weflow-cli 1.5.0 → 1.6.1

package/scripts/daily_stats.py

@@ -0,0 +1,132 @@
+#!/usr/bin/env python3
+"""公众号推送频率与日报处理频率统计。"""
+
+import argparse
+import hashlib
+import json
+import os
+import sys
+from collections import defaultdict
+from datetime import datetime, timedelta, timezone
+from pathlib import Path
+
+sys.path.insert(0, os.path.dirname(os.path.abspath(__file__)))
+from _utils import decrypt_lock, load_config
+from biz_daily import extract_article_info, get_db_keys
+
+try:
+    from sqlcipher3 import dbapi2 as sqlcipher
+except ImportError:
+    print('请安装 sqlcipher3: pip install sqlcipher3', file=sys.stderr)
+    sys.exit(1)
+
+
+def load_processed_counts(start_date: str, end_date: str) -> dict[str, int]:
+    root = Path(__file__).resolve().parent.parent / 'output' / 'biz-daily'
+    counts: dict[str, int] = defaultdict(int)
+    current = datetime.strptime(start_date, '%Y-%m-%d').date()
+    end = datetime.strptime(end_date, '%Y-%m-%d').date()
+    while current <= end:
+        path = root / current.isoformat() / '.articles.json'
+        if path.exists():
+            try:
+                payload = json.loads(path.read_text(encoding='utf-8'))
+                for article in payload.get('articles', []):
+                    source = str(article.get('source') or '').strip()
+                    if source:
+                        counts[source] += 1
+            except (OSError, ValueError, TypeError):
+                pass
+        current += timedelta(days=1)
+    return dict(counts)
+
+
+def main() -> None:
+    sys.stdout.reconfigure(encoding='utf-8', errors='replace')
+    parser = argparse.ArgumentParser(description='统计公众号推送与日报处理频率')
+    parser.add_argument('--days', type=int, default=30, help='统计最近多少天，默认 30')
+    parser.add_argument('--limit', type=int, default=30, help='最多显示多少个公众号，默认 30')
+    parser.add_argument('--json', action='store_true', help='输出 JSON')
+    args = parser.parse_args()
+    if args.days < 1 or args.days > 3650:
+        parser.error('--days 必须在 1 到 3650 之间')
+    if args.limit < 1:
+        parser.error('--limit 必须大于 0')
+
+    tz = timezone(timedelta(hours=8))
+    today = datetime.now(tz).date()
+    start_date = today - timedelta(days=args.days - 1)
+    start_ts = int(datetime.combine(start_date, datetime.min.time(), tzinfo=tz).timestamp())
+    end_ts = int(datetime.combine(today + timedelta(days=1), datetime.min.time(), tzinfo=tz).timestamp())
+
+    config = load_config()
+    keys = get_db_keys(config)
+    conn = sqlcipher.connect(keys['biz_db'])
+    conn.execute(f'PRAGMA key = "x\'{keys["biz_key"]}{keys["biz_salt"]}\'";')
+    cursor = conn.cursor()
+
+    name_map: dict[str, str] = {}
+    if keys['contact_key'] and os.path.exists(keys['contact_db']):
+        try:
+            contact = sqlcipher.connect(keys['contact_db'])
+            contact.execute(f'PRAGMA key = "x\'{keys["contact_key"]}{keys["contact_salt"]}\'";')
+            name_map = dict(contact.execute(
+                "SELECT username, COALESCE(NULLIF(remark,''), NULLIF(nick_name,''), username) "
+                "FROM contact WHERE username LIKE 'gh_%'"
+            ).fetchall())
+            contact.close()
+        except Exception:
+            pass
+
+    rows: list[dict] = []
+    users = [row[0] for row in cursor.execute(
+        "SELECT user_name FROM Name2Id WHERE user_name LIKE 'gh_%'"
+    ).fetchall()]
+    for user in users:
+        table = 'Msg_' + hashlib.md5(user.encode()).hexdigest()
+        try:
+            records = cursor.execute(
+                f'SELECT create_time, message_content FROM "{table}" '
+                'WHERE create_time >= ? AND create_time < ? ORDER BY create_time',
+                (start_ts, end_ts),
+            ).fetchall()
+        except Exception:
+            continue
+        article_count = sum(
+            1 for _, content in records
