（主题色）两层：
|
||||
骨架 工程 色条标题 + 主题色加粗 + 灰底金句卡——教程、技术、工程实践（默认）
|
||||
杂志 深色大标题 + 深灰粗体 + 装饰线金句卡——叙事、访谈、生活方式
|
||||
简报 紧凑节奏 + 主题色小标题——更新简报、快讯、清单
|
||||
色板 靛蓝 / 松绿 / 黛紫 / 绛红（--color 可自定义 #RRGGBB）
|
||||
|
||||
用法：
|
||||
python3 md2wechat.py article.md -o article.html # 工程骨架 + 靛蓝
|
||||
python3 md2wechat.py article.md --skeleton 杂志 --palette 松绿 # 指定骨架与色板
|
||||
python3 md2wechat.py article.md --list-themes # 看全部组合与适用场景
|
||||
python3 md2wechat.py article.md --preview # 三骨架对照页（375px 栏宽）
|
||||
输出用浏览器打开，全选复制 → 粘贴进公众号编辑器。
|
||||
支持的语法子集：#（跳过）/ ### / 段落 / **加粗** / `行内代码` / [链接] /  /
|
||||
> 引用（含「—— 出处」识别为金句卡）/ - 与 1. 列表 / --- 分隔 / ```代码块``` / ≤3 列表格。
|
||||
"""
|
||||
import argparse
|
||||
import html
|
||||
import re
|
||||
import sys
|
||||
|
||||
PALETTES = {"靛蓝": "#1A5CFF", "松绿": "#0F7B6C", "黛紫": "#6B4FBB", "绛红": "#B03A2E"}
|
||||
SKELETONS = {
|
||||
"工程": {"alias": "eng", "fit": "教程、技术解析、工程实践"},
|
||||
"杂志": {"alias": "mag", "fit": "叙事、访谈、生活方式、品牌"},
|
||||
"简报": {"alias": "brief", "fit": "更新简报、快讯、清单文"},
|
||||
}
|
||||
|
||||
|
||||
def esc(t):
|
||||
return html.escape(t, quote=False)
|
||||
|
||||
|
||||
def darken(hexc, f=0.72):
|
||||
hexc = hexc.lstrip("#")
|
||||
r, g, b = (int(hexc[i:i + 2], 16) for i in (0, 2, 4))
|
||||
m = lambda v: max(0, int(v * f))
|
||||
return f"#{m(r):02x}{m(g):02x}{m(b):02x}"
|
||||
|
||||
|
||||
def styles(sk, color, px):
|
||||
"""骨架差异全部参数化在这里：返回各元素的内联样式串。"""
|
||||
n = int(px[:-2])
|
||||
dark = darken(color)
|
||||
if sk == "杂志":
|
||||
r, g, b = color.lstrip("#")[0:2], color.lstrip("#")[2:4], color.lstrip("#")[4:6]
|
||||
hl = f"rgba({int(r,16)},{int(g,16)},{int(b,16)},0.12)"
|
||||
return {
|
||||
"p": f"font-size:{px};line-height:1.8;color:#333;letter-spacing:.5px;margin:0 0 22px;",
|
||||
"strong": f"color:#111;font-weight:700;background:{hl};",
|
||||
"link": f"color:{dark};text-decoration:none;border-bottom:1px solid {hl};",
|
||||
"h3": (f"margin:34px 0 16px;font-size:{n + 1}px;font-weight:700;color:#111;"
|
||||
f"border-bottom:2px solid #e8e8e8;padding-bottom:10px;line-height:1.5;"),
|
||||
"quote": (f"margin:26px 0;padding:4px 16px;border-left:2px solid #ccc;"
|
||||
f"font-size:{n - 1}px;line-height:1.8;color:#888;"),
|
||||
"golden": (f"margin:30px 0;padding:20px 10px;text-align:center;"
|
||||
f"border-top:1px solid {color};border-bottom:1px solid {color};"
|
||||
f"font-size:{n + 1}px;line-height:1.9;color:{dark};font-weight:500;"),
|
||||
"li": "margin:0 0 12px;line-height:1.8;", "li_ul": f"padding-left:1.4em;margin:0 0 22px;font-size:{px};line-height:1.8;color:#333;",
|
||||
}
|
||||
if sk == "简报":
|
||||
