Pyodbc

Fixes in this release:

* Fix incorrect size calcation for Unicode characters (654)
* Fix binding of 32-bit ints on 64-bit Linux (611)
* Fixes type mismatch for null/None with table value parameters. 596
* Fixes and cleanup of unit tests for MS Access (606), MySQL (630 632) and new tests (637 645)

Many thanks to everyone who provided these fixes. Merry Christmas.

* Use int instead of bigint when possible (based on size of data) to work with drivers that don't support bigint at all. Thanks gordthompson and v-makouz
* Support SQL Server datetime2 precision. Previously more data was passed than the column precision causing an error. Thanks PrinsEdje80, v-makouz, and v-chojas
* Make Informix unit tests work again. Thanks richard-reece
* Correct encoding error on big-endian machines for connection errors. Default to native UTF16 instead of UTF16-LE. Thanks kadler
* Fix MySQL unit tests. Thanks gordthompson

This is mostly a bug fix release.

* Issue 506 uncovered a potentially serious error where Unicode strings may not get a NULL terminator when being converted.
* Issue 504 was a double decref in the error return path of executemany. Thanks to v-chojas for this fix.

Additionally, thanks gordthompson for both adding `get_output_converter` (496) and for updating the SQL Server tests to properly bypass issues not supported with the FreeTDS driver.

Due to 506, I recommend upgrading from 4.0.25 where the issue was introduced.

Many thanks to the community for this release:

* Thanks to v-makouz and v-chojas , support for [SQL Server Table-Valued Paramaters](https://docs.microsoft.com/en-us/dotnet/framework/data/adonet/sql/table-valued-parameters) has arrived. See `test_tvp` in the SQL Server unit tests.
* Support for IBM DB/2 DECFLOAT type (472) by c-dy
* Better SQL Server and FreeTDS tests with known driver issues (469) by gordthompson.
* Fix "function sequence error" with fast executemany (450) by v-makouz.

This is a bug fix release with fixes from lots of contributors:

* Column size fixed when binding Unicode parameters - noticed when using characters more than 2-bytes wide in UTF-16. Thanks Maxim Kouznetsov.
* v-chojas fixed use of Unicode connection strings in Python 2.7. Also removed arbitrary limit of connection string lengths.
* New SQL_CONVERT_X constants added by Jordan Bolton.
* The setinputsizes method has been overloaded by skeetor and can now accept data types and sizes, not just sizes.
* v-chojas added support for non-integer-valued connection attributes
* I cleaned up SQLite unit tests which were out of date and upgraded deprecated unittest methods from all tests. v-chojas made a fix to SQL Server test_drivers test and ensured they run against Python 2.6. (Though, it might be time to officially drop support for older Python versions?)

Some changes in 4.0.22 to integrate the old parameter binding and the new executemany array binding have not handled Unicode in all cases. The integration has been reverted to ensure pyodbc is as stable as possible.

If you had issues with 4.0.22, please try 4.0.23.