Update app.py via AI Editor

app.py

@@ -33,7 +33,6 @@ uploaded_documents = {}
 uploaded_documents_fileid = {}
 uploaded_proposals = {}
 uploaded_proposals_fileid = {}
-generated_documents = {}
 shredded_documents = {}
 shredded_document = None
 generated_response = None
@@ -146,6 +145,16 @@ def save_proposal_as_docx(proposal_text, base_filename):
     memf.seek(0)
     return memf.read()
 
+def save_compliance_as_docx(compliance_text, rfp_filename):
+    doc = Document()
+    doc.add_heading(f"Compliance Check for {rfp_filename}", 0)
+    for line in compliance_text.split('\n'):
+        doc.add_paragraph(line)
+    memf = io.BytesIO()
+    doc.save(memf)
+    memf.seek(0)
+    return memf.read()
+
 def process_document(action, selected_filename=None, chat_input=None, rfp_decoded_bytes=None, cancel_event=None, stream=True, selected_generated_filename=None):
     global shredded_document, generated_response, stream_buffer
 
@@ -153,7 +162,7 @@ def process_document(action, selected_filename=None, chat_input=None, rfp_decode
 
     doc_content = None
     doc_fileid = None
-    if action in ["shred", "proposal"]:
+    if action in ["shred", "proposal", "compliance"]:
         if selected_filename and selected_filename in uploaded_documents:
             doc_content = uploaded_documents[selected_filename]
             doc_fileid = uploaded_documents_fileid.get(selected_filename)
@@ -258,7 +267,8 @@ def process_document(action, selected_filename=None, chat_input=None, rfp_decode
                 if not cancel_event or not cancel_event.is_set():
                     docx_bytes = save_proposal_as_docx(response, rfp_filename)
                     generated_docx_name = f"{os.path.splitext(rfp_filename)[0]}_proposal.docx"
-                    generated_documents[generated_docx_name] = docx_bytes
+                    uploaded_proposals[generated_docx_name] = response
+                    uploaded_proposals_fileid[generated_docx_name] = None
                     result_holder["text"] = response
                     result_holder["docx_name"] = generated_docx_name
                 else:
@@ -277,29 +287,15 @@ def process_document(action, selected_filename=None, chat_input=None, rfp_decode
         return result_holder["text"], None, result_holder["docx_name"], result_holder["text"]
 
     elif action == 'compliance':
-        # --- COMPLIANCE FUNCTIONALITY ADDED HERE ---
-        # Get selected generated doc (proposal) and selected RFP doc
-        if not selected_generated_filename or selected_generated_filename not in generated_documents:
+        if not selected_generated_filename or selected_generated_filename not in uploaded_proposals:
             return "No generated document selected for compliance.", None, None, None
         if not selected_filename or selected_filename not in uploaded_documents:
             return "No RFP/SOW/PWS/RFI document selected for compliance.", None, None, None
 
-        proposal_docx_bytes = generated_documents[selected_generated_filename]
         rfp_text = uploaded_documents[selected_filename]
+        proposal_text = uploaded_proposals[selected_generated_filename]
         logging.info(f"Compliance check: comparing proposal [{selected_generated_filename}] to RFP [{selected_filename}]")
 
