Since SSAPhiInstructions are never visited by NullPointerTransferFunctionProvider.TransferFunctionSSAVisitor,
we now respect phi instructions present at a given block by providing additional NodeTransferFunctions, improving precision.
Formerly, meets would lead to incorrect results due to incorrect initialization of initial data flow facts.
These are now properly initialized, interpreting
"State.BOTH" to mean: both "null" and "non-null" are possible values for the given variable, and
"State.UNKNOWN" to be the absurd assertion.
The initial fact at the entry block assumes variables to be BOTH, other blocks are initialy assumed unreachable and hence their variables to be UNKNOWN.
The setting should comply with the comment. Plus,
turning it on seems to lead to some unsoundness because
exception points-to sets become empty but should not be
Access is provided via corresponding methods in FieldImpl, ShrikeCTMethod and ShrikeClass.
Since we do not currently have implementation of these methods for front-ends other than Shrike, these new methods are not yet made available in the corresponding interfaces.
Some of these might have proper DTDs or XML Schema definitions
floating around somewhere that we could use. Presumably many do not.
Rather than hand-craft such definitions myself, I'm just giving each a
minimal stub DOCTYPE declaration. That's enough to satisfy Eclipse's
XML validator, which otherwise complains that these files lack grammar
constraints.
I think the "target/p2artifacts.xml" and "target/p2content.xml" files
are generated by Maven. They are well-formed XML but Eclipse's XML
validator legitimately warns that they lack grammar constraints.
Since we're not maintaining the tool that creates these files, we are
not in a position to do anything about that. Therefore, we may as
well exclude these from validation entirely. That way we can
more-clearly recognize warnings that we *can* do something about.
Ant "build.xml" files don't have a standard DTD or XML Schema; the
contents are simply too flexible for that. But we can at least
give each a stub DOCTYPE declaration. That's enough to satisfy
Eclipse's XML validator, which otherwise complains that these files
lack grammar constraints.
Most of the invalid HTML arose from bare "<" and ">" characters.
These should be escaped as "<" and ">" when not intended to
introduce HTML tags. When you have many such characters close
together, "{@literal ...}" is a nice, readable alternative that
automatically escapes its contents. If the text in question is
intended to be a code fragment, then "{@code ...}" is appropriate:
this is essentially equivalent to "<code>{@literal ...}</code>".
There were a few other HTML violations too, but none common enough to
be worth detailing here.
The contents of @author go straight into HTML, just like most other
Javadoc material. So if you want to have a "<foo@bar.com>" e-mail
address as part of the author information, the angle brackets must be
escaped. Here I've opted to do that using "{@code <foo@bar.com>}",
which has some additional styling effects that seem appropriate for
e-mail addresses. We could also have used "<foo@bar.com>" for
escaping without code styling.
Eclipse validation warns about invalid HTML content in all
Maven-generated "target/site/dependency-convergence.html" files. The
warnings are legitimate: these HTML files are indeed invalid.
However, we don't maintain the tool that generates these files, so we
are not in a position to fix them. Better, therefore, to suppress
these warnings so that we can notice and fix other problems over which
we do have control.