Michael Van Canneyt wrote 12 lines
On Mon, 13 Nov 2006, Marc Santhoff wrote:
> Hi,
>
> while planning an XML-centric program I found some related units in the
> fcl sources and many more inside the lazarus source tree.
>
> Which set of them would be the best choice?
It depends on what you want. The XML support in FPC is based on the
DOM specifications. It is also tested against the specs on W3.
Good to know. So nothing speaks against using the fcl xml suport.
Show quoted text 8 lines
>
> According to the comments in some files there is (or was) change to
> come.
>
> Do these xml parsers validate against a schema file (.xsd)?
No, the parsing is currently not validating. I think Sergei Gorelkin
planned this, but I am not sure.
I think for my purposes (detecting type of element) it'll be enough to read in the schema file and address the type attribute of the element directly from the resulting dom object. I do not need to ensure structure or completeness, the files are machine generated.
Show quoted text 7 lines
>
> My goal will be parsing (of course) and looking up types of read
> elements in .xsd files for further evaluation. Techniques to use will be
> both, sax-style stream parsing and building DOM trees. The program will
> have a gui, so lcl will be in use, too.
The FCL contains a SAX 2 compliant reader.
So my requirements are met, speed is not a big issue in this case.
Thank you both,
Marc