Files
daily-robots/daily/format_wecom.py

441 lines
15 KiB
Python

"""企微早报排版。"""
from __future__ import annotations
import re
from typing import Any
from daily.config import wecom_skill_desc_limit
from daily.localize import LocalizeJob, localize_brief_descriptions, needs_chinese
from daily.text_utils import trim_brief
ICONS = {
"header": "📰",
"highlights": "💡",
"trending": "📈",
"hot": "🔥",
"github": "🐙",
"emerging": "🌱",
"topic": "🤖",
"ainews": "🌍",
"cnainews": "🇨🇳",
"pick": "📦",
"theme": "🎯",
"file": "📄",
}
def _skill_line(rank: int, item: dict[str, Any], *, badge: str = "") -> list[str]:
source = item.get("source", "?")
installs = item.get("installs_fmt", "?")
link = item.get("link", "")
desc = item.get("desc_short", "")
badge_prefix = f"{badge} " if badge else ""
if item.get("cluster"):
count = int(item.get("cluster_count") or 1)
sample = item.get("cluster_titles") or item.get("title", "")
label = f"**{source}** · {count} skills · **{installs}**"
if link:
head = f"{rank}. {badge_prefix}[{label}]({link})"
else:
head = f"{rank}. {badge_prefix}{label}"
lines = [head]
if sample or desc:
hint = desc or sample
lines.append(f" {hint}")
return lines
title = item.get("title", "?")
if link:
head = f"{rank}. {badge_prefix}[**{title}**]({link}) · `{source}` · **{installs}**"
else:
head = f"{rank}. {badge_prefix}**{title}** · `{source}` · **{installs}**"
lines = [head]
if desc:
lines.append(f" {desc}")
return lines
def _ai_news_lines(items: list[dict[str, Any]]) -> list[str]:
lines: list[str] = []
for i, item in enumerate(items, 1):
title = item.get("title", "?")
link = item.get("link", "")
source = item.get("source_name", "?")
pub = item.get("published_fmt", "")
desc = item.get("desc_short", "")
pub_suffix = f" · {pub}" if pub else ""
if link:
head = f"{i}. [**{title}**]({link}) · `{source}`{pub_suffix}"
else:
head = f"{i}. **{title}** · `{source}`{pub_suffix}"
lines.append(head)
if desc:
lines.append(f" > {desc}")
return lines
def _github_repo_lines(repos: list[dict[str, Any]], *, show_created: bool = False) -> list[str]:
lines: list[str] = []
for i, repo in enumerate(repos, 1):
name = repo["repo"]
url = repo["url"]
lang = repo.get("language", "")
stars_today = repo.get("stars_today_fmt", "")
total = repo.get("total_stars_fmt", "")
created = repo.get("created_at", "")
meta_parts: list[str] = []
if lang:
meta_parts.append(lang)
if stars_today:
meta_parts.append(f"+{stars_today} today")
elif total:
meta_parts.append(f"{total}")
if show_created and created:
meta_parts.append(f"创建于 {created}")
meta = f" · {' · '.join(meta_parts)}" if meta_parts else ""
lines.append(f"{i}. [{name}]({url}){meta}")
desc = repo.get("desc_short") or repo.get("description", "")
if desc:
lines.append(f" > {desc}")
return lines
def _fallback_skill_desc(item: dict[str, Any]) -> str:
if item.get("cluster"):
count = int(item.get("cluster_count") or 1)
source = item.get("source") or "unknown"
