You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Describe the bug
When we are using AWSLexInteractionKit, we have noticed that for Japanese language, Lex bot response text contains question marks instead of actual Japanese characters. For European languages such as Spanish, French etc., they seem to work fine.
To Reproduce
Steps to reproduce the behavior:
Configure a Lex bot that returns text in Japanese
Use AWSLexInteractionKit to interact with the Lex bot, such as this example
Check the output text
Observed Behavior
Output text contains question marks.
Expected Behavior
Output text contains Japanese characters
Areas of the SDK you are using (AWSMobileClient, Cognito, Pinpoint, IoT, etc)?
AWSLex
Environment(please complete the following information):
SDK Version: 2.24.2
Dependency Manager: Cocoapods
Swift Version : 5
Xcode Version: 12.5.1
Device Information (please complete the following information):
Device: iPhone 12 simulator
iOS Version: iOS 14.6
Specific to simulators: n/a
Additional context
Looks like the switchModeInput.outputText is set in AWSLexInteractionKit.m, and it’s using response.message field. According to the documentation, this field can only be used for a limit set of languages.
We also noticed this commit, which indicated that message field in PostContentResponse has been deprecated, so does this mean AWSLexInteractionKit should also expose encodedMessage field via its interface?
The text was updated successfully, but these errors were encountered:
Currently, Lex returns the message as part of HTTP response headers, which restricts the character set in the response. Currently to unblock customers wanting to use special characters in speech-to-text use cases, they would need to use base64 encoded strings when creating the bot and base64 decode the responses.
Next steps:
Investigate whether this is something that we can do in the Lex interaction kit (mid-level client), or whether it has to be done at the low-level client
Investigate what, if any, backend configuration options need to be set up to support this
Investigate what, if any, backend code a customer would have to provide to support this
Describe the bug
When we are using
AWSLexInteractionKit
, we have noticed that for Japanese language, Lex bot response text contains question marks instead of actual Japanese characters. For European languages such as Spanish, French etc., they seem to work fine.To Reproduce
Steps to reproduce the behavior:
AWSLexInteractionKit
to interact with the Lex bot, such as this exampleObserved Behavior
Output text contains question marks.
Expected Behavior
Output text contains Japanese characters
Areas of the SDK you are using (AWSMobileClient, Cognito, Pinpoint, IoT, etc)?
AWSLex
Environment(please complete the following information):
Device Information (please complete the following information):
Additional context
Looks like the
switchModeInput.outputText
is set in AWSLexInteractionKit.m, and it’s usingresponse.message
field. According to the documentation, this field can only be used for a limit set of languages.We also noticed this commit, which indicated that message field in PostContentResponse has been deprecated, so does this mean AWSLexInteractionKit should also expose encodedMessage field via its interface?
The text was updated successfully, but these errors were encountered: