File size: 18,901 Bytes
7404ace
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
63d8d8e
7404ace
63d8d8e
7404ace
 
 
 
63d8d8e
7404ace
 
 
 
 
 
63d8d8e
7404ace
 
 
 
 
 
 
 
 
 
 
 
 
 
 
63d8d8e
 
 
 
 
 
7404ace
63d8d8e
7404ace
 
63d8d8e
7404ace
63d8d8e
7404ace
 
 
63d8d8e
7404ace
63d8d8e
7404ace
 
63d8d8e
 
7404ace
63d8d8e
 
 
7404ace
63d8d8e
 
7404ace
 
63d8d8e
7404ace
63d8d8e
7404ace
63d8d8e
7404ace
63d8d8e
7404ace
 
 
63d8d8e
7404ace
 
63d8d8e
 
7404ace
 
63d8d8e
 
7404ace
63d8d8e
7404ace
63d8d8e
7404ace
63d8d8e
7404ace
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
63d8d8e
7404ace
 
 
 
 
 
 
63d8d8e
 
 
7404ace
63d8d8e
 
 
 
7404ace
63d8d8e
7404ace
 
63d8d8e
7404ace
63d8d8e
 
 
 
 
 
 
 
7404ace
 
 
63d8d8e
 
7404ace
 
 
 
 
 
 
 
 
 
63d8d8e
7404ace
 
63d8d8e
7404ace
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
63d8d8e
7404ace
 
 
 
 
 
 
 
63d8d8e
 
 
 
 
7404ace
63d8d8e
7404ace
 
63d8d8e
7404ace
 
63d8d8e
7404ace
 
 
 
 
 
 
 
 
 
63d8d8e
7404ace
 
 
 
 
 
 
 
 
 
 
 
63d8d8e
 
7404ace
 
 
 
 
 
63d8d8e
 
 
 
7404ace
 
 
 
 
 
 
63d8d8e
7404ace
 
 
 
 
 
 
 
63d8d8e
7404ace
 
 
 
 
63d8d8e
7404ace
 
 
 
 
63d8d8e
 
7404ace
 
 
63d8d8e
 
7404ace
 
63d8d8e
 
7404ace
 
 
 
63d8d8e
7404ace
 
 
 
 
 
63d8d8e
7404ace
63d8d8e
7404ace
63d8d8e
7404ace
63d8d8e
7404ace
 
63d8d8e
 
7404ace
63d8d8e
7404ace
 
 
63d8d8e
7404ace
 
63d8d8e
7404ace
 
63d8d8e
 
7404ace
 
 
 
 
 
 
 
 
 
 
63d8d8e
7404ace
 
 
 
 
 
63d8d8e
7404ace
63d8d8e
7404ace
63d8d8e
7404ace
 
63d8d8e
7404ace
 
 
 
 
63d8d8e
7404ace
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
63d8d8e
 
 
7404ace
 
 
 
 
 
 
 
 
63d8d8e
 
 
 
 
7404ace
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
63d8d8e
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
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
import gradio as gr
import tempfile
import os
import requests
import json
import re
from bs4 import BeautifulSoup
from datetime import datetime
import urllib.parse


# Configuration
SPACE_NAME = "Britannica Wiki Search"
SPACE_DESCRIPTION = ""
SYSTEM_PROMPT = """You are a research aid specializing in academic literature search and analysis. Your expertise spans discovering peer-reviewed sources, assessing research methodologies, synthesizing findings across studies, and delivering properly formatted citations. When responding, anchor claims in specific sources from provided URL contexts, differentiate between direct evidence and interpretive analysis, and note any limitations or contradictory results. Employ clear, accessible language that demystifies complex research, and propose connected research directions when appropriate. Your purpose is to serve as an informed research tool supporting users through initial concept development, exploratory investigation, information collection, and source compilation."""
MODEL = "openai/gpt-4o-mini-search-preview"
GROUNDING_URLS = ["https://www.wikipedia.org/", "https://www.britannica.com/"]
# Get access code from environment variable for security
ACCESS_CODE = os.environ.get("SPACE_ACCESS_CODE", "")
ENABLE_DYNAMIC_URLS = True
ENABLE_VECTOR_RAG = False
RAG_DATA = None