return {
|
||||
"p": f"font-size:{px};line-height:1.7;color:#3a3a3a;margin:0 0 16px;",
|
||||
"strong": f"color:{color};font-weight:600;",
|
||||
"link": f"color:{color};text-decoration:none;",
|
||||
"h3": (f"margin:26px 0 12px;font-size:{n + 1}px;font-weight:600;color:{dark};"
|
||||
f"line-height:1.5;"),
|
||||
"quote": (f"margin:20px 0;padding:2px 14px;border-left:2px solid {color};"
|
||||
f"font-size:{n - 1}px;line-height:1.7;color:#777;"),
|
||||
"golden": (f"margin:24px 0;padding:12px 8px;text-align:center;"
|
||||
f"font-size:{n + 1}px;line-height:1.8;color:{dark};font-weight:500;"
|
||||
f"letter-spacing:1px;"),
|
||||
"li": "margin:0 0 8px;line-height:1.7;", "li_ul": f"padding-left:1.4em;margin:0 0 16px;font-size:{px};line-height:1.7;color:#3a3a3a;",
|
||||
}
|
||||
# 工程（默认）
|
||||
return {
|
||||
"p": f"font-size:{px};line-height:1.75;color:#3f3f3f;letter-spacing:.4px;margin:0 0 20px;",
|
||||
"strong": f"color:{color};font-weight:600;",
|
||||
"link": f"color:{color};text-decoration:none;",
|
||||
"h3": (f"margin:30px 0 16px;font-size:{n + 1}px;font-weight:600;color:{dark};"
|
||||
f"border-left:4px solid {color};padding-left:10px;line-height:1.5;"),
|
||||
"quote": (f"margin:24px 0;padding:12px 16px;border-left:3px solid #ddd;background:#f7f8fa;"
|
||||
f"font-size:{n - 1}px;line-height:1.8;color:#888;"),
|
||||
"golden": (f"margin:28px 0;padding:20px 24px;background:#f7f8fa;text-align:center;"
|
||||
f"font-size:{n + 1}px;line-height:1.9;color:{dark};font-weight:500;"),
|
||||
"li": "margin:0 0 10px;line-height:1.75;", "li_ul": f"padding-left:1.4em;margin:0 0 20px;font-size:{px};line-height:1.75;color:#3f3f3f;",
|
||||
}
|
||||
|
||||
|
||||
def inline(text, st):
|
||||
"""行内元素转换：先保护行内代码，再处理加粗/链接/图片。"""
|
||||
codes = []
|
||||
|
||||
def stash(m):
|
||||
codes.append(m.group(1))
|
||||
return f"\x00{len(codes) - 1}\x00"
|
||||
|
||||
t = re.sub(r"`([^`]+)`", stash, text)
|
||||
t = re.sub(r"!\[([^\]]*)\]\(([^)\s]+)(?:\s+\"([^\"]*)\")?\)",
|
||||
lambda m: (f'<img src="{esc(m.group(2))}" style="max-width:100%;display:block;margin:16px auto;'
|
||||
f'border-radius:6px;"'
|
||||
+ (f'<figcaption style="text-align:center;font-size:13px;color:#999;'
|
||||
f'margin:8px 0 20px;">{esc(m.group(3))}</figcaption>' if m.group(3) else "")),
|
||||
t)
|
||||
t = re.sub(r"\[([^\]]+)\]\(([^)\s]+)\)",
|
||||
rf'<a href="\2" style="{st["link"]}">\1</a>', t)
|
||||
t = re.sub(r"\*\*([^*]+)\*\*", rf'<strong style="{st["strong"]}">\1</strong>', t)
|
||||
t = re.sub(r"\x00(\d+)\x00",
|
||||
lambda m: f'<code style="background:#f2f3f5;color:#c7254e;border-radius:3px;'
