From 99798e900850863ce6337c2a0dbfd22a70cd24b3 Mon Sep 17 00:00:00 2001 From: yindaheng98 Date: Wed, 4 Sep 2024 20:22:30 -0700 Subject: [PATCH] update title hash --- citation_crawler/items.py | 3 ++- setup.py | 4 ++-- 2 files changed, 4 insertions(+), 3 deletions(-) diff --git a/citation_crawler/items.py b/citation_crawler/items.py index 76da97f..8d64aca 100644 --- a/citation_crawler/items.py +++ b/citation_crawler/items.py @@ -2,6 +2,7 @@ import re from typing import Optional, Iterable, Tuple from typing_extensions import Self +from dblp_crawler.parser import title_hash class Author(metaclass=abc.ABCMeta): @@ -44,7 +45,7 @@ def title(self) -> str: return None def title_hash(self) -> str: - return re.sub(r"[^0-9a-z\u4e00-\u9fa5]", "", self.title().lower()) + return title_hash(self.title()) @abc.abstractmethod def year(self) -> Optional[int]: diff --git a/setup.py b/setup.py index ee720ac..c277af1 100644 --- a/setup.py +++ b/setup.py @@ -15,7 +15,7 @@ setup( name='citation_crawler', - version='2.10.2', + version='2.10.3', author='yindaheng98', author_email='yindaheng98@gmail.com', url='https://github.com/yindaheng98/citation-crawler', @@ -30,7 +30,7 @@ "Operating System :: OS Independent", ], install_requires=[ - 'dblp-crawler>=2.1.4', + 'dblp-crawler>=2.1.6', 'python-dateutil>=2.8.2', 'neo4j>=5.15.0', 'typing-extensions'