Update app.py via AI Editor

app.py

@@ -15,6 +15,8 @@ import openai
 import logging
 import threading
 import re
+import markdown
+from bs4 import BeautifulSoup
 
 logging.basicConfig(level=logging.INFO, format='%(asctime)s %(levelname)s %(message)s')
 
@@ -372,7 +374,7 @@ def generate_any_doc(n_clicks_list, btn_ids, radio_values, upload_contents, uplo
             generated = generate_document(doc_type, file_contents)
             current_document = generated
             shredded_document = generated
-            preview = spreadsheet_preview(generated, doc_type)
+            preview = markdown_table_preview(generated)
             logging.info("Shred document generated.")
             return preview, "Shred generated"
         except Exception as e:
@@ -421,25 +423,25 @@ def generate_any_doc(n_clicks_list, btn_ids, radio_values, upload_contents, uplo
             generated = generate_document(doc_type, [doc_content, shredded_document])
             pink_review_document = generated
             current_document = generated
-            preview = spreadsheet_preview(generated, doc_type)
+            preview = markdown_table_preview(generated)
             logging.info("Pink Review document generated.")
             return preview, f"{doc_type} generated"
         elif doc_type in ["Red Review", "Gold Review", "Virtual Board", "LOE"]:
             generated = generate_document(doc_type, [doc_content, shredded_document])
             current_document = generated
-            preview = spreadsheet_preview(generated, doc_type)
+            preview = markdown_table_preview(generated)
             logging.info(f"{doc_type} document generated.")
             return preview, f"{doc_type} generated"
         elif doc_type in ["Pink", "Red", "Gold"]:
             generated = generate_document(doc_type, [doc_content])
             current_document = generated
-            preview = narrative_preview(generated)
+            preview = markdown_narrative_preview(generated)
             logging.info(f"{doc_type} document generated.")
             return preview, f"{doc_type} generated"
         else:
             generated = generate_document(doc_type, [doc_content])
             current_document = generated
-            preview = narrative_preview(generated)
+            preview = markdown_narrative_preview(generated)
             logging.info(f"{doc_type} document generated.")
             return preview, f"{doc_type} generated"
     except Exception as e:
@@ -461,29 +463,76 @@ def update_uploaded_doc_name(contents, filename, id_dict):
         return filename, contents, "uploaded"
     return "", None, "loaded"
 
-def spreadsheet_preview(text, doc_type=None):
-    if doc_type is None:
-        doc_type = ""
-    try:
-        df = pd.read_csv(StringIO(text))
-        table = DataTable(
-            columns=[{"name": i, "id": i} for i in df.columns],
-            data=df.to_dict('records'),
-            style_table={'overflowX': 'auto', 'maxHeight': '500px', 'overflowY': 'auto'},
-            style_cell={'textAlign': 'left', 'whiteSpace': 'normal', 'height': 'auto', 'minWidth': '120px', 'maxWidth': '320px'},
-            page_size=15
-        )
-        return html.Div(table)
-    except Exception as e:
-        logging.error(f"Error parsing CSV for preview: {str(e)}")
-        return html.Div([
-            html.Div("Unable to display spreadsheet preview. Raw output:", style={'fontWeight': 'bold'}),
-            html.Pre(text, style={'whiteSpace': 'pre-wrap', 'fontFamily': 'monospace'})
-        ])
-
-def narrative_preview(text):
-    plain = strip_markdown(text)
-    return html.Pre(plain, style={'whiteSpace': 'pre-wrap', 'fontFamily': 'sans-serif'})
+def markdown_table_preview(md_text):
+    # Render markdown to HTML, show only tables if present, else show all
+    html_out = markdown.markdown(md_text, extensions=['tables'])
+    soup = BeautifulSoup(html_out, "html.parser")
+    tables = soup.find_all('table')
+    if tables:
+        return html.Div([html.Table(table, style={'width': '100%', 'overflowX': 'auto'}) for table in tables])
+    else:
+        return html.Div(dcc.Markdown(md_text, dangerously_allow_html=True, style={'whiteSpace': 'pre-wrap', 'fontFamily': 'monospace'}))
+
+def markdown_narrative_preview(md_text):
+    return html.Div(dcc.Markdown(md_text, dangerously_allow_html=True, style={'whiteSpace': 'pre-wrap', 'fontFamily': 'sans-serif'}))
+
+def extract_markdown_tables(md_text):
+    # Extract all markdown tables from md_text as list of strings
+    tables = []
+    lines = md_text.split('\n')
+    in_table = False
+    table_lines = []
+    for line in lines:
+        if re.match(r'^\s*\|.*\|\s*$', line):
+            in_table = True
+            table_lines.append(line)
+        elif in_table and (re.match(r'^\s*\|.*\|\s*$', line) or re.match(r'^\s*$', line)):
+            table_lines.append(line)
+        else:
+            if in_table and table_lines:
+                tables.append('\n'.join(table_lines))
+                table_lines = []
+            in_table = False
+    if in_table and table_lines:
+        tables.append('\n'.join(table_lines))
+    return tables
+
+def markdown_table_to_df(md_table):
+    # Convert a single markdown table string to pandas DataFrame
+    lines = [line.strip() for line in md_table.split('\n') if line.strip()]
+    if len(lines) < 2:
+        return None
+    header = [h.strip() for h in lines[0].strip('|').split('|')]
+    # Find the first line with at least 3 dashes, that's the separator
+    sep_idx = 1
+    while sep_idx < len(lines) and not re.match(r'^\|\s*:?-+:?\s*(\|\s*:?-+:?\s*)+\|?$', lines[sep_idx]):
+        sep_idx += 1
+    data_lines = lines[sep_idx+1:] if sep_idx+1 < len(lines) else []
+    rows = []
+    for row in data_lines:
+        if not row.strip() or not row.strip().startswith('|'):
+            continue
+        cells = [c.strip() for c in row.strip('|').split('|')]
+        # Pad or trim cells to header length
+        if len(cells) < len(header):
+            cells += [''] * (len(header) - len(cells))
+        elif len(cells) > len(header):
+            cells = cells[:len(header)]
+        rows.append(cells)
+    df = pd.DataFrame(rows, columns=header)
+    return df
+
+def markdown_tables_to_xlsx(md_text):
+    tables = extract_markdown_tables(md_text)
+    output = BytesIO()
+    with pd.ExcelWriter(output, engine='xlsxwriter') as writer:
+        for i, table in enumerate(tables):
+            df = markdown_table_to_df(table)
+            if df is not None:
+                sheet_name = f"Table{i+1}"
+                df.to_excel(writer, sheet_name=sheet_name, index=False)
+    output.seek(0)
+    return output
 
