[
Date Prev][
Date Next][
Thread Prev][
Thread Next][
Date Index][
Thread Index]
[
List Home]
Re: [cdt-dev] Code completion does not propose class members declaredbelow cursor (bug 103857)
|
As for the strategy, Basically we bail out at the curse since it's very likely what comes after it will be difficult to parse and at base have lots of error recovery. Mind you we didn't try it, so maybe it's not as bad as we feared. If you want to complete things declared after the cursor, I assume you'd need to try that.
On Mon, Oct 12, 2009 at 4:11 AM, Schorn, Markus
<Markus.Schorn@xxxxxxxxxxxxx> wrote:
The token tCOMPLETION marks the incomplete (potentially
empty) name at the point where the completion
is attempted. The token can be dealt with only once, after
that the scanner will return the tEOC token, which
basically tells the parser to bail out of any grammar
rules. (e.g. it allows the parser to create the ast for a
function
definition without looking at the closing
brace).
Markus.
From: cdt-dev-bounces@xxxxxxxxxxx
[mailto:cdt-dev-bounces@xxxxxxxxxxx] On Behalf Of Sergey
Prigogin
Sent: Monday, October 12, 2009 2:25 AM
To: CDT
General developers list.
Subject: [cdt-dev] Code completion does not
propose class members declaredbelow cursor (bug 103857)
Importance:
Low
I'd like to start working on
bug
103857. As a preliminary step I'm collecting information about
current design of code completion. In particular, I'm interested
in tCOMPLETION and tEOC tokens. Why there are two separate tokens?
What role each of them plays? Do these two tokens always share the same
location? A detailed answer from somebody familiar with the current design
will be highly appreciated.
Thanks,
-sergey
_______________________________________________
cdt-dev mailing list
cdt-dev@xxxxxxxxxxx
https://dev.eclipse.org/mailman/listinfo/cdt-dev