Thai script has no word delimiter. While it's trivial for human readers to recognize word boundaries while reading, it requires some knowledge for the machine to do the same when wrapping lines or moving cursor word-wise, etc. Normally, applications need such feature to support Thai text processing.

Swath is a general-purpose utility to workaround the lack of such capability in applications. It analyzes the given Thai text by consulting a Thai word list for word boundaries, before outputting the same text with the predefined word delimiters inserted.

It can read many kinds of input, including plain text and structured documents like HTML, RTF, LaTeX and Lambda (Unicode version of LaTeX with Omega typesetter kernel). [See -f option].

For the known documents, it inserts the common word delimiters used in the corresponding formats, and pipes (|) for plain text. But the user can always override this with a preferred delimiter. [See -b option.]

Swath can also be configured to use different algorithms for the analysis. Currently, it supports two schemes: longest (greedy) matching and maximal (least words) matching. [See -m option.]

EXAMPLES

• For LaTeX (to be used with babel-thai package):

  $swath -f latex < mydoc.tex > mydoc.ttex$ latex mydoc.ttex

  Or if you composed your LaTeX source in UTF-8:

  $swath -f latex -u u,t mydoc.tex > mydoc.ttex$ latex mydoc.ttex

  This is equivalent to filtering with iconv(1):

  $iconv -f UTF-8 -t TIS-620 mydoc.tex | swath -f latex > mydoc.ttex$ latex mydoc.ttex

• For HTML (to provide web pages to web browsers that cannot wrap Thai lines properly, but support the tag):

  $ swath -f html < mydoc.html > mydoc-wbr.html