rebis-dev: Remaining issues unexpectedly not addressed

[triska]

I collect here issues that the heap-inlined and simpler string representation of rebis-dev is expected to address, but currently does not address:

• looking at a string leaks memory Partial string leak #1714 EDIT: Resolved by rebis-dev.
• character e sometimes not handled correctly The irrational number e is not always understood #2455
• 0 bytes not handled correctly 0-bytes are sometimes ignored #2554

The fact that such elementary issues are still present even in rebis-dev indicates the potential to further simplify the handling of strings in the engine so that such edge-cases do not arise.

[UWN]

#16

[triska]

#2580