Powershell – aws s3 cp files from the root of a subdirectory without getting subfolders as well? Powershell

amazon s3amazon-web-servicesaws-clipowershell

s3://bucketname/yyyymmdd/foo/bar I have some files in the root directory and I want to copy them. I do not want to copy any subdirectories to the root (foo). I'm trying this in powershell:

aws s3 cp "s3://bucketname" . --recursive --exclude "*/foo/*/*" --include "foo/*.*"  --dryrun

or

aws s3 cp "s3://bucketname" . --recursive --exclude "yy*/foo/*/*" --include "foo/*.*"  --dryrun

I get the files in foo but I also get some files stored like s3://bucketname/otherdirectory/otherfiles. I tried some alternatives:
This appears to list EVERY file in the bucket:

aws s3 cp "s3://bucketname" . --recursive --exclude "*/foo/*/*" --include "*/foo/*.*"  --dryrun

My gut says it should be

aws s3 cp source destination --recursive --exclude "*" --include "*/foo/*.*" --dryrun

meaning /directory/*anyfilename.*anyextention but this gives me every file recursively below s3://bucketname/yyyymmdd/foo/.

I'm trying to whip up a one liner without foreach patterns but I may have to resort to that.

Best Answer

Figured it out:

(get-s3object -bucket bucketname | where{$_.key -like "*/foo/*"} | where{$_.key -notlike "*/foo/*/*"}).key

This list just the files in foo with no subfolders or subfolder files. It does this regardless of the YYYYMMDD for the bucket. I can now pipe this to read-s3object to get the files. 

get-s3object -bucket bucketname | where{$_.key -like "*/foo/*"} | where{$_.key -notlike "*/foo/*/*"} | read-s3object -folder .
Related Topic