[ADAM-1361] Fix misnamed ADAM überjar.

[fnothaft]

Resolves #1361.

[coveralls]

Coverage remained the same at 82.041% when pulling 884ba69 on fnothaft:issues/1361-spark2-artifact into 2820e94 on bigdatagenomics:master.

[coveralls]

Coverage decreased (-0.1%) to 81.93% when pulling 884ba69 on fnothaft:issues/1361-spark2-artifact into 2820e94 on bigdatagenomics:master.

[AmplabJenkins]

Test PASSed.
Refer to this link for build results (access rights to CI server needed):
https://amplab.cs.berkeley.edu/jenkins//job/ADAM-prb/2050/
Test PASSed.

[heuermh]

With this

diff --git a/scripts/move_to_spark_2.sh b/scripts/move_to_spark_2.sh
index ebd9fd89..3e83ebdd 100755
--- a/scripts/move_to_spark_2.sh
+++ b/scripts/move_to_spark_2.sh
@@ -19,5 +19,6 @@ find . -name "pom.xml" -exec sed \
     -e "/adam-/ s/_2\.1/-spark2_2.1/g" \
     -e "/utils-/ $substitution_cmd" \
     -e "/adam-/ $substitution_cmd" \
+    -e "/adam_/ $substitution_cmd" \
     -e "/spark.version/ s/1.6.3/2.1.0/g" \
     -i.spark2.bak '{}' \;

the <finalName> element can stay as is. This PR adds -assembly to the assembly artifact names.

[fnothaft]

@heuermh Since the assembly jars are largely handled programmatically from the ./bin/adam-* scripts, I don't really see a benefit to changing the JAR name with <finalName>. I'd rather eliminate <finalName> and have one fewer place where things can go wrong in the move_to scripts.

[heuermh]

I don't feel strongly either way.

[heuermh]

Thank you, @fnothaft