Hua/nanobot
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

fix(feishu): parse post wrapper payload for rich text messages

Browse Source
This commit is contained in:
yzchen
2026-03-01 02:17:10 +08:00
parent f172c9f381
commit 2fc16596d0
2 changed files with 57 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files

20
nanobot/channels/feishu.py
View File

@@ -221,18 +221,32 @@ def _extract_post_content(content_json: dict) -> tuple[str, list[str]]:
text = " ".join(text_parts).strip() if text_parts else None
return text, image_keys
# Compatible with both shapes:
# 1) {"post": {"zh_cn": {...}}}
# 2) {"zh_cn": {...}} or {"title": "...", "content": [...]}
post_root = content_json.get("post") if isinstance(content_json, dict) else None
if not isinstance(post_root, dict):
post_root = content_json if isinstance(content_json, dict) else {}
# Try direct format first
if "content" in content_json:
text, images = extract_from_lang(content_json)
if "content" in post_root:
text, images = extract_from_lang(post_root)
if text or images:
return text or "", images
# Try localized format
for lang_key in ("zh_cn", "en_us", "ja_jp"):
lang_content = content_json.get(lang_key)
lang_content = post_root.get(lang_key)
text, images = extract_from_lang(lang_content)
if text or images:
return text or "", images
# Fallback: first dict-shaped child
for value in post_root.values():
if isinstance(value, dict):
text, images = extract_from_lang(value)
if text or images:
return text or "", images
return "", []