Open rly opened 1 year ago
Can you please attach a .txt. or .json of the full report? You may have to use the --detailed
flag as well to prevent aggregation
I'm having trouble reproducing the issue - the number of files displayed in that header is simply the number of unique NWB files that had messages returned: https://github.com/NeurodataWithoutBorders/nwbinspector/blob/dev/src/nwbinspector/inspector_tools.py#L118
So what I'm looking for in the full report is all the messages from all the files to confirm they appear as expected
I could technically add some directory logic to the string (like evaluate the unique parents of each file) but TBH I don't think it's all that helpful; the inspector and reports have no notion of sub-directories (or even memory of the top-level directory), only files, and files are found by recursive globbing from the main CLI calls
What happened?
When the
check_unique_identifiers
issue is raised, the directory is counted as a "file" in the output lineFound 21 issues over 3 files
. This can be confusing to users (see comment https://github.com/dandi/dandi-cli/issues/1266#issuecomment-1498924935). I'm not sure if there are other directory-level issues like this.Suggestion: change
Found 21 issues over 3 files
toFound 21 issues over 2 files and 1 directory
orFound 21 issues over 3 files and directories
Operating System
M1 macOS
Python Version
3.11
Were you streaming with ROS3?
No
Package Versions
Code of Conduct