Update app.py

app.py

@@ -3,7 +3,8 @@ import pandas as pd
 import requests
 import json
 from datetime import datetime
-from PyPDF2 import PdfReader
+from pdfminer.high_level import extract_text
+import regex as re 
 
 st.title("FHIR Converter")
 
@@ -125,31 +126,27 @@ elif conversion_type == "Clinical Notes to FHIR":
     if uploaded_pdf is not None and st.button("Convert"):
         try:
             # Extract text from PDF
-            pdf_reader = PdfReader(uploaded_pdf)
-            paragraphs = []
-            for page_num in range(len(pdf_reader.pages)):
-                page = pdf_reader.pages[page_num]
-                text = page.extract_text()
-                paragraphs.extend(text.split("\n\n"))  # Split by double newline for paragraph separation
-
+            full_text = extract_text(uploaded_pdf)
+            pattern = r'(?<!\n)\n\n(?![@#\$%\^&\*\(\)\[\]\{\};:,\.])'
+            paragraphs = re.split(pattern, full_text)
             # Send paragraphs to API
-            total_paragraphs = len(paragraphs)
+            total_paragraphs = len(paragraphs) - 1
             start_time = datetime.now()
             result_text = ""
 
-            for idx, paragraph in enumerate(paragraphs):
+            for i, paragraph in enumerate(paragraphs, 1):
                 data = {"note_content": paragraph}
                 response = requests.post("https://fhir-api-9jsn.onrender.com/convert_notes", json=data)
 
                 if response.status_code == 200:
                     result_text += response.json().get("output", "") + "\n"
                 else:
-                    st.error(f"An error occurred at paragraph {idx + 1}: {response.json().get('error', 'Unknown error')}")
+                    st.error(f"An error occurred at paragraph {i + 1}: {response.json().get('error', 'Unknown error')}")
                     break
 
                 # Update progress
                 elapsed_time = datetime.now() - start_time
-                progress_bar.progress(int(((idx + 1) / total_paragraphs) * 100))
+                progress_bar.progress(int(((i + 1) / total_paragraphs) * 100))
                 time_placeholder.text(f"Time elapsed: {elapsed_time}")
 
             if result_text: