in reply to parsing a very large array with regexps

salutations,

yes, the items of the array are words of a language, but they are all the inflected forms of the words of the language. but the user input is not or more patterns; the user input is intended to be words which are together in a phrase (or in a compound word), but glued together by euphony (sandhi) rules, like in "shivâshvah" <- shiva(Shiva-compound)+ashvah(horse-nominative), in which "a+a" becomes a long "â".

when we talk about 255 possibilities, they are only "possibilities" of division, which may be eliminated if one of the "possible" morphemes isn't found.

  • Comment on Re: parsing a very large array with regexps