Skip to main content
Find answers to common questions about using the Extrator de Tarefas Auvo application.

​
File Formats and Compatibility

The application supports three file formats from Auvo exports:
  • CSV (.csv) - Comma-separated values files
  • Excel 97-2003 (.xls) - Legacy Excel format
  • Excel 2007+ (.xlsx) - Modern Excel format
All formats are processed using pandas, with CSV files read using pd.read_csv() and Excel files using pd.read_excel() with the openpyxl engine (app.py:37-42).Note: All three formats are treated equally - the application automatically detects the format based on file extension.
Yes. The application expects Auvo export files with a specific structure:
  1. First 5 rows are skipped - These typically contain Auvo export metadata
  2. Row 6 must contain column headers with these exact names:
    • Data
    • Cliente
    • Endereco
    • OS Digital
    • Relato
  3. The Relato column must contain task descriptions where keywords will be searched
The application filters based on the Relato column and returns the five columns listed above in the results (app.py:38, app.py:49-53).
Yes, as long as your file:
  • Is in CSV, XLS, or XLSX format
  • Has 5 rows of headers/metadata before the column names (or adjust the file accordingly)
  • Contains the required columns: Data, Cliente, Endereco, OS Digital, and Relato
  • Has searchable text in the Relato column
If your source system uses different column names, you’ll need to rename them to match the expected format before uploading.
Pandas (the data processing library) handles most common encodings automatically, including:
  • UTF-8 (recommended)
  • Latin-1 (ISO-8859-1)
  • Windows-1252
If you encounter encoding errors, save your CSV file as UTF-8 using Excel, Notepad++, or another text editor.

​
Keywords and Search Functionality

You can customize keywords through the Configurações page:
  1. Navigate to the configuration page from the main menu
  2. Enter your keywords as a comma-separated list
  3. Click save
  4. Your custom keywords are stored in the session and will be used for all subsequent file uploads
Example format:
solicitar peça, quebrado, trocar cabo, instalação, manutenção
Keywords are stored in the session and persist until you clear cookies or close your browser (app.py:111-112).
The application comes with these default keywords designed for maintenance and repair tasks (app.py:115):
  • solicitar peça
  • quebrado, quebrada, quebrados
  • orçamento
  • danificada, danificado, danificados, danificadas
  • trocar cabo
  • soldar
  • trocar
  • instalar
These defaults target common action-requiring scenarios in field service reports.
No. The search is case-insensitive and finds partial matches within the Relato column text (app.py:50).For example:
  • Keyword "quebrado" will match “QUEBRADO”, “Quebrado”, “equipamento quebrado”, etc.
  • Keyword "instalar" will match “instalação”, “instalar”, “precisa instalar”, etc.
The search uses pandas’ str.contains() method with case=False parameter.
Yes! The application fully supports multi-word keywords.Examples:
  • "solicitar peça" - searches for this exact phrase
  • "trocar cabo" - finds records mentioning cable replacement
  • "equipamento danificado" - locates damaged equipment mentions
When entering multiple phrase keywords, separate them with commas:
solicitar peça, trocar cabo, equipamento danificado, sem energia
The statistics panel shows:
  1. Total de Registros - All rows in the uploaded file (after skipping 5 header rows)
  2. Tarefas Encontradas - Rows where the Relato column contains at least one keyword
  3. Taxa de Ocorrência - Percentage of records matching (app.py:55-73)
Additionally, the stats include a breakdown by keyword showing how many times each individual keyword appeared in the filtered results (app.py:67-71).Note: If a row contains multiple keywords, it’s only counted once in “Tarefas Encontradas” but appears in the per-keyword breakdown multiple times.

​
Export and Download Features

The application offers two export formats:
  1. Excel (.xlsx) - Includes two sheets:
    • Tarefas Encontradas - Filtered results with all five columns
    • Estatísticas - Summary metrics and generation timestamp
  2. PDF - Formatted report with:
    • Header with generation date
    • Statistics dashboard
    • Keywords used section
    • Results table with clickable links
Both exports use the filtered data stored in temporary session files (app.py:178-263).
No explicit limit exists in the code. The application will export all filtered results regardless of size.However, practical limitations apply:
  • Excel exports - Can handle very large datasets (tested up to 1M+ rows with pandas/openpyxl)
  • PDF exports - May become slow or timeout with thousands of rows due to WeasyPrint rendering
For result sets over 1,000 rows, we recommend using Excel export instead of PDF for better performance.
Results are stored in temporary session files, which means:You can download multiple times during the same session, but results will be lost when:
  • You close your browser
  • Your session expires
  • You clear browser cookies
  • The server restarts
When results expire, you’ll see: “Os resultados expiraram. Por favor, processe o arquivo novamente.” (app.py:183, app.py:211)
Best Practice: Download both Excel and PDF versions immediately after processing if you need them.

