Feature request: decode in-place UTF-8

Using Mercurial, sometimes broken UTF-8 appears, because Mercurial improperly detects it during file merge. Looks like this:

/*
A flag that currency is used in system.
Признак используемой в системе валюты.
*/
Used CONSTANT VARCHAR2(4) := 'Used';
/*
Validation operationId
Идентификатор операции валидации платежа в шлюзе.
*/

First comment is normal Cyrillic, but second one isn't, should be "Идентификатор операции валидации".

I want a PSPad feature to in-place decode that broken UTF-8: select the block, then choose Edit → Special conversion → Decode UTF-8. PSPad should decode selected text as UTF-8 or display confirmation when selected block cannot be properly decoded, ask "Decode anyway? (Yes/No)".