Fix null-termination bug w/ string->int cast

[jeremiah-corrado]

During AoC activities, we identified a bug involving a missing null-terminator in strings created by IO.readLine. The bug caused leftover bytes from previous readLine operations to show up in casts from string to int.

This PR modifies IO.readStringBytesData to properly terminate strings modified by readLine(ref s: string) and readLine(ref b: bytes)

It also ensures that string buffers are properly terminated before casts to integral types, by modifying _cleanupForNumericCast.

An additional test from AoC activities was added to confirm the validity of the fix.

• passing paratest

[mppf]

During AoC activities, we identified a bug involving a missing null-terminator in strings created by IO.readLine.

I was expecting that this PR would adjust IO.readLine to add a null terminator there somewhere but it does not. Why not?

[jeremiah-corrado]

Offline, @vasslitvinov suggested modifying _cleanupForNumericCast and it appears to have worked, so I decided to go with that.

I suppose adding a null terminator to readLine would be more correct tho? Maybe to protect against similar bugs on other casts?

[mppf]

Recommend if x.buffLen > 0 here around this, unless you know that x cannot be the empty string at this point.

[bradcray]

Thanks so much for picking this up quickly @jeremiah-corrado! Wish we'd discovered it prior to AoC 2022!

[jeremiah-corrado]

Sure thing, at least we got it early!

[bradcray]

True! What if it hadn't been until day 10! :O