Incorporate bash code to discover all Hadoop combinations from Spark version

[guangie88]

SPARK_VERSION_TAG=master
curl -s https://raw.githubusercontent.com/apache/spark/${SPARK_VERSION_TAG}/pom.xml | grep \<hadoop.version\> | grep -oE '[[:digit:]]\.[[:digit:]]\.[[:digit:]]'
2.7.4
3.2.0

SPARK_VERSION_TAG=v2.4.4
curl -s https://raw.githubusercontent.com/apache/spark/${SPARK_VERSION_TAG}/pom.xml | grep \<hadoop.version\> | grep -oE '[[:digit:]]\.[[:digit:]]\.[[:digit:]]'
2.6.5
2.7.3
3.1.0

SPARK_VERSION_TAG=v2.3.3
curl -s https://raw.githubusercontent.com/apache/spark/${SPARK_VERSION_TAG}/pom.xml | grep \<hadoop.version\> | grep -oE '[[:digit:]]\.[[:digit:]]\.[[:digit:]]'
2.6.5
2.7.3