The Hungarian hyphenating program is based on MorphoLogic HUMor (High speed Unification based Morphology) technology knowing all forms of approximately 100,000 roots. In case of Hungarian this means more than two billion word forms.
Besides the simple syllabication algorithm the hyphenator also uses the language segmentation provided by the morphological analyzer. This is how it overrides the syllabication rules where necessary, e.g. after verbal prefixes or at compound boundaries. This method allows a much more reliable result than the use of an algorithm and a list of exceptions alone.
The program hyphenates only words it is able to parse. (The spelling checker accepts these as correct.) If the hyphenation of unknown words was allowed, a number of mistakes would arise from the fact that neither their morphological segmentation nor their correct pronunciation is known. The latter is important in case of foreign words. Finding the mistakes resulting from this would be more complicated for the user than manually hyphenating some non-hyphenated words.
There are some homonymous words the hyphenation of which depends on the actual meaning, e.g. "legelőre". In such words the program makes the hyphenation only where it is independent from the meaning, e.g. "legelő-re".
The program handles the hyphenation of long digraphs, e.g. "asz-szony". In these words a new letter appears when hyphenated. This function is not supported by all text editors and desktop publishers. In these programs the words are hyphenated only at the other positions.
To see the list of available programs please click here!