|
||||
f'padding:2px 5px;font-size:14px;font-family:Menlo,Consolas,monospace;">'
|
||||
f'{esc(codes[int(m.group(1))])}</code>', t)
|
||||
return t
|
||||
|
||||
|
||||
def convert(md, sk, color, px):
|
||||
st = styles(sk, color, px)
|
||||
dark = darken(color)
|
||||
lines = md.split("\n")
|
||||
out, i, n = [], 0, len(lines)
|
||||
in_front = lines and lines[0].strip() == "---"
|
||||
front_done = not in_front
|
||||
|
||||
def flush_list(items, ordered):
|
||||
tag = "ol" if ordered else "ul"
|
||||
bullet = re.compile(r"^\s*(?:\d+\.|-)\s+")
|
||||
lis = "".join(
|
||||
f'<li style="{st["li"]}">' + inline(bullet.sub("", it), st) + "</li>"
|
||||
for it in items)
|
||||
out.append(f'<{tag} style="{st["li_ul"]}">{lis}</{tag}>')
|
||||
|
||||
while i < n:
|
||||
ln = lines[i]
|
||||
if not front_done:
|
||||
if i > 0 and ln.strip() == "---":
|
||||
front_done = True
|
||||
i += 1
|
||||
continue
|
||||
s = ln.strip()
|
||||
if not s:
|
||||
i += 1
|
||||
continue
|
||||
if s.startswith("```"):
|
||||
i += 1
|
||||
buf = []
|
||||
while i < n and not lines[i].strip().startswith("```"):
|
||||
buf.append(lines[i])
|
||||
i += 1
|
||||
i += 1
|
||||
code = esc("\n".join(buf))
|
||||
out.append(f'<pre style="background:#f6f8fa;border-radius:8px;padding:14px 16px;'
|
||||
f'overflow-x:auto;font-size:13.5px;line-height:1.6;margin:0 0 20px;'
|
||||
f'font-family:Menlo,Consolas,monospace;color:#24292e;">{code}</pre>')
|
||||
continue
|
||||
if s.startswith("# ") and not s.startswith("##"):
|
||||
i += 1 # 主标题不进正文（填在公众号后台）
|
||||
continue
|
||||
if s.startswith("###") and not s.startswith("####"):
|
||||
out.append(f'<h3 style="{st["h3"]}">{inline(s.lstrip("#").strip(), st)}</h3>')
|
||||
i += 1
|
||||
continue
|
||||
if re.match(r"^####+\s", s):
|
||||
out.append(f'<p style="margin:24px 0 12px;font-weight:600;color:{dark};'
|
||||
f'font-size:{px};">{inline(s.lstrip("#").strip(), st)}</p>')
|
||||
i += 1
|
||||
continue
|
||||
if re.match(r"^-{3,}$", s):
|
||||
out.append('<hr style="border:none;border-top:1px solid #eaeaea;margin:28px 0;"/>')
|
||||
i += 1
|
||||
continue
|
||||
if s.startswith(">"):
|
||||
buf = []
|
||||
while i < n and lines[i].strip().startswith(">"):
|
||||
buf.append(lines[i].strip().lstrip(">").strip())
|
||||
i += 1
|
||||
text = "<br/>".join(inline(b, st) for b in buf if b)
|
||||
key = "golden" if "——" in text else "quote"
|
||||
out.append(f'<blockquote style="{st[key]}">{text}</blockquote>')
|
||||
continue
|
||||
if s.startswith("|"):
|
||||
tbl = []
|
||||
while i < n and lines[i].strip().startswith("|"):
|
||||
tbl.append(lines[i].strip())
