fix: consistency on treating invalid strings

[muktihari]

In proto.Unmarshal, the SDK will remove any UTF-8 Null-Terminated String (0x00) from the string or []string value. To make consistency on how we should treat strings, all functions or methods that returning strings should return strings without 0x00, and all functions or methods that validate invalid strings should check for both empty string and 0x00, invalid if the result of this is true -> value == "" || value == "\x00".

Value marshaler will check whether a string is end with 0x00, if it does not end with 0x00 will automatically add it. If the users intentionally add multiple 0x00 it should allow it. In cases where message definitions differ only slightly, such as when creating FieldDescription messages with minor variations in the FieldName value (type: []string), it make sense to add 0x00 padding so that the length of the strings (the byte size of the FieldName value) become the same, this will reduce message definition interleave as it can be treated using the same message definition.

[codecov-commenter]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 100.00%. Comparing base (564b029) to head (fa46952).

Additional details and impacted files

@@            Coverage Diff            @@
##            master      #175   +/-   ##
=========================================
  Coverage   100.00%   100.00%           
=========================================
  Files           42        42           
  Lines         3471      3471           
=========================================
  Hits          3471      3471           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.