TODO (2660B)
1 # TODO 2 3 * Maybe allow opening an external program with the duplicates as arguments. 4 Alternatively, create a proper output format that deals with filenames 5 including newlines, etc., so that a separate program can parse the output 6 properly and delete the files or perform other actions. 7 * Add some sort of progress indication during signature calculation and 8 image data comparison. One idea would be to simply show the current 9 file being processed (all on the same line) during signature calculation 10 in order to give a general idea how far it has progressed. This would 11 also require the files to be sorted before adding them to the queue in 12 order for it to be useful. During the later stages, the number of the 13 current duplicate group can be printed, together with the total number 14 of groups (of course, the total number of groups can still become smaller 15 during processing). 16 * Allow adding filters to exclude filenames. 17 * Allow saving a backup of the metadata before deleting files. This would 18 also allow deleting files even if they contain metadata that is not 19 included in other files as long as the image data is duplicated somewhere. 20 * Allow configuring verbosity properly to print only some informational 21 messages. 22 * Add an option to compare metadata tags that are different between files, 23 for instance when both files contain a UserComment, but it is different 24 because it was edited at some point. Currently, that is considered to 25 be a mismatch, but it would be convenient to be able to easily choose 26 which file to keep instead of having to manually compare the tags using 27 ExifTool. 28 * Add an option to ignore specific metadata values for a certain tag. 29 For instance, I have had files where one copy contained a UserComment 30 that was just a 4-character string, possibly added automatically by 31 some program. This caused them to not be considered as duplicates even 32 though other copies contained proper UserComments. 33 * Maybe change the way ignored metadata tags are handled. Currently, they 34 are completely disregarded in the comparison, but it might be useful to 35 still consider them when deciding whether one file's metadata is a 36 superset of another file's metadata. For instance, if two images are 37 duplicates except that only one of them contains a UserComment tag, 38 and UserComment is being ignored, the current algorithm will not give 39 any preference which file should be kept and which should be deleted 40 since the UserComment tag is completely ignored. It might be better to 41 still take ignored tags into account unless doing so would cause the 42 files to not be considered as duplicates anymore.