# Get API key from environment - customizable variable name with validation
API_KEY = os.environ.get("OPENROUTER_API_KEY")
if API_KEY:
    API_KEY = API_KEY.strip()  # Remove any whitespace
    if not API_KEY:  # Check if empty after stripping
        API_KEY = None
        
# API Key validation and logging
def validate_api_key():
    """Validate API key configuration with detailed logging"""
    if not API_KEY:
        print(f"⚠️  API KEY CONFIGURATION ERROR:")
        print(f"   Variable name: OPENROUTER_API_KEY")
        print(f"   Status: Not set or empty")
        print(f"   Action needed: Set 'OPENROUTER_API_KEY' in HuggingFace Space secrets")
        print(f"   Expected format: sk-or-xxxxxxxxxx")
        return False
    elif not API_KEY.startswith('sk-or-'):
        print(f"⚠️  API KEY FORMAT WARNING:")
        print(f"   Variable name: OPENROUTER_API_KEY")
        print(f"   Current value: {API_KEY[:10]}..." if len(API_KEY) > 10 else API_KEY)
        print(f"   Expected format: sk-or-xxxxxxxxxx")
        print(f"   Note: OpenRouter keys should start with 'sk-or-'")
        return True  # Still try to use it
    else:
        print(f"βœ… API Key configured successfully")
        print(f"   Variable: OPENROUTER_API_KEY")
        print(f"   Format: Valid OpenRouter key")
        return True

# Validate on startup
try:
    API_KEY_VALID = validate_api_key()
except NameError:
    # During template generation, API_KEY might not be defined yet
    API_KEY_VALID = False

def validate_url_domain(url):
    """Basic URL domain validation"""
    try:
        from urllib.parse import urlparse
        parsed = urlparse(url)
        # Allow common domains and exclude potentially dangerous ones
        allowed_patterns = [
            'wikipedia.org', 'britannica.com', 'edu', 'gov', 'arxiv.org',
            'pubmed.ncbi.nlm.nih.gov', 'scholar.google.com', 'researchgate.net'
        ]
        return any(pattern in parsed.netloc.lower() for pattern in allowed_patterns)
    except:
        return False

def fetch_url_content(url):
    """Fetch content from URL with enhanced error handling"""
    if not validate_url_domain(url):
        return f"URL domain not in allowed list for security: {url}"
    
    try:
        headers = {
            'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36'
        }
        response = requests.get(url, headers=headers, timeout=10)
        response.raise_for_status()
        
        # Parse HTML and extract text
        soup = BeautifulSoup(response.text, 'html.parser')
        
        # Remove script and style elements
        for script in soup(["script", "style"]):
            script.decompose()
        
        # Get text and clean it up
        text = soup.get_text()
        lines = (line.strip() for line in text.splitlines())
        chunks = (phrase.strip() for line in lines for phrase in line.split("  "))
        text = ' '.join(chunk for chunk in chunks if chunk)
        
        # Limit content length
        if len(text) > 4000:
            text = text[:4000] + "..."
        
        return text
    except requests.exceptions.RequestException as e:
        return f"Error fetching {url}: {str(e)}"
    except Exception as e:
        return f"Error processing {url}: {str(e)}"

def extract_urls_from_text(text):
    """Extract URLs from text using regex"""
    url_pattern = r'http[s]?://(?:[a-zA-Z]|[0-9]|[$-_@.&+]|[!*\\(\\),]|(?:%[0-9a-fA-F][0-9a-fA-F]))+'
    urls = re.findall(url_pattern, text)
    
    # Filter out URLs that are clearly not useful
    filtered_urls = []
    for url in urls:
        # Basic filtering - must contain a dot and be reasonably long
        if '.' in url and len(url) > 10:
            filtered_urls.append(url)
    
    return filtered_urls

# Global cache for URL content to avoid re-crawling in generated spaces
_url_content_cache = {}

def get_grounding_context():
    """Fetch context from grounding URLs with caching"""
    if not GROUNDING_URLS:
        return ""
    
