- From: John Hudson <tiro@tiro.com>
- Date: Wed, 10 Jul 2013 13:53:58 -0700
- CC: public-i18n-indic@w3.org
On 10/07/13 1:26 PM, Andrew Cunningham wrote: > If and when we get to SE Asian languages, Burmese will throw a spanner > in the works. In what sense? > A few of us have been discussing first-letter, and current concensus is > it should match first orthographic syllable which can be more than one > grapheme cluster. Right, but the orthographic syllable is also the core unit of script processing for Indic and Southeast Asian script layout. Hence, the 'Basic shaping forms' layout features in each of Microsoft's font specifications for these scripts are applied at the orthographic syllable level, which means the layout engine is first applying algorithms to delimit the orthographic syllables in a word, based on patterns of characters. We do similar analysis when doing conjunct frequency analysis on text corpora. It seems to me that the same approach could be taken to select orthographic syllables as units in CSS. JH
Received on Wednesday, 10 July 2013 20:54:31 UTC