refactor: Phase 3 拆分 generate 流水线并补全测试

提取 collect/formatters/themes 等 pipeline 模块，新增 wecom 分条、RSS、delta、Agent 工作流测试。

Co-authored-by: Cursor <cursoragent@cursor.com>

daily/pipeline/formatters.py

@@ -0,0 +1,180 @@
+"""归档 / 企微格式化与摘要构建。"""
+
+from __future__ import annotations
+
+import re
+import xml.etree.ElementTree as ET
+from typing import Any
+
+import certifi
+import httpx
+
+from daily.config import full_desc_limit, wecom_skill_desc_limit
+from daily.github.auth import github_html_headers
+from daily.news.fetch import prepare_wecom_cn_news_items, prepare_wecom_news_items
+from shared.skills_data import format_installs
+
+from daily.pipeline.snapshot import skill_id
+
+
+def short_desc(text: str, limit: int = 72) -> str:
+    text = re.sub(r"\s+", " ", text or "").strip()
+    if limit <= 0 or len(text) <= limit:
+        return text
+    return text[: limit - 3] + "..."
+
+
+def archive_desc(text: str) -> str:
+    return short_desc(text, full_desc_limit())
+
+
+def wecom_desc(text: str, limit: int = 36) -> str:
+    return short_desc(text, limit)
+
+
+def prepare_skill_item(item: dict[str, Any], prev_ids: set[str], rank: int) -> dict[str, Any]:
+    badge = ""
+    sid = skill_id(item)
+    if sid not in prev_ids and prev_ids:
+        badge = "🆕"
+    elif rank == 1:
+        badge = "👑"
+    installs_fmt = item.get("installs_fmt") or format_installs(item.get("installs", 0))
+    title = item.get("source", "?") if item.get("cluster") else item.get("title", "?")
+    desc = item.get("wecom_desc") or item.get("description") or item.get("cluster_titles") or ""
+    limit = wecom_skill_desc_limit()
+    desc_short = desc if item.get("wecom_desc") or limit <= 0 else wecom_desc(desc, limit)
+    return {
+        "title": title,
+        "source": item.get("source", "?"),
+        "installs_fmt": installs_fmt,
+        "link": item.get("link", ""),
+        "desc_short": desc_short,
+        "badge": badge,
+        "cluster": bool(item.get("cluster")),
+        "cluster_count": item.get("cluster_count"),
+        "cluster_titles": item.get("cluster_titles"),
+    }
+
+
+def prepare_github_item(item: dict[str, Any]) -> dict[str, Any]:
+    return {**item, "desc_short": wecom_desc(item.get("description", ""), 40)}
+
+
+def build_highlights(
+    trending: list[dict[str, Any]],
+    hot: list[dict[str, Any]],
+    github_trending: list[dict[str, Any]],
+    github_emerging: list[dict[str, Any]],
+    ai_news: dict[str, Any] | None = None,
+    cn_ai_news: dict[str, Any] | None = None,
+) -> list[str]:
+    points: list[str] = []
+    if ai_news and ai_news.get("enabled"):
+        top_news = prepare_wecom_news_items(ai_news)
+        if top_news:
+            n0 = top_news[0]
+            pub = f" · {n0['published_fmt']}" if n0.get("published_fmt") else ""
+            points.append(
+                f"🌍 AI 时讯 [{n0['title']}]({n0['link']})（`{n0.get('source_name', '?')}`{pub}）"
+            )
+        elif ai_news.get("flat"):
+            n0 = ai_news["flat"][0]
+            pub = f" · {n0['published_fmt']}" if n0.get("published_fmt") else ""
+            points.append(f"🌍 AI 时讯 [{n0['title']}]({n0['link']})（`{n0.get('source_name', '?')}`{pub}）")
+    if cn_ai_news and cn_ai_news.get("enabled"):
+        top_cn = prepare_wecom_cn_news_items(cn_ai_news)
+        if top_cn:
+            n0 = top_cn[0]
+            pub = f" · {n0['published_fmt']}" if n0.get("published_fmt") else ""
+            points.append(
+                f"🇨🇳 国内 AI [{n0['title']}]({n0['link']})（`{n0.get('source_name', '?')}`{pub}）"
+            )
+        elif cn_ai_news.get("flat"):
+            n0 = cn_ai_news["flat"][0]
