Attempt at memory fix

[Thrameos]

Fixes #1242

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 86.95%. Comparing base (7b24a39) to head (c5ffc1d).

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #1263      +/-   ##
==========================================
+ Coverage   86.90%   86.95%   +0.05%     
==========================================
  Files         113      113              
  Lines       10336    10346      +10     
  Branches     4065     4064       -1     
==========================================
+ Hits         8982     8996      +14     
+ Misses        758      757       -1     
+ Partials      596      593       -3     

Files with missing lines	Coverage Δ
native/common/jp_context.cpp	77.89% <100.00%> (+2.33%)	⬆️

... and 3 files with indirect coverage changes

🚀 New features to boost your workflow:

• ❄ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[grauchs]

I compiled and tested this version (git pull https://github.com/Thrameos/jpype.git jniargs) and the crash reported for jpype-1.5.2 does not happen anymore, i.e. JVMstart() works fine and the issue is fixed.

[Thrameos]

@marscher I believe the source of this issue was the incorrect usage of cstr(). To fix it I pushed all of the arguments into a memory block using a copy method, then freed the block at the end. This should be a safer procedure.

I will add a note into the CHANGELOG. But I as I was never able to replicate the issue, I wasn't able to develop a test. I requested a review to make sure I didn't do something silly with the pointer work. If I don't get a review I will likely just included it in the next release as it is a stability issue.