cache & faster

yindaheng98 · Dec 15, 2023 · 08031d5 · 08031d5
1 parent 79e376e
commit 08031d5
Show file tree

Hide file tree

Showing 2 changed files with 8 additions and 2 deletions.
diff --git a/dblp_crawler/summarizer/neo4j.py b/dblp_crawler/summarizer/neo4j.py
@@ -54,8 +54,14 @@ def add_person(tx, person: DBLPPerson, added_pubs: set, added_journals: set):
            pid=person.pid(), name=person.name(),
            aff=list(person.person().affiliations()),
            orcid=orcid)
+    exist_write_papers = set([title_hash for (title_hash, ) in tx.run(
+        "MATCH (a:Person {dblp_pid: $pid})-[:WRITE]->(p:Publication) RETURN p.title_hash",
+        pid=person.pid()
+    ).values()])
     for publication in person.publications():
-        tx.run("MERGE (a:Person {dblp_pid: $pid}) "
+        if publication.title_hash() in exist_write_papers:
+            continue
+        tx.run("MATCH (a:Person {dblp_pid: $pid}) "
                "MERGE (p:Publication {title_hash: $title_hash}) "
                "MERGE (a)-[:WRITE]->(p)",
                pid=person.pid(),

diff --git a/setup.py b/setup.py
@@ -17,7 +17,7 @@
 
 setup(
     name='dblp_crawler',
-    version='1.8.9',
+    version='1.8.10',
     author='yindaheng98',
     author_email='[email protected]',
     url='https://github.com/yindaheng98/dblp-crawler',