[EMAIL PROTECTED] (Dave Rolsky) wrote: > On Wed, 12 Feb 2003, Peter J. Acklam wrote: > >> Alas, the current API is a bastard -- an odd mixture of the two >> and doesn't match *any* convention: It doesn't match the "44 BC"- >> type notation since you have to use a leading "-" rather than the >> trailing "BC", which is not very intuitive; and it doesn't match >> the astronomical convention since it doesn't allow a year zero, >> although since the year is specified simply as an integer it looks >> like the astronomical convention is used. > > Uh, speak for yourself, Mr. Christian.
Err...I used "BC" only because Eugene van der Pijll used it in his example. > I suppose we could accept both, of course, but I'm really not > convinced that its terribly useful, especially since a good > chunk of the time the input to the year function will be a > variable, not a string typed in by the programmer. But so much else in this module is ISO 8601 compliant, like the default output formats, week dates, etc. Letting the year be an exception is inconsistent. Peter -- #!/local/bin/perl5 -wp -*- mode: cperl; coding: iso-8859-1; -*- # matlab comment stripper (strips comments from Matlab m-files) s/^((?:(?:[])}\w.]'+|[^'%])+|'[^'\n]*(?:''[^'\n]*)*')*).*/$1/x;