+            if content and extract_article_info(content).get('title')
+        )
+        if article_count:
+            rows.append({'name': str(name_map.get(user, user)), 'id': user, 'pushed': article_count})
+    conn.close()
+
+    processed = load_processed_counts(start_date.isoformat(), today.isoformat())
+    for row in rows:
+        row['processed'] = processed.get(row['name'], 0)
+    rows.sort(key=lambda row: (-row['processed'], -row['pushed'], row['name']))
+
+    if args.json:
+        json_output = {
+            'success': True,
+            'period': {
+                'start': start_date.isoformat(),
+                'end': today.isoformat(),
+                'days': args.days,
+            },
+            'totalSources': len(rows),
+            'sources': rows[:args.limit],
+        }
+        print(json.dumps(json_output, ensure_ascii=False, indent=2))
+        return
+
+    print(f'公众号频率统计（{start_date.isoformat()} 至 {today.isoformat()}）')
+    print('说明：推送数是数据库中的文章数；处理数是已生成日报的文章数，不等同于打开次数。')
+    print(f'共统计 {len(rows)} 个有推送记录的公众号\n')
+    print(f'{"排名":<5}{"公众号":<28}{"推送数":>8}{"处理数":>8}{"ID"}')
+    print('-' * 72)
+    for index, row in enumerate(rows[:args.limit], 1):
+        print(f'{index:<5}{row["name"][:26]:<28}{row["pushed"]:>8}{row["processed"]:>8}  {row["id"]}')
+
+
+if __name__ == '__main__':
+    main()

package/scripts/enrich_backlinks.py

@@ -1,163 +1,163 @@
-#!/usr/bin/env python3
-"""
-双向链接增强 — 为文章添加同主题、共享概念的相关文章链接。
-
-用法:
-  python scripts/enrich_backlinks.py --date 2026-05-22
-  python scripts/enrich_backlinks.py --date 2026-05-22 --dry-run
-"""
-
-import sys, os, re
-from pathlib import Path
-from collections import defaultdict
-
-from _utils import parse_frontmatter
-
-SCRIPTS_DIR = os.path.dirname(os.path.abspath(__file__))
-PROJECT_ROOT = os.path.dirname(SCRIPTS_DIR)
-DEFAULT_SOURCE = os.path.join(PROJECT_ROOT, 'output', 'biz-daily')
-
-TOPIC_ORDER = ['AI', '学术', '新闻', '文学', '投资']
-
-
-def collect_articles(date_dir: str) -> list[dict]:
-    """收集目录下所有文章，提取元数据。"""
-    articles = []
-    for topic in TOPIC_ORDER:
-        topic_dir = Path(date_dir) / topic
-        if not topic_dir.is_dir():
-            continue
-        for md_file in sorted(topic_dir.glob('*.md')):
-            try:
-                content = md_file.read_text(encoding='utf-8')
-            except Exception:
-                continue
-            fm, body = parse_frontmatter(content)
-            # 检查是否已经有相关阅读段落
-            if '## 相关阅读' in body:
-                continue  # 已处理，跳过
-
-            articles.append({
-                'path': md_file,
-                'topic': topic,
-                'filename': f'{topic}/{md_file.name}',
-                'title': fm.get('title', md_file.stem),
-                'source': fm.get('source', ''),
-                'tags': fm.get('tags', []),
-                'concepts': fm.get('concepts', fm.get('wikilinks', [])),
-                'relevance': fm.get('relevance', ''),
-            })
-    return articles
-
-
-def find_related(article: dict, all_articles: list[dict]) -> list[dict]:
-    """查找与给定文章相关的文章。"""
-    # 同主题文章（排除自身）
-    same_topic = [a for a in all_articles
-                  if a['topic'] == article['topic']
-                  and a['filename'] != article['filename']]
-
-    # 共享概念的文章
-    my_concepts = set(article['concepts']) if isinstance(article['concepts'], list) else set()
-    shared_concept = []
-    if my_concepts:
-        shared_concept = [a for a in all_articles
-                          if a['filename'] != article['filename']
-                          and a['topic'] != article['topic']
-                          and set(a['concepts'] if isinstance(a['concepts'], list) else []) & my_concepts]
-
-    # 优先：同主题 + 高相关 + 共享概念
-    related = []
-    seen = set()
-
-    # 先加同主题高相关
-    for a in same_topic:
-        if a['relevance'] == '高':
-            related.append(a)
-            seen.add(a['filename'])