|
||||
i += 1
|
||||
rows = [r for r in tbl if not re.match(r"^\|[\s:|-]+\|$", r)]
|
||||
if rows:
|
||||
cols = max(r.count("|") - 1 for r in rows)
|
||||
if cols > 3:
|
||||
print(f"⚠️ 表格超过 3 列（{cols} 列），已跳过渲染——请先改写为列表", file=sys.stderr)
|
||||
else:
|
||||
trs = []
|
||||
for ri, r in enumerate(rows):
|
||||
cells = [c.strip() for c in r.strip("|").split("|")]
|
||||
head = (f"background:#f7f8fa;font-weight:600;color:{dark};" if ri == 0 else "")
|
||||
tds = "".join(
|
||||
f'<td style="border:1px solid #eaeaea;padding:8px 10px;'
|
||||
f'font-size:{int(px[:-2]) - 1}px;{head}">{inline(c, st)}</td>'
|
||||
for c in cells)
|
||||
trs.append(f"<tr>{tds}</tr>")
|
||||
out.append('<table style="border-collapse:collapse;margin:0 auto 20px;">'
|
||||
+ "".join(trs) + "</table>")
|
||||
continue
|
||||
if re.match(r"^\s*-\s+", ln):
|
||||
items = []
|
||||
while i < n and re.match(r"^\s*-\s+", lines[i]):
|
||||
items.append(lines[i])
|
||||
i += 1
|
||||
flush_list(items, ordered=False)
|
||||
continue
|
||||
if re.match(r"^\s*\d+\.\s+", ln):
|
||||
items = []
|
||||
while i < n and re.match(r"^\s*\d+\.\s+", lines[i]):
|
||||
items.append(lines[i])
|
||||
i += 1
|
||||
flush_list(items, ordered=True)
|
||||
continue
|
||||
out.append(f'<p style="{st["p"]}">{inline(s, st)}</p>')
|
||||
i += 1
|
||||
|
||||
return (f'<section style="max-width:578px;margin:0 auto;font-family:'
|
||||
f'-apple-system,BlinkMacSystemFont,\'PingFang SC\',\'Hiragino Sans GB\','
|
||||
f'\'Microsoft YaHei\',sans-serif;">' + "".join(out) + "</section>")
|
||||
|
||||
|
||||
SAMPLE = """### 小节标题示例
|
||||
|
||||
正文段落：这里是**主题加粗落点**，后面是[链接示例](https://example.com)与行内代码 `npm install`。
|
||||
|
||||
- **标签一**：说明文字
|
||||
- **标签二**：说明文字
|
||||
|
||||
> 值得截图转发的一句金句。 —— 出处
|
||||
|
||||
> 普通引用：补充说明的引文。
|
||||
|
||||
|
||||
|
||||
---
|
||||
|
||||
结尾段落，含**收束落点**。
|
||||
"""
|
||||
|
||||
|
||||
def list_themes():
|
||||
print("骨架 × 色板（--skeleton 骨架 --palette 色板，或 --color '#RRGGBB' 自定义）\n")
|
||||
for sk, meta in SKELETONS.items():
|
||||
print(f" {sk}（--skeleton {sk} 或 {meta['alias']}）—— {meta['fit']}")
|
||||
print(f" 色板：{' / '.join(f'{k} {v}' for k, v in PALETTES.items())}\n")
|
||||
|
||||
|
||||
def resolve_skeleton(name):
|
||||
if not name:
|
||||
return "工程"
|
||||
for sk, meta in SKELETONS.items():
|
||||
if name in (sk, meta["alias"]):
|
||||
return sk
|
||||
sys.exit(f"未知骨架 {name}，可选：{' / '.join(SKELETONS)}（或其英文别名）")
|
||||
|
||||
|
||||
def main():
|
||||
ap = argparse.ArgumentParser(description="Markdown → 微信公众号 HTML（骨架×色板，全内联样式）")
|
||||
ap.add_argument("article", nargs="?")
