cartesia 1.0.7__tar.gz → 1.0.9__tar.gz

{cartesia-1.0.7 → cartesia-1.0.9}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: cartesia
-Version: 1.0.7
+Version: 1.0.9
 Summary: The official Python library for the Cartesia API.
 Home-page: 
 Author: Cartesia, Inc.
@@ -25,6 +25,22 @@ The official Cartesia Python library which provides convenient access to the Car
 > [!IMPORTANT]
 > The client library introduces breaking changes in v1.0.0, which was released on June 24th 2024. See the [release notes](https://github.com/cartesia-ai/cartesia-python/releases/tag/v1.0.0) and [migration guide](https://github.com/cartesia-ai/cartesia-python/discussions/44). Reach out to us on [Discord](https://discord.gg/ZVxavqHB9X) for any support requests!
 
+- [Cartesia Python API Library](#cartesia-python-api-library)
+  - [Documentation](#documentation)
+  - [Installation](#installation)
+  - [Voices](#voices)
+  - [Text-to-Speech](#text-to-speech)
+    - [Server-Sent Events (SSE)](#server-sent-events-sse)
+    - [WebSocket](#websocket)
+      - [Conditioning speech on previous generations using WebSocket](#conditioning-speech-on-previous-generations-using-websocket)
+    - [Generating timestamps using WebSocket](#generating-timestamps-using-websocket)
+    - [Multilingual Text-to-Speech \[Alpha\]](#multilingual-text-to-speech-alpha)
+    - [Speed and Emotion Control \[Experimental\]](#speed-and-emotion-control-experimental)
+    - [Jupyter Notebook Usage](#jupyter-notebook-usage)
+    - [Utility methods](#utility-methods)
+      - [Output Formats](#output-formats)
+
+
 ## Documentation
 
 Our complete API documentation can be found [on docs.cartesia.ai](https://docs.cartesia.ai).
@@ -250,7 +266,7 @@ async def send_transcripts(ctx):
 
     # You can check out our models at https://docs.cartesia.ai/getting-started/available-models
     model_id = "sonic-english"
-    
+
     # You can find the supported `output_format`s at https://docs.cartesia.ai/api-reference/endpoints/stream-speech-server-sent-events
     output_format = {
         "container": "raw",
@@ -266,7 +282,7 @@ async def send_transcripts(ctx):
         "As they near Eggman's lair, our heroes charge their abilities for an epic boss battle. ",
         "Get ready to spin, jump, and sound-blast your way to victory in this high-octane crossover!"
     ]
-    
+
     for transcript in transcripts:
         # Send text inputs as they become available
         await ctx.send(
@@ -278,7 +294,7 @@ async def send_transcripts(ctx):
         )
 
     # Indicate that no more inputs will be sent. Otherwise, the context will close after 5 seconds of inactivity.
-    await ctx.no_more_inputs() 
+    await ctx.no_more_inputs()
 
 async def receive_and_play_audio(ctx):
     p = pyaudio.PyAudio()
@@ -384,7 +400,7 @@ output_stream = ctx.send(
     voice_id=voice_id,
     output_format=output_format,
 )
-    
+
 for output in output_stream:
     buffer = output["audio"]
 

{cartesia-1.0.7 → cartesia-1.0.9}/README.md

@@ -8,6 +8,22 @@ The official Cartesia Python library which provides convenient access to the Car
 > [!IMPORTANT]
 > The client library introduces breaking changes in v1.0.0, which was released on June 24th 2024. See the [release notes](https://github.com/cartesia-ai/cartesia-python/releases/tag/v1.0.0) and [migration guide](https://github.com/cartesia-ai/cartesia-python/discussions/44). Reach out to us on [Discord](https://discord.gg/ZVxavqHB9X) for any support requests!
 
+- [Cartesia Python API Library](#cartesia-python-api-library)
+  - [Documentation](#documentation)
+  - [Installation](#installation)
+  - [Voices](#voices)
+  - [Text-to-Speech](#text-to-speech)
+    - [Server-Sent Events (SSE)](#server-sent-events-sse)
+    - [WebSocket](#websocket)
+      - [Conditioning speech on previous generations using WebSocket](#conditioning-speech-on-previous-generations-using-websocket)
+    - [Generating timestamps using WebSocket](#generating-timestamps-using-websocket)
+    - [Multilingual Text-to-Speech \[Alpha\]](#multilingual-text-to-speech-alpha)
+    - [Speed and Emotion Control \[Experimental\]](#speed-and-emotion-control-experimental)
+    - [Jupyter Notebook Usage](#jupyter-notebook-usage)
+    - [Utility methods](#utility-methods)
+      - [Output Formats](#output-formats)
+
+
 ## Documentation
 
