On Thu, Jun 19, 2008 at 11:52:28AM +0100, Jonathan Kew wrote: > On 16 Jun 2008, at 5:02 pm, Apostolos Syropoulos wrote: >> This means that the \uccode for ETA WITH PSILI AND OXIA is CAPITAL ETA >> WITH PSILI AND OXIA which absolutely >> totally wrong! > > But it is defined as such by the standard: > > 1F24;GREEK SMALL LETTER ETA WITH PSILI AND OXIA;Ll;0;L;1F20 > 0301;;;;N;;;1F2C;;1F2C > > (see http://unicode.org/Public/UNIDATA/UnicodeData.txt) > >> In Greek when one capitalizes letters they all lose >> their accents.
The root meaning of "capitalize" (from "caput") refers to the practice of making the first letter in a word uppercase in proper names, at the start of a sentence, etc. In this usage, in the traditional typography of ancient Greek, capital letters do not lose their accents; they shift position. On the other hand, when an entire word is set in uppercase, these letters do lose their accents, traditionally. The Unicode standard is not well equipped to articulate this distinction -- indeed, it deliberately ignores this situation. I think that the correct level for this typographic practice to be handled is in OpenType. I am no expert on it, but I imagine that it should be possible for a font to provide mappings for text in small caps and all caps that do away with accents for text set in these typographic contexts. It is true that there are some oddities in the way the Unicode standard handles polytonic Greek capitalization on the extended Greek codepage, but in the long run these are best dealt with by moving gradually to Normalization Form D. Since text in all caps or small caps is typically found in display contexts (titles, running heads, etc.) rather than running text, I do not think that the removal of accents in these cases is of much relevance to hyphenation rules. It is much more important to get the hyphenation right for words that start with accented vowels which happen to be capitalized at the start of a sentence in running text of ancient Greek. If I am understanding all this correctly, I think the conclusion is that XeTeX should abide by the Unicode standard here. With best wishes, Peter -- Dr Peter Heslin | Durham University Tel. +44 (0)191 334 1682 Department of Classics and Ancient History Fax. +44 (0)191 334 1671 38 North Bailey | Durham DH1 3EU | England http://www.dur.ac.uk/p.j.heslin
