Refactor AI daily report pipeline

ai_daily_report/sources/labels.py

@@ -0,0 +1,78 @@
+from __future__ import annotations
+
+from urllib.parse import urlparse
+
+
+DOMAIN_LABELS = {
+    "anthropic.com": "Anthropic",
+    "arxiv.org": "arXiv",
+    "bloomberg.com": "Bloomberg",
+    "deepseek.com": "DeepSeek",
+    "github.blog": "GitHub Blog",
+    "github.com": "GitHub",
+    "huggingface.co": "Hugging Face",
+    "infoq.com": "InfoQ",
+    "mp.weixin.qq.com": "微信公众号",
+    "openai.com": "OpenAI",
+    "platform.minimaxi.com": "MiniMax：Docs",
+    "qbitai.com": "量子位",
+    "techcrunch.com": "TechCrunch",
+    "technologyreview.com": "MIT科技评论AI",
+    "theverge.com": "The Verge",
+    "x.com": "X",
+    "twitter.com": "X",
+}
+
+X_DISPLAY_NAMES = {
+    "MiniMax_AI": "MiniMax",
+    "OpenAIDevs": "OpenAI Developers",
+    "openai": "OpenAI",
+    "openclaw": "OpenClaw",
+    "xai": "xAI",
+    "krea_ai": "Krea AI",
+    "nvidia": "NVIDIA",
+    "NVIDIAAI": "NVIDIA AI",
+    "alibaba_cloud": "阿里云 / Alibaba Cloud",
+    "cb_doge": "cb_doge",
+}
+
+
+def _host(url: str) -> str:
+    host = (urlparse(url).netloc or "").lower()
+    return host[4:] if host.startswith("www.") else host
+
+
+def _domain_label(host: str) -> str:
+    for domain, label in DOMAIN_LABELS.items():
+        if host == domain or host.endswith("." + domain):
+            return label
+    return host
+
+
+def _x_handle(url: str) -> str:
+    parts = [part for part in urlparse(url).path.split("/") if part]
+    if not parts:
+        return ""
+    handle = parts[0]
+    if handle in {"i", "search", "explore", "settings", "notifications", "home", "compose"}:
+        return ""
+    return handle
+
+
+def source_label_from_url(url: str, *, fallback: str = "来源") -> str:
+    if not url:
+        return fallback
+    host = _host(url)
+    if host in {"x.com", "twitter.com"}:
+        handle = _x_handle(url)
+        if handle:
+            display = X_DISPLAY_NAMES.get(handle, handle)
+            return f"X：{display} (@{handle})"
+        return "X"
+
+    label = _domain_label(host)
+    parsed = urlparse(url)
+    path = (parsed.path or "").lower()
+    if label and ("blog" in host or "/blog" in path or "/research" in path):
+        return f"{label}：Blog"
+    return label or fallback