Developer forums (C::B DEVELOPMENT STRICTLY!) > CodeCompletion redesign

Difference from branch and trunk...

<< < (2/2)

daniloz:
Hi again,

I'm back after a while, sorry... But I managed to put together a small project to reproduce my problem (and a new one that a found as well  8) ). Here it is:


--- Code: ---//enum TNoNoCpp {ENUM_1,ENUM_2};

typedef struct tagS
{
int a;
int b;
} STest;

void main()
{
    STest z;
    STest t[4];
    STest tt[2][2];

    z.a;
    t[1].b;
    tt[2][1].;
}

--- End code ---

For the code as it is above, the code-completion founds "tagS" and "STest" and give me suggestions when I type "z." and "t[1].", but not when using the double array "tt[2][1].". This is happening on both the latest nightly build (r5731), trunk r5755, trunk r5780 and codecompletion_refactoring branch r5780.

Now, if I uncomment the enum line, the code-completion lists on the symbol window just "TNoNoCpp" and don't car about the rest of the file anymore when I use the the trunk r5780 (which contains the code-completion code from the branch) and codecompletion_refactoring branch r5780, but everything works as expected in the latest nightly build (r5731) and trunk r5755.

The last bug was the originator of my post. I happens on all files after an enum definition.

#EDIT# I've updated to trunk r5790 and now the enum issue is solved. However, I noticed another thing. Even though, I see both "tagS" and "STest" in the symbol browser, only when I click on "tagS" I can see the members, but not when I click on "STest". I'd expect to see them on both, don't you?

So, right now, I'm stuck with the latest nightly build (r5731)... ;)   
#EDIT# OK, I'm using trunk r5790 now and I'm very happy with it.  :P  But still, the first bug is there...

Once again, thk in advance

ollydbg:

--- Quote from: daniloz on September 16, 2009, 01:24:53 pm ---Even though, I see both "tagS" and "STest" in the symbol browser, only when I click on "tagS" I can see the members, but not when I click on "STest". I'd expect to see them on both, don't you?

--- End quote ---

I tested it on my local copy(latest trunk version)  and TDM-MinGW 4.4.0.

This is true. Because the STest is regarded as a derived class from tagS. If you press "ctrl + shift " + double click, then exam these class in the dialog. see below:



Currently, in Symbol browser, I don't think a derived class will show all it's ancestor's public members. :D

ollydbg:
The bug can fixed by this patch, see this post and screenshot:

http://forums.codeblocks.org/index.php/topic,11187.msg76627.html#msg76627

Navigation

[0] Message Index

[*] Previous page