Module spandex_hyphenation::score

Evaluating potential hyphenation opportunities

In Knuth–Liang hyphenation, dictionaries identify valid word breaks by searching whole words for sub-word patterns, ultimately producing a score that evaluates each interval between letters as a candidate for hyphenation.

As an example, consider the word "firkin":

    f|i|r|k|i|n

Our British English dictionary recognizes six patterns therein, each of which assigns a numeric value to one or more locations; when a location is assigned multiple values, the highest one prevails. Patterns are so compounded until a final score is produced. Using guillemets to mark word boundaries, we find:

   «f|i|r|k|i|n»
   -------------
   «f i2
      i2r
        r2k
        r5k i n
          k1i n
          k2i n»
   -------------
    f|i|r|k|i|n
     0 2 5 2 0

By convention, even values inhibit hyphenation, whereas odd values mark valid breaks. Thus, having matched these patterns, the dictionary will offer "fir·kin" as a valid hyphenation.

Traits

Score

Methods to evaluate each index in a string as an opportunity for hyphenation.