webscout 8.2.2__py3-none-any.whl → 2026.1.19__py3-none-any.whl

webscout/Provider/TTS/elevenlabs.py

@@ -1,111 +1,189 @@
-import time
-import requests
-import pathlib
-import tempfile
-from io import BytesIO
-from webscout import exceptions
-from webscout.AIbase import TTSProvider
-from webscout.litagent import LitAgent
-from concurrent.futures import ThreadPoolExecutor, as_completed
-from . import utils
-
-class ElevenlabsTTS(TTSProvider): 
-    """
-    Text-to-speech provider using the ElevenlabsTTS API.
-    """
-    # Request headers
-    headers: dict[str, str] = {
-        "User-Agent": LitAgent().random()
-    }
-    all_voices: dict[str, str] = {"Brian": "nPczCjzI2devNBz1zQrb", "Alice":"Xb7hH8MSUJpSbSDYk0k2", "Bill":"pqHfZKP75CvOlQylNhV4", "Callum":"N2lVS1w4EtoT3dr4eOWO", "Charlie":"IKne3meq5aSn9XLyUdCD", "Charlotte":"XB0fDUnXU5powFXDhCwa", "Chris":"iP95p4xoKVk53GoZ742B", "Daniel":"onwK4e9ZLuTAKqWW03F9", "Eric":"cjVigY5qzO86Huf0OWal", "George":"JBFqnCBsd6RMkjVDRZzb", "Jessica":"cgSgspJ2msm6clMCkdW9", "Laura":"FGY2WhTYpPnrIDTdsKH5", "Liam":"TX3LPaxmHKxFdv7VOQHJ", "Lily":"pFZP5JQG7iQjIQuC4Bku", "Matilda":"XrExE9yKIg1WjnnlVkGX", "Sarah":"EXAVITQu4vr4xnSDxMaL", "Will":"bIHbv24MWmeRgasZH58o"}
-
-    def __init__(self, timeout: int = 20, proxies: dict = None):
-        """Initializes the ElevenlabsTTS TTS client."""
-        self.session = requests.Session()
-        self.session.headers.update(self.headers)
-        if proxies:
-            self.session.proxies.update(proxies)
-        self.timeout = timeout
-        self.params = {'allow_unauthenticated': '1'}
-        self.temp_dir = tempfile.mkdtemp(prefix="webscout_tts_")
-
-    def tts(self, text: str, voice: str = "Brian", verbose:bool = True) -> str:
-        """
-        Converts text to speech using the ElevenlabsTTS API and saves it to a file.
-        """
-        assert (
-            voice in self.all_voices
-        ), f"Voice '{voice}' not one of [{', '.join(self.all_voices.keys())}]"
-
-        filename = pathlib.Path(tempfile.mktemp(suffix=".mp3", dir=self.temp_dir))
-
-        # Split text into sentences
-        sentences = utils.split_sentences(text)
-
-        # Function to request audio for each chunk
-        def generate_audio_for_chunk(part_text: str, part_number: int):
-            while True:
-                try:
-                    json_data = {'text': part_text, 'model_id': 'eleven_multilingual_v2'}
-                    response = self.session.post(f'https://api.elevenlabs.io/v1/text-to-speech/{self.all_voices[voice]}',params=self.params, headers=self.headers, json=json_data, timeout=self.timeout)
-                    response.raise_for_status()
-
-                    # Check if the request was successful
-                    if response.ok and response.status_code == 200:
-                        if verbose:
-                            print(f"[debug] Chunk {part_number} processed successfully")
-                        return part_number, response.content
-                    else:
-                        if verbose:
-                            print(f"[debug] No data received for chunk {part_number}. Retrying...")
-                except requests.RequestException as e:
-                    if verbose:
-                        print(f"[debug] Error for chunk {part_number}: {e}. Retrying...")
