fix

nodes/text2speech/pyproject.toml

@@ -85,13 +85,19 @@ exclude_lines = [
 max-line-length=120
 disable = [
  "missing-module-docstring",
- "fixme"
+ "fixme",
+ "R0913",
+ "W0221"
 ]
 [tool.pylint.'DESIGN']
 max-args=7
 [tool.pylint.'SIMILARITIES']
 min-similarity-lines=10
 ignore-comments=true
+[tool.pylint.'TYPECHECK']
+# List of members which are set dynamically and missed by Pylint inference
+# system, and so shouldn't trigger E1101 when accessed.
+generated-members="torch.*"
 
 [tool.pytest.ini_options]
 minversion = "6.0"

nodes/text2speech/text2speech/answer_to_speech.py

@@ -36,16 +36,19 @@ def __init__(
  The allowed parameters are:
  - audio_format: The format to save the audio into (wav, mp3, ...). Defaults to `wav`.
  Supported formats:
- - Uncompressed formats thanks to `soundfile` (see `libsndfile documentation <https://libsndfile.github.io/libsndfile/api.html>`_
+ - Uncompressed formats thanks to `soundfile` (see `libsndfile documentation
+ <https://libsndfile.github.io/libsndfile/api.html>`_
  for a list of supported formats).
  - Compressed formats thanks to `pydub`
  (uses FFMPEG: run `ffmpeg -formats` in your terminal to see the list of supported formats).
- - subtype: Used only for uncompressed formats. See `libsndfile documentation <https://libsndfile.github.io/libsndfile/api.html>`_
+ - subtype: Used only for uncompressed formats. See `libsndfile documentation
+ <https://libsndfile.github.io/libsndfile/api.html>`_
  for the complete list of available subtypes. Defaults to `PCM_16`.
  - sample_width: Used only for compressed formats. The sample width of your audio. Defaults to 2.
  - channels count: Used only for compressed formats. The number of channels your audio file has:
  1 for mono, 2 for stereo. Depends on the model, but it's often mono so it defaults to 1.
- - bitrate: Used only for compressed formats. The desired bitrate of your compressed audio. Defaults to '320k'.
+ - bitrate: Used only for compressed formats. The desired bitrate of your compressed audio.
+ Defaults to '320k'.
  - normalized: Used only for compressed formats. Normalizes the audio before compression (range 2^15)
  or leaves it untouched.
  - audio_naming_function: A function mapping the input text into the audio file name.
@@ -95,4 +98,4 @@ def run_batch(self, answers: List[List[Answer]]) -> Tuple[Dict[str, List[List[An
  for answers_list in answers:
  results["answers"].append(self.run(answers_list)[0]["answers"])
 
- return results, "output_1"
+ return results, "output_1"

nodes/text2speech/text2speech/document_to_speech.py

@@ -41,7 +41,8 @@ def __init__(
  - sample_width: Used only for compressed formats. The sample width of your audio. Defaults to 2.
  - channels count: Used only for compressed formats. The number of channels your audio file has:
  1 for mono, 2 for stereo. Depends on the model, but it's often mono so it defaults to 1.
- - bitrate: Used only for compressed formats. The desired bitrate of your compressed audio. Defaults to '320k'.
+ - bitrate: Used only for compressed formats. The desired bitrate of your compressed audio.
+ Defaults to '320k'.
  - normalized: Used only for compressed formats. Normalizes the audio before compression (range 2^15)
  or leaves it untouched.
  - audio_naming_function: The function mapping the input text into the audio file name.
@@ -77,4 +78,4 @@ def run_batch(self, documents: List[List[Document]]) -> Tuple[Dict[str, List[Lis
  for docs_list in documents:
  results["documents"].append(self.run(docs_list)[0]["documents"])
 
