Re: [cedet-semantic] Iterative visiting of all source files

[rjd <324...@gm...>]

Eric Ludlam <eric <at> siege-engine.com> writes:

> 
> On 12/15/2014 06:22 AM, RJD wrote:
> 
> >
> > I am very interested in this. What I do like with MSVS intellisense is that
> > is seems to parse all headers when you load up a new project. I am happy to
> > wait there while this happens so I can browse tags easier.
> 
> Hi,
> 
> For a given file, the command
> 
> M-x semantic-debug-idle-function RET
> 
> will parse and setup databases for everything needed for a given file. 
> There is a similar command:
> 
> M-x semantic-debug-idle-work-function RET
> 
> that will also handle all files in a directory of the same mode as the 
> current file.
> 
> > 1. With GNU TAGS - does this mean semantic does not need to parse? Does it
> > simply refer to the TAGS file produced by the gtags program
> 
> It will use GNU tags to find symbols in files not yet parsed by 
> semantic, but when Emacs visits the symbol, the regular semantic parser 
> takes over to get the extra details.
> 
> > 2. And how do you tell semantic to use the tags file instead?
> 
> This is how to configure things to use GNU Global:
> 
> http://www.randomsample.de/cedetdocs/cedet/GNU-Global.html
> 
> It doesn't disable the other tagging system, it just augments it for 
> basic symbol search.
> 
> Eric
> 
> 
Thank you Eric,

but now I ask: does this mean that Jan's version of parsing - iteratively -
all files is just an alternatively correct way to parse all files?

Does not seem to be using (semantic-debug-idle-work-function) [as you can
see in the original post].

1. If I wanted intellisense like things, I guess I could recurse all the
source files (cpp/h) and run this function on it. Similar to original post.
I also do not use EDE for the same reasons as Jan.

2. If I do write a custom lisp to parse all files from root project then I
guess I do not need to use GNUTags? Since the DB is fully formed...

Thank you so far,
learning a lot about this useful tool.