Update app.py

app.py

@@ -1,5 +1,5 @@
 import dash
-from dash import dcc, html, Input, Output, State, callback
+from dash import dcc, html, Input, Output, State
 import dash_bootstrap_components as dbc
 from dash.exceptions import PreventUpdate
 import google.generativeai as genai
@@ -11,136 +11,265 @@ import docx
 import os
 import logging
 import threading
-import base64
+from huggingface_hub import HfApi
 
 # Configure logging
 logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s')
 logger = logging.getLogger(__name__)
 
+# Initialize Dash app
 app = dash.Dash(__name__, external_stylesheets=[dbc.themes.BOOTSTRAP])
 
-# Helper functions (is_ui_file, get_file_contents, generate_guide_section, generate_guide)
-# ... (Keep these functions as they are in the original code)
+# Hugging Face API setup
+hf_api = HfApi()
 
-app.layout = dbc.Container([
-    html.H1("Automated Guide Generator", className="my-4"),
-    html.P("Generate a user guide or administration guide based on the UI-related code in a Git repository using Gemini AI. Select a Git provider, enter repository details, choose the guide type, and let AI create a comprehensive guide."),
+# Get Hugging Face variables
+GITHUB_TOKEN = os.environ.get('GITHUB_TOKEN')
+GEMINI_API_KEY = os.environ.get('GEMINI_API_KEY')
+
+def is_ui_file(filename):
+    ui_extensions = ['.erb', '.haml', '.slim', '.php', '.aspx', '.jsp', '.ftl', '.twig', '.mustache', '.handlebars', '.ejs', '.pug', '.blade.php', '.xhtml', '.fxml', '.tsx', '.jsx', '.vue', '.html', '.cshtml', '.razor', '.xaml', '.jsx']
+    return any(filename.endswith(ext) for ext in ui_extensions)
+
+def get_file_contents(git_provider, repo_url, exclude_folders):
+    file_contents = []
+    logger.info(f"Fetching files from {git_provider} repository: {repo_url}")
+    exclude_folders = [folder.strip() for folder in exclude_folders.split(',') if folder.strip()]
+    if git_provider == "GitHub":
+        g = Github(GITHUB_TOKEN)
+        repo = g.get_repo(repo_url)
+        contents = repo.get_contents("")
+        while contents:
+            file_content = contents.pop(0)
+            if file_content.type == "dir":
+                if not any(file_content.path.startswith(folder) for folder in exclude_folders):
+                    contents.extend(repo.get_contents(file_content.path))
+            elif is_ui_file(file_content.name) and not any(file_content.path.startswith(folder) for folder in exclude_folders):
+                logger.info(f"Found UI file: {file_content.path}")
+                file_contents.append((file_content.path, file_content.decoded_content.decode('utf-8', errors='ignore')))
+    elif git_provider == "GitLab":
+        gl = gitlab.Gitlab(url='https://gitlab.com', private_token=GITHUB_TOKEN)
+        project = gl.projects.get(repo_url)
+        items = project.repository_tree(recursive=True)
+        for item in items:
+            if item['type'] == 'blob' and is_ui_file(item['name']) and not any(item['path'].startswith(folder) for folder in exclude_folders):
+                logger.info(f"Found UI file: {item['path']}")
+                file_content = project.files.get(item['path'], ref='main')
+                file_contents.append((item['path'], file_content.decode().decode('utf-8', errors='ignore')))
+    elif git_provider == "Gitea":
+        base_url = "https://gitea.com/api/v1"
+        headers = {"Authorization": f"token {GITHUB_TOKEN}"}
+        def recursive_get_contents(path=""):
+            response = requests.get(f"{base_url}/repos/{repo_url}/contents/{path}", headers=headers)
+            response.raise_for_status()
+            for item in response.json():
+                if item['type'] == 'file' and is_ui_file(item['name']) and not any(item['path'].startswith(folder) for folder in exclude_folders):
+                    logger.info(f"Found UI file: {item['path']}")
+                    file_content = requests.get(item['download_url']).text
+                    file_contents.append((item['path'], file_content))
+                elif item['type'] == 'dir' and not any(item['path'].startswith(folder) for folder in exclude_folders):
+                    recursive_get_contents(item['path'])
+        recursive_get_contents()
+    else:
+        raise ValueError("Unsupported Git provider")
+    logger.info(f"Total UI files found: {len(file_contents)}")
+    return file_contents
+
+def generate_guide_section(file_path, file_content, guide_type):
+    logger.info(f"Generating {guide_type} section for file: {file_path}")
+    genai.configure(api_key=GEMINI_API_KEY)
+    model = genai.GenerativeModel('gemini-2.0-flash-lite')
     