|
||||
ap.add_argument("-o", "--output")
|
||||
ap.add_argument("--skeleton", help="骨架：工程 / 杂志 / 简报（默认 工程）")
|
||||
ap.add_argument("--palette", choices=list(PALETTES))
|
||||
ap.add_argument("--theme", choices=list(PALETTES), help="兼容旧参数，等同 --palette")
|
||||
ap.add_argument("--color", help="自定义主题色 #RRGGBB，覆盖色板")
|
||||
ap.add_argument("--font-size", default="16px")
|
||||
ap.add_argument("--list-themes", action="store_true", help="列出全部骨架与色板")
|
||||
ap.add_argument("--preview", action="store_true", help="生成三骨架对照页（用内置样文）")
|
||||
args = ap.parse_args()
|
||||
|
||||
if args.list_themes:
|
||||
list_themes()
|
||||
return
|
||||
sk = resolve_skeleton(args.skeleton)
|
||||
color = args.color or PALETTES.get(args.palette or args.theme or "靛蓝")
|
||||
if not re.match(r"^#[0-9A-Fa-f]{6}$", color):
|
||||
sys.exit(f"颜色格式应为 #RRGGBB，收到 {color}")
|
||||
|
||||
if args.preview:
|
||||
blocks = []
|
||||
for sk_name in SKELETONS:
|
||||
doc = convert(SAMPLE, sk_name, color, args.font_size)
|
||||
blocks.append(f'<h2 style="text-align:center;font-size:15px;color:#999;'
|
||||
f'margin:40px 0 12px;">骨架：{sk_name} · 色板 {color}</h2>'
|
||||
f'<div style="max-width:375px;margin:0 auto;border:1px solid #ddd;'
|
||||
f'padding:16px 14px;">{doc}</div>')
|
||||
page = ("<!DOCTYPE html><html><head><meta charset='utf-8'><title>主题预览</title></head>"
|
||||
f"<body>{''.join(blocks)}</body></html>")
|
||||
out = args.output or "theme-preview.html"
|
||||
open(out, "w", encoding="utf-8").write(page)
|
||||
print(f"✔ {out}（骨架对照页，375px 栏宽 ≈ 手机宽度）——浏览器打开选一个顺眼的")
|
||||
return
|
||||
|
||||
if not args.article:
|
||||
sys.exit("请提供 markdown 文件路径（或用 --list-themes / --preview）")
|
||||
md = open(args.article, encoding="utf-8").read()
|
||||
doc = convert(md, sk, color, args.font_size)
|
||||
full = ("<!DOCTYPE html><html><head><meta charset='utf-8'>"
|
||||
"<title>公众号文章</title></head>"
|
||||
f"<body style='background:#fff;'>{doc}</body></html>")
|
||||
out = args.output or re.sub(r"\.md$", ".html", args.article)
|
||||
open(out, "w", encoding="utf-8").write(full)
|
||||
print(f"✔ {out}（骨架 {sk} · 主题色 {color}）——浏览器打开后全选复制，粘贴进公众号编辑器")
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
main()
|
||||
@@ -0,0 +1,255 @@
|
||||
#!/usr/bin/env python3
|
||||
# -*- coding: utf-8 -*-
|
||||
"""把成稿同步到微信公众号草稿箱（只写草稿，绝不群发）。
|
||||
|
||||
前置条件：
|
||||
1. 已认证的公众号（未认证账号调用草稿接口会返回 48001）。
|
||||
2. 环境变量 WECHAT_APP_ID、WECHAT_APP_SECRET（在 mp.weixin.qq.com 设置与开发→基本配置获取）。
|
||||
3. 调用方公网 IP 已加入公众号后台的 IP 白名单（家用宽带 IP 变化后需回后台更新）。
|
||||
|
||||
用法：
|
||||
python3 publish_draft.py <文章目录> [--title 标题] [--digest 摘要] [--cover 封面图] \
|
||||
[--check] [--dry-run] [--author 作者]
|
||||
--check 只验证凭证与 IP 白名单（调 token 接口），不写任何内容
|
||||
--dry-run 不联网，打印将上传的图片与将提交的草稿参数
|
||||
退出码：0 成功；1 失败（错误码已翻译成人话）；2 凭证缺失或参数不合法。
|
||||
"""
|
||||
import argparse
|
||||
import json
|
||||
import os
|
||||
import re
|
||||
import struct
|
||||
import sys
|
||||
import urllib.parse
|
||||
import urllib.request
|
||||
