Re: ZWL, ZWNL no difference?

From: Robert A. Rosenberg (bob.rosenberg@digitscorp.com)
Date: Mon Jan 03 2000 - 16:56:46 EST


At 02:04 PM 12/30/1999 -0800, peter_constable@sil.org wrote:

> >At 09:52 AM 12/30/1999 -0800, peter_constable@sil.org wrote:
> > In this
> > scenario, processes like spell checking would not
> ignore ZWL.
> > But, this has potential to end up in an ugly mess
> given no way
> > to control users from forcing ligation using ZWL in
> cases of
> > aesthetic, non-semantic ligation, with the result that
> spell
> > checks, etc. don't work as they're supposed to.
>
>
> BR>Why should the spell-checker NOT ignore embedded ZW[N]Ls?
> They should be using the canonical form of the string
> (including unligaturing pre-composed codes] by ignoring these
> codepoints. My primary home spell-checker will treats (if I
> remember correctly) ligatures in the same way it does accents.
> I'll need to check what it says about o-ffi-ce tonight <g>.
>
> The assumption in this scenario is that ZWL is used *only*
> where ligation makes a lexical distinction, i.e. is effectively
> a core part of the orthography; the ligature is being used to
> distinguish one lexical item from another; hence a dictionary
> would include, for example, both "wachstube" and
> "wachs<ZWL>tube" (or, alternately, both "wachs<ZWNL>tube" and
> "wachstube"). In the scenario I described, a ligature in
> "office" would not require ZWL to generate a ligature; it would
> appear automatically if the font supports it (and if the
> necessary font features are enabled - if ligation is not
> enabled by default).

I beg to differ on your restriction. We are talking about having pure
unstyled Text that we want to Spell Check. We are NOT grammar checking but
just doing a "Is this string of characters one of those on our list of
'valid spellings of words'" operation. Thus the string "wachstube" can be
spelled with or without the "long s" glyph and still be valid. Unless the
use of the "long s" glyph is not appropriate in any word where s occurs (or
vice versa), the existence of the ZW[N]L should not come into play for
doing the string checking. If you want to treat "long s" as a different
character than "s" (ie: The string is only valid if the "long s" is/isn't
used) then-and-only-then should the existence of a ZW[N]L be used in the
Spell Checking validation process.

The text may include ZW[N]L codes to mark where we want the characters to
use/not-use ligature glyphs (when directly printed or input into a
WordProcessor). Thus "of[ZWL]f[ZWL]ice" is a totally correct way to "spell"
office if I want to get o[ffi]ce when I print this text file or import it
into InDesign/Office2k/name-your-favorite-WP.

> Peter
>
>



This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:57 EDT