+            pub = f" · {n0['published_fmt']}" if n0.get("published_fmt") else ""
+            points.append(
+                f"🇨🇳 国内 AI [{n0['title']}]({n0['link']})（`{n0.get('source_name', '?')}`{pub}）"
+            )
+    if trending:
+        t0 = trending[0]
+        points.append(f"📈 Skills 榜首 **{t0.get('title')}**（{format_installs(t0.get('installs', 0))}）")
+    if github_trending:
+        g0 = github_trending[0]
+        stars = g0.get("stars_today_fmt", "")
+        total = g0.get("total_stars_fmt", "")
+        star_hint = f"+{stars} today · " if stars else (f"⭐{total} · " if total else "")
+        points.append(f"🐙 GitHub Trending [{g0['repo']}]({g0['url']})（{star_hint}{g0.get('language', '')}）")
+    if github_emerging:
+        e0 = github_emerging[0]
+        points.append(f"🌱 新兴 [{e0['repo']}]({e0['url']})（⭐ {e0.get('total_stars_fmt', '?')}）")
+    elif hot:
+        h0 = hot[0]
+        points.append(f"🔥 Skills Hot 榜首 **{h0.get('title')}**（1H {format_installs(h0.get('installs', 0))}）")
+    while len(points) < 3 and len(trending) > len(points):
+        item = trending[len(points)]
+        points.append(f"✨ **{item.get('title')}** · `{item.get('source')}`")
+    return points[:3]
+
+
+def format_github_repo_section(repos: list[dict[str, Any]], *, show_created: bool = False) -> list[str]:
+    lines: list[str] = []
+    for i, repo in enumerate(repos, 1):
+        lang = repo.get("language") or "—"
+        stars_today = repo.get("stars_today_fmt") or ""
+        total = repo.get("total_stars_fmt") or ""
+        created = repo.get("created_at") or ""
+        meta_parts = [lang]
+        if stars_today:
+            meta_parts.append(f"+{stars_today} today")
+        if total:
+            meta_parts.append(f"总 ⭐ {total}")
+        if show_created and created:
+            meta_parts.append(f"创建于 {created}")
+        lines.append(f"{i}. **[{repo['repo']}]({repo['url']})** · {' · '.join(meta_parts)}")
+        desc = archive_desc(repo.get("description", ""))
+        if desc:
+            lines.append(f"   - {desc}")
+        lines.append("")
+    return lines
+
+
+def format_skill_section(items: list[dict[str, Any]], *, hot: bool = False) -> list[str]:
+    lines: list[str] = []
+    for i, item in enumerate(items, 1):
+        sid = item.get("id") or f"{item.get('source', '?')}/{item.get('title', '?')}"
+        link = item.get("link", "")
+        installs = format_installs(item.get("installs", 0))
+        meta = f"1H {installs}" if hot else f"总安装 {installs}"
+        if link:
+            lines.append(f"{i}. **[{sid}]({link})** · {meta}")
+        else:
+            lines.append(f"{i}. **{sid}** · {meta}")
+        desc = archive_desc(item.get("description", ""))
+        if desc:
+            lines.append(f"   - {desc}")
+        lines.append("")
+    return lines
+
+
+def fetch_latest_release_title(repo: str) -> str | None:
+    atom_url = f"https://github.com/{repo}/releases.atom"
+    try:
+        with httpx.Client(
+            timeout=12.0,
+            verify=certifi.where(),
+            follow_redirects=True,
+            headers=github_html_headers(),
+        ) as client:
+            resp = client.get(atom_url)
+            if resp.status_code != 200:
+                return None
+            root = ET.fromstring(resp.text)
+            ns = {"a": "http://www.w3.org/2005/Atom"}
+            entry = root.find("a:entry", ns)
+            if entry is None:
+                return None
+            title = entry.find("a:title", ns)
+            return title.text.strip() if title is not None and title.text else None
+    except Exception:
+        return None