Introduction
I am a Visual C++ guy. So why I am talking about XSLT and DHTML now?
Sometimes, I am tired of writing C++ codes. Think about how much code you
need to write when you want to sink events in ATL, how much code you need to write
when you build or manipulate collections in ATL. It's too much and boring.
But with DHTML, there are tons of building blocks you can use. I don't need
to build a grid control, I don't need to call GDI functions. And it has flexible
layout, various appearances.
This idea comes from two reasons: one is I always wanted to build a tree list view
by myself, the other is I am an XSLT learner now. ADOData.xml is the XML
file generated by ADO when you persist a recordset as XML. ADOData2Hier.xsl
is the XSLT file for transforming XML file into HTML file and does the work of grouping
data into hierarchies. You can find the grouping method in XSLT from different sources,
there is an article about it in Code Project too. Source codes are short, so just see
source code for details.
The platform I mentioned is the platform on which I built this. I think it should
be OK if you have IE5+ and XSLT (not XSL). I never used netscape because our company and I only
use IE.
The codes are just to show the idea. There are many possible improvements.
- declare global parameters in XSLT file, in the script call
addParameter
to decide which fields you want to do the grouping and summation (or other aggregation).
- any number of levels of grouping, also you can declare global parameters, but this
may require more work to do.
- there is an article "fun with tables" in MSDN, you can integrate code
from that article to do
some column reordering, column choosing, etc.
- use HTML, CSS features to make the appearance more colorful.
If somebody makes improvements, don't forget to send me a copy.
http://www.visualplant.com
