smallestai 2.1.0__py3-none-any.whl → 3.0.0__py3-none-any.whl

smallestai/atoms/rest.py

@@ -0,0 +1,258 @@
+# coding: utf-8
+
+"""
+    Agent Management API
+
+    API for managing agents, their templates, and call logs
+
+    The version of the OpenAPI document: 1.0.0
+    Generated by OpenAPI Generator (https://openapi-generator.tech)
+
+    Do not edit the class manually.
+"""  # noqa: E501
+
+
+import io
+import json
+import re
+import ssl
+
+import urllib3
+
+from smallestai.atoms.exceptions import ApiException, ApiValueError
+
+SUPPORTED_SOCKS_PROXIES = {"socks5", "socks5h", "socks4", "socks4a"}
+RESTResponseType = urllib3.HTTPResponse
+
+
+def is_socks_proxy_url(url):
+    if url is None:
+        return False
+    split_section = url.split("://")
+    if len(split_section) < 2:
+        return False
+    else:
+        return split_section[0].lower() in SUPPORTED_SOCKS_PROXIES
+
+
+class RESTResponse(io.IOBase):
+
+    def __init__(self, resp) -> None:
+        self.response = resp
+        self.status = resp.status
+        self.reason = resp.reason
+        self.data = None
+
+    def read(self):
+        if self.data is None:
+            self.data = self.response.data
+        return self.data
+
+    def getheaders(self):
+        """Returns a dictionary of the response headers."""
+        return self.response.headers
+
+    def getheader(self, name, default=None):
+        """Returns a given response header."""
+        return self.response.headers.get(name, default)
+
+
+class RESTClientObject:
+
+    def __init__(self, configuration) -> None:
+        # urllib3.PoolManager will pass all kw parameters to connectionpool
+        # https://github.com/shazow/urllib3/blob/f9409436f83aeb79fbaf090181cd81b784f1b8ce/urllib3/poolmanager.py#L75  # noqa: E501
+        # https://github.com/shazow/urllib3/blob/f9409436f83aeb79fbaf090181cd81b784f1b8ce/urllib3/connectionpool.py#L680  # noqa: E501
+        # Custom SSL certificates and client certificates: http://urllib3.readthedocs.io/en/latest/advanced-usage.html  # noqa: E501
+
+        # cert_reqs
+        if configuration.verify_ssl:
+            cert_reqs = ssl.CERT_REQUIRED
+        else:
+            cert_reqs = ssl.CERT_NONE
+
+        pool_args = {
+            "cert_reqs": cert_reqs,
+            "ca_certs": configuration.ssl_ca_cert,
+            "cert_file": configuration.cert_file,
+            "key_file": configuration.key_file,
+            "ca_cert_data": configuration.ca_cert_data,
+        }
+        if configuration.assert_hostname is not None:
+            pool_args['assert_hostname'] = (
+                configuration.assert_hostname
+            )
+
+        if configuration.retries is not None:
+            pool_args['retries'] = configuration.retries
+
+        if configuration.tls_server_name:
+            pool_args['server_hostname'] = configuration.tls_server_name
+
+
+        if configuration.socket_options is not None:
+            pool_args['socket_options'] = configuration.socket_options
+
+        if configuration.connection_pool_maxsize is not None:
+            pool_args['maxsize'] = configuration.connection_pool_maxsize
+
+        # https pool manager
+        self.pool_manager: urllib3.PoolManager
+
+        if configuration.proxy:
+            if is_socks_proxy_url(configuration.proxy):
+                from urllib3.contrib.socks import SOCKSProxyManager
+                pool_args["proxy_url"] = configuration.proxy
+                pool_args["headers"] = configuration.proxy_headers
+                self.pool_manager = SOCKSProxyManager(**pool_args)
+            else:
+                pool_args["proxy_url"] = configuration.proxy
+                pool_args["proxy_headers"] = configuration.proxy_headers
+                self.pool_manager = urllib3.ProxyManager(**pool_args)
+        else:
+            self.pool_manager = urllib3.PoolManager(**pool_args)
+
+    def request(
+        self,
+        method,
+        url,
+        headers=None,
+        body=None,
+        post_params=None,
+        _request_timeout=None
+    ):
+        """Perform requests.
+
+        :param method: http request method
+        :param url: http request url
+        :param headers: http request headers
+        :param body: request json body, for `application/json`
+        :param post_params: request post parameters,
+                            `application/x-www-form-urlencoded`
+                            and `multipart/form-data`
+        :param _request_timeout: timeout setting for this request. If one
+                                 number provided, it will be total request
+                                 timeout. It can also be a pair (tuple) of
+                                 (connection, read) timeouts.
+        """
+        method = method.upper()
+        assert method in [
+            'GET',
+            'HEAD',
+            'DELETE',