    # Create cache key from URLs
    cache_key = tuple(sorted([url for url in GROUNDING_URLS if url and url.strip()]))
    
    # Check cache first
    if cache_key in _url_content_cache:
        return _url_content_cache[cache_key]
    
    context_parts = []
    for i, url in enumerate(GROUNDING_URLS, 1):
        if url.strip():
            content = fetch_url_content(url.strip())
            context_parts.append(f"Context from URL {i} ({url}):\n{content}")
    
    if context_parts:
        result = "\n\n" + "\n\n".join(context_parts) + "\n\n"
    else:
        result = ""
    
    # Cache the result
    _url_content_cache[cache_key] = result
    return result

def export_conversation_to_markdown(conversation_history):
    """Export conversation history to markdown format"""
    if not conversation_history:
        return "No conversation to export."
    
    markdown_content = f"""# Conversation Export
Generated on: {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}
Space: {SPACE_NAME}

---

"""
    
    for i, message in enumerate(conversation_history):
        if isinstance(message, dict):
            # New format: {"role": "user", "content": "..."}
            role = message.get("role", "unknown")
            content = message.get("content", "")
            
            if role == "user":
                markdown_content += f"## πŸ‘€ User\n\n{content}\n\n"
            elif role == "assistant":
                markdown_content += f"## πŸ€– Assistant\n\n{content}\n\n"
        elif isinstance(message, (list, tuple)) and len(message) >= 2:
            # Legacy format: [user_msg, assistant_msg]
            user_msg, assistant_msg = message[0], message[1]
            if user_msg:
                markdown_content += f"## πŸ‘€ User\n\n{user_msg}\n\n"
            if assistant_msg:
                markdown_content += f"## πŸ€– Assistant\n\n{assistant_msg}\n\n"
    
    markdown_content += f"""

---

*Exported from {SPACE_NAME} on {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}*
"""
    
    return markdown_content

# Global variable to store chat history for export
chat_history_store = []

def generate_response(message, history):
    """Generate response using OpenRouter API"""
    
    # Enhanced API key validation with helpful messages
    if not API_KEY:
        error_msg = f"πŸ”‘ **API Key Required**\n\n"
        error_msg += f"Please configure your OpenRouter API key:\n"
        error_msg += f"1. Go to Settings (βš™οΈ) in your HuggingFace Space\n"
        error_msg += f"2. Click 'Variables and secrets'\n"
        error_msg += f"3. Add secret: **OPENROUTER_API_KEY**\n"
        error_msg += f"4. Value: Your OpenRouter API key (starts with `sk-or-`)\n\n"
        error_msg += f"Get your API key at: https://openrouter.ai/keys"
        print(f"❌ API request failed: No API key configured for OPENROUTER_API_KEY")
        return error_msg
    
    # Get grounding context
    grounding_context = get_grounding_context()
    
    # Add RAG context if available
    if ENABLE_VECTOR_RAG and rag_context_provider:
        rag_context = rag_context_provider.get_context(message)
        if rag_context:
            grounding_context += rag_context
    
    # If dynamic URLs are enabled, check message for URLs to fetch
    if ENABLE_DYNAMIC_URLS:
        urls_in_message = extract_urls_from_text(message)
        if urls_in_message:
            dynamic_context_parts = []
            for url in urls_in_message[:2]:  # Limit to 2 URLs to avoid overwhelming context
                content = fetch_url_content(url)
                dynamic_context_parts.append(f"\n\nDynamic context from {url}:\n{content}")
            if dynamic_context_parts:
                grounding_context += "\n".join(dynamic_context_parts)
    
    # Build enhanced system prompt with grounding context
    enhanced_system_prompt = SYSTEM_PROMPT + grounding_context
    
    # Build conversation history for API
    messages = [{
        "role": "system", 
        "content": enhanced_system_prompt
    }]
    
    # Add conversation history - Support both new messages format and legacy tuple format
    for chat in history:
        if isinstance(chat, dict):
            # New format: {"role": "user", "content": "..."}
            messages.append(chat)
        elif isinstance(chat, (list, tuple)) and len(chat) >= 2:
            # Legacy format: ("user msg", "bot msg")
            user_msg, assistant_msg = chat[0], chat[1]
            if user_msg:
                messages.append({"role": "user", "content": user_msg})
            if assistant_msg:
                messages.append({"role": "assistant", "content": assistant_msg})
    
