Update app.py

app.py

@@ -6,12 +6,7 @@ import openpyxl
 from pdfminer.high_level import extract_text
 import csv
 from pptx import Presentation
-import base64
-from io import BytesIO
-from reportlab.pdfgen import canvas
-from reportlab.lib.pagesizes import letter
-from reportlab.lib.colors import blue, red
-import streamlit.components.v1 as components
+from io import StringIO  # For handling string-based CSV
 
 # Configure Gemini API
 GOOGLE_API_KEY = os.environ.get("GOOGLE_API_KEY")
@@ -39,11 +34,11 @@ def load_and_extract_text(file_path):
         elif file_extension == 'pdf':
             text = extract_text(file_path)
         elif file_extension == 'csv':
-            with open(file_path, 'r', encoding='utf-8') as csvfile:
+            with open(file_path, 'r', encoding='utf-8') as csvfile: # Explicit encoding
                 reader = csv.reader(csvfile)
-                text = '\n'.join([' '.join(row) for row in reader])
+                text = '\n'.join([' '.join(row) for row in reader]) # Join rows with spaces
         elif file_extension == 'txt':
-            with open(file_path, 'r', encoding='utf-8') as txtfile:
+            with open(file_path, 'r', encoding='utf-8') as txtfile: #Explicit encoding
                 text = txtfile.read()
         elif file_extension == 'pptx':
             prs = Presentation(file_path)
@@ -56,8 +51,11 @@ def load_and_extract_text(file_path):
         return f"Error processing file: {e}"
 
 
-def process_document_and_answer(context, question, history=""):
-    prompt = f"Context: {context}\n\nConversation History: {history}\n\nQuestion: {question}\n\nAnswer:"
+def process_document_and_answer(extracted_text, question):
+    if "Error processing file" in extracted_text or "Unsupported file format" in extracted_text:
+        return extracted_text
+
+    prompt = f"Context: {extracted_text}\n\nQuestion: {question}\n\nAnswer:"
 
     try:
         response = model.generate_content(prompt)
@@ -66,110 +64,22 @@ def process_document_and_answer(context, question, history=""):
         return f"Error generating answer: {e}"
 
 
-def create_download_link(val, filename):
-    b64 = base64.b64encode(val)  # val looks like b'...'
-    return f'<a href="data:application/octet-stream;base64,{b64.decode()}" download="{filename}">Download conversation as PDF</a>'
-
-def move_element(element, location="bottom"):
-    js_code = f"""
-        <script>
-        window.onload = function() {{
-            var element = document.getElementById('{element}');
-            if (element) {{
-                element.style.position = 'fixed';
-                element.style.bottom = '0';
-                element.style.left = '0';
-                element.style.width = '100%';
-                element.style.backgroundColor = 'white';
-                element.style.zIndex = '9999';
-                element.style.padding = '10px';
-                element.style.boxShadow = '0px -2px 5px #888888';
-                document.body.appendChild(element);
-            }}
-        }}
-        </script>
-    """
-    components.html(js_code, height=0)
-
 # Streamlit UI
 st.title("Gemini Document Q&A")
 
-# Initialize session state
-if 'conversation_history' not in st.session_state:
-    st.session_state.conversation_history = ""
-
-if 'document_content' not in st.session_state:
-    st.session_state.document_content = None
-
-if 'question_count' not in st.session_state:
-    st.session_state.question_count = 0
-
-if 'display_answer' not in st.session_state:
-    st.session_state.display_answer = False
-
-if 'answers' not in st.session_state:
-    st.session_state.answers = []
-
-
-# File Upload
 uploaded_file = st.file_uploader("Upload a document", type=["docx", "xlsx", "pdf", "csv", "txt", "pptx"])
 
-if uploaded_file is not None:
+question = st.text_input("Ask a question about the document:")
+
+if uploaded_file is not None and question:
     # Save the uploaded file to a temporary location
     temp_file_path = "temp." + uploaded_file.name.split('.')[-1].lower()
     with open(temp_file_path, "wb") as temp_file:
         temp_file.write(uploaded_file.read())
 
-    st.session_state.document_content = load_and_extract_text(temp_file_path)
-    os.remove(temp_file_path)  # Clean up
-
-    if "Error processing file" in st.session_state.document_content:
-        st.error(st.session_state.document_content)
-        st.session_state.document_content = None
+    extracted_text = load_and_extract_text(temp_file_path)
 
-if st.session_state.document_content:
+    os.remove(temp_file_path)  # Clean up the temporary file
 
-    # Display Answers above
-    if st.session_state.answers:
-        for i, answer in enumerate(st.session_state.answers):
-            st.markdown(f"<p style='color:green;'><b>Answer {i+1}:</b> {answer}</p>", unsafe_allow_html=True)
-
-    # Question Input at bottom using javascript to stick at the bottom
-    move_element("question_input")
-    question = st.text_input("**Ask a question (type 'exit' to end):**", key="question_input", value="")
-
-    if question:
-        if question.lower() == "exit":
-            st.write("Ending conversation.")
-            st.stop() # this one is more appropiate to stop
-        else:
-            answer = process_document_and_answer(st.session_state.document_content, question, st.session_state.conversation_history)
-            st.session_state.answers.append(answer)
-            st.session_state.display_answer = True
-
-            # Update conversation history and question count
-            st.session_state.conversation_history += f"\nQuestion: {question}\nAnswer: {answer}"
-            st.session_state.question_count += 1
-            st.rerun() # Rerun to display new question at bottom and answer at top
-
-
-    # PDF Export
-    if st.session_state.conversation_history:
-        pdf_buffer = BytesIO()
-        c = canvas.Canvas(pdf_buffer, pagesize=letter)
-        textobject = c.beginText()
-        textobject.setTextOrigin(10, 730)
-        textobject.setFont("Helvetica", 12)
-
-        for line in st.session_state.conversation_history.split('\n'):
-            textobject.setFillColor(blue)
-            if line.startswith("Answer:"):
-                textobject.setFillColor(red)
-            textobject.textLine(line)
-
-        c.drawText(textobject)
-        c.save()
-        pdf_buffer.seek(0)
-
-        download_link = create_download_link(pdf_buffer.read(), "conversation.pdf")
-        st.markdown(download_link, unsafe_allow_html=True)
+    answer = process_document_and_answer(extracted_text, question)
+    st.write("Answer:", answer)