Remove extra byte in realm_string_t conversion

[nirinchev]

realm_string_t should never have been 0-terminated because it knows its own length. In fact, adding an extra empty byte will result in \u0000 being appended at the end of the string. I don't know how to test this since the only place it's used currently is for the password, but I've verified it manually by looking at the http request body.

[coveralls]

Pull Request Test Coverage Report for Build 3107262326

• 0 of 2 (100.0%) changed or added relevant lines in 1 file are covered.
• No unchanged relevant lines lost coverage.
• Overall coverage increased (+5.5%) to 93.128%

---

Totals
Change from base Build 3097932778:	5.5%
Covered Lines:	393
Relevant Lines:	422

---

💛 - Coveralls

[coveralls]

Pull Request Test Coverage Report for Build 3107262326

• 2 of 2 (100.0%) changed or added relevant lines in 1 file are covered.
• No unchanged relevant lines lost coverage.
• Overall coverage increased (+5.5%) to 93.128%

---

Totals
Change from base Build 3097932778:	5.5%
Covered Lines:	393
Relevant Lines:	422

---

💛 - Coveralls

[nielsenko]

List<int>.toCharPtr ends up calling

Pointer<Uint8> toUint8Ptr(Allocator allocator) {
    final nativeSize = length + 1;
    final result = allocator<Uint8>(nativeSize);
    final Uint8List native = result.asTypedList(nativeSize);
    native.setAll(0, this); // copy
    native.last = 0; // zero terminate
    return result.cast();
  }

i guess that need to be re-considered as well?

[nirinchev]

List<int>.toCharPtr is mapping to char* which needs null termination otherwise we can't infer the length.

[blagoev]

good catch.