Update app.py via AI Editor

app.py

@@ -12,6 +12,7 @@ from threading import Lock
 import tempfile
 import shutil
 import uuid
+import re
 
 import google.generativeai as genai
 
@@ -33,7 +34,6 @@ MAX_OUTPUT_TOKENS = 65536
 SESSION_STORE = {}
 
 def get_session_id_from_cookie(cookie_str):
-    # Parse a cookie string like "dash_session=abcd; something=xyz"
     if not cookie_str:
         return None
     for part in cookie_str.split(";"):
@@ -42,13 +42,10 @@ def get_session_id_from_cookie(cookie_str):
     return None
 
 def get_session_id(session_id=None):
-    # Always require session_id as input; never generate unless absent
     if session_id and session_id in SESSION_STORE:
         return session_id
     if session_id:
-        # New session, not yet in store
         return session_id
-    # Defensive fallback: generate a new session_id (should never hit this)
     sid = str(uuid.uuid4())
     return sid
 
@@ -157,6 +154,39 @@ def gemini_generate_content(prompt, file_id=None, chat_input=None, file_ids=None
         logging.error("Error during Gemini generate_content: %s", e)
         return f"Error during Gemini completion: {e}"
 
+def parse_markdown_table(md):
+    lines = md.split('\n')
+    table_lines = []
+    in_table = False
+    for l in lines:
+        if l.strip().startswith('|') and l.strip().endswith('|'):
+            table_lines.append(l.strip())
+            in_table = True
+        elif in_table and l.strip() == '':
+            break
+    if not table_lines:
+        raise ValueError("No markdown table found")
+    header = table_lines[0].strip('|').split('|')
+    header = [h.strip() for h in header]
+    rows = []
+    for l in table_lines[2:]:
+        r = [c.strip() for c in l.strip('|').split('|')]
+        if len(r) == len(header):
+            rows.append(r)
+    df = pd.DataFrame(rows, columns=header)
+    return df
+
+def save_markdown_as_xlsx(md_text, base_filename):
+    try:
+        df = parse_markdown_table(md_text)
+        memf = io.BytesIO()
+        df.to_excel(memf, index=False, engine='xlsxwriter')
+        memf.seek(0)
+        return memf.read()
+    except Exception as e:
+        logging.error(f"Failed to convert markdown to XLSX for {base_filename}: {e}")
+        return None
+
 def save_shredded_as_docx(shredded_text, rfp_filename):
     doc = Document()
     doc.add_heading(f"Shredded Requirements for {rfp_filename}", 0)
@@ -237,11 +267,11 @@ def process_document(sess_data, action, selected_filename=None, chat_input=None,
         prompt += doc_content
         result = gemini_generate_content(prompt, file_id=doc_fileid, chat_input=chat_input)
         if result and not result.startswith("Error"):
-            docx_bytes = save_shredded_as_docx(result, selected_filename)
-            generated_docx_name = f"{os.path.splitext(selected_filename)[0]}_shredded.docx"
-            sess_data["uploaded_documents"][generated_docx_name] = result
-            sess_data["shredded_documents"][generated_docx_name] = docx_bytes
-            return result, generated_docx_name, docx_bytes, None, None
+            xlsx_bytes = save_markdown_as_xlsx(result, selected_filename)
+            generated_xlsx_name = f"{os.path.splitext(selected_filename)[0]}_shredded.xlsx"
+            sess_data["uploaded_documents"][generated_xlsx_name] = result
+            sess_data["shredded_documents"][generated_xlsx_name] = xlsx_bytes
+            return result, generated_xlsx_name, xlsx_bytes, None, None
         else:
             return result, None, None, None, None
 
@@ -268,11 +298,11 @@ def process_document(sess_data, action, selected_filename=None, chat_input=None,
         )
         result = gemini_generate_content(prompt, file_id=None, chat_input=None)
         if result and not result.startswith("Error"):
-            docx_bytes = save_compliance_as_docx(result, selected_filename)
-            compliance_docx_name = f"{os.path.splitext(selected_filename)[0]}_compliance_check.docx"
-            sess_data["uploaded_documents"][compliance_docx_name] = result
-            sess_data["shredded_documents"][compliance_docx_name] = docx_bytes
-            return result, compliance_docx_name, docx_bytes, None, None
+            xlsx_bytes = save_markdown_as_xlsx(result, selected_filename)
+            compliance_xlsx_name = f"{os.path.splitext(selected_filename)[0]}_compliance_check.xlsx"
+            sess_data["uploaded_documents"][compliance_xlsx_name] = result
+            sess_data["shredded_documents"][compliance_xlsx_name] = xlsx_bytes
+            return result, compliance_xlsx_name, xlsx_bytes, None, None
         else:
             return result, None, None, None, None
 