-        # Decode the proposal docx to text (best-effort, using python-docx)
-        proposal_text = ""
-        try:
-            file_stream = io.BytesIO(proposal_docx_bytes)
-            doc = Document(file_stream)
-            proposal_text = "\n".join([para.text for para in doc.paragraphs])
-            if not proposal_text.strip():
-                proposal_text = "[Unable to extract text from proposal docx]"
-        except Exception as e:
-            proposal_text = "[Error decoding proposal docx: {}]".format(str(e))
-            logging.error("Error extracting text from proposal docx: %s", e)
-
         prompt = (
             "You are a proposal compliance expert. Use the following RFP/SOW/PWS/RFI and the generated proposal response. "
             "Compare the proposal to the RFP requirements and win themes. "
@@ -325,6 +321,14 @@ def process_document(action, selected_filename=None, chat_input=None, rfp_decode
                         break
                 result = stream_buffer["preview"]
                 result_holder["text"] = result
+                if not cancel_event or not cancel_event.is_set():
+                    docx_bytes = save_compliance_as_docx(result, selected_filename)
+                    compliance_docx_name = f"{os.path.splitext(selected_filename)[0]}_compliance_check.docx"
+                    uploaded_documents[compliance_docx_name] = result
+                    shredded_documents[compliance_docx_name] = docx_bytes
+                    result_holder["docx_name"] = compliance_docx_name
+                else:
+                    result_holder["docx_name"] = None
                 logging.info("Compliance check completed.")
             except Exception as e:
                 logging.error("Error during compliance check: %s", e)
@@ -334,7 +338,7 @@ def process_document(action, selected_filename=None, chat_input=None, rfp_decode
         t = Thread(target=thread_compliance)
         t.start()
         t.join()
-        return result_holder["text"], None, None, result_holder["text"]
+        return result_holder["text"], None, result_holder["docx_name"], result_holder["text"]
 
     elif action == 'recover':
         return "Recovery not implemented yet.", None, None, None
@@ -350,16 +354,14 @@ def get_uploaded_doc_list(docdict):
     doc_list = []
     for filename in docdict:
         file_content = docdict[filename]
-        if filename.lower().endswith('.docx'):
-            mime = "application/vnd.openxmlformats-officedocument.wordprocessingml.document"
-            b64 = None
-            try:
-                docx_bytes = save_shredded_as_docx(file_content, filename)
+        mime = "application/vnd.openxmlformats-officedocument.wordprocessingml.document" if filename.lower().endswith('.docx') else "text/plain"
+        try:
+            if filename.lower().endswith('.docx') and filename in shredded_documents:
+                docx_bytes = shredded_documents[filename]
                 b64 = base64.b64encode(docx_bytes).decode('utf-8')
-            except Exception:
+            else:
                 b64 = base64.b64encode(file_content.encode('utf-8')).decode('utf-8')
-        else:
-            mime = "text/plain"
+        except Exception:
             b64 = base64.b64encode(file_content.encode('utf-8')).decode('utf-8')
         download_link = html.A(
             filename,
@@ -403,16 +405,11 @@ def get_uploaded_proposal_list(docdict):
     doc_list = []
     for filename in docdict:
         file_content = docdict[filename]
-        if filename.lower().endswith('.docx'):
-            mime = "application/vnd.openxmlformats-officedocument.wordprocessingml.document"
-            b64 = None
-            try:
-                docx_bytes = save_proposal_as_docx(file_content, filename)
-                b64 = base64.b64encode(docx_bytes).decode('utf-8')
-            except Exception:
-                b64 = base64.b64encode(file_content.encode('utf-8')).decode('utf-8')
-        else:
-            mime = "text/plain"
+        mime = "application/vnd.openxmlformats-officedocument.wordprocessingml.document"
+        try:
+            docx_bytes = save_proposal_as_docx(file_content, filename)
+            b64 = base64.b64encode(docx_bytes).decode('utf-8')
+        except Exception:
             b64 = base64.b64encode(file_content.encode('utf-8')).decode('utf-8')
         download_link = html.A(
             filename,
@@ -429,46 +426,6 @@ def get_uploaded_proposal_list(docdict):
         )
     return dbc.ListGroup(doc_list, flush=True)
 
