Update app.py

app.py

@@ -30,33 +30,70 @@ def process_document(contents: str, filename: str) -> str:
     decoded = base64.b64decode(content_string)
     
     try:
-        if filename.lower().endswith('.pdf'):
-            pdf = PdfReader(io.BytesIO(decoded))
+        if filename.lower().endswith('.docx'):
+            doc = Document(BytesIO(decoded))
+            text = "\n".join([para.text for para in doc.paragraphs])
+            return text
+        elif filename.lower().endswith('.pdf'):
+            pdf = PdfReader(BytesIO(decoded))
             text = ""
             for page in pdf.pages:
                 text += page.extract_text()
-        elif filename.lower().endswith('.docx'):
-            doc = Document(io.BytesIO(decoded))
-            text = "\n".join([para.text for para in doc.paragraphs])
+            return text
         else:
             return f"Unsupported file format: {filename}. Please upload a PDF or DOCX file."
-        
-        if not text.strip():
-            return "The document appears to be empty. Please check the file and try again."
-        
-        return text
     except Exception as e:
         return f"Error processing document: {str(e)}"
 
 def generate_loe(document: str, is_file: bool = False, filename: str = "") -> Tuple[str, pd.DataFrame]:
     if is_file:
-        # Process the uploaded document
         document_text = process_document(document, filename)
         if document_text.startswith("Unsupported file format") or document_text.startswith("Error processing document"):
             return document_text, pd.DataFrame()
     else:
         document_text = document
 
+    prompt = f"""
+    Analyze the following document and provide a Level of Effort (LOE) breakdown:
+
+    Document:
+    {document_text}
+
+    For each section header in the document:
+    1. Identify the tasks to be completed
+    2. Determine the appropriate labor categories for each task
+    3. Estimate the number of hours required for each labor category to complete the task
+
+    Provide a detailed breakdown and then summarize the information in a tabular format with the following columns:
+    - Task Summary
+    - Labor Categories
+    - Hours per Labor Category
+    - Total Hours
+
+    Present the detailed breakdown first, followed by the summary table.
+    Ensure the table is properly formatted with | as column separators and a header row.
+    """
+    response = model.generate_content(prompt)
+    response_text = response.text
+    
+    # Extract the table from the response
+    table_start = response_text.find("| Task Summary |")
+    table_end = response_text.find("\n\n", table_start)
+    table_text = response_text[table_start:table_end]
+    
+    # Convert the table to a pandas DataFrame
+    try:
+        if not table_text.strip():
+            raise pd.errors.EmptyDataError("No table found in the response")
+        df = pd.read_csv(StringIO(table_text), sep='|', skipinitialspace=True).dropna(axis=1, how='all')
+        df.columns = df.columns.str.strip()
+    except pd.errors.EmptyDataError:
+        # If no table is found or it's empty, create a default DataFrame
+        df = pd.DataFrame(columns=['Task Summary', 'Labor Categories', 'Hours per Labor Category', 'Total Hours'])
+        response_text += "\n\nNote: No detailed LOE table could be generated from the AI response."
+    
+    return response_text, df
+
 def generate_outline(text: str, instructions: str) -> str:
     prompt = f"""
     Analyze the following Project Work Statement (PWS) and create an outline 
@@ -659,21 +696,19 @@ def update_loe_output(n_clicks, upload_contents, upload_filename, shred_output):
             else:
                 return "Please upload a document or complete the Shred tab first."
         
-            if isinstance(loe_text, str) and loe_text.startswith(("Unsupported file format", "Error processing document", "The document appears to be empty")):
-                return loe_text
+        if isinstance(loe_text, str) and loe_text.startswith(("Unsupported file format", "Error processing document")):
+            return loe_text
         
-            return [
-                dcc.Markdown(loe_text),
-                dash_table.DataTable(
-                    data=loe_df.to_dict('records'),
-                    columns=[{'name': i, 'id': i} for i in loe_df.columns],
-                    style_table={'overflowX': 'auto'},
-                    style_cell={'textAlign': 'left', 'padding': '5px'},
-                    style_header={'backgroundColor': 'rgb(230, 230, 230)', 'fontWeight': 'bold'}
-                )
-            ]
-        else:
-            return dash.no_update
+        return [
+            dcc.Markdown(loe_text),
+            dash_table.DataTable(
+                data=loe_df.to_dict('records'),
+                columns=[{'name': i, 'id': i} for i in loe_df.columns],
+                style_table={'overflowX': 'auto'},
+                style_cell={'textAlign': 'left', 'padding': '5px'},
+                style_header={'backgroundColor': 'rgb(230, 230, 230)', 'fontWeight': 'bold'}
+            )
+        ]
     except Exception as e:
         return f"An error occurred: {str(e)}"