fix: label cache id error

ai/classify_paper.py

@@ -158,12 +158,12 @@ class PaperCache:
         self.lock = Lock()
 
     def get(self, paper):
-        key = (paper['title'], paper.get('abstract'))
+        key = paper["id"]
         with self.lock:
             return self.cache.get(key)
 
     def set(self, paper, result):
-        key = (paper['title'], paper.get('abstract'))
+        key = paper["id"]
         with self.lock:
             self.cache[key] = result
 
@@ -188,6 +188,9 @@ def classify_papers(papers: List[Dict[str, str]]) -> Optional[List[Dict[str, Lis
         else:
             uncached_papers.append(paper)
 
+    # 输出 cache 的数量
+    print(f"Cache hit: {len(cached_results)}, Cache miss: {len(uncached_papers)}")
+
     if not uncached_papers:
         return cached_results
 

fetch_paper.py

@@ -106,6 +106,7 @@ def fetch_papers_with_daterange(start_date: Date, end_date: Date):
     preprocessed_articles = list(map(lambda article: {
         "title": article.title,
         "abstract": article.paper.summary,
+        "id": article.paper.id
     }, unique_articles))
 
     classified_articles = classify_papers(preprocessed_articles)