MCOL-4580 extent elimination for dictionary-based text/varchar types

The idea is relatively simple - encode prefixes of collated strings as
integers and use them to compute extents' ranges. Then we can eliminate
extents with strings.

The actual patch does have all the code there but miss one important
step: we do not keep collation index, we keep charset index. Because of
this, some of the tests in the bugfix suite fail and thus main
functionality is turned off.

The reason of this patch to be put into PR at all is that it contains
changes that made CHAR/VARCHAR columns unsigned. This change is needed in
vectorization work.

dbcon/execplan/simplecolumn.cpp

@@ -565,6 +565,16 @@ void SimpleColumn::evaluate(Row& row, bool& isNull)
       else
         fResult.intVal = atoll((char*)&fResult.origIntVal);
 
+      // MCOL-4580 - related, probably can be marked with XXX.
+      // This does not fail in any tests, but it is considered wrong.
+      // The reasonin behind that is that we changed signedness if characters to unsigned
+      // and it might be a case with short strings that they were copied as is using
+      // uint64ToStr encoding into int64_t values. So, potentially, unsuspecting code
+      // may use getUintVal instead of getIntVal to process short char column, getting
+      // unitialized value and give floating behavior.
+      // None of our tests failed, though.
+      fResult.uintVal = fResult.intVal;
+
       break;
     }