sample = item.get("cluster_titles") or item.get("title") or ""
return f"{count} agent skills from {source}, including {sample}"
title = item.get("title") or "skill"
source = item.get("source") or "unknown"
return f"{title} skill from {source}"
def _skill_group_key(item: dict[str, Any]) -> str:
return str(item.get("id") or f"{item.get('source')}/{item.get('title')}")
def finalize_wecom_skill_groups(
items: list[dict[str, Any]],
*,
desc_limit: int | None = None,
) -> list[dict[str, Any]]:
"""为企微 Skills 榜生成一句简要中文简介(与完整版 .md 长描述分离)。"""
if desc_limit is None:
desc_limit = wecom_skill_desc_limit()
limit = desc_limit if desc_limit > 0 else 48
copies: list[tuple[str, dict[str, Any]]] = []
jobs: list[LocalizeJob] = []
for item in items:
copy = dict(item)
desc = (copy.get("description") or "").strip()
if not desc:
desc = _fallback_skill_desc(copy)
key = _skill_group_key(copy)
job_key = f"wecom:{key}"
if needs_chinese(desc) or len(desc) > limit:
jobs.append(LocalizeJob(job_key, desc, limit))
else:
copy["wecom_desc"] = desc
copies.append((job_key, copy))
zh_map = localize_brief_descriptions(jobs, archive=True)
out: list[dict[str, Any]] = []
for job_key, copy in copies:
if job_key in zh_map:
copy["wecom_desc"] = zh_map[job_key]
elif "wecom_desc" not in copy:
copy["wecom_desc"] = _brief_fallback_desc(
(copy.get("description") or "").strip() or _fallback_skill_desc(copy),
limit,
)
out.append(copy)
return out
def _brief_fallback_desc(text: str, limit: int) -> str:
return trim_brief(text, limit)
def _grouped_skill_to_wecom_item(
item: dict[str, Any],
*,
desc_limit: int | None = None,
) -> dict[str, Any]:
if desc_limit is None:
desc_limit = wecom_skill_desc_limit()
limit = desc_limit if desc_limit > 0 else 48
installs = int(item.get("installs") or 0)
installs_fmt = item.get("installs_fmt") or str(installs)
desc = (item.get("wecom_desc") or item.get("description") or "").strip()
if not desc and item.get("cluster"):
desc = item.get("cluster_titles") or ""
if not item.get("wecom_desc"):
desc = _brief_fallback_desc(desc, limit)
return {
"title": item.get("title", ""),
"source": item.get("source", "?"),
"installs_fmt": installs_fmt,
"link": item.get("link", ""),
"desc_short": desc,
"badge": item.get("badge", ""),
"cluster": bool(item.get("cluster")),
"cluster_count": item.get("cluster_count"),
"cluster_titles": item.get("cluster_titles"),
}
def build_skills_board_section(icon_key: str, board_label: str, items: list[dict[str, Any]]) -> str:
prepared = finalize_wecom_skill_groups(items)
wecom_items = [_grouped_skill_to_wecom_item(x) for x in prepared]
lines = [f"{ICONS[icon_key]} **{board_label} Top {len(wecom_items)}**"]
for rank, item in enumerate(wecom_items, 1):
lines.extend(_skill_line(rank, item))
return "\n".join(lines)
def _move_to_wecom_skill_item(move: dict[str, Any]) -> dict[str, Any]:
installs = int(move.get("installs") or 0)
rank = move.get("rank", "?")
