D21287: Try to find always an unique visible document name without a number suffix
noreply at phabricator.kde.org
Thu May 23 06:15:54 BST 2019
loh.tar added a comment.
> the path is appended behind the document name. What about this solution?
Don't like that idea. And fit not well to the shown "Document List" problem.
> to be efficient one could hash all documents
I'm always in favor for such. But here I had no idea how to do that nicely and reliable. One Document is not the place to hold such hash, and the patch lives only there.
The "file rename case" may indeed not catch here when it goes from duplicate to unique, but how often is that the case? Guess nearby never and it is no big harm to keep the prefix in that case. "Untitled" documnets are handled as before.
> that would avoid that one needs to iterate always again over all documents to collect duplicates.
Is a loop running once over 10 to 1000 (one should rethink his workflow when he has so many files open) turns on a file name update such an issue and any optimization worth? Well, currently lead that on session restore to one million loops, how long may take that? And compared to the other needed work to open one file. Guess it's not a big thing.
To: loh.tar, #ktexteditor
Cc: dhaumann, cullmann, kwrite-devel, kde-frameworks-devel, #ktexteditor, domson, michaelh, ngraham, bruns, demsking, sars
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the KWrite-Devel