Skip files of the same name when duplicating batch conversions to avoid reloading files repeatedly

[szfz]

For batch conversion, if you choose "Skip" when the output file already exists, the program will load the file being converted before detecting whether the file with the same name already exists in the target folder and then perform the skip action, instead of comparing the file names first and then loading the file being converted after there is no file with the same name, which will take a lot of time. This repetitive loading process is very unnecessary after a program crash or when performing the conversion again after a stop.

[xnview]

with latest version?

[meteorquake]

It does so for me too, I've had to manually remove the items from the list. I came on to post about it, but see it's been mentioned here already.
Version 1.8.6 64bit W10
Certainly for filenames that don't use meta info from the picture I think it would be trivial, for those that do I presume there will be several options to approach the skipping without much overhead.
d

[xnview]

Please describe your problem and how to reproduce

[meteorquake]

This continues to be a problem for me. If you're converting 10000 images and have to stop it, then to resume it you'll want to do the same images again but skip the existing target ones. However to skip them xnview (1.9.8 tested) loads-converts them first and only then on save finds out they exist and skips them whereas it should test their existence before any action and avoid the conversion act.

To reproduce, copy say 10 image files F of size to a temporary folder T. Then batch convert F to T selecting skip and you'll find they are processed slowly and the end log reports them being loaded but the save action skipped.

[xnview]

To reproduce, copy say 10 image files F of size to a temporary folder T. Then batch convert F to T selecting skip and you'll find they are processed slowly and the end log reports them being loaded but the save action skipped.

The problem is that the output filename is only known after reading input image, you can have place holder. Which output filename do you use?

[meteorquake]

The output name is {Filename} so doesn't require converting the image
I think it would be extremely rare for the destination name to require image conversion or full loading, at most people might sometimes read the meta info for the file if their filename template called for it, but with just {Filename} not even that (these are normal JPGs)