From keld@dkuug.dk Thu Jan 16 05:51:11 1992 Received: by dkuug.dk (5.64+/8+bit/IDA-1.2.8) id AA13482; Thu, 16 Jan 92 05:51:11 +0100 Date: Thu, 16 Jan 92 05:51:11 +0100 From: Keld J|rn Simonsen Message-Id: <9201160451.AA13482@dkuug.dk> To: Teruhiko.Kurosaka@eng.sun.com, taylor%limbo.intuitive.com%xopusw@Sun.COM Subject: Re: (SC22WG14.174) Re: (XoJIG 432) (i18n.146) Re: Support for symbolic character names Cc: i18n@dkuug.dk, wg14@dkuug.dk X-Charset: ASCII X-Char-Esc: 29 > Dave Taylor writes: > |the collating / transliteration tables nice and obvious too. Indeed, > |one wonders why we don't just have everything defined that way anyway, > |so that regular C could contain tests like: > | > | if (ch == COLON || ch == EXCLAMATION_MARK || ch == ASTERISK) > | > |rather than the much less portable, and more cryptic tests like: > | > | if (ch == ':' || ch == '!' || ch == '*') > | > Interesting. To me, the former is more cryptic than the latter. The > latter is much more straight. And I don't understand why the former > is more portable than the latter. Only advantage I can see in the > former way is that you can edit a program on the machine that lacks > the character you'd like to display on the target system. Well, that is also an important issue. That allows people to write and maintain C programs on such machinery. And such machinery may be very widespread in some communities. The main problem with the above scheme is that you cannot use it in strings, else is is conceptually very much the same as the one I have proposed. There is not much difference between writing if (ch == COLON) .... and if (ch == L'\') .... The other thing is that you build on the information you have in the charmap according to my proposal, and that is the standard place to define your character set. So why invent another mechanism for something which already exist. One problem with the #define solution is also that it restricts itself to variable names, which then means that you will use some sort of natural language, eg English or Danish. This will make your program a bit cultural dependent and thus less portable. Would you like to maintain a program with German or Italian character names? The POSIX charmap scheme can then be made less cultural dependent, as \ is equally evident to a German, Danish or English - even Japanese - programmer. > BTW, is anybody trying to name all the 10,000+ Chinese characters? Yes, this has been done in SC2/WG2. They have just numbered them and then they use the number as the name. Keld