File size: 10,163 Bytes
b511d75
8d1bd48
9abec8f
19338e6
83a7945
19338e6
9abec8f
 
 
 
 
 
 
 
 
 
 
 
19338e6
 
 
 
b511d75
19338e6
 
 
9abec8f
19338e6
 
 
 
 
 
 
8d1bd48
83a7945
19338e6
 
 
 
 
 
 
 
 
 
 
 
 
83a7945
b511d75
19338e6
 
 
 
 
 
 
 
 
 
 
9abec8f
 
 
 
 
19338e6
 
9abec8f
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
19338e6
 
 
2dea97c
19338e6
 
 
 
 
 
 
 
 
 
 
 
 
 
9abec8f
83a7945
19338e6
 
9abec8f
19338e6
9abec8f
 
 
 
 
19338e6
9abec8f
 
19338e6
 
 
 
 
 
 
 
 
9abec8f
 
 
 
 
 
 
19338e6
9abec8f
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
19338e6
 
 
 
2dea97c
19338e6
 
 
 
 
 
2dea97c
19338e6
 
 
 
 
 
 
 
 
 
9abec8f
 
 
19338e6
 
9abec8f
 
 
 
 
 
 
 
19338e6
 
 
9abec8f
 
 
 
19338e6
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
9abec8f
19338e6
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
9abec8f
19338e6
9abec8f
 
19338e6
 
9abec8f
 
 
 
 
 
 
 
 
 
 
 
19338e6
 
 
 
 
 
9abec8f
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
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
import os
import gradio as gr
import requests
import json
import time
import traceback
import io
import base64
from PIL import Image, ImageEnhance, ImageFilter

# Conditional imports
try:
    import google.generativeai as genai
    GENAI_AVAILABLE = True
except ImportError:
    GENAI_AVAILABLE = False
    print("Warning: google-generativeai not installed, will attempt on-demand import")

try:
    import pyperclip
except ImportError:
    pyperclip = None

# --- Environment Configuration ---
GEMINI_KEY = os.environ.get("GEMINI_KEY", "")
DEFAULT_PORT = int(os.environ.get("PORT", 7860))
API_TIMEOUT = 30  # seconds

# --- Style Template Optimization ---
BASE_TEMPLATE = """Describe this design as a concise Flux 1.1 Pro prompt focusing on:
- Key visual elements
- Technical specifications
- Style consistency
- Functional requirements"""

STYLE_INSTRUCTIONS = {
    "General": BASE_TEMPLATE,
    "Realistic": f"{BASE_TEMPLATE}\nPHOTOREALISTIC RULES: Use photography terms, texture details, accurate lighting",
    "Kawaii": f"{BASE_TEMPLATE}\nKAWAII RULES: Rounded shapes, pastel colors, cute expressions",
    "Vector": f"{BASE_TEMPLATE}\nVECTOR RULES: Clean lines, geometric shapes, B&W gradients",
    "Silhouette": f"{BASE_TEMPLATE}\nSILHOUETTE RULES: High contrast, minimal details, strong outlines"
}

# --- Flux Configuration ---
FLUX_SPECS = {
    "aspect_ratios": ["1:1", "16:9", "4:3", "9:16"],
    "formats": ["SVG", "PNG", "PDF"],
    "color_modes": ["B&W", "CMYK", "RGB"],
    "dpi_options": [72, 150, 300, 600]
}

# --- Quality Control System ---
class QualityValidator:
    VALIDATION_TEMPLATE = """Analyze this Flux prompt:
1. Score style adherence (1-5)
2. List technical issues
3. Suggest improvements
Respond ONLY as JSON: {"score": x/10, "issues": [], "suggestions": []}"""

    @classmethod
    def validate(cls, prompt, model):
        try:
            with gr.utils.TempFiles() as temp:
                response = model.generate_content([cls.VALIDATION_TEMPLATE, prompt])
                return json.loads(response.text)
        except Exception as e:
            print(f"Validation error: {str(e)}")
            return {"score": 0, "issues": ["Validation failed"], "suggestions": []}

# --- Lazy API Initialization ---
def init_genai_api(api_key):
    """Initialize Gemini API with error handling"""
    if not GENAI_AVAILABLE:
        try:
            # Attempt dynamic import
            global genai
            import google.generativeai as genai
        except ImportError:
            raise ValueError("Failed to import google.generativeai. Install with: pip install google-generativeai")
    
    try:
        genai.configure(api_key=api_key)
        # Test connection with minimal request
        model = genai.GenerativeModel("gemini-1.5-pro")
        model.generate_content("test", request_options={"timeout": 5})
        return model
    except Exception as e:
        if "authentication" in str(e).lower():
            raise ValueError("Invalid API key or authentication error")
        elif "timeout" in str(e).lower():
            raise ValueError("API connection timeout - check your internet connection")
        else:
            raise ValueError(f"API initialization error: {str(e)}")

# --- Image Processing Pipeline ---
def preprocess_image(img):
    """Convert and enhance uploaded images"""
    try:
        if isinstance(img, str):  # Handle file paths
            img = Image.open(img)
        img = img.convert("RGB")
        img = ImageEnhance.Contrast(img).enhance(1.2)
        img = img.filter(ImageFilter.SHARPEN)
        return img
    except Exception as e:
        raise ValueError(f"Image processing error: {str(e)}")

# --- Core Generation Engine ---
def generate_prompt(image, api_key, style, creativity, neg_prompt, aspect, color_mode, dpi):
    try:
        # Validate inputs
        if not image:
            return {"error": "Please upload an image"}
            
        api_key = api_key or GEMINI_KEY
        if not api_key:
            return {"error": "API key required - set in env (GEMINI_KEY) or input field"}

