search

archivematica / Issues

Issues repository for the Archivematica project
GNU Affero General Public License v3.0
16 stars 1 forks source link

Problem: file counts are off by 1 in Backlog tab #538

Open sallain opened 5 years ago

sallain commented 5 years ago

Expected behaviour The file count column on the Backlog tab should show an accurate count of the files within a transfer.

Current behaviour The file count column looks like this: image

However, if I download the packages and run find my-package -type f | wc -l, I always get a number that is one higher than the count on the Backlog tab.

screenshot from 2019-02-28 16-28-16

I confirmed this manually by making a list of all the files in one of the transfers and manually counting them. The number agreed with the command, not with the Backlog tab file count.

Steps to reproduce

  1. Create a transfer and put it in backlog.
  2. Download the package from backlog.
  3. Count the number of files in the package.

Your environment (version of Archivematica, OS version, etc) 1.9 qa

For Artefactual use: Please make sure these steps are taken before moving this issue from Review to Verified in Waffle:

sromkey commented 5 years ago

Could the one higher be accounted for by the transfer METS?

ross-spencer commented 5 years ago

I thought I was having dejavu but I only have a similar one in Archival Storage: https://github.com/artefactual/archivematica/issues/1066 (although i should check if it still happens)

sallain commented 5 years ago

@sromkey when I count the files in a downloaded package, I'm definitely including the transfer METS in that. If Archivematica is excluding the transfer METS, it would presumably be because there's some sort of error in how Archivematica is doing the counting. There are no files with identical names, or hidden files, or anything obvious that I can see that would cause this!

sallain commented 5 years ago

@ross-spencer I looked to see if this had ever been reported and found that issue! Kind of the inverse :joy:

sromkey commented 5 years ago

I wonder if it's somehow counting all the objects in the package and also the package itself?! I'll stop speculating now :joy:

ross-spencer commented 4 years ago

Just a quick note. As I was in this code today I asked why it was happening. It is because we have elected to ignore certain files per this line. So we don't see processingMCP.xml in the final count even though it appears in the package.