Update app.py

app.py

@@ -35,7 +35,6 @@ def make_llm_api_call(prompt):
     )
     return message
 
-
 def get_llm_response(extractedtext1, extractedtext2):
     prompt = DIFFERENTIATE_PROMPT.format(text1=extractedtext1, text2=extractedtext2)
     
@@ -69,6 +68,8 @@ def main():
     st.set_page_config(layout="wide")  # Enable wide layout
     if "differences_data" not in  st.session_state:
         st.session_state.differences_data = []
+    if "display_data" not in st.session_state:
+        st.session_state.display_data = {"file1": None, "file2": None, "i": 0}
     if "file1" not in st.session_state:
         st.session_state.file1 = None
     if "file2" not in st.session_state:
@@ -91,6 +92,9 @@ def main():
         filename1 = st.session_state.file1.name
         filename2 = st.session_state.file2.name
         
+        if st.session_state.display_data["file1"] != st.session_state.file1 or st.session_state.display_data["file2"] != st.session_state.file2:
+            st.session_state.display_data = {"file1": st.session_state.file1, "file2": st.session_state.file2, "i": 0}
+        
         try:
             extracted_text1 = extract_text_with_pypdf(st.session_state.file1)
             extracted_text2 = extract_text_with_pypdf(st.session_state.file2)
@@ -103,48 +107,52 @@ def main():
             st.success(f"Content of files **{filename1}** and **{filename2}** have been extracted successfully.")
         except Exception as e:
             st.error(f"Error saving files: {str(e)}")
+        
+        try:
+            # display_text, parsed_data = process_concurrently(extracted_text1, extracted_text2)
+            # display_text, parsed_data = get_llm_response(extracted_text1, extracted_text2)
+            
+            for i,(etext1, etext2) in enumerate(zip(extracted_text1, extracted_text2)):
+                if i >= st.session_state.display_data["i"]: break
+                data = next((d for d in st.session_state.differences_data if d['etext1'] == etext1 and d['etext2'] == etext2), None)
+                if data:
+                    pdata = data['pdata']
+                    dext1 = data['dext1']
+                    dext2 = data['dext2']
+                else:
+                    pdata = get_llm_response(etext1, etext2)
+                    dext1 = etext1
+                    dext2 = etext2
+
+                    for diff in pdata:
+                        diff_text1 = diff['text1'].strip()
+                        diff_text2 = diff['text2'].strip()          
+                        if diff_text1 == "" or diff_text2 == "": continue              
+                        diff_text1_phrase = "\n".join([f"<span style='background-color: grey;'>{t}</span>" for t in diff_text1.splitlines()])
+                        diff_text2_phrase = "\n".join([f"<span style='background-color: grey;'>{t}</span>" for t in diff_text2.splitlines()])
+                        dext1 = diff_text1_phrase.join(dext1.split(diff_text1)) if diff_text1 in dext1 else dext1
+                        dext2 = diff_text2_phrase.join(dext2.split(diff_text2)) if diff_text2 in dext2 else dext2
+
+                    st.session_state.differences_data.append({"etext1": etext1, "etext2": etext2, "pdata": pdata, "dext1": dext1, "dext2": dext2})
+                    reverse_pdata = [{'text1': d['text2'], 'text2': d['text1'], 'explanation': d['explanation']} for d in pdata]
+                    st.session_state.differences_data.append({"etext1": etext2, "etext2": etext1, "pdata": reverse_pdata, "dext1": dext2, "dext2": dext1})
+                
+                display_text = "\n\n\n".join([f"**Text1:**\n\n{d['text1']}\n\n**Text2:**\n\n{d['text2']}\n\n**Explanation:**\n\n{d['explanation']}\n\n----------------------\n" for d in pdata])
+
+                with st.expander(f"**Page {i+1}** - {filename1}"):
+                    st.markdown("\n\n".join(dext1.splitlines()), unsafe_allow_html=True)
+                with st.expander(f"**Page {i+1}** - {filename2}"):
+                    st.markdown("\n\n".join(dext2.splitlines()), unsafe_allow_html=True)
+                st.markdown(display_text)
+        except Exception as e:
+            st.error(f"Error finding differences: {str(e)}")
+
 
         # Add button at the bottom to run Find Differences function
         if st.button("Find Differences"):
-            try:
-                # display_text, parsed_data = process_concurrently(extracted_text1, extracted_text2)
-                # display_text, parsed_data = get_llm_response(extracted_text1, extracted_text2)
-                
-                i = 1
-                for etext1, etext2 in zip(extracted_text1, extracted_text2):
-                    data = next((d for d in st.session_state.differences_data if d['etext1'] == etext1 and d['etext2'] == etext2), None)
-                    if data:
-                        pdata = data['pdata']
-                        dext1 = data['dext1']
-                        dext2 = data['dext2']
-                    else:
-                        pdata = get_llm_response(etext1, etext2)
-                        dext1 = etext1
-                        dext2 = etext2
-
-                        for diff in pdata:
-                            diff_text1 = diff['text1'].strip()
-                            diff_text2 = diff['text2'].strip()          
-                            if diff_text1 == "" or diff_text2 == "": continue              
-                            diff_text1_phrase = "\n".join([f"<span style='background-color: grey;'>{t}</span>" for t in diff_text1.splitlines()])
-                            diff_text2_phrase = "\n".join([f"<span style='background-color: grey;'>{t}</span>" for t in diff_text2.splitlines()])
-                            dext1 = diff_text1_phrase.join(dext1.split(diff_text1)) if diff_text1 in dext1 else dext1
-                            dext2 = diff_text2_phrase.join(dext2.split(diff_text2)) if diff_text2 in dext2 else dext2
-
-                        st.session_state.differences_data.append({"etext1": etext1, "etext2": etext2, "pdata": pdata, "dext1": dext1, "dext2": dext2})
-                        reverse_pdata = [{'text1': d['text2'], 'text2': d['text1'], 'explanation': d['explanation']} for d in pdata]
-                        st.session_state.differences_data.append({"etext1": etext2, "etext2": etext1, "pdata": reverse_pdata, "dext1": dext2, "dext2": dext1})
-                    
-                    display_text = "\n\n\n".join([f"**Text1:**\n\n{d['text1']}\n\n**Text2:**\n\n{d['text2']}\n\n**Explanation:**\n\n{d['explanation']}\n\n----------------------\n" for d in pdata])
-
-                    with st.expander(f"**Page {i}** - {filename1}"):
-                        st.markdown("\n\n".join(dext1.splitlines()), unsafe_allow_html=True)
-                    with st.expander(f"**Page {i}** - {filename2}"):
-                        st.markdown("\n\n".join(dext2.splitlines()), unsafe_allow_html=True)
-                    st.markdown(display_text)
-                    i += 1
-            except Exception as e:
-                st.error(f"Error finding differences: {str(e)}")
+            st.session_state.display_data["i"] = st.session_state.display_data["i"] + 5
+            st.rerun()
+
 
 if __name__ == "__main__":
     main()