项目初始化

daily/wecom_split.py

@@ -0,0 +1,94 @@
+"""企微 markdown 按字节上限拆分为多条消息（按区块，不截断正文）。"""
+
+from __future__ import annotations
+
+import re
+
+_SECTION_START = re.compile(r"^[📰💡🎯🌍📈🔥🐙🌱🤖📦📄]")
+
+
+def _utf8_len(text: str) -> int:
+    return len(text.encode("utf-8"))
+
+
+def _split_lines_by_budget(text: str, limit: int) -> list[str]:
+    lines = text.splitlines()
+    chunks: list[str] = []
+    buf: list[str] = []
+    for line in lines:
+        candidate = "\n".join(buf + [line]) if buf else line
+        if _utf8_len(candidate) <= limit:
+            buf.append(line)
+            continue
+        if buf:
+            chunks.append("\n".join(buf))
+            buf = []
+        if _utf8_len(line) <= limit:
+            buf = [line]
+        else:
+            encoded = line.encode("utf-8")
+            start = 0
+            while start < len(encoded):
+                piece = encoded[start : start + limit].decode("utf-8", errors="ignore")
+                chunks.append(piece)
+                start += len(piece.encode("utf-8"))
+    if buf:
+        chunks.append("\n".join(buf))
+    return chunks
+
+
+def _split_sections(text: str) -> list[str]:
+    sections: list[str] = []
+    current: list[str] = []
+    for line in text.splitlines():
+        if _SECTION_START.match(line) and current:
+            sections.append("\n".join(current))
+            current = [line]
+        else:
+            current.append(line)
+    if current:
+        sections.append("\n".join(current))
+    return sections
+
+
+def split_wecom_messages(text: str, limit: int = 4096) -> list[str]:
+    """超长时拆成多条；每条不超过 limit 字节，按区块边界优先。"""
+    text = text.strip()
+    if not text or _utf8_len(text) <= limit:
+        return [text] if text else []
+
+    footer_reserve = 40
+    pack_limit = max(512, limit - footer_reserve)
+
+    sections: list[str] = []
+    for sec in _split_sections(text):
+        if _utf8_len(sec) <= pack_limit:
+            sections.append(sec)
+        else:
+            sections.extend(_split_lines_by_budget(sec, pack_limit))
+
+    packed: list[str] = []
+    buf: list[str] = []
+    for sec in sections:
+        candidate = "\n\n".join(buf + [sec]) if buf else sec
+        if _utf8_len(candidate) <= pack_limit:
+            buf.append(sec)
+        else:
+            if buf:
+                packed.append("\n\n".join(buf))
+            buf = [sec]
+    if buf:
+        packed.append("\n\n".join(buf))
+
+    total = len(packed)
+    if total <= 1:
+        return packed
+
+    result: list[str] = []
+    for i, chunk in enumerate(packed, 1):
+        suffix = f"\n\n> 📄 {i}/{total}"
+        body = chunk
+        while body and _utf8_len(body + suffix) > limit:
+            body = body.rsplit("\n", 1)[0] if "\n" in body else body[:-1]
+        result.append(body + suffix)
+    return result