​
Session and Data Persistence

Custom keywords are stored in the Flask session, which persists as long as:
  • Your browser remains open
  • You don’t clear cookies
  • The session cookie doesn’t expire (default is until browser closes)
When you close your browser or clear cookies, keywords reset to the defaults.
If you want permanent custom keywords, you would need to modify the application code to store them in a configuration file or database (not currently supported).
No. Here’s what happens to your files:
  1. Upload - Your original file is processed in memory (app.py:140)
  2. Temporary storage - Filtered results are saved to temp/ folder with a unique UUID filename (app.py:145-149)
  3. Session reference - Only the temporary filename is stored in the session (app.py:152)
  4. Automatic cleanup - Files remain until manually deleted
The application does NOT automatically delete temporary files. The temp/ folder will accumulate files over time. Periodic manual cleanup is recommended.
To clean up:
rm temp/*.csv  # Linux/macOS
del temp\*.csv  # Windows
Yes! The application maintains a session-based history of the last 10 processed files (app.py:75-84).Access it through the Histórico page, which shows:
  • Original filename
  • Processing date and time
  • Number of tasks found
  • Total records processed
Like custom keywords, this history is session-based and cleared when you close your browser.
Security considerations:Session isolation:
  • Each user session has a unique SECRET_KEY-signed cookie
  • Temporary files use random UUID names (app.py:145)
  • Other users cannot access your session data or files
Data persistence:
  • Original uploads are NOT saved to disk
  • Only filtered results are temporarily stored
  • No database or permanent storage is used
Recommendations for production:
  • Use HTTPS to encrypt data in transit
  • Set appropriate file system permissions on the temp/ folder
  • Implement automatic temporary file cleanup
  • Use environment-specific SECRET_KEY (never commit .env to version control)
  • Consider adding authentication for multi-user deployments
The default setup (app.py:266) runs with debug=True, which is NOT secure for production use. Always set debug=False in production environments.

​
Browser Compatibility

The application uses standard web technologies and works with all modern browsers:
  • Chrome/Edge (recommended) - Version 90+
  • Firefox - Version 88+
  • Safari - Version 14+
  • Opera - Version 76+
The interface uses Bootstrap 5 for styling, which has excellent cross-browser compatibility.Required browser features:
  • JavaScript enabled (for table search and interactivity)
  • Cookies enabled (for session management)
  • File upload support
Yes, the interface is responsive thanks to Bootstrap 5. However:Mobile browsers:
  • File uploads work on iOS Safari and Android Chrome
  • Table viewing is functional but may require horizontal scrolling
  • Downloads work on all major mobile browsers
Recommended usage:
  • Desktop/laptop - Full functionality, best experience
  • Tablet - Good for viewing and basic uploads
  • Phone - Works but challenging for large tables
For the best experience processing large CSV files and viewing results, use a desktop or laptop computer.
No. The application requires:
  • An active Flask server running (python app.py)
  • Network connection to localhost (127.0.0.1:5000)
  • Browser access to the running server
For offline use, you would need to:
  1. Run the Flask server locally on your machine
  2. Access it via http://127.0.0.1:5000 in your browser
  3. Process files while the server is running
This is “offline” in the sense of not requiring internet, but you still need the local server running.

​
Technical Questions

The application requires Python 3.10 or higher (as documented in README.md).This requirement comes from dependencies like pandas 2.3.2 and numpy 2.3.2, which require modern Python versions.To check your version:
python --version
If you have an older version, download Python 3.10+ from https://www.python.org/downloads/
The current setup (app.py:266) uses Flask’s development server with debug=True, which is not suitable for production.For production deployment, you should:
  1. Change debug mode: 
    if __name__ == '__main__':
    app.run(debug=False)
  2. Use a production WSGI server:
    • Gunicorn (Linux/macOS): gunicorn app:app
    • Waitress (Windows): waitress-serve --listen=*:5000 app:app
  3. Add security measures:
    • File upload size limits
    • Rate limiting
    • HTTPS/SSL certificates
    • Authentication/authorization
    • CSRF protection
  4. Implement file cleanup:
    • Cron job to delete old temp files
    • Automatic cleanup on session expiration
  5. Set environment variables properly:
    • Strong SECRET_KEY
    • Production database (if adding one)
The columns are hardcoded in app.py:52-53. To modify:
  1. Change the search column (currently Relato): 
    coluna_descricao = 'YourColumnName'  # Line 49
  2. Change result columns (currently 5 columns): 
    colunas_resultado = ['Column1', 'Column2', 'Column3']  # Line 52
  3. Adjust header skip rows (currently 5): 
    df = pd.read_csv(file, skiprows=X)  # Line 38, change X
df = pd.read_excel(file, skiprows=X, engine='openpyxl')  # Line 42
Restart the application after making changes.
The application currently only provides a web interface. To add API endpoints:
  1. Add JSON response routes: 
    @app.route('/api/process', methods=['POST'])
def api_process():
    # Process file and return JSON
    return jsonify(results)
  2. Use Flask-RESTful or Flask-RESTX for structured API development
  3. Implement authentication (JWT, API keys, etc.)
  4. Add CORS headers if calling from different domains: 
    from flask_cors import CORS
CORS(app)
API integration requires code modifications and is not available out of the box.

​
Getting Started

Follow this quick start guide:
  1. Install dependencies (see Installation guide) 
    python -m venv venv
source venv/bin/activate
pip install -r requirements.txt
  2. Configure environment
    • Create .env file with SECRET_KEY
    • Create temp/ folder
  3. Start the application 
    python app.py
  4. Open in browser: http://127.0.0.1:5000
  5. Upload a test file
    • Use a small Auvo CSV export (< 1000 rows)
    • Process with default keywords
  6. Explore features
    • View statistics
    • Download Excel/PDF
    • Customize keywords in Configurações
    • Check Histórico
Start with a small test file to familiarize yourself with the interface before processing large production exports.
If you encounter issues:
  1. Check the Troubleshooting guide - Covers common problems and solutions
  2. Review error messages - Look at the terminal where you ran python app.py
  3. Read this FAQ - Many questions are answered here
  4. Verify your setup: 
    python --version  # Should be 3.10+
pip list  # Verify all packages installed
ls temp/  # Check temp folder exists
  5. Enable debug output - The application runs with debug=True by default, showing detailed errors
When seeking help, always include:
  • Python version
  • Operating system
  • Full error message
  • Steps to reproduce

Build docs developers (and LLMs) love

Get started for freeTalk to us