        # Initialize model with proper error handling
        try:
            model = init_genai_api(api_key)
        except ValueError as e:
            return {"error": str(e)}

        # Process image with timeout protection
        start_time = time.time()
        img = preprocess_image(image)
        img_bytes = io.BytesIO()
        img.save(img_bytes, format="PNG")
        img_b64 = base64.b64encode(img_bytes.getvalue()).decode()

        # Build instruction
        instruction = f"{STYLE_INSTRUCTIONS[style]}\nAVOID: {neg_prompt}\n"
        instruction += f"ASPECT: {aspect}, COLORS: {color_mode}, DPI: {dpi}\n"

        # Generate prompt with timeout protection
        try:
            response = model.generate_content(
                contents=[instruction, {"mime_type": "image/png", "data": img_b64}],
                generation_config={"temperature": creativity},
                request_options={"timeout": API_TIMEOUT}
            )
            raw_prompt = response.text
        except requests.exceptions.Timeout:
            return {"error": "API request timed out (>30s). Try a smaller image or check your connection."}
        except Exception as e:
            return {"error": f"Generation failed: {str(e)}"}

        # Quality validation (skip if taking too long)
        validation = {"score": 8, "issues": [], "suggestions": []}
        if time.time() - start_time < 20:  # Only validate if we have time
            try:
                validation = QualityValidator.validate(raw_prompt, model)
                if validation.get("score", 0) < 7:
                    response = model.generate_content(
                        f"Improve this prompt: {raw_prompt}\nIssues: {validation['issues']}",
                        request_options={"timeout": 10}
                    )
                    raw_prompt = response.text
            except:
                # Continue even if validation fails
                pass

        # Token tracking
        input_tokens = len(img_b64) // 4  # Approximate base64 token count
        output_tokens = len(raw_prompt.split())
        
        return {
            "prompt": raw_prompt,
            "validation": validation,
            "stats": {"input": input_tokens, "output": output_tokens}
        }

    except Exception as e:
        traceback.print_exc()
        return {"error": str(e)}

# --- UI Components ---
def create_advanced_controls():
    with gr.Accordion("โš™๏ธ Advanced Settings", open=False):
        with gr.Row():
            creativity = gr.Slider(0.0, 1.0, 0.7, label="Creativity Level")
            neg_prompt = gr.Textbox(label="๐Ÿšซ Negative Prompts", placeholder="What to avoid")
        with gr.Row():
            aspect = gr.Dropdown(FLUX_SPECS["aspect_ratios"], value="1:1", label="Aspect Ratio")
            color_mode = gr.Dropdown(FLUX_SPECS["color_modes"], value="RGB", label="Color Mode")
            dpi = gr.Dropdown([str(d) for d in FLUX_SPECS["dpi_options"]], value="300", label="Output DPI")
    return [creativity, neg_prompt, aspect, color_mode, dpi]

# --- UI Response Formatting ---
def format_generation_response(result):
    """Format the response from generate_prompt for the UI"""
    if "error" in result:
        return result["error"], None, None
    else:
        return result.get("prompt", ""), result.get("validation", {}), result.get("stats", {})

# --- Main Interface ---
def build_interface():
    with gr.Blocks(title="Flux Pro Generator", theme=gr.themes.Soft()) as app:
        # Header
        gr.Markdown("# ๐ŸŽจ Flux Pro Prompt Generator")
        gr.Markdown("Generate optimized design prompts from images using Google's Gemini")
        
        # Security Section
        api_key = gr.Textbox(
            label="๐Ÿ”‘ Gemini API Key",
            value=GEMINI_KEY,
            type="password",
            info="Set GEMINI_KEY environment variable for production"
        )

        # Main Workflow
        with gr.Row(variant="panel"):
            with gr.Column(scale=1):
                img_input = gr.Image(
                    label="๐Ÿ–ผ๏ธ Upload Design",
                    type="pil",
                    sources=["upload"],
                    interactive=True
                )
                style = gr.Dropdown(
                    list(STYLE_INSTRUCTIONS.keys()),
                    value="General",
                    label="๐ŸŽจ Target Style"
                )
                adv_controls = create_advanced_controls()
                gen_btn = gr.Button("โœจ Generate Prompt", variant="primary")
                status_msg = gr.Textbox(label="Status", visible=True)

            with gr.Column(scale=2):
                prompt_output = gr.Textbox(
                    label="๐Ÿ“ Optimized Prompt",
                    lines=8,
                    interactive=False
                )
                with gr.Row():
                    copy_btn = gr.Button("๐Ÿ“‹ Copy")
                quality_report = gr.JSON(
                    label="๐Ÿ” Quality Report",
                    visible=True
                )
                token_stats = gr.JSON(
                    label="๐Ÿงฎ Token Usage",
                    visible=True
                )

        # Event Handling
        gen_btn.click(
            lambda *args: format_generation_response(generate_prompt(*args)),
            inputs=[img_input, api_key, style] + adv_controls,
            outputs=[prompt_output, quality_report, token_stats],
            api_name="generate"
        )

        if pyperclip:
            copy_btn.click(
                lambda x: pyperclip.copy(x) if x else None,
                inputs=prompt_output,
                outputs=None
            )
        else:
            copy_btn.click(
                lambda: "Copy functionality not available (pyperclip not installed)",
                inputs=None,
                outputs=status_msg
            )

    return app

# --- Production Launch ---
if __name__ == "__main__":
    app = build_interface()
    app.launch(server_name="0.0.0.0", server_port=DEFAULT_PORT, share=False)