remove header from pdf_text before checking for quotes

app.py

@@ -22,12 +22,15 @@ anthropic_api_key = os.environ["ANTHROPIC_API_KEY"]
 client_anthropic = Anthropic(api_key=anthropic_api_key)
 
 def parse_quotes(input_string, pdf_text):
+
+    pdf_text_wo_header = re.sub(r'(?m)^Under review as a conference paper at ICLR 2024*$', '', pdf_text)
+    
     # Find all matches of <quote>...</quote> and extract the content between the tags
     matches = re.findall(r'<quote>(.*?)</quote>', input_string)
     
     count = len(matches)
     extracted_texts = matches
-    match_count = sum(1 for text in extracted_texts if text in pdf_text)
+    match_count = sum(1 for text in extracted_texts if text in pdf_text_wo_header)
     
     return count, match_count
 
@@ -503,7 +506,7 @@ if user_input:
         # Column 4: Display revised feedback (from critic)
         with col4:
             st.subheader('Revised Feedback')
-            st.write(f"Quotes found verbatim in pdf text: {match_count}/{count}")
+            if count > 0: st.write(f"Quotes found verbatim in pdf text: {match_count}/{count}")
             st.write(revised_feedback)
     else:
 
@@ -527,7 +530,7 @@ if user_input:
         # Column 3: Display revised feedback (from critic)
         with col3:
             st.subheader('Revised Feedback')
-            st.write(f"Quotes found verbatim in pdf text: {match_count}/{count}")
+            if count > 0: st.write(f"Quotes found verbatim in pdf text: {match_count}/{count}")
             st.write(revised_feedback)
 
 else: