21.2 Character strings

Bigloo Text function: hyphenate word hyphens

The function hyphenate accepts as input a single word and returns as output a list of subwords. The argument hyphens is an opaque data structure obtained by calling the function load-hyphens or make-hyphens.

Example:

(hyphenate "software" (load-hyphens 'en)) ⇒ ("soft" "ware")

Bigloo Text function: load-hyphens obj

Loads an hyphens table and returns a data structure suitable for hyphenate. The variable obj can either be a file name containing an hyphens table or a symbol denoting a pre-defined hyphens table. Currently, Bigloo supports two tables: en for an English table and fr for a French table. The procedure load-hyphens invokes make-hyphens to build the hyphens table.

Example:

(define (hyphenate-text text lang)
   (let ((table (with-handler 
                   (lambda (e)               
                      (unless (&io-file-not-found-error? e)
                         (raise e)))
                   (load-hyphens lang)))
         (words (string-split text " ")))
      (if table 
          (append-map (lambda (w) (hyphenate w table)) words)
           words)))

The procedure hyphenate-text hyphenates the words of the text according to the rules for the language denoted by its code lang if there is a file lang-hyphens.sch. If there is no such file, the text remains un-hyphenated.

Bigloo Text function: make-hyphens [:language] [:exceptions] [:patterns]

Creates an hyphens table out of the arguments exceptions and patterns.

The implementation of the table of hyphens created by make-hyphens follows closely Frank Liang’s algorithm as published in his doctoral dissertation Word Hy-phen-a-tion By Com-pu-ter available on the TeX Users Group site here: http://www.tug.org/docs/liang/. This table is a trie (see http://en.wikipedia.org/wiki/Trie for a definition and an explanation).

Most of this implementation is borrowed from Phil Bewig’s work available here: http://sites.google.com/site/schemephil/, along with his paper describing the program from which the Bigloo implementation is largely borrowed.

exceptions must be a non-empty list of explicitly hyphenated words.

Explicitly hyphenated words are like the following: "as-so-ciate", "as-so-ciates", "dec-li-na-tion", where the hyphens indicate the places where hyphenation is allowed. The words in exceptions are used to generate hyphenation patterns, which are added to patterns (see next paragraph).

patterns must be a non-empty list of hyphenation patterns.

Hyphenation patterns are strings of the form ".anti5s", where a period denotes the beginning or the end of a word, an odd number denotes a place where hyphenation is allowed, and an even number a place where hyphenation is forbidden. This notation is part of Frank Liang’s algorithm created for Donald Knuth’s TeX typographic system.

This document was generated on March 31, 2014 using texi2html 5.0.