From 4c2ac3ab5674bd46075641d3ad40f88b8181e75f Mon Sep 17 00:00:00 2001
From: Beast <beast@beast-trader.local>
Date: Sun, 14 Jun 2026 13:29:25 +0800
Subject: [PATCH] feat: Chinese news sources, AI summary, todo extraction

---
 scripts/extract_all_todos.py | 68 ++++++++++++++++++++++++++++++++++++
 scripts/fetch_news.py        | 10 +++---
 scripts/push_todos.py        | 17 +++++++++
 3 files changed, 91 insertions(+), 4 deletions(-)
 create mode 100644 scripts/extract_all_todos.py
 create mode 100644 scripts/push_todos.py

diff --git a/scripts/extract_all_todos.py b/scripts/extract_all_todos.py
new file mode 100644
index 0000000..c41fd36
--- /dev/null
+++ b/scripts/extract_all_todos.py
@@ -0,0 +1,68 @@
+"""Extract ALL todos from ALL existing inspiration digests and push to todo API."""
+import os, re, requests
+
+DIGEST_DIR = os.path.expanduser("~/inspiration-collector/ai-insights")
+TODO_API = "http://127.0.0.1:9001/api/todos"
+
+
+def get_all_digests():
+    """Recursively find all digest files."""
+    result = []
+    for root, dirs, files in os.walk(DIGEST_DIR):
+        for f in files:
+            if f.endswith("_digest.md") and not f.endswith(".bak.md"):
+                result.append(os.path.join(root, f))
+    result.sort()
+    return result
+
+
+def extract_todos(filepath):
+    with open(filepath) as f:
+        content = f.read()
+    todos = []
+    in_todo_section = False
+    for line in content.split("\n"):
+        # Markdown checkbox: - [ ] something
+        m = re.match(r"\s*[-*]\s*\[\s*[ ]\s*\]\s*(.+)", line)
+        if m:
+            todos.append(m.group(1).strip())
+            continue
+        # Numbered todos under "待办事项" section
+        m2 = re.match(r"\s*\d+\.\s*\*{0,2}(.+?)\*{0,2}\s*$", line)
+        if m2 and in_todo_section:
+            text = m2.group(1).strip()
+            if text and len(text) > 5:
+                todos.append(text)
+        # Track if we're in the todo section
+        if "待办" in line or "TODO" in line:
+            in_todo_section = True
+        elif line.strip() == "" and in_todo_section:
+            in_todo_section = False
+    return todos
+
+
+def push_todos(todos):
+    try:
+        resp = requests.get(TODO_API, timeout=5)
+        existing = [t["text"] for t in resp.json().get("todos", [])]
+    except:
+        existing = []
+    count = 0
+    for text in todos:
+        if text not in existing:
+            requests.post(TODO_API, json={"text": text, "source": "inspiration"}, timeout=5)
+            count += 1
+    return count
+
+
+if __name__ == "__main__":
+    digests = get_all_digests()
+    print(f"Found {len(digests)} digest files")
+    total = 0
+    for fp in digests:
+        todos = extract_todos(fp)
+        if todos:
+            pushed = push_todos(todos)
+            total += pushed
+            print(f"  {os.path.basename(fp)}: {len(todos)} found, {pushed} new")
+    print(f"\nTotal: {total} new todos pushed")
diff --git a/scripts/fetch_news.py b/scripts/fetch_news.py
index e093b99..f25dbbc 100644
--- a/scripts/fetch_news.py
+++ b/scripts/fetch_news.py
@@ -7,16 +7,18 @@ from datetime import datetime, timezone, timedelta
 OUTPUT = "/var/www/nav/data/news.json"
 tz = timezone(timedelta(hours=8))
 
-# RSS feeds - balanced selection of world news sources
+# RSS feeds - balanced selection of world news sources (Chinese weighted)
 FEEDS = [
     {"name": "BBC World", "url": "http://feeds.bbci.co.uk/news/world/rss.xml", "lang": "en"},
+    {"name": "\u65b0\u6d6a\u65b0\u95fb", "url": "https://rss.sina.com.cn/news/china.xml", "lang": "zh"},
+    {"name": "\u6d8c\u62a5", "url": "https://www.ybai.com/wp-json/wp/v2/posts?per_page=10", "lang": "zh"},
     {"name": "BBC \u4e2d\u6587", "url": "https://www.bbc.com/zhongwen/simp/index.xml", "lang": "zh"},
-    {"name": "Reuters", "url": "https://www.reutersagency.com/feed/", "lang": "en"},
     {"name": "HN", "url": "https://hnrss.org/frontpage?count=10", "lang": "en"},
+    {"name": "Reuters", "url": "https://www.reutersagency.com/feed/", "lang": "en"},
 ]
 
-MAX_PER_FEED = 8
-MAX_TOTAL = 30
+MAX_PER_FEED = 10
+MAX_TOTAL = 40
 
 
 def fetch_news():
diff --git a/scripts/push_todos.py b/scripts/push_todos.py
new file mode 100644
index 0000000..56d2961
--- /dev/null
+++ b/scripts/push_todos.py
@@ -0,0 +1,17 @@
+"""Push 4 key todos from the existing library to todo API."""
+import requests
+
+todos = [
+    "为心理黑箱绘制决策流程图，复盘日线震荡却做多的那一单",
+    "为服务器AI设置即时调用接口，支持命令行快捷提问",
+    "修复Dashboard的初始金额和策略编号显示错误",
+    "为倒计时建立仪式感：在Memos创建告别笔记标签",
+]
+
+for text in todos:
+    r = requests.post("http://127.0.0.1:9001/api/todos",
+        json={"text": text, "source": "inspiration"}, timeout=5)
+    print(f"  {r.status_code}: {text[:40]}")
+
+r = requests.get("http://127.0.0.1:9001/api/todos", timeout=5)
+print(f"\nTotal todos: {len(r.json()['todos'])}")