File size: 7,787 Bytes
1778e91
3d47604
1778e91
 
bd07fc2
1778e91
1a75585
3d47604
1778e91
 
 
 
 
 
 
 
3404e9a
94bdc12
 
 
8b6570a
7e5d5c2
94bdc12
 
3404e9a
 
 
 
 
 
 
 
 
1778e91
 
3404e9a
 
1778e91
3404e9a
8b6570a
 
1778e91
 
3404e9a
1778e91
 
 
 
 
 
 
 
 
 
bd07fc2
1778e91
 
0b6d459
1778e91
 
bd07fc2
 
 
 
 
 
0b6d459
1778e91
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
e94fe41
3d47604
 
e94fe41
3d47604
 
 
 
 
 
 
 
 
a352c42
e94fe41
a352c42
3d47604
a352c42
3d47604
 
a352c42
e94fe41
a352c42
1778e91
3d47604
 
 
 
 
 
 
 
 
f71cd77
bd07fc2
 
 
3d47604
bd07fc2
3d47604
 
 
 
 
bd07fc2
3d47604
8b6570a
1778e91
 
 
 
 
 
 
e9f6da1
3404e9a
e9f6da1
 
 
 
 
8660783
8e0a4a4
e9f6da1
 
 
3d47604
 
 
 
 
 
e9f6da1
 
 
bd07fc2
3d47604
2ca982d
1778e91
 
 
 
 
 
 
 
3d47604
1778e91
 
 
 
 
 
 
 
 
 
 
 
 
 
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
from functools import lru_cache
from typing import List, Tuple, Optional
import aiohttp
import elevenlabs
import time
from contextlib import asynccontextmanager
from logger import setup_logger, log_execution_time, log_async_execution_time
from models import OpenRouterModel

logger = setup_logger("api_clients")

class OpenRouterClient:
    """Handles OpenRouter API interactions with comprehensive logging and error tracking"""
    
    def __init__(self, api_key: str):
        logger.info("Initializing OpenRouter client")
        self.api_key = api_key
        self.base_url = "https://openrouter.ai/api/v1"
        self.headers = {
            "Authorization": f"Bearer {api_key}",
            "HTTP-Referer": "https://localhost:7860",  # Required by OpenRouter
            "X-Title": "URL to Podcast Generator",  # Required by OpenRouter
            "Content-Type": "application/json"
        }
        logger.debug("OpenRouter client initialized successfully")
    
    @property
    def api_key(self):
        return self._api_key

    @api_key.setter
    def api_key(self, value: str):
        if not value or len(value) < 32:
            logger.error("Invalid API key format")
            raise ValueError("Invalid OpenRouter API key")
        self._api_key = value
        # Update headers when API key changes
        self.headers = {
            "Authorization": f"Bearer {value}",
            "HTTP-Referer": "https://localhost:7860",
            "X-Title": "URL to Podcast Generator",
            "Content-Type": "application/json",
        }
        logger.info("OpenRouter API key updated successfully")
    
    @asynccontextmanager
    async def get_session(self):
        logger.debug("Creating new aiohttp session")
        async with aiohttp.ClientSession(headers=self.headers) as session:
            yield session
    
    @lru_cache(maxsize=1)
    async def get_models(self) -> List[Tuple[str, str]]:
        """
        Fetch available models from OpenRouter API using pydantic models
        
        Returns:
            List of tuples containing (model_id, model_id) where both values are the same
        """
        logger.info("Fetching available models from OpenRouter")
        async with self.get_session() as session:
            async with session.get(f"{self.base_url}/models") as response:
                response.raise_for_status()
                data = await response.json()
                models = [OpenRouterModel(**model) for model in data["data"]]
                logger.info(f"Successfully fetched {len(models)} models")
                return [(model.name, model.id) for model in models]

    @log_async_execution_time(logger)
    async def generate_script(self, content: str, prompt: str, model_id: str) -> str:
        """
        Generate a podcast script with detailed progress tracking and validation
        