@@ -298,11 +328,11 @@ def process_document(sess_data, action, selected_filename=None, chat_input=None,
         )
         result = gemini_generate_content(prompt, file_id=None, chat_input=None)
         if result and not result.startswith("Error"):
-            docx_bytes = save_virtual_board_as_docx(result, selected_filename)
-            board_docx_name = f"{os.path.splitext(selected_filename)[0]}_evaluation_board.docx"
-            sess_data["uploaded_documents"][board_docx_name] = result
-            sess_data["shredded_documents"][board_docx_name] = docx_bytes
-            return result, board_docx_name, docx_bytes, None, None
+            xlsx_bytes = save_markdown_as_xlsx(result, selected_filename)
+            board_xlsx_name = f"{os.path.splitext(selected_filename)[0]}_evaluation_board.xlsx"
+            sess_data["uploaded_documents"][board_xlsx_name] = result
+            sess_data["shredded_documents"][board_xlsx_name] = xlsx_bytes
+            return result, board_xlsx_name, xlsx_bytes, None, None
         else:
             return result, None, None, None, None
 
@@ -402,12 +432,12 @@ def process_document(sess_data, action, selected_filename=None, chat_input=None,
         prompt += f"\n---\nProposal Document ({selected_proposal_filename}):\n{proposal_text}\n"
         result = gemini_generate_content(prompt, file_id=None, chat_input=chat_input)
         if result and not result.startswith("Error"):
-            loe_docx_name = f"{proposal_base_name}_loe.docx"
-            sess_data["proposals"][loe_docx_name] = result
-            sess_data["proposals_fileid"][loe_docx_name] = None
-            docx_bytes = save_loe_as_docx(result, proposal_base_name)
-            logging.info(f"LOE generated and saved as {loe_docx_name}")
-            return result, None, None, loe_docx_name, docx_bytes
+            loe_xlsx_name = f"{proposal_base_name}_loe.xlsx"
+            sess_data["proposals"][loe_xlsx_name] = result
+            sess_data["proposals_fileid"][loe_xlsx_name] = None
+            xlsx_bytes = save_markdown_as_xlsx(result, proposal_base_name)
+            logging.info(f"LOE generated and saved as {loe_xlsx_name}")
+            return result, None, None, loe_xlsx_name, xlsx_bytes
         else:
             return result, None, None, None, None
 
@@ -417,7 +447,7 @@ def get_documents_list(docdict, shreddedict):
     all_docs = {}
     for filename, text in docdict.items():
         all_docs[filename] = text
-    for filename, docx_bytes in shreddedict.items():
+    for filename, doc_bytes in shreddedict.items():
         if filename not in all_docs:
             all_docs[filename] = None
     if not all_docs:
@@ -425,12 +455,16 @@ def get_documents_list(docdict, shreddedict):
     doc_list = []
     for filename in all_docs:
         truncated = truncate_filename(filename)
-        if filename.lower().endswith('.docx') and filename in shreddedict:
+        ext = filename.lower().split('.')[-1]
+        if ext == "xlsx" and filename in shreddedict:
+            b64 = base64.b64encode(shreddedict[filename]).decode('utf-8')
+            mime = "application/vnd.openxmlformats-officedocument.spreadsheetml.sheet"
+        elif ext == "docx" and filename in shreddedict:
             b64 = base64.b64encode(shreddedict[filename]).decode('utf-8')
             mime = "application/vnd.openxmlformats-officedocument.wordprocessingml.document"
         else:
             content = docdict.get(filename, "")
-            b64 = base64.b64encode(content.encode('utf-8')).decode('utf-8')
+            b64 = base64.b64encode((content.encode('utf-8') if isinstance(content, str) else b"")).decode('utf-8')
             mime = "text/plain"
         download_link = html.A(
             truncated,
@@ -453,14 +487,21 @@ def get_proposals_list(proposaldict):
     doc_list = []
     for filename in proposaldict:
         truncated = truncate_filename(filename)
+        ext = filename.lower().split('.')[-1]
         file_content = proposaldict[filename]
         try:
-            if filename.lower().endswith('_loe.docx'):
+            if ext == "xlsx":
+                from io import BytesIO
+                b64 = base64.b64encode(save_markdown_as_xlsx(file_content, filename)).decode('utf-8')
+                mime = "application/vnd.openxmlformats-officedocument.spreadsheetml.sheet"
+            elif filename.lower().endswith('_loe.docx'):
                 docx_bytes = save_loe_as_docx(file_content, filename)
+                b64 = base64.b64encode(docx_bytes).decode('utf-8')
+                mime = "application/vnd.openxmlformats-officedocument.wordprocessingml.document"
             else:
                 docx_bytes = save_proposal_as_docx(file_content, filename)
-            b64 = base64.b64encode(docx_bytes).decode('utf-8')
-            mime = "application/vnd.openxmlformats-officedocument.wordprocessingml.document"
+                b64 = base64.b64encode(docx_bytes).decode('utf-8')
+                mime = "application/vnd.openxmlformats-officedocument.wordprocessingml.document"
         except Exception:
             b64 = base64.b64encode(file_content.encode('utf-8')).decode('utf-8')
             mime = "text/plain"
@@ -481,9 +522,9 @@ def get_proposals_list(proposaldict):
 