+            'POST',
+            'PUT',
+            'PATCH',
+            'OPTIONS'
+        ]
+
+        if post_params and body:
+            raise ApiValueError(
+                "body parameter cannot be used with post_params parameter."
+            )
+
+        post_params = post_params or {}
+        headers = headers or {}
+
+        timeout = None
+        if _request_timeout:
+            if isinstance(_request_timeout, (int, float)):
+                timeout = urllib3.Timeout(total=_request_timeout)
+            elif (
+                    isinstance(_request_timeout, tuple)
+                    and len(_request_timeout) == 2
+                ):
+                timeout = urllib3.Timeout(
+                    connect=_request_timeout[0],
+                    read=_request_timeout[1]
+                )
+
+        try:
+            # For `POST`, `PUT`, `PATCH`, `OPTIONS`, `DELETE`
+            if method in ['POST', 'PUT', 'PATCH', 'OPTIONS', 'DELETE']:
+
+                # no content type provided or payload is json
+                content_type = headers.get('Content-Type')
+                if (
+                    not content_type
+                    or re.search('json', content_type, re.IGNORECASE)
+                ):
+                    request_body = None
+                    if body is not None:
+                        request_body = json.dumps(body)
+                    r = self.pool_manager.request(
+                        method,
+                        url,
+                        body=request_body,
+                        timeout=timeout,
+                        headers=headers,
+                        preload_content=False
+                    )
+                elif content_type == 'application/x-www-form-urlencoded':
+                    r = self.pool_manager.request(
+                        method,
+                        url,
+                        fields=post_params,
+                        encode_multipart=False,
+                        timeout=timeout,
+                        headers=headers,
+                        preload_content=False
+                    )
+                elif content_type == 'multipart/form-data':
+                    # must del headers['Content-Type'], or the correct
+                    # Content-Type which generated by urllib3 will be
+                    # overwritten.
+                    del headers['Content-Type']
+                    # Ensures that dict objects are serialized
+                    post_params = [(a, json.dumps(b)) if isinstance(b, dict) else (a,b) for a, b in post_params]
+                    r = self.pool_manager.request(
+                        method,
+                        url,
+                        fields=post_params,
+                        encode_multipart=True,
+                        timeout=timeout,
+                        headers=headers,
+                        preload_content=False
+                    )
+                # Pass a `string` parameter directly in the body to support
+                # other content types than JSON when `body` argument is
+                # provided in serialized form.
+                elif isinstance(body, str) or isinstance(body, bytes):
+                    r = self.pool_manager.request(
+                        method,
+                        url,
+                        body=body,
+                        timeout=timeout,
+                        headers=headers,
+                        preload_content=False
+                    )
+                elif headers['Content-Type'].startswith('text/') and isinstance(body, bool):
+                    request_body = "true" if body else "false"
+                    r = self.pool_manager.request(
+                        method,
+                        url,
+                        body=request_body,
+                        preload_content=False,
+                        timeout=timeout,
+                        headers=headers)
+                else:
+                    # Cannot generate the request from given parameters
+                    msg = """Cannot prepare a request message for provided
+                             arguments. Please check that your arguments match
+                             declared content type."""
+                    raise ApiException(status=0, reason=msg)
+            # For `GET`, `HEAD`
+            else:
+                r = self.pool_manager.request(
+                    method,
+                    url,
+                    fields={},
+                    timeout=timeout,
+                    headers=headers,
+                    preload_content=False
+                )
+        except urllib3.exceptions.SSLError as e:
+            msg = "\n".join([type(e).__name__, str(e)])
+            raise ApiException(status=0, reason=msg)
+
+        return RESTResponse(r)