-def get_generated_doc_list(docdict):
-    if not docdict:
-        return html.Div("No generated documents yet.", style={"wordWrap": "break-word"})
-    doc_list = []
-    for filename in docdict:
-        docx_bytes = docdict[filename]
-        b64 = base64.b64encode(docx_bytes).decode('utf-8')
-        download_link = html.A(
-            filename,
-            href=f"data:application/vnd.openxmlformats-officedocument.wordprocessingml.document;base64,{b64}",
-            download=filename,
-            target="_blank",
-            style={"wordWrap": "break-word", "marginRight": "10px", "textDecoration": "underline"}
-        )
-        doc_list.append(
-            dbc.ListGroupItem([
-                download_link,
-                dbc.Button("Delete", id={'type': 'delete-generated-btn', 'index': filename, 'group': 'generated'}, size="sm", color="danger", className="float-end ms-2")
-            ], className="d-flex justify-content-between align-items-center")
-        )
-    return dbc.ListGroup(doc_list, flush=True)
-
-def get_generated_rfp_download_section(selected_value):
-    if not selected_value or selected_value not in generated_documents:
-        return html.Div("No generated document selected.", style={"wordWrap": "break-word"})
-    docx_bytes = generated_documents[selected_value]
-    b64 = base64.b64encode(docx_bytes).decode('utf-8')
-    download_link = html.A(
-        f"Download {selected_value}",
-        href=f"data:application/vnd.openxmlformats-officedocument.wordprocessingml.document;base64,{b64}",
-        download=selected_value,
-        target="_blank",
-        style={"wordWrap": "break-word"}
-    )
-    return html.Div([
-        html.H6("RFP/SOW/PWS/RFI", style={"marginTop": "10px"}),
-        html.Div(download_link, style={"marginBottom": "15px"}),
-        html.Div("Preview not available for docx. Download to view.", style={"wordWrap": "break-word"})
-    ])
-
 app.layout = dbc.Container([
     dcc.Store(id='shred-store', data={'text': None, 'docx_name': None}),
     dcc.Store(id='proposal-store', data={'text': None, 'docx_name': None}),
@@ -507,15 +464,14 @@ app.layout = dbc.Container([
                         multiple=False
                     ),
                     html.Hr(style={"marginTop": "20px", "marginBottom": "10px"}),
-                    html.H6("Shredded Requirements"),
+                    html.H6("Shredded/Compliance Documents"),
                     html.Div(get_shredded_doc_list(shredded_documents), id='shredded-doc-list')
                 ])
             ], className="mb-3"),
-
             dbc.Card([
                 dbc.CardHeader(html.H5("Proposal")),
                 dbc.CardBody([
-                    html.H6("Uploaded Documents"),
+                    html.H6("Generated Proposals"),
                     html.Div(get_uploaded_proposal_list(uploaded_proposals), id='uploaded-proposal-list'),
                     dcc.Dropdown(
                         id='select-proposal-dropdown',
@@ -544,22 +500,6 @@ app.layout = dbc.Container([
                     )
                 ])
             ], className="mb-3"),
-
-            dbc.Card([
-                dbc.CardHeader(html.H5("Generated Documents")),
-                dbc.CardBody([
-                    dcc.Dropdown(
-                        id='select-generated-dropdown',
-                        options=[{'label': fn, 'value': fn} for fn in generated_documents.keys()],
-                        placeholder="Select a generated document",
-                        value=next(iter(generated_documents), None),
-                        style={"marginBottom": "10px"}
-                    ),
-                    html.Div(id='generated-rfp-section'),
-                    html.Hr(style={"marginTop": "20px", "marginBottom": "10px"}),
-                    html.Div(get_generated_doc_list(generated_documents), id='generated-doc-list'),
-                ])
-            ], className="mb-3")
         ], style={'minWidth': '260px', 'width':'30vw','maxWidth':'30vw'}, width=3),
 
