On 28/01/2004 08:22, Doug Ewell wrote:

Peter Kirk <peterkirk at qaya dot org> wrote:



Thank you, Doug. This is a useful test page for IPA support. It would
be even more useful if it attempted to display 02E5-02E9, perhaps with
a suitable note about limited support.



Just to clarify for the list: The page does display the tone letters U+02E5 through U+02E9 *in isolation*, as shown in the Unicode charts. What Peter is suggesting is that it should also attempt to display the contour marks using suitable combinations of the tone letters, "perhaps with a suitable note about limited support."



Indeed. More specifically, I would suggest an attempt to render the right half of the image in http://www.arts.gla.ac.uk/IPA/tones.html. The ligated forms based on 02E5-02E9 can be encoded in Unicode now, and can be displayed correctly in any font in which the ligature is encoded. There is more of a problem with the contour tone diacritics, above the e's in the right half of the image, as three of these combining marks are indeed missing from Unicode. This is similar to the list quoted by Peter Constable in N2646 (http://std.dkuug.dk/jtc1/sc2/wg2/docs/n2646.pdf) figure 9.

It is interesting that Peter considers treating these contour tone combining marks as ligatures of simple combining marks, but rejects this on the grounds that ZWJ and ZWNJ cannot be used with combining marks. If this restriction is removed as proposed in http://www.unicode.org/review/pr-27.html (the review period for which has just expired), this mechanism again becomes a possibility; and this may be an additional argument in favour of this review issue. Without such a ligation mechanism, there is a possibility that not only the six new combining marks mentioned in N2646 may be needed, but a set of around 100 corresponding to the around 100 Chao tone ligatures proposed in N2646.

WG2 documents N2626 and N2646 are useful starting points for
understanding the relationship between the five basic tone letters and
the numerous contour marks that can be constructed from them.



N2646 seems more useful. A more basic reference is http://www.unicode.org/versions/Unicode4.0.0/ch07.pdf p.185.

I note in passing that in http://web.uvic.ca/ling/resources/ipa/charts/unicode_ipa-chart.htm the tie bar between k and p is not rendered at all in Mozilla 1.6 on Windows 2000.

--
Peter Kirk
[EMAIL PROTECTED] (personal)
[EMAIL PROTECTED] (work)
http://www.qaya.org/




Reply via email to