Skip to content

Releases: auroraapi/aurora-python

Python SDK v0.2.0

22 Jun 05:27
e05d027
Compare
Choose a tag to compare

This release adds a major feature to the SDK: Dialog Builder support. Although it isn't complete, it supports some of the major features.

Added

  • Dialog Builder support
    • Create a Dialog
    • Access Dialog context updates through on_context_update
    • Set UDFs using Dialog.set_function
    • Set and access the Dialog context data through context functions
      • context.get_step_data(id) gets the step data for the given step id
      • context.get_user_data(id) gets the user data for the given id
      • context.set_user_data(id, value) sets the given user data for the given id
      • context.get_current_step() gets the current step in the dialog
    • Run a created dialog using Dialog.run()

Python SDK v0.1.2

13 Mar 19:15
Compare
Choose a tag to compare

Changed

  • Audio recording now keeps about 1.25s of audio before actual talking begins to prevent the beginning of the speech from appearing cut off. This fixes an issue where the STT doesn't recognize the first word.
  • This release has been collapsed into v0.1.1, as the change is the same.

Python Beta SDK

27 Feb 03:38
Compare
Choose a tag to compare

This release represents the first beta version of the Aurora Python SDK. For all 0.1.x versions, the interface will not change. Bugs will be fixed and features will be added in a backwards-compatible way. This release features the following:

  • Developer interfaces
    • Text object, for text-to-speech and interpret
    • Interpret object (as a result of calling Text(...).interpret())
    • Speech object (as a result of calling Text(...).speech()) which can be created from a recording or a stored audio WAV file.
      • speech.listen – listen once and return a Speech object
      • speech.continuously_listen – keep listening and yield Speech objects until break
      • speech.listen_and_transcribe – listen once, but stream audio to server while listening and continuously transcribe until stream ends; returns a Text object
      • speech.continuously_listen_and_transcribe – a combination of continuously_listen and listen_and_transcribe
  • Audio
    • AudioFile class for high-level audio manipulation, recording, and playback
  • API
    • Functions to send API requests to the backend
  • Tests
    • 100% coverage with 50 tests

You can install this with pip install --upgrade auroraapi.