|
||||
API = "https://api.weixin.qq.com"
|
||||
ERR = {
|
||||
40001: "AppSecret 错误或 access_token 已过期——核对 WECHAT_APP_SECRET，或稍后重试",
|
||||
40013: "AppID 无效——核对 WECHAT_APP_ID",
|
||||
40164: "本机公网 IP 不在白名单——把提示里的 IP 加到公众号后台「基本配置→IP白名单」后重试",
|
||||
40125: "AppSecret 无效或已重置——到后台重置 Secret 后更新环境变量",
|
||||
45009: "接口当日调用额度用完——明天再试",
|
||||
48001: "该账号无此 API 权限——草稿箱接口需要已认证公众号；未认证个人订阅号用不了",
|
||||
40113: "图片格式被素材接口拒绝（部分 PNG 编码会被挑剔）——转成 jpg 再试："
|
||||
"sips -s format jpeg cover.png --out cover.jpg（macOS 自带），"
|
||||
"然后用 --cover cover.jpg 重跑",
|
||||
53401: "封面图校验失败——换一张（建议 900×383 以上的 2.35:1 图）",
|
||||
53402: "封面 media_id 无效或不是图片素材",
|
||||
}
|
||||
|
||||
|
||||
def die(msg, code=1):
|
||||
print(f"❌ {msg}", file=sys.stderr)
|
||||
sys.exit(code)
|
||||
|
||||
|
||||
def wx_err(r):
|
||||
"""把微信返回体翻译成人话；成功返回 None。"""
|
||||
code = r.get("errcode", 0)
|
||||
if not code:
|
||||
return None
|
||||
hint = ERR.get(code, r.get("errmsg", "未知错误"))
|
||||
return f"errcode {code}：{hint}（{r.get('errmsg', '')}）"
|
||||
|
||||
|
||||
def http_json(url, data=None, timeout=30):
|
||||
"""GET（data=None）或 JSON POST，返回解析后的 dict。"""
|
||||
req = urllib.request.Request(url, method="POST" if data is not None else "GET")
|
||||
body = None
|
||||
if data is not None:
|
||||
body = json.dumps(data, ensure_ascii=False).encode("utf-8")
|
||||
req.add_header("Content-Type", "application/json; charset=utf-8")
|
||||
with urllib.request.urlopen(req, body, timeout=timeout) as resp:
|
||||
return json.loads(resp.read().decode("utf-8"))
|
||||
|
||||
|
||||
def multipart(field, filename, content):
|
||||
b = "----LeWeChatBoundary7MA4YWxk"
|
||||
parts = [
|
||||
f"--{b}\r\nContent-Disposition: form-data; name=\"{field}\"; "
|
||||
f"filename=\"{filename}\"\r\nContent-Type: application/octet-stream\r\n\r\n"
|
||||
]
|
||||
body = parts[0].encode("utf-8") + content + f"\r\n--{b}--\r\n".encode("utf-8")
|
||||
return body, b
|
||||
|
||||
|
||||
def post_file(url, path):
|
||||
with open(path, "rb") as f:
|
||||
content = f.read()
|
||||
body, boundary = multipart("media", os.path.basename(path), content)
|
||||
req = urllib.request.Request(url, method="POST", data=body)
|
||||
req.add_header("Content-Type", f"multipart/form-data; boundary={boundary}")
|
||||
with urllib.request.urlopen(req, timeout=60) as resp:
|
||||
return json.loads(resp.read().decode("utf-8"))
|
||||
|
||||
|
||||
def img_size(path):
|
||||
"""读图片像素尺寸：PNG（IHDR）/ JPEG（SOF 帧），标准库解析。"""
|
||||
with open(path, "rb") as f:
|
||||
head = f.read(32)
|
||||
f.seek(0)
|
||||
data = f.read()
|
||||
if head.startswith(b"\x89PNG"):
|
||||
w, h = struct.unpack(">II", data[16:24])
|
||||
return w, h
|
||||
if head.startswith(b"\xff\xd8"): # JPEG：找第一帧 SOF0/2
|
||||
i = 2
|
||||
while i < len(data) - 9:
|
||||
if data[i] != 0xFF:
|
||||
i += 1
|
||||
continue
|
||||
marker = data[i + 1]
|
||||
if marker in (0xC0, 0xC2):
|
||||
h, w = struct.unpack(">HH", data[i + 5:i + 9])
|
||||
return w, h
|
||||
