Saxon

Legacy ID: #5197982 Legacy Poster: Scott A. Colcord (sacolcor)

<http://www.saxonica.com/documentation/conformance/xslt20.html>, item #13 (xsl:number support), says that Saxon provides support for Kanji digits through codepoint (x4e00). The digits themselves are correct. However, Saxon (as of v9.0) appears to use them as though they were ordinary decimal digits, producing the following sequence: 1: 一 2: 二 3: 三 4: 四 5: 五 6: 六 7: 七 8: 八 9: 九 10: 一〇 11: 一一 12: 一二 13: 一三 14: 一四 15: 一五 16: 一六 17: 一七 18: 一八 19: 一九 20: 二〇 21: 二一 22: 二二 23: 二三 But in CJK languages, number counting is done using a marking and grouping algorithm, and should, I believe, generate the following sequence: 1: 一 2: 二 3: 三 4: 四 5: 五 6: 六 7: 七 8: 八 9: 九 10: 十 11: 十一 (10 + 1) 12: 十二 (10 + 2) 13: 十三 14: 十四 15: 十五 16: 十六 17: 十七 18: 十八 19: 十九 20: 二十 (210) 21: 二十一 (210 + 1) 22: 二十二 (210 + 2) 23: 二十三 (210 + 3) A description of this algorithm is at <http://www.w3.org/TR/css3-lists/#cjk-ideographic>. Note, however, that while the algorithm there appears to be correct, a message at <http://lists.w3.org/Archives/Public/www-style/2003Apr/0063.html> indicates that several of the CJK character sets on that page are not. Would it be possible to use the CJK algorithm when numbering using codepoint U+4E00? Even just handling numbers 1-99 correctly would cover the majority of cases. Thanks, Scott Colcord (PTC/Arbortext)