fix: assign unique names to new data sets #294

	name: Regression Label

	on:
	pull_request:
	types: [ labeled ]
	jobs:
	re_run:
	# only continue if a push or PR labeled with 'run regression'
	if: github.event.label.name == 'run regression'
	runs-on: ubuntu-latest
	permissions:
	actions: write
	steps:
	- name: Re Run last push
	# get the last regression run triggered by a pull, and get this run's id
	# then rerun the run
	# if the run id can't be found the rerun command will fail which should
	# fail the job
	# When the run hasn't finished yet, we try to cancel the run
	# and wait 30s for it to finish canceling before trying to re-run it.
	# If the run isn't complete the rerun command prints this message:
	# returned: run 6410886572 cannot be rerun; its workflow file may be broken
	run: \|
	run_id=$(gh run list -e push -b ${{github.head_ref}} -w 'CI v3 Regression' -L 1 --json databaseId -q '.[0].databaseId')
	run_status=$(gh run view $run_id --json status -q '.status')
	if [[ "$run_status" != "completed" ]]
	then
	echo "run $run_id is $run_status, trying to cancel"
	gh run cancel $run_id
	count=0
	until [[ "$run_status" == "completed" \|\| $count -gt 15 ]]
	do
	sleep 2
	count=$((count+1))
	run_status=$(gh run view $run_id --json status -q '.status')
	echo "run: $run_id, status: $run_status, try: $count"
	done
	fi
	echo "rerunning $run_id"
	gh run rerun $run_id
	env:
	GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
	# set repository so we don't have to check out all of the code
	GH_REPO: ${{github.repository}}

Provide feedback