-                    time.sleep(1)
-        try:
-            # Using ThreadPoolExecutor to handle requests concurrently
-            with ThreadPoolExecutor() as executor:
-                futures = {executor.submit(generate_audio_for_chunk, sentence.strip(), chunk_num): chunk_num 
-                        for chunk_num, sentence in enumerate(sentences, start=1)}
-                
-                # Dictionary to store results with order preserved
-                audio_chunks = {}
-
-                for future in as_completed(futures):
-                    chunk_num = futures[future]
-                    try:
-                        part_number, audio_data = future.result()
-                        audio_chunks[part_number] = audio_data  # Store the audio data in correct sequence
-                    except Exception as e:
-                        if verbose:
-                            print(f"[debug] Failed to generate audio for chunk {chunk_num}: {e}")
-
-            # Combine audio chunks in the correct sequence
-            combined_audio = BytesIO()
-            for part_number in sorted(audio_chunks.keys()):
-                combined_audio.write(audio_chunks[part_number])
-                if verbose:
-                    print(f"[debug] Added chunk {part_number} to the combined file.")
-
-            # Save the combined audio data to a single file
-            with open(filename, 'wb') as f:
-                f.write(combined_audio.getvalue())
-            if verbose:
-                print(f"[debug] Final Audio Saved as {filename}")
-            return filename.as_posix()
-
-        except requests.exceptions.RequestException as e:
-            if verbose:
-                print(f"[debug] Failed to perform the operation: {e}")
-            raise exceptions.FailedToGenerateResponseError(
-                f"Failed to perform the operation: {e}"
-            )
-
-# Example usage
-if __name__ == "__main__":
-    elevenlabs = ElevenlabsTTS()
-    text = "This is a test of the ElevenlabsTTS text-to-speech API. It supports multiple sentences and advanced logging."
-
-    print("[debug] Generating audio...")
-    audio_file = elevenlabs.tts(text, voice="Brian")
-    print(f"Audio saved to: {audio_file}")
+##################################################################################
+##  ElevenLabs TTS Provider                                                      ##
+##################################################################################
+import os
+import pathlib
+import tempfile
+from concurrent.futures import ThreadPoolExecutor, as_completed
+from typing import Any, Optional, Union, cast
+
+import requests
+from litprinter import ic
+
+from webscout import exceptions
+from webscout.litagent import LitAgent
+
+try:
+    from . import utils
+    from .base import BaseTTSProvider
+except ImportError:
+    # Handle direct execution
+    import os
+    import sys
+    sys.path.insert(0, os.path.join(os.path.dirname(__file__), '..', '..', '..'))
+    from webscout.Provider.TTS import utils
+    from webscout.Provider.TTS.base import BaseTTSProvider
+
+class ElevenlabsTTS(BaseTTSProvider):
+    """
+    Text-to-speech provider using the ElevenLabs API.
+
+    This provider supports both authenticated (with API key) and
+    unauthenticated (limited) usage if available.
+    """
+    required_auth = True
+
+    # Supported models
+    SUPPORTED_MODELS = [
+        "eleven_multilingual_v2",
+        "eleven_flash_v2_5",
+        "eleven_flash_v2",
+        "eleven_turbo_v2_5",
+        "eleven_turbo_v2",
+        "eleven_monolingual_v1"
+    ]
+
+    # Request headers
+    headers: dict[str, str] = {
+        "User-Agent": LitAgent().random(),
+        "Accept": "audio/mpeg",
+        "Content-Type": "application/json",
+    }
+
+    # ElevenLabs voices
+    SUPPORTED_VOICES = [
+        "brian", "alice", "bill", "callum", "charlie", "charlotte",
+        "chris", "daniel", "eric", "george", "jessica", "laura",
+        "liam", "lily", "matilda", "sarah", "will"
+    ]
+
+    # Voice mapping
+    voice_mapping = {
+        "brian": "nPczCjzI2devNBz1zQrb",
+        "alice": "Xb7hH8MSUJpSbSDYk0k2",
+        "bill": "pqHfZKP75CvOlQylNhV4",
+        "callum": "N2lVS1w4EtoT3dr4eOWO",
+        "charlie": "IKne3meq5aSn9XLyUdCD",
+        "charlotte": "XB0fDUnXU5powFXDhCwa",
+        "chris": "iP95p4xoKVk53GoZ742B",
+        "daniel": "onwK4e9ZLuTAKqWW03F9",
+        "eric": "cjVigY5qzO86Huf0OWal",
+        "george": "JBFqnCBsd6RMkjVDRZzb",
+        "jessica": "cgSgspJ2msm6clMCkdW9",
+        "laura": "FGY2WhTYpPnrIDTdsKH5",
+        "liam": "TX3LPaxmHKxFdv7VOQHJ",
+        "lily": "pFZP5JQG7iQjIQuC4Bku",
+        "matilda": "XrExE9yKIg1WjnnlVkGX",
+        "sarah": "EXAVITQu4vr4xnSDxMaL",
+        "will": "bIHbv24MWmeRgasZH58o"
+    }
+
+    def __init__(self, api_key: Optional[str] = None, timeout: int = 30, proxies: Optional[dict] = None):
+        """
+        Initialize the ElevenLabs TTS client.
