Allow constructing a microsoft.sql.DateTimeOffset instance from a java.time.OffsetDateTime value

[arashi01]

Operates similarly to existing constructor however avoids additional expense of creating Timestamp instances and increased timezone related error potential, by directly converting java.time.OffsetDateTime instances.

Reverse of existing DateTimeOffset.getOffsetDateTime() method.

Rounds precision to closest 100ns matching existing semantics of Timestamp based constructor. Values where nanos are within 50ns of the next second are rounded upwards to the next second; as per current semantics.

Supercedes #2339

[tkyc]

FYI, I did a force push to fix the pipeline issue.

[arashi01]

Also just noticed the second unit test (valueOfOffsetDateTimeRounding()) can cause a false test failure if the test is run at any time with a 59 seconds value, since it uses OffsetDateTime.now() for it's seed data.

Will fix it now. Rather than give the test value a wholly static value I think it's probably better to just fix the "seconds" value, similar to the "nanos" value.

[arashi01]

Fixed the test. Fixed the test value to OffsetDateTime.now().withSecond(58).

[arashi01]

@tkyc Just a heads up. The current HEAD doesn't include the updated test. See https://github.com/microsoft/mssql-jdbc/compare/ae407c4ff5be6062fc63536c90edb532ffb45a0b..859c32db2a2d9822595d9d8b73849ae2962270b7

[tkyc]

@arashi01

My mistake, could you push again? I had to do another force push which overwrote your last commit.

[lilgreenbird]

the reason the force-push was done by @tkyc was to get rid of the previous commits which was tied to the old versions of the pipelines for some reason and we wanted to remove CI-mssql-jdbc pipeline from the PR validation check. But now it seems with your latest force-push this came right back :(
Well I guess we can just ignore the failures on this particular pipeline.

[lilgreenbird]

pls use formatter to format all files

[arashi01]

pls use formatter to format all files

Done. Although used Intellij's Eclipse code style import support to use the mssql-jdbc_formatter.xml. If theres still an issue I'll install Eclipse and rerun.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 50.20%. Comparing base (cfb018a) to head (9395bcc).

❗ Current head 9395bcc differs from pull request most recent head 46bbb6d. Consider uploading reports for the commit 46bbb6d to get more accurate results

Additional details and impacted files

@@              Coverage Diff              @@
##               main    #2340       +/-   ##
=============================================
- Coverage     73.60%   50.20%   -23.41%     
+ Complexity     6082     3793     -2289     
=============================================
  Files           142      143        +1     
  Lines         33123    33149       +26     
  Branches       5623     5630        +7     
=============================================
- Hits          24380    16641     -7739     
- Misses         6234    14122     +7888     
+ Partials       2509     2386      -123     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[barryw-mssql]

Looks fine