Fix the calculation for when to create OoM heapdump

Previously, process.memoryUsage() was used to determine when the heap was reaching its limit.
After testing this thoroughly, it seems that "heapTotal" is not representing the memory limit that was set with the "max_old_space_size" flag, but is just showing the total size the heap can grow to at that time. So, "heapTotal" doesn't seem to reflect the limit.

Changed the calculation so the "after" values of the "gc-stats" package are used to determine when the limit is (almost) reached. "usedHeapSize" and "heapSizeLimit" are used, where heapSizeLimit does seem to reflect the value set by "max_old_space_size".

Changed the default threshold to 70%, as this value seems to work better with this new calculation.

Improved error handling

Improved error handling

Added CPU profile functionality

Added CPU profile functionality

Add 'addTimestamp' option

Add addTimestamp option.
Error handling. Updated README.md.

Add limit

Add limit

Made heapsize calculation more precise and stable

1.0.5

BC-5586 Investigate possibilities to detect OutOfMemory errors on pro…

Updated documentation

Change name to path.

Updated documentation and create heapdumps on request

Documentation update, feature update for creating heapdumps on request.
Updated examples.

1.0.1 - First version

Initial version.