Skip to main content

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
RE: [cdt-core-dev] CDT Symbol Table implementation`


The index only stores the information required to enable features like Search and Open Type.  It serves purely as a table of relationships between different C/C++ elements in the code.  There is a user preference on Content Assist that allows it to also query the index, as this helps us out in the world of C where you do not have to include a header file in order to compile a translation unit using types defined in that header.   The indexer uses an IParser to generate an AST to figure out these relationships.   The AST interfaces serve as a facade to the underlying symbol-table implementation.  

We are looking at different ways of improving the performance and scalability of the index and parser, particularly in areas that it is noticable in the UI.  Unfortunately at this point, we are not caching include files for COMPLETE_PARSE modes, as this is a significant problem that requires a decent amount of investigation.  

JohnC
www.eclipse.org/cdt


cdt-core-dev-admin@xxxxxxxxxxx wrote on 05/25/2004 01:24:41 AM:

> As you told me that there is a .index file which is stored in
> workspace\.metadata\.plugins\ , is this file maintaining just an
> index of identifiers for searching and refactoring purpose or does
> it also play a part in content assist and text tool tip.Where i can
> find the details on the format of the file, if there is any specific.

> And can you guide me to any document that explains how the ast has
> been created in the cdt. Actually I may be able to understand that
> studying the code but this is one of the things I hate most in life,
> reading existing code (undocumented and uncommented if i may add).
> One thing is troubling me that why isn't there a search option in
> the mailing lists or newsgroups?
>
> ----- Original Message -----
> From: "Keithen Hayenga"
> Date: Mon, 24 May 2004 15:40:43 -0700
> To:
> Subject: RE: [cdt-core-dev] CDT Symbol Table implementation`

> The parser can do either a QUICK or COMPLETE parse. The main
> difference is that the quick parse does not have to follow all of
> the include paths. When you type a keystroke or more, a quick parse
> will be kicked off. This can be used to rapidly point out errors as
> you type for that nice incremental feel. At other times, such as
> when you save a file that had changed, a complete parse will happen
> which rebuilds the symbol index. The editor does have the concept of
> regions of text, but in my small test cases i only see parsing
> starting from the beginning of the file i am editing.

>
> --

> __________________________________________________________
> Sign-up for Ads Free at Mail.com
> http://www.mail.com/?sr=signup

> _______________________________________________ cdt-core-dev mailing
> list cdt-core-dev@xxxxxxxxxxx http://dev.eclipse.
> org/mailman/listinfo/cdt-core-dev

Back to the top