|
![](/i/fill.gif) |
Darren New wrote:
> Bernd Fuhrmann wrote:
>
>> Yes, it is. There are a lot of features missing that are used in a lot
>> of modern programming languages. Examples:
>> * Structs/Classes
>> * References, esp. to functions
>> * Namespaces
>> * strict type checking
>
>
> There are plenty of large high-performance production servers as well as
> end user desktop software written in languages that lack most or all of
> these features. Their lack doesn't make programming languages "not real".
>
>> <xsl:if test="$IndX<MaxX/2">
>
>
> The problem with doing something like this is, you then get to ask a
> question like "transform this scene to remove items that fail this
> condition." How about "select in this scene objects that don't
> intersect the sphere"?
This is not possible without giving XSLT some information about what a
sphere is. This is possible, but it will get
> If you're not going to code everything in XML, anything you don't code
> in XML is going to be opaque to whatever transforms you want to do. You
> won't even be able to say "shift this element 10 units right."
Right.
>> Still, this does not solve those namespace and data access problems.
>> There still needs something to be done. I just have to think about
>> another way...
>
> Write a parser for SDL. Since the help files describe the SDL using
> basically BNF notation, you can use yacc/bison to write a parser for it.
> I'm not understanding why it would be difficult, unless the docs don't
> match the reality, in which case you have a bug report. I mean, you
> even get the source to an existing parser, yes? :-)
I could write my own parser, sure. But that wouldn't fix the problem.
I'd still have unprocessable code.
As mentioned earlier: I will have to think about something different to
solve my problem. XML is a solution but due to certain limits of XSLT
not the best one.
Regards and thanks for all your comments,
Bernd Fuhrmann
Post a reply to this message
|
![](/i/fill.gif) |