if marker in (0xD8, 0x01) or 0xD0 <= marker <= 0xD7:
|
||||
i += 2
|
||||
continue
|
||||
seg_len = struct.unpack(">H", data[i + 2:i + 4])[0]
|
||||
i += 2 + seg_len
|
||||
raise ValueError(f"无法识别图片尺寸（支持 PNG/JPEG）：{path}")
|
||||
|
||||
|
||||
def crop_fractions(w, h, tw, th):
|
||||
"""最大居中裁剪框 → 微信 pic_crop 格式 x1_y1_x2_y2（0–1 百分比）。"""
|
||||
if w / h >= tw / th:
|
||||
ch, cw = h, h * tw / th
|
||||
else:
|
||||
cw, ch = w, w * th / tw
|
||||
x1, y1 = (w - cw) / 2 / w, (h - ch) / 2 / h
|
||||
x2, y2 = x1 + cw / w, y1 + ch / h
|
||||
return f"{x1:.4f}_{y1:.4f}_{x2:.4f}_{y2:.4f}"
|
||||
|
||||
|
||||
def get_token(app_id, secret):
|
||||
q = urllib.parse.urlencode({"grant_type": "client_credential",
|
||||
"appid": app_id, "secret": secret})
|
||||
try:
|
||||
r = http_json(f"{API}/cgi-bin/token?{q}")
|
||||
except urllib.error.URLError as e:
|
||||
die(f"网络请求失败：{e}")
|
||||
err = wx_err(r)
|
||||
if err:
|
||||
die(f"获取 access_token 失败——{err}")
|
||||
return r["access_token"]
|
||||
|
||||
|
||||
def main():
|
||||
ap = argparse.ArgumentParser(description="同步成稿到公众号草稿箱（不群发）")
|
||||
ap.add_argument("dir", help="文章目录（含 article.html）")
|
||||
ap.add_argument("--title", required=not any(a in sys.argv for a in ("--check", "--dry-run")))
|
||||
ap.add_argument("--digest", default="")
|
||||
ap.add_argument("--author", default="")
|
||||
ap.add_argument("--cover", help="封面图路径（默认在文章目录找 cover.png/jpg/jpeg/webp）")
|
||||
ap.add_argument("--check", action="store_true", help="只验证凭证与 IP 白名单")
|
||||
ap.add_argument("--no-crop", action="store_true", help="不提交封面裁剪框参数")
|
||||
ap.add_argument("--dry-run", action="store_true", help="不联网，打印执行计划")
|
||||
args = ap.parse_args()
|
||||
|
||||
app_id = os.environ.get("WECHAT_APP_ID", "")
|
||||
secret = os.environ.get("WECHAT_APP_SECRET", "")
|
||||
|
||||
# --check 只验凭证与 IP，不需要文章产物
|
||||
if args.check:
|
||||
if not (app_id and secret):
|
||||
print("❌ 缺少环境变量 WECHAT_APP_ID / WECHAT_APP_SECRET\n"
|
||||
" 获取路径：mp.weixin.qq.com → 设置与开发 → 基本配置\n"
|
||||
" 设置方式：export WECHAT_APP_ID=xxx WECHAT_APP_SECRET=yyy（建议写进 ~/.zshrc）\n"
|
||||
" 别把密钥写进任何 git 仓库。", file=sys.stderr)
|
||||
sys.exit(2)
|
||||
get_token(app_id, secret) # 失败时内部已 die 并翻译错误码
|
||||
print("✔ 凭证与 IP 白名单验证通过，可以同步草稿箱")
|
||||
return
|
||||
|
||||
# ---- 收集产物（dry-run 也走这段，但不联网） ----
|
||||
html_path = os.path.join(args.dir, "article.html")
|
||||
if not os.path.isfile(html_path):
|
||||
die(f"找不到 {html_path}——先跑 md2wechat.py 生成 HTML", 2)
|
||||
html_text = open(html_path, encoding="utf-8").read()
|
||||
|
||||
if args.title and len(args.title) > 64:
|
||||
die(f"标题 {len(args.title)} 字，微信上限 64 字", 2)
|
||||
if len(args.digest) > 120:
|
||||
die(f"摘要 {len(args.digest)} 字，微信上限 120 字", 2)
|
||||
|
||||
cover = args.cover or next(
|
||||
(os.path.join(args.dir, f"cover{ext}")
|
||||
for ext in (".jpg", ".jpeg", ".png", ".webp")
|
||||
if os.path.isfile(os.path.join(args.dir, f"cover{ext}"))), None)
|
||||
|
||||
# 正文里的本地图片 → 逐个经 uploadimg 换成微信域名 URL
|
||||