         dbc.Col([
@@ -598,10 +538,9 @@ app.layout = dbc.Container([
     Output('select-document-dropdown', 'options'),
     Output('select-document-dropdown', 'value'),
     Output('shredded-doc-list', 'children'),
-    Output('generated-doc-list', 'children'),
-    Output('select-generated-dropdown', 'options'),
-    Output('select-generated-dropdown', 'value'),
-    Output('generated-rfp-section', 'children'),
+    Output('uploaded-proposal-list', 'children'),
+    Output('select-proposal-dropdown', 'options'),
+    Output('select-proposal-dropdown', 'value'),
     Output('stream-status', 'data'),
     Output('stream-interval', 'disabled'),
     [
@@ -616,15 +555,12 @@ app.layout = dbc.Container([
         State('upload-proposal', 'filename'),
         Input({'type': 'delete-proposal-btn', 'index': ALL, 'group': 'proposal'}, 'n_clicks'),
         State('select-proposal-dropdown', 'value'),
-        Input({'type': 'delete-generated-btn', 'index': ALL, 'group': 'generated'}, 'n_clicks'),
-        State('select-generated-dropdown', 'value'),
         Input({'type': 'delete-shredded-btn', 'index': ALL, 'group': 'shredded'}, 'n_clicks'),
         State('shredded-doc-list', 'children'),
-        Input('select-generated-dropdown', 'value'),
+        State('select-proposal-dropdown', 'value'),
         State('chat-input', 'value'),
         State('select-document-dropdown', 'value'),
         State('select-proposal-dropdown', 'value'),
-        State('select-generated-dropdown', 'value'),
         Input('cancel-action-btn', 'n_clicks'),
         Input('stream-interval', 'n_intervals'),
         State('stream-status', 'data')
@@ -635,10 +571,9 @@ def master_callback(
     shred_clicks, proposal_clicks, compliance_clicks,
     rfp_content, rfp_filename, rfp_delete_clicks, selected_doc,
     proposal_content, proposal_filename, proposal_delete_clicks, selected_proposal,
-    generated_delete_clicks, selected_generated,
     shredded_delete_clicks, shredded_doc_children,
-    select_generated_value,
-    chat_input, selected_filename, selected_proposal_dropdown, selected_generated_dropdown_state,
+    select_proposal_dropdown_value,
+    chat_input, selected_filename, selected_proposal_dropdown, selected_proposal_dropdown_state,
     cancel_clicks,
     stream_n_intervals, stream_status
 ):
@@ -659,8 +594,7 @@ def master_callback(
 
     rfp_delete_clicks = safe_get_n_clicks(ctx, 5)
     proposal_delete_clicks = safe_get_n_clicks(ctx, 9)
-    generated_delete_clicks = safe_get_n_clicks(ctx, 11)
-    shredded_delete_clicks = safe_get_n_clicks(ctx, 13)
+    shredded_delete_clicks = safe_get_n_clicks(ctx, 11)
 
     uploaded_rfp_decoded_bytes = None
 
@@ -668,7 +602,7 @@ def master_callback(
 
     if triggered_id == 'stream-interval':
         if not stream_status or not stream_status.get('streaming'):
-            return dash.no_update, dash.no_update, dash.no_update, dash.no_update, dash.no_update, dash.no_update, dash.no_update, dash.no_update, dash.no_update, dash.no_update, dash.no_update, stream_status, True
+            return dash.no_update, dash.no_update, dash.no_update, dash.no_update, dash.no_update, dash.no_update, dash.no_update, dash.no_update, dash.no_update, dash.no_update, stream_status, True
         preview = stream_buffer.get("preview", "")
         still_streaming = gemini_lock.locked()
         if not still_streaming:
@@ -678,20 +612,18 @@ def master_callback(
             doc_value = selected_doc if selected_doc in uploaded_documents else (next(iter(uploaded_documents), None) if uploaded_documents else None)
             shredded_doc_list_items = get_shredded_doc_list(shredded_documents)
             uploaded_doc_list = get_uploaded_doc_list(uploaded_documents)
-            generated_doc_list = get_generated_doc_list(generated_documents)
-            generated_doc_options = [{'label': fn, 'value': fn} for fn in generated_documents.keys()]
-            generated_doc_value = select_generated_value if select_generated_value in generated_documents else (next(iter(generated_documents), None) if generated_documents else None)
-            generated_rfp_section = get_generated_rfp_download_section(generated_doc_value)
+            uploaded_proposal_list = get_uploaded_proposal_list(uploaded_proposals)
+            proposal_options = [{'label': fn, 'value': fn} for fn in uploaded_proposals.keys()]
+            proposal_value = select_proposal_dropdown_value if select_proposal_dropdown_value in uploaded_proposals else (next(iter(uploaded_proposals), None) if uploaded_proposals else None)
             return (
                 shred_store, proposal_store, output_data_upload,
                 uploaded_doc_list, doc_options, doc_value,
                 shredded_doc_list_items,
-                generated_doc_list, generated_doc_options, generated_doc_value,
-                generated_rfp_section,
+                uploaded_proposal_list, proposal_options, proposal_value,
                 stream_status,
                 True
             )
-        return dash.no_update, dash.no_update, dcc.Markdown(preview, style={"whiteSpace": "pre-wrap", "wordWrap": "break-word"}), dash.no_update, dash.no_update, dash.no_update, dash.no_update, dash.no_update, dash.no_update, dash.no_update, dash.no_update, stream_status, False
+        return dash.no_update, dash.no_update, dcc.Markdown(preview, style={"whiteSpace": "pre-wrap", "wordWrap": "break-word"}), dash.no_update, dash.no_update, dash.no_update, dash.no_update, dash.no_update, dash.no_update, dash.no_update, stream_status, False
 
