Fix[ntcd_machine]: data corruption and out-of-bounds write

[678098]

• Fix data corruption and out-of-bounds write when receiving into a bsl::string in ntcd::Packet::dequeueData
• Added a test function to verify receiving data into a string with spare capacity.

[678098]

Fails in main:

 684/1013 Test  #684: ntc ntcd::MachineTest::verifyReceiveIntoString ......................................Subprocess aborted***Exception:   0.13 sec
[ 2026-09-18T20:58:51.514384Z ][ FATAL ][ 7F939D8A6780 ][ ntcd_machine.t.cpp:4433 ][ BSLS.LOG ]:
Assertion failed: remoteData == CLIENT_DATA
Found:     (remoteData)
Expected: HELLOWORLD (CLIENT_DATA) 
99% tests passed, 1 tests failed out of 684
Total Test time (real) =  67.36 sec

https://github.com/bloomberg/ntf-core/actions/runs/35394079728/job/105758779335?pr=417

[678098]

The old code copied the bytes into the string first, then called resize() to make the string long enough. But growing a string fills the new space with zeros — so it immediately erased the bytes it had just copied. The caller got the right byte count but all zeros, and the packet had already been discarded, so the data was gone for good.

So the fix is to resize the string first, then copy into it.

[678098]

There was also a problem here. When the string had no spare room at all, the old code tried to make room with resize(d_data.length()) - but that might shrink the string rather than enlarging its buffer, so the copy ran off the end of the allocated memory.

For example, with
data->capacity() == data->size() == 4096
position == 4096
d_data.length() == 128

This leads to data->resize(128) and trying to write past the buffer at data-data() + 4096