[Interest] Depth-first filtering for QAbstractProxyModel

[André Somers]

Op 06/09/2016 om 21:20 schreef Thompson, Adam B.:
>
> All,
>
> I co-develop an application that uses a QTreeView to display the 
> hierarchical contents of external files. The contents of the tree can 
> become quite a chore to sift through, so I added a means of filtering 
> via a QSortFilterProxyModel. This works as expected, but the user may 
> want to specify a pattern that matches a node deep in the model’s 
> structure but not any of the node’s ancestors. As such, the 
> breadth-first filtering mechanism provided in QSortFilterProxyModel 
> isn’t appropriate for our needs.
>
> I modified filterAcceptsRow to work in a depth-first manner to verify 
> my desired filtering effect would work, which it does. However, as 
> expected, it yields a huge performance penalty since it performs a 
> depth-first search for each row in the model. This is because the 
> (recursive) depth-first logic is called for the same nodes many times. 
> I added a crude caching scheme which really doesn’t properly mitigate 
> the performance issue.
>
> My question: can anyone provide some tips on how to properly subclass 
> QAbstractProxyModel so I can provide some custom depth-first filtering 
> capabilities? (Sorting isn’t really necessary at the moment, really 
> just filtering.)
>
>
Perhaps it is possible to do this in a more efficient way using the 
underlying data structure instead? That is: make it possible to query 
each node in your underlying structure directly instead of using the 
model API to iterate down the tree each time? How big is your model in 
terms of depth and number of nodes?

André

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.qt-project.org/pipermail/interest/attachments/20160907/7ced34cd/attachment.html>