local_imgs = re.findall(r'src="([^"]+)"', html_text)
|
||||
local_imgs = [s for s in local_imgs if not s.startswith(("http://", "https://", "data:"))]
|
||||
|
||||
if args.dry_run:
|
||||
print("—— DRY RUN（不联网，不需要凭证）——")
|
||||
print(f"标题：{args.title or '（未提供）'}")
|
||||
print(f"摘要：{args.digest or '（空）'}（{len(args.digest)}/120 字）")
|
||||
print(f"封面：{cover or '❌ 未找到——用 --cover 指定或放 cover.png 进文章目录'}")
|
||||
print(f"正文本地图片 {len(local_imgs)} 张：")
|
||||
for s in local_imgs:
|
||||
print(f" - {s}")
|
||||
crop_note = "跳过（--no-crop）" if args.no_crop else f"2.35:1={crop_fractions(*img_size(cover), 2.35, 1)} · 1:1={crop_fractions(*img_size(cover), 1, 1)}"
|
||||
print(f"裁剪框：{crop_note}")
|
||||
print("执行动作：上传封面拿 thumb_media_id → 逐张上传正文图并替换 src → 提交 draft/add")
|
||||
return
|
||||
|
||||
if not (app_id and secret):
|
||||
print("❌ 缺少环境变量 WECHAT_APP_ID / WECHAT_APP_SECRET\n"
|
||||
" 获取路径：mp.weixin.qq.com → 设置与开发 → 基本配置\n"
|
||||
" 设置方式：export WECHAT_APP_ID=xxx WECHAT_APP_SECRET=yyy（建议写进 ~/.zshrc）\n"
|
||||
" 别把密钥写进任何 git 仓库。", file=sys.stderr)
|
||||
sys.exit(2)
|
||||
|
||||
if not args.title:
|
||||
die("正式同步必须提供 --title", 2)
|
||||
if not cover:
|
||||
die("找不到封面图——先按封面 brief 生成 cover.png 放进文章目录，或用 --cover 指定", 2)
|
||||
|
||||
token = get_token(app_id, secret)
|
||||
|
||||
print("上传封面…")
|
||||
r = post_file(f"{API}/cgi-bin/material/add_material?type=image&access_token={token}", cover)
|
||||
err = wx_err(r)
|
||||
if err:
|
||||
die(f"封面上传失败——{err}")
|
||||
thumb_media_id = r["media_id"]
|
||||
print(f" ✔ thumb_media_id={thumb_media_id}")
|
||||
|
||||
for s in local_imgs:
|
||||
p = os.path.join(args.dir, s)
|
||||
if not os.path.isfile(p):
|
||||
die(f"正文图片不存在：{p}")
|
||||
print(f"上传正文图 {s}…")
|
||||
r = post_file(f"{API}/cgi-bin/media/uploadimg?access_token={token}", p)
|
||||
err = wx_err(r)
|
||||
if err:
|
||||
die(f"正文图上传失败（{s}）——{err}")
|
||||
html_text = html_text.replace(f'src="{s}"', f'src="{r["url"]}"')
|
||||
print(f" ✔ {s} → {r['url']}")
|
||||
|
||||
print("提交草稿…")
|
||||
article = {"title": args.title, "content": html_text,
|
||||
"thumb_media_id": thumb_media_id,
|
||||
"need_open_comment": 0, "only_fans_can_comment": 0}
|
||||
if not args.no_crop:
|
||||
w, h = img_size(cover)
|
||||
article["pic_crop_235_1"] = crop_fractions(w, h, 2.35, 1)
|
||||
article["pic_crop_1_1"] = crop_fractions(w, h, 1, 1)
|
||||
print(f" 裁剪框：2.35:1={article['pic_crop_235_1']} · 1:1={article['pic_crop_1_1']}")
|
||||
if args.digest:
|
||||
article["digest"] = args.digest
|
||||
if args.author:
|
||||
article["author"] = args.author
|
||||
r = http_json(f"{API}/cgi-bin/draft/add?access_token={token}", {"articles": [article]})
|
||||
err = wx_err(r)
|
||||
if err:
|
||||
die(f"草稿提交失败——{err}")
|
||||
print(f"✔ 已进入草稿箱（media_id={r['media_id']}）\n"
|
||||
f" 下一步：mp.weixin.qq.com → 内容与互动 → 发表 → 草稿，人工预览确认后再发表。\n"
|
||||
f" 本 skill 到草稿箱为止，不执行群发。")
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
main()
|
||||
Reference in New Issue
Block a user