Accentor doesn't always return all of the possible variants

[DinoTheDinosaur]

When setting the parameter mode to 'many'
For example:
дорогой -> доро+гой and дорого+й
BUT!
дела -> дела+, окна -> о+кна
Although the warning ' Word has too many accent variants!' appears in both cases.

[DinoTheDinosaur]

Seems to be the trouble with the homograph dict.

[DinoTheDinosaur]

Accents_new.json doesn't include some variants of words, although it always includes precise morphological form of the existing variant. It is better to use the combination of homographs_old.json and wiktionary parsing for dictionary generation. For transcription without variants it is better to use homographs.json. This will be implemented in a switch homograph_dict='old' or homograph_dict='new' accordingly.