Feature request

[IPJ]

I've been searching for a duplicate finder that works inside archives for some time, AllDup almost does what I want but not quite, hence these feature requests:

1) Support deleting a duplicate file from within an archive when marked in the duplicate result list.
I have a set of 300K Zip files some of which contain duplicate files or are duplicated in a normal folder tree. I need to be able to locate duplicates within the archives and sometimes be able to delete the file in the archive rather than the file in the folder tree. So far, I can use AllDup to see the duplicates in the zip files and mark them for deletion but AllDup doesn't delete the files marked within the zip

2) When selecting to search for duplicates in zips or Rars, allow the CRC32 to be used rather than SHA or MD5 when comparing file contents.
Both Zips and Rars store the CRC32 value of the files stored in the archive as part of the meta data for the file. If you use CRC32 it would be no longer necessary to extract the individual files to a temp folder in order to calculate a file finger print. Just read the CRC32 value for the file while listing the files. This would massively speed up finding duplicates stored in archives. It would also avoid a major issue working with archives in AllDup see below.

BUG/Issue: when working with Zips or Rars, AllDup generates a lot of errors when trying to extract files from zips to calculate SHA values. Typically this occurs with archives that have very long names or have embedded zipped files with long names or paths. I believe that when extracting these files to the temp directory specified in the options you create a file/path name that exceeds Windows and or archive code limits on file/path name length. Even if I set the temp path to be "c:\t", I still generate 'cannot extract' or 'cannot calculate checksum' errors for these long filenames/paths. If you read the CRC32 from within the zip (having checked the zip has no CRC errors) then no extraction of files would be required and so these errors would largely be avoided plus offering a significant performance boost vs the extract/calculate SHA value approach.

BUG/Issue: Odd behaviour in the results list when using arrow keys to navigate through the list. If I try and use the down arrow key to move between entries, the selection will often refuse to move or refuse to leave one file group and move into the next file group but it sometimes works. I can't work out when it will or wont work but would like to be able to move to any row in the results just using the down/up arrow keys.
This also implies that left arrow should expand a closed file group and right arrow should close the file group. Additionally it would be desirable to support pressing Spacebar to mark a file for deletion. This is fairly normal arrow key/selection key behaviour in other applications including windows explorer.

[Administrator]

With the next update it will be possible to delete files from inside a zip file.

your suggestion about the use of the stored CRC32 from inside zip or rar files has been noted on our ToDo-List.

it would be desirable to support pressing Spacebar to mark a file for deletion

See search result menu bar -> options -> "edit hot keys" -> "Select/deselect file"...