On 12/10/2005, at 23.41, Sune Foldager wrote:
I propose that it should scan down and when it sees the second start fold marker at the same depth, add 1 to a counter; when it sees the first close fold marker at that depth, it checks the counter...since it is non-zero, it subtracts one. Finally, when it sees the last fold marker, and the counter is at zero, then it uses that as the correct, paired end fold.
It's conceptually done like that in latest build (I took a bit of freedom wrt the counter, since I already push/pop the markers on a stack ;) ).
How about specifying folding markers in pairs? So it would only match the correct start to the correct end.
Most languages only fold on { and }, or use a generic 'end' keyword to end different start blocks, so not much would be gained from this -- HTML comes to mind as one where there are different paired elements, but one could still have nested divs with same indent etc.