Subject: Re: XSL is difficult to...? From: "Don Park" <donpark@xxxxxxxxx> Date: Thu, 29 Apr 1999 08:06:21 -0700 |
> My own view is that the XSL vocabulary is sufficiently small that it should > be possible for most of the commands to be single English words. I do find > it inelegant to have such a mix of different styles in one language: > > - ultra-verbose keywords (from-ancestor-or-self) > - long keywords (apply-templates, preserve-space, function-available) > - consise keywords (choose, if, match, select, position) > - abbreviated words (param, concat, id, docref, qname, pi) > - ultra-terse punctuation (//.[@*='?']/..) Very nicely summarized. If we can improve the first two, ultra-verbose and long keywords, I will be happier, er, quieter. One of the things that got me to raise these issues is your XSL compiler XSL file which was rather painful for me to figure out how it works. One of the problems, I think, was that I was reading the XSL file with the output in mind. Locating Java literal output in your XSL file by eyeball-only was very tiring so I ended up hopping around the doc (which is disorienting) while searching for specific keywords. Perhaps what we need is a coding guideline for XSL. For example, if the literal output tags and text were positioned at column 0, they might be easier to locate than if they were indented along with XSL tags. Best, Don XSL-List info and archive: http://www.mulberrytech.com/xsl/xsl-list
Current Thread |
---|
|
<- Previous | Index | Next -> |
---|---|---|
RE: XSL is difficult to...?, Kay Michael | Thread | Specifying output encoding type in , Yoichiro Hirano |
Re: Accessible FOs (was: Re: Transf, Paul Prescod | Date | RE: text() Pattern, Rob Nichols |
Month |