bq. How can it not be a true error?!
If you @link to a package-private class in a javadoc, it DOES NOT create a
link, but rather just writes the plain class's name. In fact, the result of
{@link PackagePrivateClass} is the same as {@code PackagePrivateClass}. At
least, I ran 'ant javadocs' and viewed the HTML, and the result was like
the latter.
Uwe, I'm not talking about the method's signature which took a
package-private class. That DOES create a broken link, and it's good that
documentation-lint found it, because I agree with you that it's wrong.
I agree there's not much value to link to package-private classes in
javadocs (I'm not arguing about it). But I don't know if documentation-lint
failed on the method signature's broken link, or the {@link} in the
javadocs. Let's hope that it will catch both.
Shai
On Mon, Jan 7, 2013 at 2:23 PM, Robert Muir <[email protected]> wrote:
> On Mon, Jan 7, 2013 at 1:38 AM, Shai Erera <[email protected]> wrote:
> > I don't like this failure. CategoryPathUtils contains some {@link
> > CharBlockArray} mentions, while the latter is package-private. When I run
> > 'ant javadocs' the generated javadocs just include the text
> CharBlockArray
> > with no link. I don't think that's a true error?
>
> How can it not be a true error?!
>
> This script only checks for broken links: because you @link to
> package-private shit, then it resutls with html with broken links.
>
> its a true error by definition since your html contains broken links:
> its just that simple. There is no need to discuss whether its "true"
> or "false" error.
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [email protected]
> For additional commands, e-mail: [email protected]
>
>
