ls in awscli returns "PRE". Why and how to get rid of it

you are able to do that with something like

aws s3 ls s3://directory --recursive | awk '{print $4}' | grep .json

What is "PRE"?

You can think of prefixes as a way to organize your data in a similar way to directories. However, prefixes are not directories.

How do I get rid of "PRE" in the output?

Use the aws s3api list-objects-v2 command which supports querying and output formatting instead of the aws s3 ls command.

aws s3api list-objects-v2 --bucket <mybucket_name> --prefix <path> --query "Contents[?contains(Key, '.json')].Key"

Note that <bucketname> here is just the name of the bucket. Do not include s3:// or any additional /.

The --query parameter is what gives you the powerful output.

For example

--query Contents[].Key would print all of the keys.
--query Contents[?contains(Key, '.json')].Key specifies additional criteria on the Key (the Key contains .json (e.g. the extension)) and then the final .Key tells it what attribute of the json output to return.

Related