 Our complete API documentation can be found [on docs.cartesia.ai](https://docs.cartesia.ai).
@@ -233,7 +249,7 @@ async def send_transcripts(ctx):
 
     # You can check out our models at https://docs.cartesia.ai/getting-started/available-models
     model_id = "sonic-english"
-    
+
     # You can find the supported `output_format`s at https://docs.cartesia.ai/api-reference/endpoints/stream-speech-server-sent-events
     output_format = {
         "container": "raw",
@@ -249,7 +265,7 @@ async def send_transcripts(ctx):
         "As they near Eggman's lair, our heroes charge their abilities for an epic boss battle. ",
         "Get ready to spin, jump, and sound-blast your way to victory in this high-octane crossover!"
     ]
-    
+
     for transcript in transcripts:
         # Send text inputs as they become available
         await ctx.send(
@@ -261,7 +277,7 @@ async def send_transcripts(ctx):
         )
 
     # Indicate that no more inputs will be sent. Otherwise, the context will close after 5 seconds of inactivity.
-    await ctx.no_more_inputs() 
+    await ctx.no_more_inputs()
 
 async def receive_and_play_audio(ctx):
     p = pyaudio.PyAudio()
@@ -367,7 +383,7 @@ output_stream = ctx.send(
     voice_id=voice_id,
     output_format=output_format,
 )
-    
+
 for output in output_stream:
     buffer = output["audio"]
 

cartesia-1.0.9/cartesia/__init__.py

@@ -0,0 +1,3 @@
+from cartesia.client import AsyncCartesia, Cartesia
+
+__all__ = ["Cartesia", "AsyncCartesia"]

{cartesia-1.0.7 → cartesia-1.0.9}/cartesia/_types.py

@@ -1,4 +1,5 @@
 from typing import List, TypedDict
+
 from cartesia.utils.deprecated import deprecated
 
 

{cartesia-1.0.7 → cartesia-1.0.9}/cartesia/client.py

@@ -1,41 +1,47 @@
 import asyncio
 import base64
-from collections import defaultdict
 import json
+import logging
 import os
 import uuid
+from collections import defaultdict
 from types import TracebackType
 from typing import (
     Any,
     AsyncGenerator,
-    Iterator,
+    Callable,
     Dict,
     Generator,
+    Iterator,
     List,
     Optional,
+    Set,
     Tuple,
     Union,
-    Callable,
-    Set,
 )
 
 import aiohttp
 import httpx
-import logging
 import requests
-from websockets.sync.client import connect
-from iterators import TimeoutIterator
 
-from cartesia.utils.retry import retry_on_connection_error, retry_on_connection_error_async
+try:
+    from websockets.sync.client import connect
+
+    IS_WEBSOCKET_SYNC_AVAILABLE = True
+except ImportError:
+    IS_WEBSOCKET_SYNC_AVAILABLE = False
+
 from cartesia._types import (
+    DeprecatedOutputFormatMapping,
     EventType,
     OutputFormat,
     OutputFormatMapping,
-    DeprecatedOutputFormatMapping,
     VoiceControls,
     VoiceMetadata,
 )
-
+from cartesia.utils.retry import retry_on_connection_error, retry_on_connection_error_async
+from iterators import TimeoutIterator
+from websockets.sync.client import connect
 
 DEFAULT_MODEL_ID = "sonic-english"  # latest default model
 MULTILINGUAL_MODEL_ID = "sonic-multilingual"  # latest multilingual model
@@ -207,38 +213,27 @@ class Voices(Resource):
 
         return response.json()
 
-    def clone(self, filepath: Optional[str] = None, link: Optional[str] = None) -> List[float]:
-        """Clone a voice from a clip or a URL.
+    def clone(self, filepath: Optional[str] = None, enhance: str = True) -> List[float]:
+        """Clone a voice from a clip.
 
         Args:
             filepath: The path to the clip file.
-            link: The URL to the clip
+            enhance: Whether to enhance the clip before cloning the voice (highly recommended). Defaults to True.
 
