[SVN] "Insert Matching Start Bracket" broken

[Gerd Knops]

On May 2, 2006, at 3:43 PM, Allan Odgaard wrote:

> On 2/5/2006, at 18:06, Gerd Knops wrote:
>
>> On May 2, 2006, at 9:50 AM, Rob Rix wrote:
>>> Yes, that's what it's supposed to do. There's basically no way to  
>>> tell
>>> the difference between which of these two cases should be  
>>> correct, so
>>> we have to pick a default, and the default is the nearest to the  
>>> end.
>> Makes sense, this one is ambiguous and no amount of parsing will  
>> fix that one. But then the example in the source should be modified.
>
> One idea we have though is to extract all candidates, and then  
> compare these with a list of known method prototypes. That way, at  
> least when using known methods, it will be able to pick the proper  
> one.

Cool. I do have a quick-and-dirty method parser in perl which I use  
to auto-generate header files. Would that be helpful?

Gerd