Create gen_changelog.py

添加脚本

Author: YangMo0524

.github/scripts/gen_changelog.py

@@ -0,0 +1,139 @@
+#!/usr/bin/env python3
+"""
+If version already exists in changelog.md, exit 0 without modification.
+Usage: python .github/scripts/gen_changelog.py v4.10.9
+"""
+from __future__ import annotations
+import sys, json, re, html, urllib.request
+from datetime import datetime, timezone, timedelta
+from pathlib import Path
+
+API_URL = "https://community.fit2cloud.com/v1/products/dataease/releases"
+CHANGELOG = Path('docs/changelog.md')
+MARKER = '## 2 更新内容'
+TZ = timezone(timedelta(hours=8))
+
+TITLE_MAP = {
+    '安全漏洞修复': ('Warning', '**安全漏洞修复**', 'fix'),
+    '新增功能': ('Abstract', '新增功能 :star2:', 'feat'),
+    '功能优化': ('Abstract', '功能优化 :sunflower:', 'refactor'),
+    '问题修复': ('Abstract', '问题修复 :palm_tree:', 'fix'),
+    "What's new": ('Abstract', '新增功能 :star2:', 'feat'),
+    'Improvements': ('Abstract', '功能优化 :sunflower:', 'refactor'),
+    'Bug fixes': ('Abstract', '问题修复 :palm_tree:', 'fix'),
+}
+H2_UL = re.compile(r'<h2><a[^>]*></a>([^<]+)</h2>\n?(<ul>.*?</ul>)', re.S)
+# 【新增】用于抓取 h2 后面紧跟的 p 标签 (感谢信息)
+# 逻辑：匹配 <h2>...</h2> 后面可能有的换行，然后匹配 <p>...</p>
+H2_P_THANKS = re.compile(r'<h2><a[^>]*></a>[^<]*</h2>\n?(<p>.*?</p>)', re.S)
+LI = re.compile(r'<li>(.*?)</li>', re.S)
+
+
+def fetch():
+    with urllib.request.urlopen(API_URL, timeout=30) as r:
+        return json.loads(r.read().decode())
+
+
+def normalize(v: str) -> str:
+    return v.replace('-lts','')
+
+
+def find_release(target: str, data):
+    for rel in data:
+        if normalize(rel.get('version','')) == target:
+            return rel
+    return None
+
+
+def build_block(rel: dict):
+    version = normalize(rel['version'])
+    ts = rel.get('publishTime')
+    dt = datetime.fromtimestamp(ts/1000, tz=TZ) if ts else datetime.now(tz=TZ)
+    date_str = f"{dt.year}年{dt.month}月{dt.day}日"
+    html_content = rel.get('releaseNoteH', '')
+    sections = []
+
+    for title, ul_html in H2_UL.findall(html_content):
+        items = LI.findall(ul_html)
+        cleaned = []
+        for it in items:
+            if not it.strip():
+                continue
+            # 处理 CVE 链接：将 <a href="url">CVE-XXX</a> 转换为 [(CVE-XXX)](url)
+            def replace_cve_link(match):
+                url = match.group(1)
+                cve_text = html.unescape(re.sub(r'<[^>]+>', '', match.group(2)).strip())
+                return f'[({cve_text})]({url})'
+            # 先处理 CVE 链接
+            processed = re.sub(r'<a\s+href="([^"]+)"[^>]*>(.*?)</a>', replace_cve_link, it)
+            # 再清理其他 HTML 标签
+            processed = html.unescape(re.sub(r'\s+', ' ', re.sub(r'<[^>]+>', '', processed)).strip())
+            if processed:
+                cleaned.append(processed)
+        thanks_note = ""
+        # 简单有效的正则：查找当前 title 对应的 h2 后面的 p
+        # 构造动态正则： <h2>...Title...</h2> ... <ul>...</ul> ... <p>(内容)</p>
+        # 由于 ul_html 内容可能很长且有特殊字符，直接用 title 定位最安全
+        safe_title = re.escape(title.strip())
+        # 正则解释：匹配包含 title 的 h2，后面任意字符(非贪婪)，然后匹配一个 <p>...</p>
+        p_pattern = re.compile(r'<h2><a[^>]*></a>' + safe_title + r'</h2>.*?<p>(.*?)</p>', re.S)
+        p_match = p_pattern.search(html_content)
+        if p_match:
+            p_content = p_match.group(1)
+            # 清理 HTML 标签，只留文本
+            clean_p = html.unescape(re.sub(r'<[^>]+>', '', p_content)).strip()
+            # 只有当内容包含"感谢"或者标题包含"安全"/"漏洞"时，才采纳
+            if "感谢" in clean_p and "漏洞" in title:
+                thanks_note = f"\n    {clean_p}"
+        if not cleaned:
+            continue
+        admon, nice, tag = TITLE_MAP.get(title, ('info', title, 'note'))
+        lines = '\n'.join(f"    - {i}" for i in cleaned)
+        if thanks_note:
+            if lines:
+                lines += "\n" + thanks_note
+        sections.append(f"!!! {admon} \"{nice}\"\n\n{lines}\n")
+    if not sections:
+        clean = html.unescape(re.sub(r'<[^>]+>','', html_content)).strip()
+        if clean:
+            sections.append(f"!!! info \"发布说明\"\n    - note: {clean}\n")
+    block = '\n'.join([f"### {version}", date_str, ''] + sections)
+    return block
+
+
+def main():
+    if len(sys.argv) < 2:
+        print('Version arg required, e.g. v2.10.1 or v2.10.1-lts or 2.10.1', file=sys.stderr)
+        return 1
+    raw = sys.argv[1].strip()
+    # Accept forms: v2.10.1 or v2.10.1-lts or 2.10.1
+    if not raw.startswith('v'):
+        raw = 'v' + raw
+    target = normalize(raw)  # remove -lts suffix if present
+    data = fetch()
+    rel = find_release(target, data)
+    if not rel:
+        print(f'Target version {raw} (normalized {target}) not found in API list', file=sys.stderr)
+        return 1
+    if not CHANGELOG.exists():
+        print('Changelog file missing.')
+        return 1
+    content = CHANGELOG.read_text(encoding='utf-8')
+    # 【修复】使用正则严格匹配 Markdown 标题行 (例如: ### v2.10.19)
+    # ^ 表示行首，#+ 表示一个或多个#，\s* 表示可选空格，re.escape 防止版本号中的点被当作通配符
+    pattern = re.compile(r'^#+\s*' + re.escape(target) + r'\s*$', re.MULTILINE)
+
+    if pattern.search(content):
+        print(f'⚠️ Version {target} already exists in changelog.md (Detected by regex). Skip.')
+        return 0
+    block = build_block(rel)
+    if MARKER in content:
+        new_content = content.replace(MARKER, MARKER + '\n\n' + block, 1)
+    else:
+        new_content = MARKER + '\n\n' + block + content
+    CHANGELOG.write_text(new_content, encoding='utf-8')
+    print('Inserted changelog for', target)
+    return 0
+
+if __name__ == '__main__':
+    raise SystemExit(main())