Re: undo and combining characters

From: Andrew Dunbar (hippietrail@yahoo.com)
Date: Wed Apr 24 2002 - 21:26:24 EDT

  • Next message: Andrew Dunbar: "Re: undo and combining characters"

     --- Karl Ove Hufthammer <huftis@bigfoot.com> wrote: >
    Andrew Dunbar <hippietrail@yahoo.com> wrote in
    >
    news:20020424151710.65680.qmail@web9605.mail.yahoo.com:
    >
    > > If we are going to do auto-ligating, we really
    > should
    > > provide an option for it.
    >
    > This can be done in three ways:
    >
    > 1 Store ligating preference in the document
    > 2 Store ligating preferences in the AbiWord
    > preferences
    > (must only affect *new* documents; existing
    > documents
    > should use the preference stored in the document)
    > 3 On a case by case basis. E.g. the user
    > right-clicks on
    > a ligature and choses 'don't ligate', or selects a
    > non-ligated character sequence (where ligating has
    > been
    > disabled and chooses 'ligate'). This can be done
    > with

    Without this option it wouldn't be possible to
    produce an AbiWord document which demonstrated the
    difference between "fi" and the "fi" ligature!
    So I'm very in favour of it.

    > zero-width joiners and non-joiners. See section
    > 13.2
    > at <URL:
    > http://www.unicode.org/unicode/reports/tr27/ >.

    Zero-width characters in all their flavours are also
    something important we should be thinking about while
    designing a new layout engine. There's a number of
    them in Unicode and they'd be hell to retrofit.

    I don't think the ZWJ and ZWNJ are the answer here
    though. They're good for Arabic and Devanagari which
    always ligate by default. But they're very confusing
    for both users and developers.
    For English ligatures (optional ligatures) I would
    suggest an inline <ligate> tag which would do nothing
    if put around non-ligatable characters. A <noligate>
    tag also if we decide to go with an "auto-ligate"
    option.

    Andrew Dunbar.

    > > Note also that it will be
    > > font-dependent. It's always going to be possible
    > to
    > > use the 'fi' ligature codepoint but we might have
    > some
    > > fonts with the ligature and some without it.
    >
    > Yes, a ligature glyph should of course *only* be
    > displayed if it
    > exists.
    >
    > --
    > Karl Ove Hufthammer

    =====
    http://linguaphile.sourceforge.net http://www.abisource.com

    __________________________________________________
    Do You Yahoo!?
    Everything you'll ever need on one web page
    from News and Sport to Email and Music Charts
    http://uk.my.yahoo.com

    This archive was generated by hypermail 2.1.4 : Wed Apr 24 2002 - 21:27:55 EDT