Behavior: Advanced->Files->Search for existing data files

[mblauser]

I tested the behavior found under Advanced->Files->Search for existing data files which led to a surprising and disappointing result. There is room for improvement of this feature. I am not sure of the intended behavior of the feature, so I don't know if I've found a legitimate bug.

Setup:

1) Downloaded seven files within a .torrent containing eighty-four files.

2) Stopped the .torrent

3) Renamed and sanitized the directories and filenames

4) Placed the files within a subdirectory for recursive testing

5) Invoked Advanced->Files->Search for existing data files using a base directory containing my subdirectory

Structure:

~/torrents

~/torrents/ExistingFilesBaseDir

~/torrents/ExistingFilesBaseDir/ExistingFilesSubDir

~/torrents/ExistingFilesBaseDir/ExistingFilesSubDir/NewFilename1

~/torrents/ExistingFilesBaseDir/ExistingFilesSubDir/NewFilename2

~/torrents/ExistingFilesBaseDir/ExistingFilesSubDir/NewFilename3

~/torrents/ExistingFilesBaseDir/ExistingFilesSubDir/NewFilename4

~/torrents/ExistingFilesBaseDir/ExistingFilesSubDir/NewFilename5

~/torrents/ExistingFilesBaseDir/ExistingFilesSubDir/NewFilename6

~/torrents/ExistingFilesBaseDir/ExistingFilesSubDir/NewFilename7

Output:

1/4/18 2:35 PM: Enumerating files in /Users/me/torrents/ExistingFilesBaseDir

Found 7 files with 7 distinct sizes

Processing 'MyTestTorrent', piece size=4.19 MB

    Matched=0, complete=29, no candidates=55, remaining=0 (total=84)

1/4/18 2:35 PM: Complete, downloads updated=0

My findings:

• First of all, I am pleased that this feature searches recursively.
  
• Although this feature identified all of the files in my target directory, it did not match on any of them. 
• This feature doesn't seem to work if the files have been renamed, even when using the quick rename function on each of the files within the downloaded torrent to match the filenames on disk.

Enhancement Opportunity:

• Begin by looking for matching filenames on disk, then compare via hash
• Continue by looking for matching filesizes on disk [ignoring filenames], then compare via hash (works for completed files). This would identify completed files that have been renamed.

Background:

Please take a lot of creative liberties (or laziness) in naming files. All too often I find myself downloading some that I already have because the filename was misleading or not informative. Using this feature above would save on a lot of duplicate downloads, or help seed the swarm by identifying files that I may already have elsewhere and renamed for sanity (instead of downloading duplicates).

This feature may help alleviate a gripe that I have with torrents in general. Often a torrent will contain a bunch of files that exist up to a certain point in time, but there isn't a mechanism for adding additional files as they become available after that point. Therefore, a new torrent is created containing all of the originals plus new content. It would be nice to identify the files which already exist, then download only the missing (new) files.