"""Real image generation tests using actual API calls.
These tests make real API calls to test the complete functionality.
They require valid API keys and will incur costs.
Run with: pytest tests/integration/test_real_image_generation.py -v -s
"""
import asyncio
import json
from pathlib import Path
import pytest
from image_gen_mcp.config.settings import Settings
from image_gen_mcp.storage.manager import ImageStorageManager
from image_gen_mcp.tools.image_generation import ImageGenerationTool
from image_gen_mcp.types.enums import (
ImageQuality,
ImageSize,
ImageStyle,
)
from image_gen_mcp.utils.cache import CacheManager
@pytest.mark.integration
@pytest.mark.slow
class TestRealImageGeneration:
"""Integration tests for real image generation."""
@pytest.fixture
async def real_tool(self, tmp_path):
"""Create a real image generation tool with actual settings."""
settings = Settings()
# Use project storage directory to persist generated images
from image_gen_mcp.config.settings import StorageSettings
from pathlib import Path
project_storage = Path(__file__).parent.parent.parent / "storage"
storage_settings = StorageSettings(
base_path=str(project_storage),
retention_days=30,
max_size_gb=10.0,
cleanup_interval_hours=24,
)
storage_manager = ImageStorageManager(settings=storage_settings)
cache_manager = CacheManager(settings=settings.cache)
tool = ImageGenerationTool(
storage_manager=storage_manager,
cache_manager=cache_manager,
settings=settings,
)
yield tool
# Cleanup
if hasattr(storage_manager, 'close'):
await storage_manager.close()
if hasattr(cache_manager, 'close'):
await cache_manager.close()
async def test_list_available_models(self, real_tool):
"""Test listing available models."""
print("\n" + "=" * 80)
print("TEST: List Available Models")
print("=" * 80)
models_info = real_tool.get_supported_models()
print(json.dumps(models_info, indent=2))
assert isinstance(models_info, dict)
assert "providers" in models_info or len(models_info) > 0
available_providers = real_tool.get_available_providers()
print(f"\nAvailable Providers: {available_providers}")
assert isinstance(available_providers, list)
async def test_basic_vivid_generation(self, real_tool):
"""Test basic image generation with vivid style."""
print("\n" + "=" * 80)
print("TEST: Basic Vivid Style Generation")
print("=" * 80)
prompt = (
"A serene Japanese zen garden with cherry blossoms, "
"koi pond, and traditional stone lanterns at sunset, "
"digital art style"
)
print(f"Prompt: {prompt}")
result = await real_tool.generate(
prompt=prompt,
style=ImageStyle.VIVID,
quality=ImageQuality.AUTO,
size=ImageSize.LANDSCAPE,
)
print(f"\nImage ID: {result['image_id']}")
print(f"Task ID: {result['task_id']}")
print(f"Image URL: {result['image_url']}")
print(f"Resource URI: {result['resource_uri']}")
print(f"\nMetadata:")
print(json.dumps(result['metadata'], indent=2))
# Assertions
assert result['image_id']
assert result['task_id']
assert result['image_url']
assert result['metadata']['style'] == 'vivid'
assert result['metadata']['prompt'] == prompt
async def test_natural_style_generation(self, real_tool):
"""Test natural style generation."""
print("\n" + "=" * 80)
print("TEST: Natural Style Generation")
print("=" * 80)
prompt = (
"A cozy coffee shop interior with wooden furniture, "
"warm lighting, plants, and people working on laptops"
)
print(f"Prompt: {prompt}")
result = await real_tool.generate(
prompt=prompt,
style=ImageStyle.NATURAL,
quality=ImageQuality.AUTO,
size=ImageSize.LANDSCAPE,
)
print(f"\nImage ID: {result['image_id']}")
print(f"Image URL: {result['image_url']}")
print(f"Style: {result['metadata']['style']}")
assert result['image_id']
assert result['metadata']['style'] == 'natural'
async def test_portrait_orientation(self, real_tool):
"""Test portrait orientation image."""
print("\n" + "=" * 80)
print("TEST: Portrait Orientation")
print("=" * 80)
prompt = (
"A majestic snow leopard standing on a rocky mountain peak, "
"dramatic clouds in background, wildlife photography style"
)
print(f"Prompt: {prompt}")
result = await real_tool.generate(
prompt=prompt,
style=ImageStyle.VIVID,
quality=ImageQuality.AUTO,
size=ImageSize.PORTRAIT,
)
print(f"\nImage ID: {result['image_id']}")
print(f"Size: {result['metadata']['size']}")
assert result['image_id']
assert result['metadata']['size'] == '1024x1536'
async def test_square_size(self, real_tool):
"""Test square size image."""
print("\n" + "=" * 80)
print("TEST: Square Size")
print("=" * 80)
prompt = (
"A futuristic cyberpunk city street at night with neon signs, "
"flying cars, and rain-soaked pavement"
)
print(f"Prompt: {prompt}")
result = await real_tool.generate(
prompt=prompt,
style=ImageStyle.VIVID,
quality=ImageQuality.AUTO,
size=ImageSize.SQUARE,
)
print(f"\nImage ID: {result['image_id']}")
print(f"Size: {result['metadata']['size']}")
assert result['image_id']
assert result['metadata']['size'] == '1024x1024'
async def test_high_quality(self, real_tool):
"""Test high quality setting."""
