[prev in list] [next in list] [prev in thread] [next in thread] 

List:       kdevelop-devel
Subject:    Re: Please, raise your hand if you'd like...
From:       Adam Treat <manyoso () yahoo ! com>
Date:       2005-03-11 18:19:55
Message-ID: 200503111319.56180.manyoso () yahoo ! com
[Download RAW message or body]

On Thursday 10 March 2005 10:32 am, Anders Lund wrote:
> Yes, optionally. We have discussed a way to allow entries from several
> sources as well, so both icons and sorting could come in handy here. The
> idea is that it should be possible for sevral objects to use the
> completion, so that for example kdevelop code completion does not disable
> word completion etc. + in some cases, entries for more sources could be
> shown all at once maybe.

Wow, that would be cool.

> > * The ability to specify what parts of the entry are "completeable" eg,
> > if I'm completing a method decl I don't necessarily want to key the
> > return type?
>
> Doesn't we support that allready? there should be a prefix and a suffix
> that is not completed. The docs in ktexteditor/codecompletioninterface.h
> says:

Yes, but the prefix and postfix are not inserted upon completion.  only 
entry.text.  I'm looking for the ability to have prefix and postfix inserted, 
but I still don't want them part of the completion keying.  Make sense?  Like 
if I want to complete a method declaration, I don't want to have to key the 
return type before I get the method name, but I still need the return type 
inserted upon completion..

> > * KTE::CompletionEntry needs methods for specifying an icon to go with
> > the entry and comparison operator overloads to allow sorting by entry
> > type. It also needs some way of specifying that a prefix/postfix should
> > be inserted upon completion, but not part of the actual keying.
>
> This can be achieved using the 'userdata', as specified in the docs.

I'll look at the docs, but I don't know what part of the above it is supposed 
to help ;)

> > * KTE::CodeCompletionInterface needs methods for determining if the
> > current completion box isvisible and hiding methods.
>
> isCompletionBoxVisible() could be a valid addition. I don't understand why
> the hiding is required.

Hiding is required in some cases where we don't want the box any longer.  I 
want to abort it programmatically.  For instance, I don't want the completion 
box to stay up if I've pressed the delete key and no word is remaining to 
complete, but it does.  See the recent cvs commits in kdevelop for an 
example.

_______________________________________________
KDevelop-devel mailing list
KDevelop-devel@barney.cs.uni-potsdam.de
http://barney.cs.uni-potsdam.de/mailman/listinfo/kdevelop-devel

[prev in list] [next in list] [prev in thread] [next in thread]