Thank you, this is exactly what I need and works well. Sometimes the simplest things are right under ones nose.
The reason I need is that I am working on the code for wisent-php at intervals and then use the updates work.
Its not so much bad parsing, rather it is my bad elisp skills. :-) Periodically some rather odd things are passed to semantic-tag-expand-function for example and this triggers an error, which I do not always have time to fix.
From: Eric M. Ludlam [mailto:firstname.lastname@example.org]
Sent: Fri 4/17/2009 4:46 PM
To: Mike Rowse
Subject: Re: Turn off semantic parsing for a buffer
It should be fine to edit code that Semantic cannot parse. Is there
something that it is doing that makes Emacs not usable? If so, I'd
rather fix that problem.
For a quick fix, you can just disable the idle scheduler for that
In just that buffer. I haven't tried that in a long time though so
I'm not certain what all the consequences are. You could also disable
it globally. Most things can be enabled/disabled on a per-buffer
basis independently of Semantic's parser installation state. It
should be possible to find the offender, and disable it.
>>> "Mike Rowse" <MR@waccglobal.org> seems to think that:
>I sent this along to cedet-devel, but I think it may have been eaten up =
>by a spam filter.=20
[ ... ]
>From: Mike Rowse
>Sent: Thu 4/16/2009 12:53 PM
>Subject: Turn off semantic parsing for a buffer
>I am wondering what the easiest way is to turn off semantic parsing for =
>a buffer? I have php files that have parsing errors that I do not always =
>have time to fix in a given moment. It would be nice to be able to =
>continue working on parse-able files with CEDET/Semantic along side =
>I tried to do the following:
> "A hack way to load a file in php-mode without CEDET"
> (make-local-variable 'php-mode-hook)
> (make-local-variable 'c-mode-hook)
> (remove-hook 'php-mode-hook #'wisent-php-default-setup t)
> (remove-hook 'c-mode-hook 'semantic-default-c-setup t)
>When I EDEBUG this function it appears to remove the wisent-php function =
>from a buffer-local version. Semantic kicks right back again (with =
>determination and persistence :-) ). I am baffled, I suspect it may =
>have to do with the idle parser or compiled code somewhere. I figure =
>there must be an easier way, but it is not obvious to me.