# Format Demo
This sample demonstrates the different output formats supported by Genkit Python.
## Formats Demonstrated
1. **Text (`text`)**: Raw text output.
2. **JSON (`json`)**: Structured JSON object output.
3. **Array (`array`)**: JSON array of items.
4. **Enum (`enum`)**: Single value from a predefined list.
5. **JSONL (`jsonl`)**: Newline-delimited JSON objects (great for streaming).
## Running
1. Set your `GEMINI_API_KEY` environment variable.
2. Run the sample:
### Monitoring and Running
For an enhanced development experience, use the provided `run.sh` script to start the sample with automatic reloading:
```bash
./run.sh
```
This script uses `watchmedo` to monitor changes in:
- `src/` (Python logic)
- `../../packages` (Genkit core)
- `../../plugins` (Genkit plugins)
- File patterns: `*.py`, `*.prompt`, `*.json`
Changes will automatically trigger a restart of the sample. You can also pass command-line arguments directly to the script, e.g., `./run.sh --some-flag`.
```bash
export GEMINI_API_KEY=your-key
./run.sh
```
Then use the Genkit Developer UI to test the flows.
## Testing This Demo
1. **Open DevUI** at http://localhost:4000
2. **Test each format** (inputs pre-populated with defaults):
- [ ] `generate_haiku_text` - Text format (topic: "coding")
- [ ] `get_country_info_json` - JSON format (country: "Japan")
- [ ] `recommend_books_array` - Array format (genre: "Fantasy")
- [ ] `classify_sentiment_enum` - Enum format (review text)
- [ ] `create_story_characters_jsonl` - JSONL format (theme: "Space Opera")
3. **Expected behavior**:
- Text: Returns plain haiku text
- JSON: Returns structured country info object
- Array: Returns list of book recommendations
- Enum: Returns single sentiment classification
- JSONL: Returns newline-delimited character objects
4. **Verify output structure**:
- JSON outputs should be valid JSON
- Array outputs should be valid JSON arrays
- Enum outputs should match predefined values
