It was magical in my last project. :)
the text and tables myself. So put the data in XML format, and use XSLT to
display it in a HTML file. Works like a charm, saved me months of work, and
it looks good on everyone's computer regardless of resolution or DPI. This
I remember around 10 years ago when I was first learning about XML. We
were using it for database integration. Someone told me that XML describes
the data structure, and HTML describes what the data looks like. When it
was first being thrown about it was espoused as some mysterious solution to
everyone's problem, but no one really knew what it did. The reason for
that is that XML doesn't really "do" anything. It just describes some data
in a sort of readable format. We use Xerces for parsing since our
requirement is to use DTDs, but I wouldn't recomment it for anyone getting
started these days. I usually point people to the CMarkup classes. Really
easy to use and implement and they work well. You can get a free version
or you can pay for a more feature rich version.
Tom
"Joseph M. Newcomer" <newcomer@flounder.com> wrote in message
news:m9se33h2n8ol8asel6tln6e5525untebv4@4ax.com...
XML files are just files. There's nothing magic about them. You can use
the very complex
Microsoft XML support, even if it is only produces a DOM (one of the more
useless designs
for data representation), or roll your own with any of the open-source
XML libraries
(google will give you a ton of them, and not all are GNU-licensed so you
can use them in
products whose source you don't want to disclose).
Essentially what you do is read the files in and create representations
in memory of the
data. When you want to write them out you convert them back to XML. I
typically use
Serialize methods on my structures, which recursively call serialize
methods of the
substructures. If the data is massive, you may have to deal with
incremental loading, and
if it is very small, you can probably roll-your-own representation; I
once did one with
the actual text, and methods such as "GetName" which parsed the
substructure in real time
to retrieve the name (for small databases and slow interaction rates,
this works really
well). But it's just a file, and whatever meaning you want to apply to
it is up to you.
joe