Spaces:
Runtime error
Runtime error
File size: 6,657 Bytes
8cc9c71 4ee0746 8cc9c71 4ee0746 8cc9c71 4ee0746 8cc9c71 4ee0746 8cc9c71 4ee0746 8cc9c71 4ee0746 8cc9c71 4ee0746 8cc9c71 4ee0746 8cc9c71 4ee0746 8cc9c71 4ee0746 8cc9c71 4ee0746 8cc9c71 4ee0746 8cc9c71 80f27e2 8cc9c71 4ee0746 8cc9c71 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 |
import os
import json
from pathlib import Path
import asyncio
import base64
from openai import AsyncOpenAI
from dotenv import load_dotenv
load_dotenv()
VISION_MODEL = "gpt-4o-2024-08-06"
client = AsyncOpenAI()
async def analyze_screenshot(design_id: str, design_path: Path):
"""Analyze screenshots and return description, categories, and visual characteristics"""
try:
# Check files exist
metadata_path = design_path / "metadata.json"
desktop_img = design_path / "screenshot_desktop.png"
mobile_img = design_path / "screenshot_mobile.png"
if not all(f.exists() for f in [metadata_path, desktop_img, mobile_img]):
print(f"Missing required files for design {design_id}")
return design_id, None, None, None
# Load existing metadata
with open(metadata_path, "r") as f:
metadata = json.load(f)
# Read both images
try:
with open(desktop_img, "rb") as f:
desktop_base64 = base64.b64encode(f.read()).decode('utf-8')
with open(mobile_img, "rb") as f:
mobile_base64 = base64.b64encode(f.read()).decode('utf-8')
except Exception as e:
print(f"Error reading images for design {design_id}: {str(e)}")
return design_id, None, None, None
print(f"Analyzing design {design_id}...")
# Get response first
response = await client.chat.completions.create(
model=VISION_MODEL,
messages=[
{
"role": "system",
"content": """You are an expert graphic designer analyzing print and web designs for aesthetics, functionality, audience appeal, and potential applications.
The design should be considered from a visual standpoint. Use chain of thought to consider color palette, visual layout, typography, artistic style, mood, and potential applications.
Consider gradients, texture, background effects, and the use of images.
Treat all text content as placeholder Lorem Ipsum.
Provide analysis in clean JSON format with these exact keys:
{
"description": "A one-paragraph summary highlighting exceptional features of the design",
"categories": ["category1", "category2"],
"visual_characteristics": ["characteristic1", "characteristic2"]
}
Provide 4-6 categories and 4-6 visual characteristics most relevant to the style and feel of the design. Do not reference css or web design directly because this analysis is primarily about design. These lists should describe the design to another LLM that will use this data to generate a UI.
"""
},
{
"role": "user",
"content": [
{
"type": "text",
"text": "Analyze this visual design. Output only the JSON object."
},
{
"type": "image_url",
"image_url": {
"url": f"data:image/png;base64,{desktop_base64}",
"detail": "high"
}
},
{
"type": "image_url",
"image_url": {
"url": f"data:image/png;base64,{mobile_base64}",
"detail": "high"
}
}
]
}
],
max_tokens=1000
)
# Then get the content
response_content = response.choices[0].message.content.strip()
# Ensure the response is not empty
if not response_content:
print(f"Empty response for design {design_id}")
return design_id, None, None, None
# Extract JSON content from markdown code block
if "```json" in response_content:
# Remove the ```json prefix and ``` suffix
response_content = response_content.split("```json")[1].split("```")[0].strip()
# Parse the JSON response
try:
analysis = json.loads(response_content)
# Update metadata with all fields
metadata.update(analysis)
# Save updated metadata
with open(metadata_path, "w") as f:
json.dump(metadata, f, indent=2)
print(f"Successfully analyzed design {design_id}")
# Return visual_characteristics as fourth element
return design_id, analysis["description"], analysis["categories"], analysis["visual_characteristics"]
except json.JSONDecodeError as e:
print(f"Error parsing JSON response for design {design_id}: {str(e)}")
return design_id, None, None, None
except Exception as e:
print(f"Error processing design {design_id}: {str(e)}")
return design_id, None, None, None
async def main():
designs_dir = Path("designs")
if not designs_dir.exists():
print("Designs directory not found!")
return
# Get all design directories
design_dirs = [d for d in designs_dir.iterdir() if d.is_dir()]
if not design_dirs:
print("No design directories found!")
return
print(f"Found {len(design_dirs)} designs to analyze")
# Create list of design IDs to analyze (001-050)
design_ids = [f"{i:03d}" for i in range(1, 51)]
# Analyze all designs
tasks = []
for design_dir in design_dirs:
design_id = design_dir.name
tasks.append(analyze_screenshot(design_id, design_dir))
# Run analyses concurrently
results = await asyncio.gather(*tasks)
# Print summary
successful = 0
for design_id, desc, cats, _ in results:
if desc is not None:
successful += 1
print(f"\nDesign {design_id}:")
print(f"Description: {desc}")
print(f"Categories: {', '.join(cats)}")
print(f"\nSuccessfully analyzed {successful} out of {len(design_dirs)} designs")
if __name__ == "__main__":
asyncio.run(main()) |