Re: [xsl] Unix Philosophy Applied to XSLT Development

Subject: Re: [xsl] Unix Philosophy Applied to XSLT Development
From: "Steven D. Majewski steve.majewski@xxxxxxxxx" <xsl-list-service@xxxxxxxxxxxxxxxxxxxxxx>
Date: Sat, 23 Jul 2022 20:13:12 -0000
I often run my xslt:transform from XQuery and use the => operator to chain
transforms together.
This also only parses and serializes only at the start and end of the process.
Also, I probably structure my XQuery code with more small functions. I know
you can write functions in XSLT, but the XQuery syntax just leads me to
organize things by function rather than template.

Also, XQuery feels (to me) more like a scripting language where you can
interactively try things out, where XSLT feels like it requires more up-front
thinking and analysis.

I would also add the push style programming with templates in XSLT, already
provides a different method of separation of concerns, in that each template
may only do one simple thing - the complex part is handled my the dispatcher
doing the pattern matching. However, Ibm not sure that the way to structure
templates to make the easily extensible comes easily b or at least, doing so
isnbt (in my experience) how itbs taught or commonly done. Ibm thinking
of things like using apply-imports to allow customizations of templates by
adding another module import.


b Steve M.


> On Jul 23, 2022, at 11:17 AM, Norm Tovey-Walsh ndw@xxxxxxxxxx
<xsl-list-service@xxxxxxxxxxxxxxxxxxxxxx> wrote:
>
>> To recap: Identify general tools. Implement the tools. Use the tools
>> to solve many problems, not just one specific problem. Arrange the
>> tools in any order.
>
> Therebs a certain elegant simplicity to running everything through pipes
> on the Unix command line, but be aware that youbre serializing at the
> end of each step and parsing again at the beginning of each step.
>
> This can be expensive and may involve loss of information (schema types,
> for example, from a validated instance in the first step wonbt be
> preserved in the second unless you re-validate after re-parsing).
>
> What you want is an XProc pipeline :-)
>
>                                        Be seeing you,
>                                          norm
>
> --
> Norman Tovey-Walsh <ndw@xxxxxxxxxx>
> https://nwalsh.com/
>
>> It is as easy for the strong man to be strong, as it is for the weak to
>> be weak.--Emerson

Current Thread