print("\n" + "=" * 80)
print("TEST: High Quality")
print("=" * 80)
prompt = (
"An incredibly detailed macro photograph of a monarch butterfly "
"on a purple flower, with water droplets and bokeh background"
)
print(f"Prompt: {prompt}")
result = await real_tool.generate(
prompt=prompt,
style=ImageStyle.VIVID,
quality=ImageQuality.HIGH,
size=ImageSize.LANDSCAPE,
)
print(f"\nImage ID: {result['image_id']}")
print(f"Quality: {result['metadata']['quality']}")
print(f"Cost: ${result['metadata'].get('cost_estimate', 'N/A')}")
assert result['image_id']
assert result['metadata']['quality'] == 'high'
@pytest.mark.integration
@pytest.mark.slow
class TestCreativeUseCases:
"""Test various creative use case scenarios."""
@pytest.fixture
async def real_tool(self, tmp_path):
"""Create a real image generation tool with actual settings."""
settings = Settings()
# Use project storage directory to persist generated images
from image_gen_mcp.config.settings import StorageSettings
from pathlib import Path
project_storage = Path(__file__).parent.parent.parent / "storage"
storage_settings = StorageSettings(
base_path=str(project_storage),
retention_days=30,
max_size_gb=10.0,
cleanup_interval_hours=24,
)
storage_manager = ImageStorageManager(settings=storage_settings)
cache_manager = CacheManager(settings=settings.cache)
tool = ImageGenerationTool(
storage_manager=storage_manager,
cache_manager=cache_manager,
settings=settings,
)
yield tool
# Cleanup
if hasattr(storage_manager, 'close'):
await storage_manager.close()
if hasattr(cache_manager, 'close'):
await cache_manager.close()
async def test_product_photography(self, real_tool):
"""Test product photography style."""
print("\n" + "=" * 80)
print("TEST: Product Photography")
print("=" * 80)
prompt = (
"Professional product photography of a luxury smartwatch "
"on a marble surface with dramatic studio lighting and reflections"
)
print(f"Prompt: {prompt}")
result = await real_tool.generate(
prompt=prompt,
style=ImageStyle.VIVID,
quality=ImageQuality.AUTO,
size=ImageSize.LANDSCAPE,
)
print(f"\nImage ID: {result['image_id']}")
print(f"URL: {result['image_url']}")
assert result['image_id']
assert result['metadata']['style'] == 'vivid'
async def test_social_media_graphic(self, real_tool):
"""Test social media graphic generation."""
print("\n" + "=" * 80)
print("TEST: Social Media Graphic")
print("=" * 80)
prompt = (
"Modern minimalist Instagram post background with pastel gradient, "
"geometric shapes, and space for text overlay"
)
print(f"Prompt: {prompt}")
result = await real_tool.generate(
prompt=prompt,
style=ImageStyle.NATURAL,
quality=ImageQuality.AUTO,
size=ImageSize.SQUARE,
)
print(f"\nImage ID: {result['image_id']}")
print(f"URL: {result['image_url']}")
assert result['image_id']
assert result['metadata']['size'] == '1024x1024'
async def test_concept_art(self, real_tool):
"""Test fantasy concept art."""
print("\n" + "=" * 80)
print("TEST: Fantasy Concept Art")
print("=" * 80)
prompt = (
"Fantasy concept art of a floating island with waterfalls, "
"ancient ruins, and bioluminescent plants, magical atmosphere"
)
print(f"Prompt: {prompt}")
result = await real_tool.generate(
prompt=prompt,
style=ImageStyle.VIVID,
quality=ImageQuality.AUTO,
size=ImageSize.LANDSCAPE,
)
print(f"\nImage ID: {result['image_id']}")
print(f"URL: {result['image_url']}")
assert result['image_id']
async def test_batch_generation(self, real_tool):
"""Test generating multiple images in sequence."""
print("\n" + "=" * 80)
print("TEST: Batch Generation")
print("=" * 80)
prompts = [
"A peaceful mountain lake at sunrise with mist",
"A modern workspace with plants and natural light",
"Abstract geometric pattern in blue and gold",
]
results = []
for i, prompt in enumerate(prompts, 1):
print(f"\n{i}. Generating: {prompt}")
result = await real_tool.generate(
prompt=prompt,
style=ImageStyle.VIVID,
quality=ImageQuality.AUTO,
size=ImageSize.LANDSCAPE,
)
print(f" Image ID: {result['image_id']}")
results.append(result)
assert len(results) == len(prompts)
assert all(r['image_id'] for r in results)
# Summary
total_cost = sum(
r['metadata'].get('cost_estimate', 0) for r in results
)
print(f"\nTotal Cost: ${total_cost:.4f}")
if __name__ == "__main__":
"""Allow running tests directly with python."""
pytest.main([__file__, "-v", "-s"])
