* Test stability fixes
* disable audio duration check due to service issus
* Increase translation timeout, format & add try / catch to synthesis test
* Change to a new version of Jest that supports retries. Modify tests to work with new version. Allow 3 retries.
* package-lock update
* Re-install with the same npm version as the ADO agents are using.
* Remove gulp-parent from forced resolution list
* npm audit auto fixes
* Increase error timeouts
* Increase timeouts
* Didn't copy the env change from the package.json to the jest config
* Add logging and increase token timeout.
Co-authored-by: Yulin Li <liyulin@pku.edu.cn>
Some dependencies require a forced version update, others were automatic.
Also, change negative translation case to use an invalid language code, as the random one chosen before now works.
Disable one OCSP test as the test cert is expired and the site publisher is waiting on CA renewal.
* Fix connection leak on bad audio buffer
* Fix connection leak in simple scenarios
* remove unused import
* make typescript happy
* Update FileAudioSource to take a buffer input (allowing Node usage), enable more testing against it.
* package updates, ts => 3.8, jest => latest
* clean up recognition correctly on audio input error
* remove unnecessary parameter
* revert to pre-PR
* Extend OCSP certificate verification to include a memory and disk based cache.
The disk cache will by default be located in TMPDIR/<username>/ but can be overridden.
When the cached OCSP response is within 1 day of expiring, or 1/2 way through it's vaility period (whichever is shorter) a background
task is queued to refresh it while the cached value is used.
* More fixes
* Fix error case
* Fix typos
* Create more idiomatic RestConnection class and methods
* Fix lint for new classes
* Initial implementation of create profile
* Fix broken conversation translator test
* Fixes to return actual VoiceProfile object
* Build fixes
* Clean up callback code
* Initial version of voice enrollment
* Add delete profile to client, add rest connection class in case needed for speaker recognizer.
* Add blob promise to IAudioSource interface in place of file
* Add reset profile method
* Initial implementation of verifySpeaker and identifySpeaker
* Fix create model bug, rename public methods to align with SDK conventions
* SpeakerRecognition recognizeOnceAsync() works now for Identification models/profiles
* speaker verify now working
* Add CancellationDetail classes for SR results
* Handle service errors and result property setting
* Better error handling for VoiceProfile, VoiceProfileResult callbacks
* Bug fixes, misspelled detail fields
* Fix error creation for Enrollments
* Add tests for speaker recognition
* Generate VoiceProfileResult errors using response statusText
* Remove currently unused RestConnection class
* Use "reason" getter for result consistency
* Use PromiseHelper for instant promise fulfillment
* add xmlhttprequest dependency for node
* Node support for XMLHttpRequest primitive
* Allow CancellationDetails to be seen externally
* Allow PushAudioInputStream to be read as Blob for REST connections
* Address code review issues, simplify callback handling in VoiceProfileClient, lengthen rest timeout
* More code review corrections
* Fix weird node promise issue, clean up promise code
* Another fix from input stream
* Use Buffer when sending AudioPushStream with node.js
* Add china host detection, use Buffer.from instead of deprecated Buffer constructor
* Use PromiseHelper to send error promise for unimplemented blob getter
* Add new static test variables for Speaker ID tests
* Implemented DialogConnectorFactory.
* WIP
* begin add dialog tests
* connect impl and first test
* headers for browser
* expose connect
* initial listenOnce
* repalce websocket use
* listenOnce with browser connection support
* listenOnce with activity responses
* expose get/setter props
* cleanup
* generic send fnc
* extracting functionality from base reco
* Save the day
* First incarnation of the BaseAudioPlayer class.
# Conflicts:
# tests/DialogServiceConnectorTests.ts
* Fix bugs in BaseAudioPlayer and added manual tests for it.
