Multilingual I/O and Text Manipulation System(2) : The Structure of the Output Method Drawing the World's Writing Scripts beyond ISO 2022

概要

論文の詳細を見る
To realize the Multilingual Output Mechanism required by the first paper of this series, indepedencies of Locale Model, code set designs, codepoint extension methods and of writing conventions are essential. One codepoint does not always stand for one glyph in Graphic Character Set (GCS)s[1]. For example, in Thai in TIS 620-2533:1990, multiple codepoints determine one final glyph with Thai writing convention, and in Arabic in ISO 8859-6:1987, one codepoint has several glyph candidates, from which the final glyph is determined according to its position of the word. Thus, final glyphs cannot be determined by only ISO 2022[2] and ISO 6429[3]. In X11R5, GCS may be added to a locale table for multilingual drawing, but, X11R5 can draw only the GCSs from left to right only when each codepoint in a GCS is mapped to one glyph in a font file. X11R6 was implemented as multiple Output Method (OM)s binding corresponding OM dependent locale tables and as one of the OMs loaded when a locale was specified. Thus, it spends more resources and is fully locale dependent, and one OM in it does not ensure the same drawing manner on different systems because there are no drawing rules for extensions/writing conventions in the locale table.Based on Multi-locale Model and Global IOTMC Model[5], our OM newly developed satisfies all requirements to be multilingual OM.
一般社団法人情報処理学会の論文
1994-09-20