+
+        Args:
+            api_key (str): ElevenLabs API key. If None, tries to use unauthenticated endpoint.
+            timeout (int): Request timeout in seconds.
+            proxies (dict): Proxy configuration.
+        """
+        super().__init__()
+        self.api_key = api_key or os.getenv("ELEVENLABS_API_KEY")
+        self.api_url = "https://api.elevenlabs.io/v1/text-to-speech"
+        self.session = requests.Session()
+        self.session.headers.update(self.headers)
+        if self.api_key:
+            self.session.headers.update({"xi-api-key": self.api_key})
+        if proxies:
+            self.session.proxies.update(proxies)
+        self.timeout = timeout
+        self.default_voice = "brian"
+
+    def tts(
+        self,
+        text: str,
+        voice: str | None = None,
+        verbose: bool = False,
+        **kwargs
+    ) -> str:
+        """
+        Convert text to speech using ElevenLabs API.
+        """
+        model = kwargs.get("model", "eleven_multilingual_v2")
+        response_format = kwargs.get("response_format", "mp3")
+        if voice is None:
+            voice = "brian"
+        if verbose is False:
+            verbose = kwargs.get("verbose", True)
+        if not text:
+            raise ValueError("Input text must be a non-empty string")
+
+        voice_id = self.voice_mapping.get(voice.lower(), voice)
+
+        file_extension = f".{response_format}"
+        temp_file = tempfile.NamedTemporaryFile(suffix=file_extension, dir=self.temp_dir, delete=False)
+        temp_file.close()
+        filename = pathlib.Path(temp_file.name)
+
+        sentences = utils.split_sentences(text)
+        if verbose:
+            ic.configureOutput(prefix='DEBUG| ')
+            ic(f"ElevenlabsTTS: Processing {len(sentences)} chunks")
+
+        def generate_chunk(part_text: str, part_num: int):
+            payload = {
+                "text": part_text,
+                "model_id": model,
+                "voice_settings": {
+                    "stability": 0.5,
+                    "similarity_boost": 0.5
+                }
+            }
+            url = f"{self.api_url}/{voice_id}"
+            params = {}
+            if not self.api_key:
+                # Some public endpoints might still work without key but they are very restricted
+                params['allow_unauthenticated'] = '1'
+
+            response = self.session.post(url, json=payload, params=params, timeout=self.timeout)
+            if response.status_code == 401 and not self.api_key:
+                raise exceptions.FailedToGenerateResponseError("ElevenLabs requires an API key for this request.")
