Le lun. 7 juin 2021 à 15:37, Elliotte Rusty Harold <[email protected]> a écrit :
> On Mon, Jun 7, 2021 at 1:21 PM Guillaume Nodet <[email protected]> wrote: > > > > The generated xml readers are used to actually read the pom. That's an > > important part :-) > > I spent quite some time profiling those and XPP3 is really fast, so I'd > > really like to avoid using JAXB for that. > > > > > > I agree with not using JAXB. It's a poorly designed API that causes > more problems than it solves, though I could say the same for all the > other data binding APIs out there. Not just for XML either. JSON and > SQL data binding APIs suffer from the same flawed goal of avoiding > writing the code necessary to understand the data. It never works and > it always ends up costing more in the end. > When I was working on an alternative to JAXB for Camel, I ended up writing a simple generator that generates the reader + xsd from POJOs. > > There are current issues in our pom.xml parsing, particularly around > namespaces, that aren't going to be resolved until we get rid of or > fix Modello, and possibly that's going to require fixing XPP3. I don't > know. However the XML parser should be an implementation detail. > Which parsing problems exactly ? I used the MXParser / XPP3 for the Camel reader and it has full support for namespaces afaik. > > -- > Elliotte Rusty Harold > [email protected] > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [email protected] > For additional commands, e-mail: [email protected] > > -- ------------------------ Guillaume Nodet