 app.layout = dbc.Container([
     dcc.Store(id='preview-window-state', data='expanded'),
-    dcc.Store(id='session-id-store', storage_type='session'),  # Session ID per browser session
-    html.Div(id='set-session-cookie', style={'display': 'none'}),  # dummy div for JS callback
-    dcc.Location(id='dummy-url', refresh=False),  # For JS init on page load
+    dcc.Store(id='session-id-store', storage_type='session'),
+    html.Div(id='set-session-cookie', style={'display': 'none'}),
+    dcc.Location(id='dummy-url', refresh=False),
     dbc.Row([
         dbc.Col([
             dbc.Card([
@@ -572,6 +613,11 @@ app.layout = dbc.Container([
                         children=html.Div(
                             id="output-preview-container",
                             children=html.Div(id="output-data-upload"),
+                            style={
+                                "height": "300px",  # Adjust this value as needed to triple the original height
+                                "overflowY": "auto",  # Adds vertical scrollbar when content exceeds height
+                                "overflowX": "auto",  # Adds horizontal scrollbar if needed
+                            }
                         ),
                         style={"textAlign": "center"}
                     )
@@ -581,7 +627,6 @@ app.layout = dbc.Container([
     ], style={'marginTop':'20px'})
 ], fluid=True)
 
-# JS callback: Set session cookie and session-id-store on first page load if not present
 app.clientside_callback(
     """
     function(n, dummy_url) {
@@ -662,7 +707,6 @@ def master_callback(
     chat_input, cancel_clicks, preview_window_state,
     session_id
 ):
-    # Always get session_id from dcc.Store, never generate
     sid = get_session_id(session_id)
     sess_data = get_session_data(sid)
     ctx = callback_context
@@ -793,17 +837,17 @@ def master_callback(
         try:
             if triggered_id == "shred-action-btn":
                 action_name = "shred"
-                result, generated_filename, generated_docx_bytes, _, _ = process_document(sess_data, action_name, doc_value, chat_input, uploaded_rfp_decoded_bytes, None)
+                result, generated_filename, generated_xlsx_bytes, _, _ = process_document(sess_data, action_name, doc_value, chat_input, uploaded_rfp_decoded_bytes, None)
                 output_data_upload = dcc.Markdown(result, style={"whiteSpace": "pre-wrap", "wordWrap": "break-word"})
             elif triggered_id == "compliance-action-btn":
                 action_name = "compliance"
-                result, generated_filename, generated_docx_bytes, _, _ = process_document(
+                result, generated_filename, generated_xlsx_bytes, _, _ = process_document(
                     sess_data, action_name, doc_value, chat_input, uploaded_rfp_decoded_bytes, proposal_value
                 )
                 output_data_upload = dcc.Markdown(result, style={"whiteSpace": "pre-wrap", "wordWrap": "break-word"})
             elif triggered_id == "board-action-btn":
                 action_name = "virtual_board"
-                result, generated_filename, generated_docx_bytes, _, _ = process_document(
+                result, generated_filename, generated_xlsx_bytes, _, _ = process_document(
                     sess_data, action_name, doc_value, chat_input, uploaded_rfp_decoded_bytes, proposal_value
                 )
                 output_data_upload = dcc.Markdown(result, style={"whiteSpace": "pre-wrap", "wordWrap": "break-word"})
@@ -822,7 +866,7 @@ def master_callback(
                 output_data_upload = dcc.Markdown(result, style={"whiteSpace": "pre-wrap", "wordWrap": "break-word"})
             elif triggered_id == "loe-action-btn":
                 action_name = "loe"
-                result, _, _, generated_filename, generated_docx_bytes = process_document(
+                result, _, _, generated_filename, generated_xlsx_bytes = process_document(
                     sess_data, action_name, None, chat_input, None, proposal_value
                 )
                 output_data_upload = dcc.Markdown(result, style={"whiteSpace": "pre-wrap", "wordWrap": "break-word"})