Section 8.1.2 defines stereotypes CORBAAnonymousSequence and CORBAAnonymousArray for mapping sequence/array attributes of struct/union/valuetype/interface without a dedicated typedef.
However, an equivalent CORBAAnonymousString stereotype is not defined.
Thus it is not obvious how to map e.g.

  struct struct_with_anon_bounded_str {
    string<32> anon_bounded_str;
  };

  union union_with_anon_bounded_str switch (boolean) {
    case TRUE:
      string<32> anon_bounded_str;
  };

  valuetype valuetype_with_anon_bounded_str {
    public string<32> anon_bounded_str;
  };

  interface interface_with_anon_bounded_str {
    attribute string<32> anon_bounded_str;
  };

The UML Profile for CORBA and CCM on page 47 states:

The extended UML metamodel contains an abstract stereotype <<CORBATemplate>>, which is a generalization of <<CORBAString>>, <<CORBAWstring>>, and <<CORBASequence>> stereotypes.

The part about «CORBAString» and «CORBAWstring» inheriting from «CORBATemplate» does not fit with figure 8.7 on page 42 which shows that those stereotypes inherit from «CORBABounded».

According to the class diagram, the sentence should be:

The extended UML metamodel contains an abstract stereotype <<CORBATemplate>>, which is a generalization of <<CORBAArray>> and <<CORBASequence>> stereotypes.

Chapter 9 should be clearly marked as non-normative. Noted during smsc review. Thus, formal document number not available

P. 51 . The CORBAUses stereotype has a property "multiple" of type PrimitiveKind P. 53 . The CORBAUses stereotype has a property "isMultiple" of type boolean. This looks like a naming and type inconsistency between diagram and tabular representation of the stereotype. If this is the case and isMultiple is really typed boolean, PrimitiveKind becomes useless (I did not find any other use of PrimitiveKind in the profile). If not (I mean PrimitiveKind is the wanted type), it may be better to type the "multiple" property with the "CORBAPrimitive" stereotype instead of creating and using an enumeration like PrimitiveKind.

Litteral of the enumeration PrimitiveKind are prefixed with "CORBA" or "Corba", instead of "CORBA" (upper case) for all litterals

For the stereotype designating IDL "typedef", the UML Profile for CORBA
v1.0 (02-04-01) uses the capitalization "CORBATypedef".
The UML Profile for CORBA and CCM v1.0 (08-04-07) sometimes uses the
same capitalization but also uses the capitalization "CORBATypeDef"
(notice the capital "D".)
For compatibility with the UML Profile for CORBA, I suggest replacing
all occurrences of "CORBATypeDef" by "CORBATypedef".