AVRO-3486: Set namespace with protocol fullName

[opwvhk]

When parsing or creating a protocol, use both the name and namespace
given to determine the actual name, namespace and fullName.

Reason: the spec says "The name and namespace qualification rules
defined for schema objects apply to protocols as well."

Make sure you have checked all steps below.

Jira

• My PR addresses the following Avro Jira issues and references them in the PR title. For example, "AVRO-1234: My Avro PR"
  • https://issues.apache.org/jira/browse/AVRO-3486
  • In case you are adding a dependency, check if the license complies with the ASF 3rd Party License Policy.

Tests

• My PR adds updates the following unit tests OR does not need testing for this extremely good reason:
  org.apache.avro.compiler.idl.TestIdl#runTests() for import.avdl and nestedimport.avdl by updating bar.avpr

Commits

• My commits all reference Jira issues in their subject lines. In addition, my commits follow the guidelines from "How to write a good git commit message":
  1. Subject is separated from body by a blank line
  2. Subject is limited to 50 characters (not including Jira issue reference)
  3. Subject does not end with a period
  4. Subject uses the imperative mood ("add", not "adding")
  5. Body wraps at 72 characters
  6. Body explains "what" and "why", not "how"

Documentation

• In case of new functionality, my PR adds documentation that describes how to use it.
  • All the public functions and the classes in the PR contain Javadoc that explain what it does

[Christophe-LeSaec]

Good code,
A proposition for code for very specific corner case;
I also suggest to add unit test directly on TestProtocol class (it's more direct than to pass by avro file).

[Christophe-LeSaec]

What if namespace is "java" and name is "lang.String"
Here, a suggestion :

        StringBuilder namespaceBuilder = new StringBuilder((namespace != null) ? namespace : "");
        if (namespaceBuilder.length() > 0 && namespaceBuilder.charAt(namespaceBuilder.length() - 1) != '.') {
            namespaceBuilder.append('.');
        }
        namespaceBuilder.append(name.substring(0, lastDot));
       this.namespace = namespaceBuilder.toString();

[KalleOlaviNiemitalo]

If a protocol has namespace "java" and name "lang.String", then its fullname should be "lang.String" for sure (not "java.lang.String"). But it is not clear from the spec whether types defined in this protocol should then inherit the "java" namespace (as declared) or the "lang" namespace (based on the fullname of the protocol). I don't see any justification for making nested types inherit the "java.lang" namespace.

Relevant parts of the spec:

avro/doc/src/content/xdocs/spec.xml

Lines 296 to 316 in 4e1fefc

	<p>In record, enum and fixed definitions, the fullname is
	determined in one of the following ways:</p>
	<ul>
	<li>A name and namespace are both specified. For example,
	one might use <code>"name": "X", "namespace":
	"org.foo"</code> to indicate the
	fullname <code>org.foo.X</code>.</li>
	<li>A fullname is specified. If the name specified contains
	a dot, then it is assumed to be a fullname, and any
	namespace also specified is ignored. For example,
	use <code>"name": "org.foo.X"</code> to indicate the
	fullname <code>org.foo.X</code>.</li>
	<li>A name only is specified, i.e., a name that contains no
	dots. In this case the namespace is taken from the most
	tightly enclosing schema or protocol. For example,
	if <code>"name": "X"</code> is specified, and this occurs
	within a field of the record definition
	of <code>org.foo.Y</code>, then the fullname
	is <code>org.foo.X</code>. If there is no enclosing
	namespace then the null namespace is used.</li>
	</ul>

avro/doc/src/content/xdocs/spec.xml

Lines 850 to 851 in 4e1fefc

	<p>The name and namespace qualification rules defined for schema objects
	apply to protocols as well.</p>

In particular:

For example, if "name": "X" is specified, and this occurs within a field of the record definition of org.foo.Y, then the fullname is org.foo.X.

If that is intended to apply only when there is a "namespace":"org.foo" property in the record definition or in some outer definition, and not when the record definition has "name":"org.foo.Y", "namespace":"something_else", then the specification should say so.

[opwvhk]

The intent is to use the same namespace the the definition itself, and any enclosed definitions that don't specify one themselves. This is a lot easier to comprehend than using different namespaces.

[clesaec]

LGTM