smallestai/waves/__init__.py

@@ -0,0 +1,5 @@
+from smallestai.waves.waves_client import WavesClient
+from smallestai.waves.async_waves_client import AsyncWavesClient
+from smallestai.waves.stream_tts import TextToAudioStream
+
+__all__ = ["WavesClient", "AsyncWavesClient", "TextToAudioStream"]

smallest/async_tts.py → smallestai/waves/async_waves_client.py

@@ -4,14 +4,14 @@ import json
 import aiohttp
 import aiofiles
 import requests
-from typing import Optional, Union, List
+from typing import Optional, Union, List, AsyncIterator
 
-from smallest.exceptions import TTSError, APIError
-from smallest.utils import (TTSOptions, validate_input, preprocess_text, add_wav_header, chunk_text,
+from smallestai.waves.exceptions import TTSError, APIError
+from smallestai.waves.utils import (TTSOptions, validate_input, preprocess_text, add_wav_header, chunk_text,
                      get_smallest_languages, get_smallest_models, ALLOWED_AUDIO_EXTENSIONS, API_BASE_URL)
 
 
-class AsyncSmallest:
+class AsyncWavesClient:
     def __init__(
         self,
         api_key: str = None,
@@ -19,6 +19,9 @@ class AsyncSmallest:
         sample_rate: Optional[int] = 24000,
         voice_id: Optional[str] = "emily",
         speed: Optional[float] = 1.0,
+        consistency: Optional[float] = 0.5,
+        similarity: Optional[float] = 0.0,
+        enhancement: Optional[int] = 1,
         add_wav_header: Optional[bool] = True
     ) -> None:
         """
@@ -34,6 +37,9 @@ class AsyncSmallest:
         - sample_rate (int): The sample rate for the audio output.
         - voice_id (TTSVoices): The voice to be used for synthesis.
         - speed (float): The speed of the speech synthesis.
+        - consistency (float): This parameter controls word repetition and skipping. Decrease it to prevent skipped words, and increase it to prevent repetition. Only supported in `lightning-large` model. Range - [0, 1]
+        - similarity (float): This parameter controls the similarity between the synthesized audio and the reference audio. Increase it to make the speech more similar to the reference audio. Only supported in `lightning-large` model. Range - [0, 1]
+        - enhancement (int): Enhances speech quality at the cost of increased latency. Only supported in `lightning-large` model. Range - [0, 2].
         - add_wav_header (bool): Whether to add a WAV header to the output audio.
 
         Methods:
@@ -45,7 +51,7 @@ class AsyncSmallest:
         self.api_key = api_key or os.environ.get("SMALLEST_API_KEY")
         if not self.api_key:
             raise TTSError()
-        if model == "lightning-large":
+        if model == "lightning-large" and voice_id is None:
             voice_id = "lakshya"
 
         self.chunk_size = 250
@@ -56,7 +62,10 @@ class AsyncSmallest:
             voice_id=voice_id,
             api_key=self.api_key,
             add_wav_header=add_wav_header,
-            speed=speed
+            speed=speed,
+            consistency=consistency,
+            similarity=similarity,
+            enhancement=enhancement
         )
         self.session = None
 
@@ -121,27 +130,25 @@ class AsyncSmallest:
     async def synthesize(
             self,
             text: str,
-            consistency: Optional[float] = 0.5,
-            similarity: Optional[float] = 0,
-            enhancement: Optional[bool] = False,
+            stream: Optional[bool] = False,
             save_as: Optional[str] = None,
             **kwargs
-        ) -> Union[bytes, None]:
+        ) -> Union[bytes, None, AsyncIterator[bytes]]:
         """
         Asynchronously synthesize speech from the provided text.
 
