Fix bug when a result set has more than one row of [n]varchar(max) values

[LHCGreg]

If the value in the first row is 'foo', and the value in the second is 'bar', and the third is 'baz', the results returned would be 'foo', 'foobar', and 'foobarbaz'. That is, each row's value gets concatenated to the previous rows.

I'm not familiar enough with the codebase to know if this is a good way to fix it or if truncating the stream in _DefaultChunkedHandler when getting the end value might have any unexpected effects.

https://github.com/denisenkom/pytds/pull/95 must be merged first in order for tests to pass.

Fixes #91

[LHCGreg]

As expected, appveyor tests failed because #95 (Fix bulk inserting [n]varchar(max) values) had not been merged yet at the time the tests ran.

[codecov[bot]]

Codecov Report

Merging #96 into master will increase coverage by <.01%. The diff coverage is 100%.

@@            Coverage Diff             @@
##           master      #96      +/-   ##
==========================================
+ Coverage   94.02%   94.02%   +<.01%     
==========================================
  Files          26       26              
  Lines        7818     7821       +3     
==========================================
+ Hits         7351     7354       +3     
  Misses        467      467

Impacted Files	Coverage Δ
tests/connected_test.py	98.55% <100%> (ø)	:arrow_up:
src/pytds/tds_types.py	92.98% <100%> (+0.01%)	:arrow_up:

---

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 e11fffd...4144483. Read the comment docs.

[LHCGreg]

Rebased on new master after some other fixes were merged, tests pass now.

[denisenkom]

wow, how come nobody noticed this before, thanks!