         Returns:
             The embedding of the cloned voice as a list of floats.
         """
-        # TODO: Python has a bytes object, use that instead of a filepath
-        if not filepath and not link:
-            raise ValueError("At least one of 'filepath' or 'link' must be specified.")
-        if filepath and link:
-            raise ValueError("Only one of 'filepath' or 'link' should be specified.")
-        if filepath:
-            url = f"{self._http_url()}/voices/clone/clip"
-            with open(filepath, "rb") as file:
-                files = {"clip": file}
-                headers = self.headers.copy()
-                headers.pop("Content-Type", None)
-                response = httpx.post(url, headers=headers, files=files, timeout=self.timeout)
-                if not response.is_success:
-                    raise ValueError(f"Failed to clone voice from clip. Error: {response.text}")
-        elif link:
-            url = f"{self._http_url()}/voices/clone/url"
-            params = {"link": link}
+        if not filepath:
+            raise ValueError("Filepath must be specified.")
+        url = f"{self._http_url()}/voices/clone/clip"
+        with open(filepath, "rb") as file:
+            files = {"clip": file}
+            files["enhance"] = str(enhance).lower()
             headers = self.headers.copy()
-            headers.pop("Content-Type")  # The content type header is not required for URLs
-            response = httpx.post(url, headers=self.headers, params=params, timeout=self.timeout)
+            headers.pop("Content-Type", None)
+            response = httpx.post(url, headers=headers, files=files, timeout=self.timeout)
             if not response.is_success:
-                raise ValueError(f"Failed to clone voice from URL. Error: {response.text}")
+                raise ValueError(f"Failed to clone voice from clip. Error: {response.text}")
 
         return response.json()["embedding"]
 
@@ -469,6 +464,10 @@ class _WebSocket:
         Raises:
             RuntimeError: If the connection to the WebSocket fails.
         """
+        if not IS_WEBSOCKET_SYNC_AVAILABLE:
+            raise ImportError(
+                "The synchronous WebSocket client is not available. Please ensure that you have 'websockets>=12.0' or compatible version installed."
+            )
         if self.websocket is None or self._is_websocket_closed():
             route = "tts/websocket"
             try:

cartesia-1.0.9/cartesia/version.py

@@ -0,0 +1 @@
+__version__ = "1.0.9"

{cartesia-1.0.7 → cartesia-1.0.9}/cartesia.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: cartesia
-Version: 1.0.7
+Version: 1.0.9
 Summary: The official Python library for the Cartesia API.
 Home-page: 
 Author: Cartesia, Inc.
@@ -25,6 +25,22 @@ The official Cartesia Python library which provides convenient access to the Car
 > [!IMPORTANT]
 > The client library introduces breaking changes in v1.0.0, which was released on June 24th 2024. See the [release notes](https://github.com/cartesia-ai/cartesia-python/releases/tag/v1.0.0) and [migration guide](https://github.com/cartesia-ai/cartesia-python/discussions/44). Reach out to us on [Discord](https://discord.gg/ZVxavqHB9X) for any support requests!
 
+- [Cartesia Python API Library](#cartesia-python-api-library)
+  - [Documentation](#documentation)
+  - [Installation](#installation)
+  - [Voices](#voices)
+  - [Text-to-Speech](#text-to-speech)
+    - [Server-Sent Events (SSE)](#server-sent-events-sse)
+    - [WebSocket](#websocket)
+      - [Conditioning speech on previous generations using WebSocket](#conditioning-speech-on-previous-generations-using-websocket)
+    - [Generating timestamps using WebSocket](#generating-timestamps-using-websocket)
+    - [Multilingual Text-to-Speech \[Alpha\]](#multilingual-text-to-speech-alpha)
+    - [Speed and Emotion Control \[Experimental\]](#speed-and-emotion-control-experimental)
+    - [Jupyter Notebook Usage](#jupyter-notebook-usage)
+    - [Utility methods](#utility-methods)
+      - [Output Formats](#output-formats)
+
+
 ## Documentation
 
 Our complete API documentation can be found [on docs.cartesia.ai](https://docs.cartesia.ai).
@@ -250,7 +266,7 @@ async def send_transcripts(ctx):
 
     # You can check out our models at https://docs.cartesia.ai/getting-started/available-models
     model_id = "sonic-english"
-    
+
     # You can find the supported `output_format`s at https://docs.cartesia.ai/api-reference/endpoints/stream-speech-server-sent-events
     output_format = {
         "container": "raw",
@@ -266,7 +282,7 @@ async def send_transcripts(ctx):
         "As they near Eggman's lair, our heroes charge their abilities for an epic boss battle. ",
         "Get ready to spin, jump, and sound-blast your way to victory in this high-octane crossover!"
     ]
-    
+
     for transcript in transcripts:
         # Send text inputs as they become available
         await ctx.send(
@@ -278,7 +294,7 @@ async def send_transcripts(ctx):
         )
 
     # Indicate that no more inputs will be sent. Otherwise, the context will close after 5 seconds of inactivity.
-    await ctx.no_more_inputs() 
+    await ctx.no_more_inputs()
 
 async def receive_and_play_audio(ctx):
     p = pyaudio.PyAudio()
@@ -384,7 +400,7 @@ output_stream = ctx.send(
     voice_id=voice_id,
     output_format=output_format,
 )
-    
+
 for output in output_stream:
     buffer = output["audio"]
 