    # Add current message
    messages.append({"role": "user", "content": message})
    
    try:
        # Make API request to OpenRouter
        response = requests.post(
            url="https://openrouter.ai/api/v1/chat/completions",
            headers={
                "Authorization": f"Bearer {API_KEY}",
                "Content-Type": "application/json",
                "HTTP-Referer": "https://huggingface.co",  # Required by some providers
                "X-Title": "HuggingFace Space"  # Helpful for tracking
            },
            json={
                "model": MODEL,
                "messages": messages,
                "temperature": 0.7,
                "max_tokens": 1500,
                "stream": False
            },
            timeout=30
        )
        
        if response.status_code == 200:
            try:
                response_data = response.json()
                assistant_response = response_data['choices'][0]['message']['content']
                print(f"βœ… API request successful")
                return assistant_response
            except (KeyError, IndexError, json.JSONDecodeError) as e:
                print(f"❌ Failed to parse API response: {str(e)}")
                return f"API Error: Failed to parse response - {str(e)}"
        elif response.status_code == 401:
            error_msg = f"πŸ” **Authentication Error**\n\n"
            error_msg += f"Your API key appears to be invalid or expired.\n\n"
            error_msg += f"**Troubleshooting:**\n"
            error_msg += f"1. Check that your **OPENROUTER_API_KEY** secret is set correctly\n"
            error_msg += f"2. Verify your API key at: https://openrouter.ai/keys\n"
            error_msg += f"3. Ensure your key starts with `sk-or-`\n"
            error_msg += f"4. Check that you have credits on your OpenRouter account"
            print(f"❌ API authentication failed: {response.status_code} - {response.text[:200]}")
            return error_msg
        elif response.status_code == 429:
            error_msg = f"⏱️ **Rate Limit Exceeded**\n\n"
            error_msg += f"Too many requests. Please wait a moment and try again.\n\n"
            error_msg += f"If this persists, check your OpenRouter plan limits."
            print(f"❌ Rate limit exceeded: {response.status_code}")
            return error_msg
        else:
            error_msg = f"🚫 **API Error {response.status_code}**\n\n"
            error_msg += f"An unexpected error occurred. Please try again.\n\n"
            error_msg += f"If this persists, the issue may be with the OpenRouter service."
            print(f"❌ API error: {response.status_code} - {response.text[:200]}")
            return error_msg
            
    except requests.exceptions.Timeout:
        error_msg = f"⏰ **Request Timeout**\n\n"
        error_msg += f"The request took too long to complete. Please try again with a shorter message."
        print(f"❌ Request timeout")
        return error_msg
    except requests.exceptions.ConnectionError:
        error_msg = f"🌐 **Connection Error**\n\n"
        error_msg += f"Unable to connect to the AI service. Please check your internet connection and try again."
        print(f"❌ Connection error")
        return error_msg
    except Exception as e:
        error_msg = f"πŸ’₯ **Unexpected Error**\n\n"
        error_msg += f"An unexpected error occurred: {str(e)}\n\n"
        error_msg += f"Please try again or contact support if this persists."
        print(f"❌ Unexpected error: {str(e)}")
        return error_msg

# Global state for access control
access_granted = gr.State(False)

def verify_access_code(code):
    """Verify the access code"""
    global _access_granted_global
    if not ACCESS_CODE:
        return gr.update(visible=False), gr.update(visible=True), True
    
    if code.strip() == ACCESS_CODE.strip():
        _access_granted_global = True
        return gr.update(visible=False), gr.update(visible=True), True
    else:
        return gr.update(value="❌ Incorrect access code. Please try again.", visible=True), gr.update(visible=False), False

def protected_generate_response(message, history):
    """Wrapper that checks access before generating response"""
    global _access_granted_global
    if ACCESS_CODE and not _access_granted_global:
        return "Access denied. Please enter the correct access code."
    return generate_response(message, history)

def store_and_generate_response(message, history):
    """Wrapper that stores conversation and generates response"""
    global chat_history_store
    