board_badge = (move.get("badge") or "").strip()
badge = f"[新入 #{rank}]"
if board_badge:
badge = f"{badge} {board_badge}"
return {
"title": move.get("title", "?"),
"source": move.get("source", "?"),
"installs_fmt": move.get("installs_fmt") or str(installs),
"link": move.get("link", ""),
"description": (move.get("description") or "").strip(),
"badge": badge,
}
def build_skills_delta_sections(
trending_moves: list[dict[str, Any]],
hot_moves: list[dict[str, Any]],
) -> str:
sections: list[str] = []
if trending_moves:
prepared = finalize_wecom_skill_groups([_move_to_wecom_skill_item(m) for m in trending_moves])
items = [_grouped_skill_to_wecom_item(x) for x in prepared]
lines = [f"{ICONS['trending']} **Skills Trending 变化**"]
for rank, item in enumerate(items, 1):
lines.extend(_skill_line(rank, item, badge=item.get("badge", "")))
sections.append("\n".join(lines))
if hot_moves:
prepared = finalize_wecom_skill_groups([_move_to_wecom_skill_item(m) for m in hot_moves])
items = [_grouped_skill_to_wecom_item(x) for x in prepared]
lines = [f"{ICONS['hot']} **Skills Hot 变化**"]
for rank, item in enumerate(items, 1):
lines.extend(_skill_line(rank, item, badge=item.get("badge", "")))
sections.append("\n".join(lines))
return "\n\n".join(sections)
def _github_move_to_repo(move: dict[str, Any]) -> dict[str, Any]:
return {
"repo": move.get("repo", "?"),
"url": move.get("url", ""),
"language": move.get("language", ""),
"stars_today_fmt": move.get("stars_today_fmt", ""),
"total_stars_fmt": move.get("total_stars_fmt", ""),
"created_at": move.get("created_at", ""),
"description": move.get("description", ""),
"desc_short": (move.get("description") or "").strip(),
"badge": f"[新入 #{move.get('rank', '?')}]",
}
def build_github_delta_sections(movement: dict[str, Any], *, topic_name: str) -> str:
sections: list[str] = []
mapping = [
("github_trending_moves", "github", "GitHub Trending 变化", False),
("github_emerging_moves", "emerging", "GitHub 新兴 变化", True),
("github_topic_moves", "topic", f"Topic `{topic_name}` 变化", False),
]
for key, icon_key, label, show_created in mapping:
moves = movement.get(key) or []
if not moves:
continue
lines = [f"{ICONS[icon_key]} **{label}**"]
for i, move in enumerate(moves, 1):
repo = _github_move_to_repo(move)
badge = repo.pop("badge", "")
chunk = _github_repo_lines([repo], show_created=show_created)
if chunk:
chunk[0] = f"{i}. {badge} " + chunk[0].split(". ", 1)[-1]
lines.extend(chunk)
sections.append("\n".join(lines))
return "\n\n".join(sections)
_SKILL_SECTIONS = re.compile(
r"📈 \*\*Skills Trending.*?(?=🐙 \*\*GitHub Trending)",
re.DOTALL,
)
_SKILL_TRENDING_BLOCK = re.compile(r"📈 \*\*Skills Trending[^\n]*\n(?:.*?\n)*?(?=\n🔥 \*\*Skills Hot|\n🐙 |\n🌱 |\n🤖 |\Z)", re.DOTALL)
_SKILL_HOT_BLOCK = re.compile(r"🔥 \*\*Skills Hot[^\n]*\n(?:.*?\n)*?(?=\n🐙 |\n🌱 |\n🤖 |\Z)", re.DOTALL)
_GITHUB_SECTIONS = re.compile(r"🐙 \*\*GitHub Trending.*", re.DOTALL)
def _strip_board_sections(md: str) -> str:
md = _SKILL_TRENDING_BLOCK.sub("", md)
md = _SKILL_HOT_BLOCK.sub("", md)
if _GITHUB_SECTIONS.search(md):
md = _GITHUB_SECTIONS.sub("", md)
return re.sub(r"\n{3,}", "\n\n", md).rstrip()
def replace_wecom_board_sections(
md: str,
*,
mode: str,
movement: dict[str, Any],
trending: list[dict[str, Any]],
hot: list[dict[str, Any]],
topic_name: str,
) -> str:
from daily.delta import partition_skill_moves_for_wecom
if mode == "delta":
t_moves, h_moves = partition_skill_moves_for_wecom(
movement.get("skills_trending_moves") or [],
movement.get("skills_hot_moves") or [],
)
skills_sec = build_skills_delta_sections(t_moves, h_moves)
github_sec = build_github_delta_sections(movement, topic_name=topic_name)