-    dbc.Card([
-        dbc.CardBody([
-            dcc.Dropdown(
-                id='git-provider',
-                options=[
-                    {'label': 'GitHub', 'value': 'GitHub'},
-                    {'label': 'GitLab', 'value': 'GitLab'},
-                    {'label': 'Gitea', 'value': 'Gitea'}
-                ],
-                placeholder="Select Git Provider"
-            ),
-            dbc.Input(id='repo-url', placeholder="Repository URL (owner/repo)", type="text", className="mt-3"),
-            dcc.RadioItems(
-                id='guide-type',
-                options=[
-                    {'label': 'User Guide', 'value': 'User Guide'},
-                    {'label': 'Administration Guide', 'value': 'Administration Guide'}
-                ],
-                className="mt-3"
-            ),
-            dbc.Input(id='exclude-folders', placeholder="Exclude Folders (comma-separated)", type="text", className="mt-3"),
-            dbc.Button("Generate Guide", id="generate-button", color="primary", className="mt-3"),
-        ])
-    ], className="mb-4"),
+    if guide_type == "User Guide":
+        prompt = f"""Based on the following UI-related code file, generate a section for a user guide:
+
+        File: {file_path}
+        Content:
+        {file_content}
+
+        Please focus on:
+        1. The specific features and functionality this UI component provides to the end users
+        2. Step-by-step instructions on how to use these features
+        3. Any user interactions or inputs required
+        4. Expected outcomes or results for the user
+
+        Important formatting instructions:
+        - The output should be in plain text no markdown for example do not use * or ** or # or ##.  Instead use numbers like 1., 2. for bullets
+        - Use clear section titles
+        - Follow this numbering heirarchy (1.0, 1.1, 1.2), (2.0, 2.1, 2.2), (3.0, 3.1, 3.2) 
+        - Explain the purpose and benefit of each feature for non-technical users
+        - This is an end user manual, not a system administration manual so focus on the end user components
+        """
+    else:  # Administration Guide
+        prompt = f"""Based on the following UI-related code file, generate a section for an System guide:
+
+        File: {file_path}
+        Content:
+        {file_content}
+
+        Please focus on explaining what that component is and does:
+        1. Any configuration options or settings related to this UI component
+        2. Security considerations or access control related to this feature
+        3. How to monitor or troubleshoot issues with this component
+        4. Best practices for managing and maintaining this part of the system
+
+        Important formatting instructions:
+        - The output should be in plain text no markdown for example for example do not use * or ** or # or ##.  Instead use numbers like 1., 2. for bullets
+        - Use clear section titles
+        - Use clear section titles that has the name of the file in parenthesis
+        - Follow this numbering heirarchy (1.0, 1.1, 1.2), (2.0, 2.1, 2.2), (3.0, 3.1, 3.2) 
+        - Explain the purpose and implications of each component
+        """
+
+    response = model.generate_content(prompt)
+    logger.info(f"Generated {guide_type} section for {file_path}")
+    return response.text
+
+def generate_guide(git_provider, repo_url, guide_type, exclude_folders):
+    try:
+        logger.info(f"Starting guide generation for {repo_url}")
+        file_contents = get_file_contents(git_provider, repo_url, exclude_folders)
+        
+        guide_sections = []
+        for file_path, content in file_contents:
+            section = generate_guide_section(file_path, content, guide_type)
+            guide_sections.append(section)
+            logger.info(f"Added section for {file_path}")
+        
+        full_guide = f"# {guide_type}\n\n" + "\n\n".join(guide_sections)
+        
+        logger.info("Creating DOCX file")
+        doc = docx.Document()
+        doc.add_heading(guide_type, 0)
+        
+        for line in full_guide.split('\n'):
+            line = line.strip()
+            if line.startswith('# '):
+                doc.add_heading(line[2:], level=1)
+            elif line.startswith('## '):
+                doc.add_heading(line[3:], level=2)
+            elif line.startswith('Step'):
+                doc.add_paragraph(line, style='List Number')
+            else:
+                doc.add_paragraph(line)
+        
+        with tempfile.NamedTemporaryFile(delete=False, suffix='.docx') as temp_docx:
+            doc.save(temp_docx.name)
+            docx_path = temp_docx.name
+        logger.info(f"DOCX file saved: {docx_path}")
+
+        logger.info("Creating Markdown file")
+        with tempfile.NamedTemporaryFile(delete=False, suffix='.md', mode='w', encoding='utf-8') as temp_md:
+            temp_md.write(full_guide)
+            md_path = temp_md.name
+        logger.info(f"Markdown file saved: {md_path}")
+        
+        logger.info("Guide generation completed successfully")
+        return full_guide, docx_path, md_path
     