-
-    # 共享概念
-    for a in shared_concept:
-        if a['filename'] not in seen:
-            related.append(a)
-            seen.add(a['filename'])
-
-    # 补充同主题中相关
-    for a in same_topic:
-        if a['filename'] not in seen and a['relevance'] == '中':
-            related.append(a)
-            seen.add(a['filename'])
-
-    # 最后补任意来源
-    for a in same_topic:
-        if a['filename'] not in seen:
-            related.append(a)
-            seen.add(a['filename'])
-
-    return related[:5]
-
-
-def enrich_article(article: dict, related: list[dict], dry_run=False):
-    """为文章末尾追加相关阅读段落。"""
-    if not related:
-        return 0
-
-    lines = ['', '---', '', '## 相关阅读', '']
-    for a in related:
-        label = f"📎 {a['title']}"
-        if a.get('source'):
-            label += f" — {a['source']}"
-        # 使用相对路径 Wikilink
-        link_path = a['filename'].replace('/', '/')
-        lines.append(f'- [[{link_path}|{label}]]')
-
-    appendix = '\n'.join(lines) + '\n'
-
-    if dry_run:
-        print(f'  → {article["title"][:40]}')
-        print(f'    相关: {", ".join(a["title"][:30] for a in related)}')
-        return len(related)
-
-    filepath = article['path']
-    current = filepath.read_text(encoding='utf-8')
-    filepath.write_text(current.rstrip() + '\n' + appendix, encoding='utf-8')
-    return len(related)
-
-
-def main():
-    sys.stdout.reconfigure(encoding='utf-8', errors='replace')
-    import argparse
-    parser = argparse.ArgumentParser(description='为文章添加双向链接（相关阅读）')
-    parser.add_argument('--date', required=True, help='日期 YYYY-MM-DD')
-    parser.add_argument('--dry-run', action='store_true', help='仅预览不写入')
-    parser.add_argument('--source', default=DEFAULT_SOURCE, help='源目录')
-    args = parser.parse_args()
-
-    date_dir = os.path.join(args.source, args.date)
-    if not os.path.isdir(date_dir):
-        print(f'[ERROR] 目录不存在: {date_dir}')
-        sys.exit(1)
-
-    articles = collect_articles(date_dir)
-    print(f'找到 {len(articles)} 篇文章')
-    print(f'主题分布: {dict((t, len([a for a in articles if a["topic"]==t])) for t in TOPIC_ORDER if any(a["topic"]==t for a in articles))}')
-    print()
-
-    enriched = 0
-    skipped = 0
-    for article in articles:
-        related = find_related(article, articles)
-        if not related:
-            skipped += 1
-            continue
-        n = enrich_article(article, related, args.dry_run)
-        if n > 0:
-            enriched += 1
-
-    print(f'\n✓ 完成: {enriched} 篇已增强, {skipped} 篇无关联文章')
-    if args.dry_run:
-        print('  (--dry-run 模式，未实际写入)')
-
-
-if __name__ == '__main__':
-    main()
+#!/usr/bin/env python3
+"""
+双向链接增强 — 为文章添加同主题、共享概念的相关文章链接。
+
+用法:
+  python scripts/enrich_backlinks.py --date 2026-05-22
+  python scripts/enrich_backlinks.py --date 2026-05-22 --dry-run
+"""
+
+import sys, os, re
+from pathlib import Path
+from collections import defaultdict
+
+from _utils import parse_frontmatter
+
+SCRIPTS_DIR = os.path.dirname(os.path.abspath(__file__))
+PROJECT_ROOT = os.path.dirname(SCRIPTS_DIR)
+DEFAULT_SOURCE = os.path.join(PROJECT_ROOT, 'output', 'biz-daily')
+
+TOPIC_ORDER = ['AI', '学术', '新闻', '文学', '投资', '政治']
+
+
+def collect_articles(date_dir: str) -> list[dict]:
+    """收集目录下所有文章，提取元数据。"""
+    articles = []
+    for topic in TOPIC_ORDER:
+        topic_dir = Path(date_dir) / topic
+        if not topic_dir.is_dir():
+            continue
+        for md_file in sorted(topic_dir.glob('*.md')):
+            try:
+                content = md_file.read_text(encoding='utf-8')
+            except Exception:
+                continue
+            fm, body = parse_frontmatter(content)
+            # 检查是否已经有相关阅读段落
+            if '## 相关阅读' in body:
+                continue  # 已处理，跳过
+
+            articles.append({
+                'path': md_file,
+                'topic': topic,
+                'filename': f'{topic}/{md_file.name}',