     if triggered_id == 'cancel-action-btn':
         stream_event.set()
@@ -701,16 +633,14 @@ def master_callback(
         doc_value = selected_doc if selected_doc in uploaded_documents else (next(iter(uploaded_documents), None) if uploaded_documents else None)
         shredded_doc_list_items = get_shredded_doc_list(shredded_documents)
         uploaded_doc_list = get_uploaded_doc_list(uploaded_documents)
-        generated_doc_list = get_generated_doc_list(generated_documents)
-        generated_doc_options = [{'label': fn, 'value': fn} for fn in generated_documents.keys()]
-        generated_doc_value = select_generated_value if select_generated_value in generated_documents else (next(iter(generated_documents), None) if generated_documents else None)
-        generated_rfp_section = get_generated_rfp_download_section(generated_doc_value)
+        uploaded_proposal_list = get_uploaded_proposal_list(uploaded_proposals)
+        proposal_options = [{'label': fn, 'value': fn} for fn in uploaded_proposals.keys()]
+        proposal_value = select_proposal_dropdown_value if select_proposal_dropdown_value in uploaded_proposals else (next(iter(uploaded_proposals), None) if uploaded_proposals else None)
         return (
             shred_store, proposal_store, output_data_upload,
             uploaded_doc_list, doc_options, doc_value,
             shredded_doc_list_items,
-            generated_doc_list, generated_doc_options, generated_doc_value,
-            generated_rfp_section,
+            uploaded_proposal_list, proposal_options, proposal_value,
             {'streaming': False},
             True
         )
@@ -786,23 +716,10 @@ def master_callback(
                     upload_triggered = True
                     break
 
-    if triggered_id and isinstance(generated_delete_clicks, list):
-        for i, n_click in enumerate(generated_delete_clicks):
-            if n_click:
-                btn_id = ctx.inputs_list[11][i]['id']
-                del_filename = btn_id['index']
-                if del_filename in generated_documents:
-                    del generated_documents[del_filename]
-                    logging.info(f"Generated doc deleted: {del_filename}")
-                    if selected_generated == del_filename:
-                        selected_generated = None
-                    upload_triggered = True
-                    break
-
     if triggered_id and isinstance(shredded_delete_clicks, list):
         for i, n_click in enumerate(shredded_delete_clicks):
             if n_click:
-                btn_id = ctx.inputs_list[13][i]['id']
+                btn_id = ctx.inputs_list[11][i]['id']
                 del_filename = btn_id['index']
                 if del_filename in shredded_documents:
                     del shredded_documents[del_filename]
@@ -814,10 +731,9 @@ def master_callback(
     doc_value = selected_doc if selected_doc in uploaded_documents else (next(iter(uploaded_documents), None) if uploaded_documents else None)
     shredded_doc_list_items = get_shredded_doc_list(shredded_documents)
     uploaded_doc_list = get_uploaded_doc_list(uploaded_documents)
-    generated_doc_list = get_generated_doc_list(generated_documents)
-    generated_doc_options = [{'label': fn, 'value': fn} for fn in generated_documents.keys()]
-    generated_doc_value = select_generated_value if select_generated_value in generated_documents else (next(iter(generated_documents), None) if generated_documents else None)
-    generated_rfp_section = get_generated_rfp_download_section(generated_doc_value)
+    uploaded_proposal_list = get_uploaded_proposal_list(uploaded_proposals)
+    proposal_options = [{'label': fn, 'value': fn} for fn in uploaded_proposals.keys()]
+    proposal_value = select_proposal_dropdown_value if select_proposal_dropdown_value in uploaded_proposals else (next(iter(uploaded_proposals), None) if uploaded_proposals else None)
 
     output_data_upload = html.Div("No action taken yet.", style={"wordWrap": "break-word"})
 