-    dbc.Spinner(
-        dcc.Loading(
-            id="loading-output",
-            children=[
-                html.Div(id="output-area"),
-                dcc.Download(id="download-docx"),
-                dcc.Download(id="download-md")
-            ],
-            type="default",
-        )
+    except Exception as e:
+        logger.error(f"An error occurred: {str(e)}", exc_info=True)
+        return f"An error occurred: {str(e)}", None, None
+
+# App layout
+app.layout = dbc.Container([
+    dbc.Navbar(
+        dbc.Container([
+            html.A(
+                dbc.Row([
+                    dbc.Col(html.Img(src="/assets/logo.png", height="30px")),
+                    dbc.Col(dbc.NavbarBrand("Automated Guide Generator", className="ms-2")),
+                ],
+                align="center",
+                className="g-0",
+                ),
+                href="/",
+                style={"textDecoration": "none"},
+            )
+        ]),
+        color="primary",
+        dark=True,
     ),
     
-    dcc.Store(id='guide-store'),
-])
+    dbc.Row([
+        dbc.Col([
+            html.H1("Automated Guide Generator", className="text-center my-4"),
+            html.P("Generate a user guide or administration guide based on the UI-related code in a Git repository using Gemini AI. Select a Git provider, enter repository details, choose the guide type, and let AI create a comprehensive guide.", className="text-center mb-4"),
+            
+            dbc.Card([
+                dbc.CardBody([
+                    dbc.Form([
+                        dbc.Select(
+                            id="git-provider",
+                            options=[
+                                {"label": "GitHub", "value": "GitHub"},
+                                {"label": "GitLab", "value": "GitLab"},
+                                {"label": "Gitea", "value": "Gitea"}
+                            ],
+                            placeholder="Select Git Provider",
+                        ),
+                        dbc.Input(id="repo-url", type="text", placeholder="Repository URL (owner/repo)"),
+                        dbc.RadioItems(
+                            id="guide-type",
+                            options=[
+                                {"label": "User Guide", "value": "User Guide"},
+                                {"label": "Administration Guide", "value": "Administration Guide"}
+                            ],
+                            inline=True,
+                        ),
+                        dbc.Input(id="exclude-folders", type="text", placeholder="Exclude Folders (comma-separated)"),
+                        dbc.Button("Generate Guide", id="generate-button", color="primary", className="mt-3"),
+                    ])
+                ])
+            ], className="mb-4"),
+            
+            dbc.Spinner(
+                dbc.Card([
+                    dbc.CardBody([
+                        html.H4("Generated Guide", className="card-title"),
+                        html.Div(id="generated-guide", style={"whiteSpace": "pre-wrap"}),
+                        html.Div([
+                            dbc.Button("Download DOCX", id="download-docx", color="secondary", className="me-2"),
+                            dbc.Button("Download Markdown", id="download-md", color="secondary"),
+                        ], className="mt-3")
+                    ])
+                ], className="mt-4"),
+                color="primary",
+            ),
+        ], width=12)
+    ])
+], fluid=True)
 