+                'title': fm.get('title', md_file.stem),
+                'source': fm.get('source', ''),
+                'tags': fm.get('tags', []),
+                'concepts': fm.get('concepts', fm.get('wikilinks', [])),
+                'relevance': fm.get('relevance', ''),
+            })
+    return articles
+
+
+def find_related(article: dict, all_articles: list[dict]) -> list[dict]:
+    """查找与给定文章相关的文章。"""
+    # 同主题文章（排除自身）
+    same_topic = [a for a in all_articles
+                  if a['topic'] == article['topic']
+                  and a['filename'] != article['filename']]
+
+    # 共享概念的文章
+    my_concepts = set(article['concepts']) if isinstance(article['concepts'], list) else set()
+    shared_concept = []
+    if my_concepts:
+        shared_concept = [a for a in all_articles
+                          if a['filename'] != article['filename']
+                          and a['topic'] != article['topic']
+                          and set(a['concepts'] if isinstance(a['concepts'], list) else []) & my_concepts]
+
+    # 优先：同主题 + 高相关 + 共享概念
+    related = []
+    seen = set()
+
+    # 先加同主题高相关
+    for a in same_topic:
+        if a['relevance'] == '高':
+            related.append(a)
+            seen.add(a['filename'])
+
+    # 共享概念
+    for a in shared_concept:
+        if a['filename'] not in seen:
+            related.append(a)
+            seen.add(a['filename'])
+
+    # 补充同主题中相关
+    for a in same_topic:
+        if a['filename'] not in seen and a['relevance'] == '中':
+            related.append(a)
+            seen.add(a['filename'])
+
+    # 最后补任意来源
+    for a in same_topic:
+        if a['filename'] not in seen:
+            related.append(a)
+            seen.add(a['filename'])
+
+    return related[:5]
+
+
+def enrich_article(article: dict, related: list[dict], dry_run=False):
+    """为文章末尾追加相关阅读段落。"""
+    if not related:
+        return 0
+
+    lines = ['', '---', '', '## 相关阅读', '']
+    for a in related:
+        label = f"📎 {a['title']}"
+        if a.get('source'):
+            label += f" — {a['source']}"
+        # 使用相对路径 Wikilink
+        link_path = a['filename'].replace('/', '/')
+        lines.append(f'- [[{link_path}|{label}]]')
+
+    appendix = '\n'.join(lines) + '\n'
+
+    if dry_run:
+        print(f'  → {article["title"][:40]}')
+        print(f'    相关: {", ".join(a["title"][:30] for a in related)}')
+        return len(related)
+
+    filepath = article['path']
+    current = filepath.read_text(encoding='utf-8')
+    filepath.write_text(current.rstrip() + '\n' + appendix, encoding='utf-8')
+    return len(related)
+
+
+def main():
+    sys.stdout.reconfigure(encoding='utf-8', errors='replace')
+    import argparse
+    parser = argparse.ArgumentParser(description='为文章添加双向链接（相关阅读）')
+    parser.add_argument('--date', required=True, help='日期 YYYY-MM-DD')
+    parser.add_argument('--dry-run', action='store_true', help='仅预览不写入')
+    parser.add_argument('--source', default=DEFAULT_SOURCE, help='源目录')
+    args = parser.parse_args()
+
+    date_dir = os.path.join(args.source, args.date)
+    if not os.path.isdir(date_dir):
+        print(f'[ERROR] 目录不存在: {date_dir}')
+        sys.exit(1)
+
+    articles = collect_articles(date_dir)
+    print(f'找到 {len(articles)} 篇文章')
+    print(f'主题分布: {dict((t, len([a for a in articles if a["topic"]==t])) for t in TOPIC_ORDER if any(a["topic"]==t for a in articles))}')
+    print()
+
+    enriched = 0
+    skipped = 0
+    for article in articles:
+        related = find_related(article, articles)
+        if not related:
+            skipped += 1
+            continue
+        n = enrich_article(article, related, args.dry_run)
+        if n > 0:
+            enriched += 1
+
+    print(f'\n✓ 完成: {enriched} 篇已增强, {skipped} 篇无关联文章')
+    if args.dry_run:
+        print('  (--dry-run 模式，未实际写入)')
+
+
+if __name__ == '__main__':
+    main()