@@ -829,8 +745,7 @@ def master_callback(
                 shred_store, proposal_store, output_data_upload,
                 uploaded_doc_list, doc_options, doc_value,
                 shredded_doc_list_items,
-                generated_doc_list, generated_doc_options, generated_doc_value,
-                generated_rfp_section,
+                uploaded_proposal_list, proposal_options, proposal_value,
                 {'streaming': False},
                 True
             )
@@ -843,56 +758,41 @@ def master_callback(
             finally:
                 gemini_lock.release()
         action_name = "shred" if triggered_id=="shred-action-btn" else ("proposal" if triggered_id=="proposal-action-btn" else "compliance")
-        t = Thread(target=stream_gemini_thread, args=(action_name, doc_value, chat_input, uploaded_rfp_decoded_bytes, generated_doc_value))
+        # For compliance, select_proposal_dropdown_value is the proposal doc
+        t = Thread(target=stream_gemini_thread, args=(action_name, doc_value, chat_input, uploaded_rfp_decoded_bytes, select_proposal_dropdown_value))
         t.daemon = True
         t.start()
         output_data_upload = dcc.Markdown("Starting Gemini operation...", style={"whiteSpace": "pre-wrap", "wordWrap": "break-word"})
+        # After action is started, future stream-interval will update the lists
         return (
             shred_store, proposal_store, output_data_upload,
             uploaded_doc_list, doc_options, doc_value,
             shredded_doc_list_items,
-            generated_doc_list, generated_doc_options, generated_doc_value,
-            generated_rfp_section,
+            uploaded_proposal_list, proposal_options, proposal_value,
             {'streaming': True},
             False
         )
 
-    if triggered_id == 'select-generated-dropdown':
-        generated_rfp_section = get_generated_rfp_download_section(select_generated_value)
-        return (
-            shred_store, proposal_store, output_data_upload,
-            uploaded_doc_list, doc_options, doc_value,
-            shredded_doc_list_items,
-            generated_doc_list, generated_doc_options, generated_doc_value,
-            generated_rfp_section,
-            {'streaming': False},
-            True
-        )
-
     if upload_triggered:
         doc_value = doc_value if doc_value in uploaded_documents else (next(iter(uploaded_documents), None) if uploaded_documents else None)
-        generated_doc_value = generated_doc_value if generated_doc_value in generated_documents else (next(iter(generated_documents), None) if generated_documents else None)
-        generated_rfp_section = get_generated_rfp_download_section(generated_doc_value)
+        proposal_value = proposal_value if proposal_value in uploaded_proposals else (next(iter(uploaded_proposals), None) if uploaded_proposals else None)
         output_data_upload = html.Div("Upload/Delete completed.", style={"wordWrap": "break-word"})
         return (
             shred_store, proposal_store, output_data_upload,
             uploaded_doc_list, doc_options, doc_value,
             shredded_doc_list_items,
-            generated_doc_list, generated_doc_options, generated_doc_value,
-            generated_rfp_section,
+            uploaded_proposal_list, proposal_options, proposal_value,
             {'streaming': False},
             True
         )
 
     doc_value = doc_value if doc_value in uploaded_documents else (next(iter(uploaded_documents), None) if uploaded_documents else None)
-    generated_doc_value = generated_doc_value if generated_doc_value in generated_documents else (next(iter(generated_documents), None) if generated_documents else None)
-    generated_rfp_section = get_generated_rfp_download_section(generated_doc_value)
+    proposal_value = proposal_value if proposal_value in uploaded_proposals else (next(iter(uploaded_proposals), None) if uploaded_proposals else None)
     return (
         shred_store, proposal_store, output_data_upload,
         uploaded_doc_list, doc_options, doc_value,
         shredded_doc_list_items,
-        generated_doc_list, generated_doc_options, generated_doc_value,
-        generated_rfp_section,
+        uploaded_proposal_list, proposal_options, proposal_value,
         {'streaming': False},
         True
     )