- return results, "output_1"
+ return results, "output_1"

nodes/text2speech/text2speech/utils/text_to_speech.py

@@ -83,13 +83,19 @@ def text_to_audio_file(
  :param generated_audio_dir: The folder to save the audio file to.
  :param audio_format: The format to save the audio into (wav, mp3, ...).
  Supported formats:
- - Uncompressed formats thanks to `soundfile` (see `libsndfile documentation <https://libsndfile.github.io/libsndfile/api.html>`_ for a list of supported formats)
- - Compressed formats thanks to `pydub` (uses FFMPEG: run `ffmpeg -formats` in your terminal to see the list of supported formats).
- :param subtype: Used only for uncompressed formats. See https://libsndfile.github.io/libsndfile/api.html for the complete list of available subtypes.
+ - Uncompressed formats thanks to `soundfile` (see `libsndfile documentation
+ <https://libsndfile.github.io/libsndfile/api.html>`_ for a list of supported formats)
+ - Compressed formats thanks to `pydub` (uses FFMPEG: run `ffmpeg -formats` in your terminal to see the
+ list of supported formats).
+ :param subtype: Used only for uncompressed formats. See https://libsndfile.github.io/libsndfile/api.html for
+ the complete list of available subtypes.
  :param sample_width: Used only for compressed formats. The sample width of your audio. Defaults to 2.
- :param channels count: Used only for compressed formats. THe number of channels your audio file has: 1 for mono, 2 for stereo. Depends on the model, but it's often mono so it defaults to 1.
- :param bitrate: Used only for compressed formats. The desired bitrate of your compressed audio. Defaults to '320k'.
- :param normalized: Used only for compressed formats. Normalizes the audio before compression (range 2^15) or leaves it untouched.
+ :param channels count: Used only for compressed formats. THe number of channels your audio file has:
+ 1 for mono, 2 for stereo. Depends on the model, but it's often mono so it defaults to 1.
+ :param bitrate: Used only for compressed formats. The desired bitrate of your compressed audio.
+ Defaults to '320k'.
+ :param normalized: Used only for compressed formats. Normalizes the audio before compression (range 2^15) or
+ leaves it untouched.
  :param audio_naming_function: A function mapping the input text into the audio file name.
  By default, the audio file gets the name from the MD5 sum of the input text.
  :return: The path to the generated file.
@@ -112,7 +118,7 @@ def text_to_audio_file(
  self.compress_audio(
  data=audio_data,
  path=file_path,
- format=audio_format,
+ audio_format=audio_format,
  sample_rate=self.model.fs,
  sample_width=sample_width,
  channels_count=channels_count,
@@ -138,15 +144,16 @@ def text_to_audio_data(self, text: str, _models_output_key: str = "wav") -> np.a
  output = prediction.get(_models_output_key, None)
  if output is None:
  raise AudioNodeError(
- f"The model returned no output under the {_models_output_key} key. The available output keys are {prediction.keys()}. Make sure you selected the right key."
+ f"The model returned no output under the {_models_output_key} key."
+ f"The available output keys are {prediction.keys()}. Make sure you selected the right key."
  )
  return output.cpu().numpy()
 
  def compress_audio(
  self,
  data: np.array, # type: ignore [valid-type]
  path: Path,
- format: str,
+ audio_format: str,
  sample_rate: int,
  sample_width: int = 2,
  channels_count: int = 1,
@@ -158,13 +165,15 @@ def compress_audio(
 
  :param data: The audio data to compress.
  :param path: The path to save the compressed audio at.
- :param format: The format to compress the data into ('mp3', 'wav', 'raw', 'ogg' or other ffmpeg/avconv supported files).
+ :param audio_format: The format to compress the data into ('mp3', 'wav', 'raw', 'ogg' or other ffmpeg/avconv
+ supported files).
  :param sample_rate: The sample rate of the audio. Depends on the model.
  :param sample_width: The sample width of your audio. Defaults to 2.
- :param channels count: The number of channels your audio file has: 1 for mono, 2 for stereo. Depends on the model, but it's often mono so it defaults to 1.
+ :param channels count: The number of channels your audio file has: 1 for mono, 2 for stereo.
+ Depends on the model, but it's often mono so it defaults to 1.
  :param bitrate: The desired bitrate of your compressed audio. Default to '320k'.
  :param normalized: Normalizes the audio before compression (range 2^15) or leaves it untouched.
  """
  data = np.int16((data * 2**15) if normalized else data)
  audio = AudioSegment(data.tobytes(), frame_rate=sample_rate, sample_width=sample_width, channels=channels_count)
- audio.export(path, format=format, bitrate=bitrate)
+ audio.export(path, format=audio_format, bitrate=bitrate)