Adding locale fix for turkey locale issue when lowercasing an "i"

[peterbae]

using toLowerCase() without any locale makes it implicitly toLowerCase(Locale.getDefault()). This is fine in most counties but in Turkey, the default locale converts uppercase "I" to a dotless "i", which is not the same as the lowercase "i" in English. Passing Locale.ENGLISH (or Locale.US) in toLowerCase() solves this problem. I have made these changes to all the applicable places in the codebase.

Fixes issue #381

[msftclas]

@peterbae,
Thanks for your contribution as a Microsoft full-time employee or intern. You do not need to sign a CLA.
Thanks,
Microsoft Pull Request Bot

[codecov-io]

Codecov Report

Merging #384 into dev will increase coverage by 0.02%.
The diff coverage is 33.33%.

@@             Coverage Diff              @@
##                dev     #384      +/-   ##
============================================
+ Coverage     40.16%   40.18%   +0.02%     
- Complexity     1892     1894       +2     
============================================
  Files           107      107              
  Lines         24482    24481       -1     
  Branches       4038     4038              
============================================
+ Hits           9833     9838       +5     
+ Misses        12815    12800      -15     
- Partials       1834     1843       +9

Flag	Coverage Δ	Complexity Δ
#JDBC41	40.07% <33.33%> (+0.02%)	1886 <0> (+3)	⬆️
#JDBC42	40.02% <33.33%> (ø)	1886 <0> (+1)	⬆️

Impacted Files	Coverage Δ	Complexity Δ
...om/microsoft/sqlserver/jdbc/SQLServerBulkCopy.java	49.35% <0%> (-0.15%)	212 <0> (+2)
...in/java/com/microsoft/sqlserver/jdbc/IOBuffer.java	45.14% <0%> (+0.22%)	0 <0> (ø)	⬇️
...QLServerColumnEncryptionAzureKeyVaultProvider.java	0% <0%> (ø)	0 <0> (ø)	⬇️
...c/main/java/com/microsoft/sqlserver/jdbc/Util.java	47.51% <100%> (ø)	63 <0> (ø)	⬇️
.../microsoft/sqlserver/jdbc/SQLServerConnection.java	45.06% <66.66%> (ø)	267 <0> (ø)	⬇️
src/main/java/microsoft/sql/DateTimeOffset.java	37.14% <0%> (-2.86%)	8% <0%> (-2%)
...om/microsoft/sqlserver/jdbc/ReaderInputStream.java	46.06% <0%> (-2.25%)	17% <0%> (ø)
...om/microsoft/sqlserver/jdbc/SimpleInputStream.java	47.4% <0%> (-1.49%)	9% <0%> (-1%)
...m/microsoft/sqlserver/jdbc/SQLServerStatement.java	59.06% <0%> (-0.31%)	129% <0%> (-1%)
...ncurrentlinkedhashmap/ConcurrentLinkedHashMap.java	42.2% <0%> (-0.22%)	46% <0%> (ø)
... and 5 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 3bd68b5...ab0dfa3. Read the comment docs.

[nsidhaye]

Few Suggestions:

1. Instead of using direct "".toLowerCase you can add new wrapper function(s) in StringUtil class. If locale is null then will use current default implementations.

[peterbae]

Hi Nikhil, thanks for the suggestion. I think providing a wrapper function for toLowerCase() and adding more locale flexibility is a good idea, but currently we only want to explicitly use English as our locale to avoid case conversion problems, and I think this change solves that problem. I'll come back to that feature later when it becomes necessary.