avoid unchecked casts in AbstractGremlinResultReader#readProperty

[harti2006]

Description

This is a small improvement of the readProperty method: I replaced the unchecked cast with some more robust ones that don't rely on a specific implementation of java.util.Map. This allows me to use this library with Amazon Neptune and the Serializers.GRAPHSON_V1D0 (org.apache.tinkerpop.gremlin.driver.ser.GraphSONMessageSerializerGremlinV1d0).
The recommended Serializers.GRAPHSON (org.apache.tinkerpop.gremlin.driver.ser.GraphSONMessageSerializerV1d0) did not work in my project, because it produced the same data structure as handled in #243.

Todos

• Tests
• Documentation

Steps to Test

• set up an instance of Amazon Neptune (without IAM role authentication)
• configure this library with Serializers.GRAPHSON_V1D0
• create a Vertex class and respective Repository
• use repository.save or repository.findAll

[codecov-io]

Codecov Report

Merging #256 into master will increase coverage by 0.01%.
The diff coverage is 75%.

@@            Coverage Diff             @@
##           master     #256      +/-   ##
==========================================
+ Coverage   86.21%   86.22%   +0.01%     
==========================================
  Files          60       60              
  Lines        1357     1358       +1     
  Branches      232      232              
==========================================
+ Hits         1170     1171       +1     
  Misses         74       74              
  Partials      113      113

Impacted Files	Coverage Δ
...conversion/result/AbstractGremlinResultReader.java	88.88% <75%> (+1.38%)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update a1c230f...13f6c81. Read the comment docs.

[harti2006]

The changed code is already covered by some unit tests.

[targeter21]

Could a reviewer look at this in the interest of it being merged? This seems to be all that's needed for working with AWS Neptune.