Merge putDtype and putType; remove duplicated code

[Noremos]

This is also a preparation for the JSON Type implementation

[asfernandes]

If you not choose better names for methods you created to not duplicate code, I prefer to have the code duplicated.
I understand nothing what the refactored methods do by their names.

[Noremos]

If you not choose better names for methods you created to not duplicate code, I prefer to have the code duplicated. I understand nothing what the refactored methods do by their names.

I tried to follow the style of the original method names. putType and putDtype are not the most transparent names either. So it would be nice if you could suggest alternative names for my methods

[asfernandes]

What means THasTableName?

[Noremos]

This indicates the presence of typeOfTable.* fields in the TypeClause& type. That is, we use blr_column_name (i.e., the table name with typeOfTable) or blr_domain_name (with typeOfName).

[dyemanov]

I'd question the need of the T prefix, as it usually means some type used as the template argument, not the variable. Simple HasTableName or maybe IsTypeOfTable is IMHO better readable.

[Noremos]

It would be nice to have some explicit coding style for such cases. The existing one suggests using a single letter, which is not the clearest name possible in this case. Personally, I use the T prefix to make the template argument easier to recognize in the code, especially with if constexpr statements

[asfernandes]

I'd question the need of the T prefix, as it usually means some type used as the template argument, not the variable. Simple HasTableName or maybe IsTypeOfTable is IMHO better readable.

Yes.

[asfernandes]

Why create objects in heap and leak them instead of use local objects and let RAII destruct them?

[Noremos]

Oops, I forgot to fix my raw temporary implementation

[Noremos]

Oh, this is why I am leaking memory:

protected:
	// DsqlCompilerScratch should never be destroyed using delete.
	// It dies together with it's pool.
	~DsqlCompilerScratch()
	{
	}

I can create a derived class and use it, but I'm not sure if it's safe to delete it, since the DsqlCompilerScratch has a strict prohibition

[asfernandes]

I'd question the need of the T prefix, as it usually means some type used as the template argument, not the variable. Simple HasTableName or maybe IsTypeOfTable is IMHO better readable.

Yes.