|
Re: cached models [message #835737 is a reply to message #835659] |
Tue, 03 April 2012 15:58 |
Sebastian Zarnekow Messages: 3118 Registered: July 2009 |
Senior Member |
|
|
Hi Vlad,
you may want to try the features that are available here:
https://github.com/mark-christiaens/xtext
It will not make it into Xtext 2.3, but we have are currently working on
the necessary infrastructure changes to allow clients to implement that
without forking Xtext: https://bugs.eclipse.org/bugs/show_bug.cgi?id=359824
Regards,
Sebastian
--
Need professional support for Eclipse Modeling?
Go visit: http://xtext.itemis.com
Am 03.04.12 16:09, schrieb Vlad Dumitrescu:
> Another weird question from yours truly :)
> We have a large code base and it takes quite a while to parse everything
> when a project is open. Besides, most of the time people will navigate
> the code in read-only mode. To me, it makes sense to make sure that the
> serialized model for the baseline is available offline and let the
> framework use it directly instead of building it again and again.
>
> Is there support to do that in Xtext? More precisely, when opening a
> resource I want to be able to skip parsing it and use a serialized model
> from somewhere.
>
> My guess would be that I need to override the XtextResourceFactory to
> provide my special XtextResource that overrides doLoad, where I can fake
> calling the parser. Would that be enough?
>
> best regards,
> Vlad
>
|
|
|
|
|
Re: cached models [message #836430 is a reply to message #836423] |
Wed, 04 April 2012 13:35 |
|
Hi,
the index is persited in the workspace by default. see org.eclipse.xtext.builder.builderState.PersistedStateProvider.
so i do not get your point.
But: of course the parsed file is not stored in the index. btw parsing is usually not the problem. linking and validation usually cost more time.
=> you have to patch/fork xtext or live with what is there.
~Christian
Twitter : @chrdietrich
Blog : https://www.dietrich-it.de
|
|
|
|
|
|
Powered by
FUDForum. Page generated in 0.03340 seconds