Q: Decompositions

From: Michael Mehlich (mmehlich@semdesigns.com)
Date: Wed May 21 1997 - 13:30:54 EDT


Hello,

I have some problems regarding the decomposition of characters:

1. The standard contains a character block for CJK compatibility
   ideographs that are used in other standards for indicating
   alternative pronounciations. The character block name indicates
   that these characters are compatibility equivalent to their
   counterparts in the CJK unified ideographs block. However, as
   they are only alternative pronounciation for the same character
   they may also be considered as canonical equivalent. I could
   not figure out from the standard which of these two variants is
   intended.

2. There are decomposition rules for hangul syllables into hangul
   jamos. Is this decomposition considered to be a canonical
   decomposition, a compatibility decomposition, or does it have to
   be considered as a different kind of decomposition?

3. It is possible to build a character "latin capital letter DZ with
   diaeresis" by building the sequence
      "latin capital letter DZ"; "combining accute accent"
   which would be rendered as a DZ with a diaresis over DZ placed in
   the middle.
   Applying the compatibility decomposition of "latin capital letter DZ"
   results in the sequence
      "latin capital letter D"; "latin capital letter Z";
      "combining accute accent"
   which would be rendered as a D followed by a Z with an acute only
   on top of Z.
   I would consider these two renderings as having different semantics.
   This means that a compatibility decomposition changes more than
   just formatting informations (though one might argue that a
   DZ with a ACUTE above is not a character used in any language; but
   then consider the appearance of new languages ...).

--
Michael Mehlich                    email: mmehlich@semdesigns.com
Semantic Designs, Inc.             voice: 512-250-1018
12636 Research Blvd. #C-214        fax:   512-250-1191
Austin, Texas 78759                www:   http://www.semdesigns.com



This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:34 EDT