+            response.raise_for_status()
+            return part_num, response.content
+
+        try:
+            with ThreadPoolExecutor(max_workers=2) as executor:
+                futures = {executor.submit(generate_chunk, s.strip(), i): i for i, s in enumerate(sentences)}
+                audio_chunks = {}
+                for future in as_completed(futures):
+                    idx, data = future.result()
+                    audio_chunks[idx] = data
+
+                with open(filename, 'wb') as f:
+                    for i in sorted(audio_chunks.keys()):
+                        f.write(audio_chunks[i])
+
+                if verbose:
+                    ic.configureOutput(prefix='INFO| ')
+                    ic(f"Audio saved to {filename}")
+                return str(filename)
+        except Exception as e:
+            raise exceptions.FailedToGenerateResponseError(f"ElevenLabs TTS failed: {e}")
+
+    def create_speech(
+        self,
+        input_text: str,
+        model: Optional[str] = None,
+        voice: Optional[str] = None,
+        response_format: Optional[str] = None,
+        instructions: Optional[str] = None,
+        verbose: bool = False,
+        **kwargs: Any
+    ) -> str:
+        return self.tts(
+            text=input_text,
+            model=model or "eleven_multilingual_v2",
+            voice=voice or "brian",
+            response_format=response_format or "mp3",
+            verbose=verbose
+        )

webscout/Provider/TTS/freetts.py

@@ -0,0 +1,218 @@
+##################################################################################
+##  FreeTTS Provider                                                             ##
+##################################################################################
+import pathlib
+import tempfile
+import time
+from typing import Any, Optional, Union, cast
+
+import requests
+from litprinter import ic
+
+from webscout import exceptions
+from webscout.litagent import LitAgent
+from webscout.Provider.TTS.base import BaseTTSProvider
+
+
+class FreeTTS(BaseTTSProvider):
+    """
+    Text-to-speech provider using the FreeTTS.ru API.
+
+    Features:
+    - Multiple languages (Russian, English, Ukrainian, etc.)
+    - High-quality neural voices
+    - Supports long texts via polling
+    """
+    required_auth = False
+
+    headers = {
+        "Accept": "*/*",
+        "Accept-Language": "en-US,en;q=0.9",
+        "Content-Type": "application/json",
+        "Origin": "https://freetts.ru",
+        "Referer": "https://freetts.ru/",
+        "User-Agent": LitAgent().random()
+    }
+
+    # Supported formats
+    SUPPORTED_FORMATS = ["mp3"]
+
+    def __init__(self, lang="ru-RU", timeout: int = 30, proxies: Optional[dict] = None):
+        """
+        Initialize the FreeTTS TTS client.
+
+        Args:
+            lang (str): Language code for voice filtering
+            timeout (int): Request timeout in seconds
+            proxies (dict): Proxy configuration
+        """
+        super().__init__()
+        self.lang = lang
+        self.base_url = "https://freetts.ru"
+        self.api_url = f"{self.base_url}/api/synthesis"
+        self.list_url = f"{self.base_url}/api/list"
+        self.history_url = f"{self.base_url}/api/history"
+
+        self.session = requests.Session()
+        self.session.headers.update(self.headers)
+        if proxies:
+            self.session.proxies.update(proxies)
+        self.timeout = timeout
+
+        self.voices = {}
+        self.load_voices()
+        # Set default voice to first available for requested lang
+        self.default_voice = self._get_default_voice()
+
+    def load_voices(self):
+        """Load voice data from the API."""
+        try:
+            response = self.session.get(self.list_url, timeout=self.timeout)
+            if response.status_code == 200:
+                data = response.json()
+                if data.get("status") == "success":
+                    voices_list = data.get("data", {}).get("voices", [])
+                    for voice in voices_list:
+                        v_id = voice.get("id")
+                        v_lang = voice.get("lang")
+                        v_name = voice.get("name")
+                        if v_id:
+                            self.voices[v_id] = {
+                                "lang": v_lang,
+                                "name": v_name,
+                                "sex": voice.get("sex")
+                            }
+                            if v_id not in self.SUPPORTED_VOICES:
+                                self.SUPPORTED_VOICES.append(v_id)
+        except Exception as e:
+            ic.configureOutput(prefix='WARNING| ')
+            ic(f"Error loading FreeTTS voices: {e}")
+
+    def _get_default_voice(self):
+        for v_id, info in self.voices.items():
+            if info["lang"] == self.lang:
+                return v_id
+        return next(iter(self.voices.keys())) if self.voices else None
+
+    def get_available_voices(self):
+        """Return all available voices for the current language."""