 @app.callback(
-    Output('guide-store', 'data'),
-    Output('output-area', 'children'),
-    Input('generate-button', 'n_clicks'),
-    State('git-provider', 'value'),
-    State('repo-url', 'value'),
-    State('guide-type', 'value'),
-    State('exclude-folders', 'value'),
-    prevent_initial_call=True
+    [Output("generated-guide", "children"),
+     Output("download-docx", "href"),
+     Output("download-md", "href")],
+    [Input("generate-button", "n_clicks")],
+    [State("git-provider", "value"),
+     State("repo-url", "value"),
+     State("guide-type", "value"),
+     State("exclude-folders", "value")]
 )
-def generate_guide_callback(n_clicks, git_provider, repo_url, guide_type, exclude_folders):
-    if not all([git_provider, repo_url, guide_type]):
+def update_output(n_clicks, git_provider, repo_url, guide_type, exclude_folders):
+    if n_clicks is None:
         raise PreventUpdate
-
+    
     def generate_guide_thread():
-        try:
-            guide_text, docx_path, md_path = generate_guide(
-                git_provider, repo_url, "", "", guide_type, exclude_folders
-            )
-            
-            with open(docx_path, 'rb') as docx_file, open(md_path, 'rb') as md_file:
-                docx_content = base64.b64encode(docx_file.read()).decode('utf-8')
-                md_content = base64.b64encode(md_file.read()).decode('utf-8')
-            
-            os.unlink(docx_path)
-            os.unlink(md_path)
-            
-            return {
-                'guide_text': guide_text,
-                'docx_content': docx_content,
-                'md_content': md_content
-            }
-        except Exception as e:
-            logger.error(f"An error occurred: {str(e)}", exc_info=True)
-            return {'error': str(e)}
-
+        nonlocal guide_text, docx_path, md_path
+        guide_text, docx_path, md_path = generate_guide(git_provider, repo_url, guide_type, exclude_folders)
+    
+    guide_text, docx_path, md_path = None, None, None
     thread = threading.Thread(target=generate_guide_thread)
     thread.start()
     thread.join()
-
-    result = thread.result() if hasattr(thread, 'result') else None
-
-    if result and 'error' not in result:
-        output = [
-            html.H3("Generated Guide"),
-            html.Pre(result['guide_text']),
-            dbc.Button("Download DOCX", id="btn-download-docx", color="secondary", className="me-2"),
-            dbc.Button("Download Markdown", id="btn-download-md", color="secondary")
-        ]
-        return result, output
+    
+    if docx_path and md_path:
+        docx_url = f"/download?path={docx_path}"
+        md_url = f"/download?path={md_path}"
     else:
-        error_message = result['error'] if result and 'error' in result else "An unknown error occurred"
-        return None, html.Div(f"Error: {error_message}", style={'color': 'red'})
-
-@app.callback(
-    Output("download-docx", "data"),
-    Input("btn-download-docx", "n_clicks"),
-    State('guide-store', 'data'),
-    prevent_initial_call=True
-)
-def download_docx(n_clicks, data):
-    if data and 'docx_content' in data:
-        return dict(content=data['docx_content'], filename="guide.docx", base64=True)
-    raise PreventUpdate
-
-@app.callback(
-    Output("download-md", "data"),
-    Input("btn-download-md", "n_clicks"),
-    State('guide-store', 'data'),
-    prevent_initial_call=True
-)
-def download_md(n_clicks, data):
-    if data and 'md_content' in data:
-        return dict(content=data['md_content'], filename="guide.md", base64=True)
-    raise PreventUpdate
+        docx_url = md_url = "#"
+    
+    return guide_text, docx_url, md_url
 
 if __name__ == '__main__':
     print("Starting the Dash application...")