board_block = "\n\n".join(x for x in [skills_sec, github_sec] if x)
md = _strip_board_sections(md)
if board_block:
return md + "\n\n" + board_block + "\n"
return md + "\n"
trending_sec = build_skills_board_section("trending", "Skills Trending", trending)
hot_sec = build_skills_board_section("hot", "Skills Hot", hot)
replacement = f"{trending_sec}\n\n{hot_sec}\n\n"
if _SKILL_SECTIONS.search(md):
return _SKILL_SECTIONS.sub(replacement, md)
github_marker = "🐙 **GitHub Trending"
idx = md.find(github_marker)
if idx >= 0:
return md[:idx] + replacement + md[idx:]
return md.rstrip() + "\n\n" + replacement
def replace_wecom_skill_sections(
md: str,
*,
trending: list[dict[str, Any]],
hot: list[dict[str, Any]],
mode: str = "full",
movement: dict[str, Any] | None = None,
topic_name: str = "llm",
) -> str:
"""用 Python 合并后的 Skills 榜替换或插入 Agent 早报中的对应区块。"""
return replace_wecom_board_sections(
md,
mode=mode,
movement=movement or {},
trending=trending,
hot=hot,
topic_name=topic_name,
)
def _format_pick_link(pick_command: str, *, title: str = "", url: str = "") -> str:
cmd = pick_command.strip()
if not cmd:
return ""
label = title.strip()
if cmd.startswith("http://") or cmd.startswith("https://"):
if not label:
m = re.match(r"https?://github\.com/([^/\s#?]+/[^/\s#?]+)", cmd)
label = m.group(1) if m else cmd
return f"[{label}]({cmd})"
m = re.match(r"npx skills add (\S+)", cmd)
if m:
skill_path = m.group(1)
if not label:
label = skill_path.split("/")[-1]
href = url.strip() or f"https://skills.sh/{skill_path}"
return f"[{label}]({href})"
return f"`{cmd}`"
def build_wecom_report(
*,
date_str: str,
time_str: str,
updated: str,
highlights: list[str],
theme_line: str,
trending: list[dict[str, Any]],
hot: list[dict[str, Any]],
repos: list[dict[str, Any]],
emerging: list[dict[str, Any]],
topic_name: str,
topic_repos: list[dict[str, Any]],
ai_news: list[dict[str, Any]] | None = None,
cn_ai_news: list[dict[str, Any]] | None = None,
pick_command: str,
pick_why: str = "",
pick_title: str = "",
pick_url: str = "",
include_boards: bool = True,
) -> str:
lines = [
f"{ICONS['header']} **早报 · {date_str}**",
f"> ⏱ {time_str} · 数据截至 {updated}",
"",
f"{ICONS['highlights']} **今日速览**",
]
for point in highlights[:3]:
lines.append(f"> {point}")
lines.append("")
lines.append(f"{ICONS['theme']} {theme_line}")
lines.append("")
if ai_news:
lines.append(f"{ICONS['ainews']} **国际 AI 时讯 Top {len(ai_news)}**")
lines.extend(_ai_news_lines(ai_news))
lines.append("")
if cn_ai_news:
lines.append(f"{ICONS['cnainews']} **国内 AI 时讯 Top {len(cn_ai_news)}**")
lines.extend(_ai_news_lines(cn_ai_news))
lines.append("")
if include_boards:
lines.append(f"{ICONS['trending']} **Skills Trending Top {len(trending)}**")
for rank, item in enumerate(trending, 1):
lines.extend(_skill_line(rank, item, badge=item.get("badge", "")))
lines.append("")
lines.append(f"{ICONS['hot']} **Skills Hot Top {len(hot)}**")
for rank, item in enumerate(hot, 1):
lines.extend(_skill_line(rank, item, badge=item.get("badge", "")))
lines.append("")
if repos:
lines.append(f"{ICONS['github']} **GitHub Trending Top {len(repos)}**")
lines.extend(_github_repo_lines(repos))
lines.append("")
if emerging:
lines.append(f"{ICONS['emerging']} **新兴项目 Top {len(emerging)}**")
lines.extend(_github_repo_lines(emerging, show_created=True))
lines.append("")
if topic_repos:
lines.append(f"{ICONS['topic']} **Topic `{topic_name}` Top {len(topic_repos)}**")
lines.extend(_github_repo_lines(topic_repos))
lines.append("")
lines.append(f"{ICONS['pick']} **今日首推**")
lines.append(_format_pick_link(pick_command, title=pick_title, url=pick_url))
if pick_why:
lines.append(f"> {pick_why}")
return "\n".join(lines)