    # Generate response
    response = protected_generate_response(message, history)
    
    # Store in global history for export (using new format)
    chat_history_store = history + [{"role": "user", "content": message}, {"role": "assistant", "content": response}]
    
    return response

def export_current_conversation():
    """Export the current conversation"""
    if not chat_history_store:
        return gr.update(visible=False)
    
    markdown_content = export_conversation_to_markdown(chat_history_store)
    
    # Save to temporary file
    with tempfile.NamedTemporaryFile(mode='w', suffix='.md', delete=False) as f:
        f.write(markdown_content)
        temp_file = f.name
    
    return gr.update(value=temp_file, visible=True)

def export_conversation(history):
    """Export conversation from gradio history"""
    if not history:
        return "No conversation to export."
    
    return export_conversation_to_markdown(history)

def get_configuration_status():
    """Get current configuration status for display"""
    status_parts = []
    
    if API_KEY_VALID:
        status_parts.append("βœ… **API Key:** Configured and valid")
    else:
        status_parts.append("❌ **API Key:** Not configured - Set `OPENROUTER_API_KEY` in Space secrets")
    
    status_parts.append(f"πŸ€– **Model:** {MODEL}")
    status_parts.append(f"🌑️ **Temperature:** 0.7")
    status_parts.append(f"πŸ“ **Max Tokens:** 1500")
    
    if GROUNDING_URLS:
        status_parts.append(f"πŸ”— **URL Grounding:** {len(GROUNDING_URLS)} URLs configured")
    
    if ENABLE_DYNAMIC_URLS:
        status_parts.append("πŸ”„ **Dynamic URLs:** Enabled")
        
    if ENABLE_VECTOR_RAG:
        status_parts.append("πŸ“š **Document RAG:** Enabled")
    
    if ACCESS_CODE:
        status_parts.append("πŸ” **Access Control:** Enabled")
    else:
        status_parts.append("🌐 **Access:** Public")
    
    return "\n".join(status_parts)

# Global access tracking
_access_granted_global = not bool(ACCESS_CODE)  # If no access code, grant access

# Create interface with access code protection
with gr.Blocks(title=SPACE_NAME) as demo:
    gr.Markdown(f"# {SPACE_NAME}")
    gr.Markdown(SPACE_DESCRIPTION)
    
    # Configuration status (always visible)
    with gr.Accordion("πŸ“Š Configuration Status", open=not API_KEY_VALID):
        gr.Markdown(get_configuration_status())
    
    # Access code section (only visible if ACCESS_CODE is set)
    if ACCESS_CODE:
        with gr.Column(visible=True) as access_section:
            gr.Markdown("πŸ” **Access Required**")
            gr.Markdown("Please enter the access code to use this space.")
        
        access_input = gr.Textbox(
            label="Access Code",
            placeholder="Enter access code...",
            type="password"
        )
        access_btn = gr.Button("Submit", variant="primary")
        access_error = gr.Markdown(visible=False)
    
    # Main chat interface (hidden until access granted)
    with gr.Column(visible=not bool(ACCESS_CODE)) as chat_section:
        chat_interface = gr.ChatInterface(
            fn=store_and_generate_response,  # Use wrapper function to store history
            title="",  # Title already shown above
            description="",  # Description already shown above
            examples=None,
            type="messages"  # Use modern message format for better compatibility
        )
        
        # Export functionality
        with gr.Row():
            export_btn = gr.Button("πŸ“₯ Export Conversation", variant="secondary", size="sm")
            export_file = gr.File(label="Download Conversation", visible=False)
        
        # Connect export functionality
        export_btn.click(
            export_current_conversation,
            outputs=[export_file]
        )
    
    # Connect access verification
    if ACCESS_CODE:
        access_btn.click(
            verify_access_code,
            inputs=[access_input],
            outputs=[access_error, chat_section, access_granted]
        )
        access_input.submit(
            verify_access_code,
            inputs=[access_input],
            outputs=[access_error, chat_section, access_granted]
        )

if __name__ == "__main__":
    demo.launch()