- server.js - web socket streaming long audio file (node server.js from the folder)
- AudioPlayerBaseTest.html lauch http-server in the folder and then load the page
* lint cleanup
* Added missed server.js file for audio player manual tests
* Added audio output stream
* TSLint fixes
* AudioOutputStreamTests
* update message loop
* AudioOutputStream fixes
Jsonification of activity
Added AudioOutputStream to activity
ActivityPayloadResponse class
* refactor message loop and add sendActivity
* Turn management classes
* Add integration with turn management
* Few fixes - turn.end is not coming
Merge branch 'brandom/dialog' of https://github.com/microsoft/cognitive-services-speech-sdk-js into gghizila/TTSStreamIntegration
# Conflicts:
# src/common.speech/DialogServiceAdapter.ts
* Fixed the traces for turns to report "debugturn" to help with turn.end investigation missing
* Added timeouts as workaround to not routed messages to 2s.
Added debugturn logging for investingating the messages not routed.
Added timestamps to console.info
* remove promisehelper wait, rm cancel recos
* minor code restructuring
* * The last audio message is null. When we encounter it we closed the audio stream.
* Added isClosed to the audio output stream impl.
* Changed tests to defer the shutdown to avoid random red herrings.
* * The last audio message is null. When we encounter it we closed the audio stream.
* Added isClosed to the audio output stream impl.
* Changed tests to defer the shutdown to avoid random red herrings.
* Added try / catch for audio message handling.
* restrict config to once per connection
* Upgrade https-proxy-agent to 2.2.3 after the last alert.
* fix reco end, remove speech context on turn.end
* Fix machine in the middle vulnerability for "https-proxy-agent" (#97)
- changes generated by npm audit fix --force
- upgrade "https-proxy-agent" to 3.0.0 as there is inconsistency on reports as what version fixes the vulnerability
* fix for mic staying open
* split config into BF and Commands
* add temporary ApplicationId to tests
* add reconnect on websocket timeout, config parity
* remove authHeader from botFramework config
* Implemented DialogConnectorFactory.
* WIP
* begin add dialog tests
* connect impl and first test
* headers for browser
* initial listenOnce
* expose connect
* repalce websocket use
* listenOnce with browser connection support
* listenOnce with activity responses
* expose get/setter props
* cleanup
* generic send fnc
* extracting functionality from base reco
* Save the day
* First incarnation of the BaseAudioPlayer class.
# Conflicts:
# tests/DialogServiceConnectorTests.ts
* Fix bugs in BaseAudioPlayer and added manual tests for it.
- server.js - web socket streaming long audio file (node server.js from the folder)
- AudioPlayerBaseTest.html lauch http-server in the folder and then load the page
* lint cleanup
* Added missed server.js file for audio player manual tests
* Added audio output stream
* TSLint fixes
* AudioOutputStreamTests
* update message loop
* AudioOutputStream fixes
Jsonification of activity
Added AudioOutputStream to activity
ActivityPayloadResponse class
* refactor message loop and add sendActivity
* Turn management classes
* Add integration with turn management
* Few fixes - turn.end is not coming
Merge branch 'brandom/dialog' of https://github.com/microsoft/cognitive-services-speech-sdk-js into gghizila/TTSStreamIntegration
# Conflicts:
# src/common.speech/DialogServiceAdapter.ts
* Fixed the traces for turns to report "debugturn" to help with turn.end investigation missing
* Added timeouts as workaround to not routed messages to 2s.
Added debugturn logging for investingating the messages not routed.
Added timestamps to console.info
* remove promisehelper wait, rm cancel recos
* minor code restructuring
* * The last audio message is null. When we encounter it we closed the audio stream.
* Added isClosed to the audio output stream impl.
* Changed tests to defer the shutdown to avoid random red herrings.
* * The last audio message is null. When we encounter it we closed the audio stream.
* Added isClosed to the audio output stream impl.
* Changed tests to defer the shutdown to avoid random red herrings.
* Added try / catch for audio message handling.
* restrict config to once per connection
* fix reco end, remove speech context on turn.end
* fix for mic staying open
* split config into BF and Commands
* add temporary ApplicationId to tests
* add reconnect on websocket timeout, config parity
* remove authHeader from botFramework config
* merge
* restore code coverage
* extend gitattributes
- changes generated by npm audit fix --force
- upgrade "https-proxy-agent" to 3.0.0 as there is inconsistency on reports as what version fixes the vulnerability