        Performance metrics and content analysis are logged at each step.
        """
        logger.info(f"Starting script generation with model: {model_id}")
        logger.debug(f"Input metrics - Content: {len(content)} chars, Prompt: {len(prompt)} chars")
        
        # Validate inputs
        if not content or len(content) < 100:
            logger.error("Content too short for meaningful script generation")
            raise ValueError("Insufficient content for script generation")
            
        if not prompt or len(prompt) < 10:
            logger.error("Prompt too short or missing")
            raise ValueError("Please provide a more detailed prompt")
        
        system_prompt = """DO NOT WRITE ASIDES OR ACTION DESCRIPTIONS, YOU WRITE DIALOG ONLY!!. You are an expert podcast dialog writer with these specific requirements:
1. Start the content immediately - no introductions, timestamps, or meta-commentary
2. Write in a natural, conversational tone suitable for speaking
3. Structure the podcast dialog with clear paragraphs and natural pauses
4. Use informal language while maintaining professionalism
5. Focus on narrative flow and engaging delivery
6. Keep technical terms simple and explained
7. Include vocal variety cues through punctuation
8. Write as if speaking directly to the listener
9. Use storytelling techniques to maintain interest
10. Do not add muscial queues or sound effects
11. Add host and show intros, outros, and transitions as needed
"""

        user_prompt = f"""Write podcast dialog for a single person based on the following content. Make it engaging and easy to follow.

Context: {prompt if prompt else 'Create an informative and engaging podcast episode'}

Content:
{content}

Format the dialog  in a clear, readable way with appropriate spacing. Do not add asides or action descriptions. Only add spoken dialog."""

        try:
            request_data = {
                "model": model_id,
                "messages": [
                    {"role": "system", "content": system_prompt},
                    {"role": "user", "content": user_prompt}
                ],
                "temperature": 0.7,
                "max_tokens": 2000
            }
            
            async with self.get_session() as session:
                async with session.post(
                    f"{self.base_url}/chat/completions",
                    json=request_data
                ) as response:
                    if response.status != 200:
                        error_text = await response.text()
                        logger.error(f"OpenRouter API error: {error_text}")
                        raise ValueError(f"API request failed: {error_text}")
                    
                    data = await response.json()
                    return data['choices'][0]['message']['content']
                    
        except Exception as e:
            logger.error(f"Script generation failed", exc_info=True)
            raise

class ElevenLabsClient:
    def __init__(self, api_key: str):
        self.api_key = api_key
        elevenlabs.set_api_key(api_key)

    def get_voices(self) -> List[Tuple[str, str]]:
        """
        Synchronously get available voices from ElevenLabs
        
        Returns:
            List of tuples containing (voice_id, display_name)
            where display_name shows the name and description but not the ID
        """
        try:
            voices = elevenlabs.voices()
            return [(
                
                f"{voice.name} ({voice.labels.get('accent', 'No accent')})" + 
                (f" - {voice.description[:50]}..." if voice.description else ""),
                voice.voice_id  # Value (hidden from user)
            ) for voice in voices]
        except Exception as e:
            logger.error("Failed to fetch voices from ElevenLabs", exc_info=True)
            raise

    def generate_audio(self, text: str, voice_id: str):
        """Generate audio synchronously"""
        logger.info(f"Starting audio generation with voice: {voice_id}")
        logger.debug(f"Input text length: {len(text)} chars")
        
        if len(text) > 5000:
            logger.warning(f"Long text detected ({len(text)} chars), may impact performance")
        
        try:
            start_time = time.time()
            audio = elevenlabs.generate(
                text=text,
                voice=voice_id,
                model="eleven_monolingual_v1"
            )
            
            duration = time.time() - start_time
            audio_size = len(audio)
            logger.info(f"Audio generated: {audio_size} bytes in {duration:.2f} seconds")
            logger.debug(f"Audio generation rate: {len(text)/duration:.2f} chars/second")
            
            return audio
        except Exception as e:
            logger.error("Audio generation failed", exc_info=True)
            raise