How to use TessBaseAPI::GetUnichar() safely?

[db4]

I'm trying to get the actual charset loaded by Tesseract from a .traineddata file. I can call TessBaseAPI::GetUnichar(id++) in a loop but how to find what the maximum acceptable id is? TessBaseAPI::GetUnichar() implementation is just

 

const char* TessBaseAPI::GetUnichar(int unichar_id) {

 return tesseract_->unicharset.id_to_unichar(unichar_id);

}

that in turn calls

  

const char* UNICHARSET::id_to_unichar(UNICHAR_ID id) const {

 if (id == INVALID_UNICHAR_ID) {

  return INVALID_UNICHAR;

 }

 ASSERT_HOST(id < this->size());

 return unichars[id].representation;

}

I assume it will crash if id is out of bounds. And UNICHARSET::size() is not exposed via TessBaseAPI. Any idea how one could use TessBaseAPI::GetUnichar() to get chars known to Tesseract?