655 Fix to return correct column type for spatial datatypes

[peterbae]

Fixes issue #655. Also added test cases for it.

Also removed an unused import from Geometry/Geography classes.

[peterbae]

this appveyor failure is unrelated to this PR, it's a bug I found during internal testing that occurs only with SQL Server 2008. I will merge this PR after that issue has been solved.

[peterbae]

Actually I'm going to use a workaround for this problem, it seems like it could take a while to solve.

[codecov-io]

Codecov Report

Merging #657 into dev will decrease coverage by 0.05%.
The diff coverage is 57.14%.

@@             Coverage Diff              @@
##                dev     #657      +/-   ##
============================================
- Coverage     48.12%   48.07%   -0.06%     
+ Complexity     2578     2576       -2     
============================================
  Files           113      113              
  Lines         26574    26579       +5     
  Branches       4429     4432       +3     
============================================
- Hits          12790    12778      -12     
- Misses        11660    11671      +11     
- Partials       2124     2130       +6

Flag	Coverage Δ	Complexity Δ
#JDBC42	47.94% <57.14%> (ø)	2568 <0> (-2)	⬇️
#JDBC43	47.92% <57.14%> (-0.13%)	2574 <0> (-2)

Impacted Files	Coverage Δ	Complexity Δ
...in/java/com/microsoft/sqlserver/jdbc/Geometry.java	71.09% <ø> (ø)	33 <0> (ø)	⬇️
...n/java/com/microsoft/sqlserver/jdbc/Geography.java	69.53% <ø> (ø)	33 <0> (ø)	⬇️
...oft/sqlserver/jdbc/SQLServerResultSetMetaData.java	40.93% <57.14%> (-0.04%)	24 <0> (+1)
...om/microsoft/sqlserver/jdbc/ReaderInputStream.java	44.94% <0%> (-2.25%)	16% <0%> (-1%)
...rc/main/java/com/microsoft/sqlserver/jdbc/DDC.java	43.84% <0%> (-1.12%)	107% <0%> (ø)
...ncurrentlinkedhashmap/ConcurrentLinkedHashMap.java	38.54% <0%> (-0.86%)	42% <0%> (-1%)
...m/microsoft/sqlserver/jdbc/SQLServerResultSet.java	32.89% <0%> (-0.61%)	244% <0%> (-3%)
...rc/main/java/com/microsoft/sqlserver/jdbc/dtv.java	63.35% <0%> (-0.07%)	0% <0%> (ø)
...in/java/com/microsoft/sqlserver/jdbc/IOBuffer.java	54.23% <0%> (+0.06%)	0% <0%> (ø)	⬇️
...n/java/com/microsoft/sqlserver/jdbc/DataTypes.java	78.01% <0%> (+0.16%)	5% <0%> (+1%)	⬆️
... and 3 more

---

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 2039522...8ad8252. Read the comment docs.

[rene-ye]

I don't really have a problem with this, but why are we doing type check in 3 different ways. VARIANT is being checked separately outside the switch, and now GEOMETRY/GEOGRAPHY will be checked by name. Can we maybe stick to one form? This Microsoft overview of spatial datatypes seems to imply that it's only available in SQL Server 2012 and later, so shouldn't this be in the isKatamaiOrLater() block?

[peterbae]

We can't stick to one form. If you debug this part of the code, you'll see that the getSSType will give SQL_VARIANT or UDT. If it's UDT, we need to look at the SSTypeName to find out if it's Geometry or Geography, so this is the only way to check for these datatypes.

The overview page mentions that new spatial features have been introduced in SQL Server 2012, but this does not mean that spatial datatypes are supported starting from 2012 - it's supported starting from 2008. so we need this check for 2008 and up.

[cheenamalhotra]

Please create Constants for these strings and reference them here.

[cheenamalhotra]

Please create Constants for these strings and reference them here.