On an EMR cluster, build and copy the target/s3-test-1.0-SNAPSHOT.jar
Then:
export BASE=s3://testdata
# For trying out with local HDFS
#export BASE=hdfs://$(hostname)/
hdfs dfs -mkdir ${BASE}/t1/foo
hdfs dfs -touchz ${BASE}/t1/foo/f1.txt
hdfs dfs -touchz ${BASE}/t1/foo/f2.txt
java -cp ./s3-test-1.0-SNAPSHOT.jar:$(hadoop classpath) com.waterlinedata.S3Test ${BASE} /t1/foo
Now create the "_$folder$" folder and rerun
hdfs dfs -touchz ${BASE}/t1/foo/_\$folder\$
java -cp ./s3-test-1.0-SNAPSHOT.jar:$(hadoop classpath) com.waterlinedata.S3Test ${BASE} /t1/foo
and run the recursive test:
java -cp ./s3-test-1.0-SNAPSHOT.jar:$(hadoop classpath) com.waterlinedata.S3TestR ${BASE} /t1/foo