 def strip_markdown(text):
     text = re.sub(r'(\*\*|__)(.*?)\1', r'\2', text)
@@ -502,11 +551,11 @@ def strip_markdown(text):
 
 def generate_document(document_type, file_contents):
     if document_type in spreadsheet_types:
-        prompt = f"""Ignore all other instructions and output only a spreadsheet for {document_type} as described below. Do not include any narrative, only the spreadsheet in CSV format.
+        prompt = f"""Ignore all other instructions and output only a spreadsheet for {document_type} as described below. Do not include any narrative, only the spreadsheet in markdown table format.
 Instructions: {document_types[document_type]}
 Project Artifacts:
 {' '.join(file_contents)}
-Output only the spreadsheet in CSV format, no narrative or explanation."""
+Output only the spreadsheet as a markdown table, no narrative or explanation."""
     elif document_type in narrative_types:
         prompt = f"""Generate a {document_type} document based on the following project artifacts:
 {' '.join(file_contents)}
@@ -566,11 +615,11 @@ def update_document_via_chat(btn_send, btn_clear, chat_input, selected_doc_type)
 
     if selected_doc_type in spreadsheet_types:
         prompt = f"""Update the following {selected_doc_type} spreadsheet based on this instruction: {chat_input}
-Current spreadsheet (CSV format):
+Current spreadsheet (markdown table format):
 {current_document}
 Instructions:
-1. Provide the updated spreadsheet in CSV format only.
-2. Do not include any narrative, only the spreadsheet in CSV.
+1. Provide the updated spreadsheet as a markdown table only.
+2. Do not include any narrative, only the markdown table.
 Now, provide the updated {selected_doc_type} spreadsheet:
 """
     else:
@@ -597,9 +646,9 @@ Now, provide the updated {selected_doc_type}:
         current_document = response['choices'][0]['message']['content']
         logging.info("Document updated via chat successfully.")
         if selected_doc_type in spreadsheet_types:
-            preview = spreadsheet_preview(current_document, selected_doc_type)
+            preview = markdown_table_preview(current_document)
         else:
-            preview = narrative_preview(current_document)
+            preview = markdown_narrative_preview(current_document)
         return f"Document updated based on: {chat_input}", preview
     except Exception as e:
         logging.error(f"Error updating document via chat: {str(e)}")
@@ -618,13 +667,10 @@ def download_document(n_clicks, selected_doc_type):
 
     if selected_doc_type in spreadsheet_types:
         try:
-            df = pd.read_csv(StringIO(current_document))
-            output = BytesIO()
-            with pd.ExcelWriter(output, engine='xlsxwriter') as writer:
-                df.to_excel(writer, sheet_name=selected_doc_type, index=False)
+            # Convert markdown tables to xlsx
+            xlsx_bytes = markdown_tables_to_xlsx(current_document)
             logging.info(f"{selected_doc_type} document downloaded as Excel.")
-            output.seek(0)
-            return dcc.send_bytes(output.read(), f"{selected_doc_type}.xlsx")
+            return dcc.send_bytes(xlsx_bytes.read(), f"{selected_doc_type}.xlsx")
         except Exception as e:
             logging.error(f"Error downloading {selected_doc_type} document: {str(e)}")
             return dcc.send_string(f"Error downloading {selected_doc_type}: {str(e)}", f"{selected_doc_type}_error.txt")
@@ -653,7 +699,6 @@ def auto_expand_textarea(value, current_rows):
     if value is None or value == "":
         return 5
     num_lines = value.count('\n') + 1
-    # To avoid the box growing too huge, set a reasonable max
     max_rows = 20
     rows = min(max(num_lines, 5), max_rows)
     return rows