Upload entityName2link.json

extract_Name2link.py

@@ -0,0 +1,31 @@
+import json
+import re
+
+with open('./entity2id.json') as entity2id, open('./entityName2id.json') as entityName2id:
+    entity2id = json.load(entity2id)
+    entityName2id = json.load(entityName2id)
+
+    id2entity = {k: v for v, k in entity2id.items()}
+    print('entity2id has:', len(entity2id))
+    print('id2entity has:', len(id2entity))
+    print('entityName2id has:', len(entityName2id))
+
+
+    entityName2link = {en : id2entity[entityName2id[en]] for en in entityName2id}
+    json.dump(entityName2link, open('entityName2link.json', 'w'))
+
+    # entityName2link = {}
+    # film_pattern = re.compile(r'/([^/]*\((\d+_)?film\))>')
+    # for link, id in entity2id.items():
+    #     match = film_pattern.search(link)
+    #     if match:
+    #         name = match.group(1).replace('_', ' ')
+    #         name = ' '.join(name.split(' ')).strip(' ')
+    #         film_start = name.rfind('film')
+    #         name = name[:film_start-1] + name[film_start+4:]
+    #         if name[-2] == '(':
+    #             name = name[:-2]
+    #         entityName2link[name] = link
+    # print('entity2link has:', len(entityName2link))
+    # json.dump(entityName2link, open('entity2link.json', 'w'))
+

inspired_unicrs.py

@@ -4,7 +4,7 @@ import datasets
 import re
 
 ENTITY = 'entity'
-ENTITY_PATTERN = r'<entity> {} </entity>'
+ENTITY_PATTERN = r'<entity>{}</entity>'
 
 
 def markup_entity(utt: str, entities: List[str]):