Speech initialization error: 2147483665

The code is simple as this:

let synthesizer: MVSpeechSynthesizer = MVSpeechSynthesizer.sharedSynthSize()

And it tells me that AnyObject? cannot be converted to MVSpeechSynthesizer. Shouldn't sharedSynthSize() return MVSpeechSynthesizer? Why AnyObject? instead?

Please put the MIT license into the project.

If I just initialize the synthesizer like this:

var mvSynthesizer = MVSpeechSynthesizer()

And I set the speechString and call startRead. It reads it really slowly. I have to explicitly set the uRate to AVSpeechUtteranceDefaultSpeechRate to get the normal speech rate. Why don't you use AVSpeechUtteranceDefaultSpeechRate as the default value?

If I just create an MVSpeechSynthesizer and use all the default settings, it doesn't seem to automatically detect and speak Chinese. Try telling it to read "你好". It produces no sound at all.

The sample app crashes when you try to synthesize speech for only one word.

Crash occurs on [ViewController startRead:] line 87.

Issue: speechLanguage is nil.

It seems that NSLinguisticTagger does not find a language for string containing only one word, so [MVSpeechSyntesizer detectLanguage:] returns @"und". As a result _speechLanguage is not found in supported languages and not set in [MVSpeechSyntesizer setSpeechString:] method.

Solution:
The crash occurs when updating Current language labels, so quick and temporay fix would be to add
if (!speechLanguage) {
return ;
}
to mvSpeech.speechStartBlock implementation in [ViewController startRead:]. However to avoid confusion it should be imo covered earlier.

This can be either fixed by setting default language in [MVSpeechSyntesizer detectLanguage:] method (result checking) or in [MVSpeechSyntesizer setSpeechString:] after enumeration.