Tweak how papers are referenced

corpus_query.py

@@ -130,6 +130,8 @@ def process_query(query, model=llm_ver):
     for sn in response.source_nodes:
         # print(sn)
         sd = sn.metadata["source document"]
+        if 'et_al_' in sd:
+            sd = sd.replace('WormAtlas Handbook:', 'Paper: ')
 
         if sd not in files_used:
             if len(files_used) == 0 or sn.score >= cutoff:
@@ -158,7 +160,10 @@ def run_query(query):
         v = metadata[k]
         if SOURCE_DOCUMENT in v:
             if v[SOURCE_DOCUMENT] not in files_used:
-                files_used.append(v[SOURCE_DOCUMENT])
+                sd = v[SOURCE_DOCUMENT]
+                if 'et_al_' in sd:
+                    sd = sd.replace('WormAtlas Handbook:', 'Paper: ')
+                files_used.append(sd)
 
     srcs = '\n- '.join(files_used)
     answer = f"""{response_text}