+        return [v_id for v_id, info in self.voices.items() if info["lang"] == self.lang]
+
+    def tts(
+        self,
+        text: str,
+        voice: Optional[str] = None,
+        verbose: bool = False,
+        **kwargs
+    ) -> str:
+        """
+        Convert text to speech.
+        """
+        response_format = kwargs.get("response_format", "mp3")
+        if not text:
+            raise ValueError("Input text must be a non-empty string")
+
+        voice = voice or self.default_voice
+        if not voice:
+            raise ValueError("No voices available")
+
+        payload = {
+            "text": text,
+            "voiceid": voice,
+            "ext": response_format
+        }
+
+        try:
+            # Step 1: Start synthesis
+            if verbose:
+                ic.configureOutput(prefix='DEBUG| ')
+                ic(f"FreeTTS: Starting synthesis for voice {voice}")
+
+            response = self.session.post(self.api_url, json=payload, timeout=self.timeout)
+            response.raise_for_status()
+
+            # Step 2: Poll for completion
+            max_polls = 20
+            poll_interval = 2
+
+            for i in range(max_polls):
+                poll_resp = self.session.get(self.api_url, timeout=self.timeout)
+                poll_resp.raise_for_status()
+                poll_data = poll_resp.json()
+
+                if poll_data.get("status") == 200 or poll_data.get("message") == "Обработка: 100%":
+                    if verbose:
+                        ic.configureOutput(prefix='DEBUG| ')
+                        ic("FreeTTS: Synthesis completed")
+                    break
+
+                if verbose:
+                    ic.configureOutput(prefix='DEBUG| ')
+                    ic(f"FreeTTS: {poll_data.get('message', 'Processing...')}")
+
+                time.sleep(poll_interval)
+            else:
+                raise exceptions.FailedToGenerateResponseError("FreeTTS synthesis timed out")
+
+            # Step 3: Get result from history
+            hist_resp = self.session.get(self.history_url, timeout=self.timeout)
+            hist_resp.raise_for_status()
+            hist_data = hist_resp.json()
+
+            if hist_data.get("status") == "success":
+                history = hist_data.get("data", [])
+                if history:
+                    # Find matching item in history
+                    # For simplicity, take the first one as it's the latest
+                    audio_url = history[0].get("url")
+                    if audio_url:
+                        if not audio_url.startswith("http"):
+                            audio_url = self.base_url + audio_url
+
+                        # Download the file
+                        audio_file_resp = self.session.get(audio_url, timeout=self.timeout)
+                        audio_file_resp.raise_for_status()
+
+                        # Save to temp file
+                        temp_file = tempfile.NamedTemporaryFile(suffix=f".{response_format}", dir=self.temp_dir, delete=False)
+                        temp_file.close()
+                        filename = pathlib.Path(temp_file.name)
+                        with open(filename, "wb") as f:
+                            f.write(audio_file_resp.content)
+
+                        if verbose:
+                            ic.configureOutput(prefix='DEBUG| ')
+                            ic(f"FreeTTS: Audio saved to {filename}")
+                        return str(filename)
+
+            raise exceptions.FailedToGenerateResponseError("Failed to retrieve audio URL from history")
+
+        except Exception as e:
+            raise exceptions.FailedToGenerateResponseError(f"FreeTTS failed: {e}")
+
+    def create_speech(
+        self,
+        input_text: str,
+        model: Optional[str] = None,
+        voice: Optional[str] = None,
+        response_format: Optional[str] = None,
+        instructions: Optional[str] = None,
+        verbose: bool = False,
+        **kwargs: Any
+    ) -> str:
+        return self.tts(
+            text=input_text,
+            model=model,
+            voice=voice,
+            response_format=response_format or "mp3",
+            instructions=instructions,
+            verbose=verbose
+        )
+
+if __name__ == "__main__":
+    tts = FreeTTS()
+    try:
+        path = tts.tts("Привет, это проверка FreeTTS.", verbose=True)
+        print(f"Saved to: {path}")
+    except Exception as e:
+        print(f"Error: {e}")