         Args:
         - text (str): The text to be converted to speech.
+        - stream (Optional[bool]): If True, returns an iterator yielding audio chunks instead of a full byte array.
         - save_as (Optional[str]): If provided, the synthesized audio will be saved to this file path.
                                    The file must have a .wav extension.
-        - consistency (Optional[float]): This parameter controls word repetition and skipping. Decrease it to prevent skipped words, and increase it to prevent repetition. Only supported in `lightning-large` model.
-        - similarity (Optional[float]): This parameter controls the similarity between the synthesized audio and the reference audio. Increase it to make the speech more similar to the reference audio. Only supported in `lightning-large` model.
-        - enhancement (Optional[bool]): Enhances speech quality at the cost of increased latency. Only supported in `lightning-large` model.
         - kwargs: Additional optional parameters to override `__init__` options for this call.
 
         Returns:
-        - Union[bytes, None]: The synthesized audio content in bytes if `save_as` is not specified;
-                              otherwise, returns None after saving the audio to the specified file.
+        - Union[bytes, None, Iterator[bytes]]: 
+            - If `stream=True`, returns an iterator yielding audio chunks.
+            - If `save_as` is provided, saves the file and returns None.
+            - Otherwise, returns the synthesized audio content as bytes.
 
         Raises:
         - TTSError: If the provided file name does not have a .wav extension when `save_as` is specified.
@@ -165,44 +172,50 @@ class AsyncSmallest:
             for key, value in kwargs.items():
                 setattr(opts, key, value)
 
-            validate_input(preprocess_text(text), opts.model, opts.sample_rate, opts.speed, consistency, similarity, enhancement)
+            text = preprocess_text(text)
+            validate_input(text, opts.model, opts.sample_rate, opts.speed, opts.consistency, opts.similarity, opts.enhancement)
 
             self.chunk_size = 250
             if opts.model == 'lightning-large':
                 self.chunk_size = 140
 
             chunks = chunk_text(text, self.chunk_size)
-            audio_content = b""
-
-            for chunk in chunks:
-                payload = {
-                    "text": preprocess_text(chunk),
-                    "sample_rate": opts.sample_rate,
-                    "voice_id": opts.voice_id,
-                    "add_wav_header": False,
-                    "speed": opts.speed,
-                    "model": opts.model
-                }
-                
-                if opts.model == "lightning-large":
-                    if consistency:
-                        payload["consistency"] = consistency
-                    if similarity:
-                        payload["similarity"] = similarity
-                    if enhancement:
-                        payload["enhancement"] = enhancement
-
-
-                headers = {
-                    "Authorization": f"Bearer {self.api_key}",
-                    "Content-Type": "application/json",
-                }
-
-                async with self.session.post(f"{API_BASE_URL}/{opts.model}/get_speech", json=payload, headers=headers) as res:
-                    if res.status != 200:
-                        raise APIError(f"Failed to synthesize speech: {await res.text()}. For more information, visit https://waves.smallest.ai/")
-
-                    audio_content += await res.read()
+
+            async def audio_stream():
+                for chunk in chunks:
+                    payload = {
+                        "text": chunk,
+                        "sample_rate": opts.sample_rate,
+                        "voice_id": opts.voice_id,
+                        "add_wav_header": False,
+                        "speed": opts.speed,
+                        "model": opts.model
+                    }
+                    
+                    if opts.model == "lightning-large":
+                        if opts.consistency is not None:
+                            payload["consistency"] = opts.consistency
+                        if opts.similarity is not None:
+                            payload["similarity"] = opts.similarity
+                        if opts.enhancement is not None:
+                            payload["enhancement"] = opts.enhancement
+
+
+                    headers = {
+                        "Authorization": f"Bearer {self.api_key}",
+                        "Content-Type": "application/json",
+                    }
+
+                    async with self.session.post(f"{API_BASE_URL}/{opts.model}/get_speech", json=payload, headers=headers) as res:
+                        if res.status != 200:
+                            raise APIError(f"Failed to synthesize speech: {await res.text()}. For more information, visit https://waves.smallest.ai/")
+
+                        yield await res.read()
+            
+            if stream:
+                return audio_stream()
+
+            audio_content = b"".join([chunk async for chunk in audio_stream()])
 
             if save_as:
                 if not save_as.endswith(".wav"):