Update main.py

main.py

@@ -75,7 +75,7 @@ def preprocess(query: str, k: int) -> tuple[str, str]:
         "ANSWER:\n"
     )
 
-    references = []
+    references = {}
     research_abstracts = ""
 
     for i in range(k):
@@ -86,18 +86,21 @@ def preprocess(query: str, k: int) -> tuple[str, str]:
         text = top_five["text"].values[i]
 
         research_abstracts += str(i + i) + ". This excerpt is from: '" + title + "':\n" + text + "\n"
-        references.append(
-            "["
+        header = "["
             + title.title()
             + "]("
             + url
             + ")\n"
-        )
+
+        if header not in references.keys():
+            references[header] = []
+
+        references[header].append(text)
         
     prompt = prompt.replace("{{ABSTRACTS_GO_HERE}}", research_abstracts)
     prompt = prompt.replace("{{QUERY_GOES_HERE}}", query)
 
-    return prompt, "\n\n### References\n\n"+"\n".join([str(i+1)+". "+ref for i, ref in enumerate(list(set(references)))])
+    return prompt, "\n\n### References\n\n"+"\n".join([str(i+1)+". "+ref for i, ref in references.keys()])
     
 def postprocess(response: str, bypass_from_preprocessing: str) -> str:
     """