Update app.py

app.py

@@ -8,6 +8,9 @@ import openai
 import os
 import time
 from dash.exceptions import PreventUpdate
+import PyPDF2
+import docx
+import chardet
 
 # Initialize the Dash app
 app = dash.Dash(__name__, external_stylesheets=[dbc.themes.BOOTSTRAP])
@@ -16,7 +19,7 @@ app = dash.Dash(__name__, external_stylesheets=[dbc.themes.BOOTSTRAP])
 openai.api_key = os.environ.get('OPENAI_API_KEY')
 
 # Global variables
-uploaded_files = []
+uploaded_files = {}
 current_matrix = None
 matrix_type = None
 
@@ -92,6 +95,27 @@ app.layout = dbc.Container([
     ])
 ], fluid=True)
 
+def parse_file_content(contents, filename):
+    content_type, content_string = contents.split(',')
+    decoded = base64.b64decode(content_string)
+    try:
+        if filename.endswith('.pdf'):
+            with io.BytesIO(decoded) as pdf_file:
+                reader = PyPDF2.PdfReader(pdf_file)
+                return ' '.join([page.extract_text() for page in reader.pages])
+        elif filename.endswith('.docx'):
+            with io.BytesIO(decoded) as docx_file:
+                doc = docx.Document(docx_file)
+                return ' '.join([para.text for para in doc.paragraphs])
+        elif filename.endswith('.txt') or filename.endswith('.rtf'):
+            encoding = chardet.detect(decoded)['encoding']
+            return decoded.decode(encoding)
+        else:
+            return "Unsupported file format"
+    except Exception as e:
+        print(f"Error processing file {filename}: {str(e)}")
+        return "Error processing file"
+
 @app.callback(
     Output('file-list', 'children'),
     Input('upload-files', 'contents'),
@@ -99,18 +123,57 @@ app.layout = dbc.Container([
     State('file-list', 'children')
 )
 def update_output(list_of_contents, list_of_names, existing_files):
+    global uploaded_files
     if list_of_contents is not None:
-        new_files = [
-            html.Div([
-                html.Button('×', id={'type': 'remove-file', 'index': i}, style={'marginRight': '5px', 'fontSize': '10px'}),
-                html.Span(filename)
-            ]) for i, filename in enumerate(list_of_names)
-        ]
+        new_files = []
+        for i, (content, name) in enumerate(zip(list_of_contents, list_of_names)):
+            file_content = parse_file_content(content, name)
+            uploaded_files[name] = file_content
+            new_files.append(html.Div([
+                html.Button('×', id={'type': 'remove-file', 'index': name}, style={'marginRight': '5px', 'fontSize': '10px'}),
+                html.Span(name)
+            ]))
         if existing_files is None:
             existing_files = []
         return existing_files + new_files
     return existing_files
 
+@app.callback(
+    Output('file-list', 'children', allow_duplicate=True),
+    Input({'type': 'remove-file', 'index': dash.ALL}, 'n_clicks'),
+    State('file-list', 'children'),
+    prevent_initial_call=True
+)
+def remove_file(n_clicks, existing_files):
+    global uploaded_files
+    ctx = dash.callback_context
+    if not ctx.triggered:
+        raise PreventUpdate
+    removed_file = ctx.triggered[0]['prop_id'].split(',')[0].split(':')[-1].strip('}')
+    uploaded_files.pop(removed_file, None)
+    return [file for file in existing_files if file['props']['children'][1]['props']['children'] != removed_file]
+
+def generate_matrix_with_gpt(matrix_type, file_contents):
+    prompt = f"Generate a {matrix_type} based on the following project artifacts:\n\n"
+    prompt += "\n\n".join(file_contents)
+    prompt += f"\n\nCreate a {matrix_type} in a format that can be represented as a pandas DataFrame."
+
+    response = openai.ChatCompletion.create(
+        model="gpt-3.5-turbo",
+        messages=[
+            {"role": "system", "content": "You are a helpful assistant that generates project management matrices."},
+            {"role": "user", "content": prompt}
+        ]
+    )
+
+    matrix_text = response.choices[0].message.content
+    # Parse the matrix_text into a pandas DataFrame
+    # This is a simplified parsing, you might need to adjust based on the actual output format
+    lines = matrix_text.strip().split('\n')
+    headers = lines[0].split('|')
+    data = [line.split('|') for line in lines[2:]]
+    return pd.DataFrame(data, columns=headers)
+
 @app.callback(
     Output('matrix-preview', 'children'),
     Output('loading-output', 'children'),
@@ -125,13 +188,12 @@ def generate_matrix(*args):
     button_id = ctx.triggered[0]['prop_id'].split('.')[0]
     matrix_type = button_id.replace('btn-', '').replace('-', ' ').title()
     
-    # Simulate matrix generation (replace with actual OpenAI call)
-    time.sleep(2)  # Simulating API call delay
-    current_matrix = pd.DataFrame({
-        'Column 1': ['Row 1', 'Row 2', 'Row 3'],
-        'Column 2': ['Data 1', 'Data 2', 'Data 3'],
-        'Column 3': ['Info 1', 'Info 2', 'Info 3']
-    })
+    if not uploaded_files:
+        return html.Div("Please upload project artifacts before generating a matrix."), ""
+
+    file_contents = list(uploaded_files.values())
+    
+    current_matrix = generate_matrix_with_gpt(matrix_type, file_contents)
     
     return dbc.Table.from_dataframe(current_matrix, striped=True, bordered=True, hover=True), f"{matrix_type} generated"
 
@@ -144,15 +206,29 @@ def generate_matrix(*args):
 )
 def update_matrix_via_chat(n_clicks, chat_input):
     global current_matrix
-    if not chat_input:
+    if not chat_input or current_matrix is None:
         raise PreventUpdate
     
-    # Simulate chat interaction and matrix update (replace with actual OpenAI call)
-    time.sleep(1)  # Simulating API call delay
-    chat_response = f"Updated the matrix based on: {chat_input}"
-    current_matrix = current_matrix.append({'Column 1': 'New Row', 'Column 2': 'New Data', 'Column 3': 'New Info'}, ignore_index=True)
+    prompt = f"Update the following {matrix_type} based on this instruction: {chat_input}\n\n"
+    prompt += current_matrix.to_string()
+    
+    response = openai.ChatCompletion.create(
+        model="gpt-3.5-turbo",
+        messages=[
+            {"role": "system", "content": "You are a helpful assistant that updates project management matrices."},
+            {"role": "user", "content": prompt}
+        ]
+    )
+
+    updated_matrix_text = response.choices[0].message.content
+    # Parse the updated_matrix_text into a pandas DataFrame
+    # This is a simplified parsing, you might need to adjust based on the actual output format
+    lines = updated_matrix_text.strip().split('\n')
+    headers = lines[0].split('|')
+    data = [line.split('|') for line in lines[2:]]
+    current_matrix = pd.DataFrame(data, columns=headers)
     
-    return chat_response, dbc.Table.from_dataframe(current_matrix, striped=True, bordered=True, hover=True)
+    return f"Matrix updated based on: {chat_input}", dbc.Table.from_dataframe(current_matrix, striped=True, bordered=True, hover=True)
 
 @app.callback(
     Output("download-matrix", "data"),