Installation of hyphenation patterns

Jeffrey H. Kingston
22 December 1992

Basser Lout Version 2.05 incorporates automatic hyphenation using the method
introduced by TeX (see Appendix H of the TeXBook by D. E. Knuth).  No special
action is required to install hyphenation unless it is desired to change the
hyphenation patterns that control it.

The first time on any run that hyphenation is required, Lout will search the
@SysInclude directories for a binary file called lout.hyph.pk, which
contains a memory image of the TeX hyphenation patterns, modified so that
the file may be safely shared by big-endian and little-endian machines.  If
this file cannot be found, Lout then searches for the text file lout.hyph
instead, and uses it to construct lout.hyph.pk.  To change the hyphenation
patterns, delete lout.hyph.pk and modify lout.hyph; the rest is automatic.

File lout.hyph mainly contains the long list of TeX hyphenation patterns,
but it begins with a sequence of character classes, one class per line,
terminated by a blank line:

    @!$%^&*()_-+=~`{[}]:;'|<,.>?/0123456789
    aA
    bB
    cC
    ...
    yY
    zZ

Briefly, hyphenation of a word works like this.  If the word contains a
character not found in any of these classes, it will not be hyphenated.
Otherwise the word is analysed into sequences of letters separated by
sequences of non-letters (the characters in the first class are by
definition the non-letters, and the remaining characters are the letters).
Each sequence of five or more letters is then matched with the hyphenation
patterns and hyphenated.  The matching process treats the members of each
class as identical (so the classes above ensure that the distinction
between upper and lower case is ignored).  The hyphen character "-" is
treated specially.  Apart from the weighting digits, the patterns should
contain only letters.

Extreme lengths were resorted to to compress the .pk file as much as
possible, and attempts to increase the number of classes or patterns
significantly are likely to fail.  Some words are not hyphenated well by
the standard patterns and must be treated as exceptions; see the
DocumentLayout package implementation (file dl in lout/include) for
the standard list of these exceptions and how to deal with them.