{cartesia-1.0.7 → cartesia-1.0.9}/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.ruff]
 # Add more rule codes as needed
-extend-select = [
+lint.extend-select = [
     "D",    # pydocstyle - to replace docformatter
 ]
 
@@ -43,14 +43,14 @@ line-length = 100
 # Enable the count of violations
 output-format = "full"
 
-[tool.ruff.isort]
-force-wrap-aliases = true
-combine-as-imports = true
-force-sort-within-sections = true
-known-first-party = []
-known-third-party = []
-known-local-folder = []
-lines-after-imports = 2
-
-[tool.ruff.pydocstyle]
+[tool.ruff.lint.pydocstyle]
 convention = "google"
+
+[tool.isort]
+profile = "black"
+multi_line_output = 3
+include_trailing_comma = true
+force_grid_wrap = 0
+use_parentheses = true
+ensure_newline_before_comments = true
+line_length = 100

{cartesia-1.0.7 → cartesia-1.0.9}/tests/test_deprecated.py

@@ -1,8 +1,7 @@
-from packaging.version import Version
-
 import cartesia as Cartesia
-from cartesia.utils.deprecated import _DEPRECATED_FUNCTION_STATS
 import cartesia.version as version
+from cartesia.utils.deprecated import _DEPRECATED_FUNCTION_STATS
+from packaging.version import Version
 
 
 def test_deprecated_to_remove_by_version():

{cartesia-1.0.7 → cartesia-1.0.9}/tests/test_tts.py

@@ -5,17 +5,18 @@ different results. Therefore, we cannot test for complete correctness but rather
 general correctness.
 """
 
+import asyncio
 import logging
 import os
 import sys
-from cartesia import AsyncCartesia, Cartesia
-from cartesia.client import DEFAULT_MODEL_ID, MULTILINGUAL_MODEL_ID
-from cartesia._types import VoiceControls, VoiceMetadata
+import uuid
 from typing import AsyncGenerator, Generator, List
+
 import numpy as np
 import pytest
-import uuid
-import asyncio
+from cartesia import AsyncCartesia, Cartesia
+from cartesia._types import VoiceControls, VoiceMetadata
+from cartesia.client import DEFAULT_MODEL_ID, MULTILINGUAL_MODEL_ID
 
 THISDIR = os.path.dirname(__file__)
 sys.path.insert(0, os.path.dirname(THISDIR))
@@ -79,19 +80,17 @@ def test_get_voice_from_id(client: Cartesia):
     voices = client.voices.list()
     assert voice in voices
 
-# Does not work currently, LB issue
-# def test_clone_voice_with_link(client: Cartesia):
-#     url = "https://youtu.be/g2Z7Ddd573M?si=P8BM_hBqt5P8Ft6I&t=69"
-#     logger.info("Testing voices.clone with link")
-#     cloned_voice_embedding = client.voices.clone(link=url)
-#     assert isinstance(cloned_voice_embedding, list)
-#     assert len(cloned_voice_embedding) == 192
-
 def test_clone_voice_with_file(client: Cartesia):
     logger.info("Testing voices.clone with file")
     output = client.voices.clone(filepath=os.path.join(RESOURCES_DIR, "sample-speech-4s.wav"))
     assert isinstance(output, list)
 
+@pytest.mark.parametrize("enhance", [True, False])
+def test_clone_voice_with_file_enhance(client: Cartesia, enhance: bool):
+    logger.info("Testing voices.clone with file")
+    output = client.voices.clone(filepath=os.path.join(RESOURCES_DIR, "sample-speech-4s.wav"), enhance=enhance)
+    assert isinstance(output, list)
+
 def test_create_voice(client: Cartesia):
     logger.info("Testing voices.create")
     embedding = np.ones(192).tolist()

cartesia-1.0.7/cartesia/__init__.py

@@ -1,3 +0,0 @@
-from cartesia.client import Cartesia, AsyncCartesia
-
-__all__ = ["Cartesia", "AsyncCartesia"]

cartesia-1.0.7/cartesia/version.py

@@ -1 +0,0 @@
-__version__ = "1.0.7"

{cartesia-1.0.7 → cartesia-1.0.9}/cartesia/utils/retry.py

@@ -1,9 +1,9 @@
-import time
-
-from aiohttp.client_exceptions import ServerDisconnectedError
 import asyncio
+import time
 from functools import wraps
 from http.client import RemoteDisconnected
+
+from aiohttp.client_exceptions import ServerDisconnectedError
 from httpx import TimeoutException
 from requests.exceptions import ConnectionError