chinese papers

yindaheng98 · May 18, 2024 · 4cde227 · 4cde227
1 parent c2fc27e
commit 4cde227
Show file tree

Hide file tree

Showing 3 changed files with 4 additions and 5 deletions.
diff --git a/citation_crawler/graph.py b/citation_crawler/graph.py
@@ -1,8 +1,7 @@
 import abc
 import logging
-import asyncio
 from tqdm.asyncio import tqdm
-from typing import Tuple, Optional, AsyncIterable, List
+from typing import Tuple, Optional, AsyncIterable, List, Dict
 import random
 from dblp_crawler.gather import gather
 from .items import Paper
@@ -150,7 +149,7 @@ async def _init_papers(self):
         async for paper, news in tqdm(gather(*tasks), desc="Writing init papers", total=len(tasks)):
             yield paper, news
 
-    async def _bfs_once(self) -> int:
+    async def _bfs_once(self):
         # 初始化
         if not self.inited:
             async for paper, news in self._init_papers():

diff --git a/citation_crawler/items.py b/citation_crawler/items.py
@@ -44,7 +44,7 @@ def title(self) -> str:
         return None
 
     def title_hash(self) -> str:
-        return re.sub(r"[^0-9a-z]", "", self.title().lower())
+        return re.sub(r"[^0-9a-z\u4e00-\u9fa5]", "", self.title().lower())
 
     @abc.abstractmethod
     def year(self) -> Optional[int]:

diff --git a/setup.py b/setup.py
@@ -15,7 +15,7 @@
 
 setup(
     name='citation_crawler',
-    version='2.10.1',
+    version='2.10.2',
     author='yindaheng98',
     author_email='[email protected]',
     url='https://github.com/yindaheng98/citation-crawler',