feat: Chinese news sources, AI summary, todo extraction

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")