On 09/19/2013 10:46 AM, Jonathan Gibbons wrote:
-- does the change affect the generated docs.
... and the answer is "yes". :-)
The order of some parts of the generated docs differs before/after the
patch. This is an issue which needs to be fixed.
At first, I thought it was due to the use
On 09/19/2013 10:46 AM, Jonathan Gibbons wrote:
There are two criteria for a change like this:
-- the obvious one -- do all the javadoc regression tests pass. These
are the tests
langtools/test/com/sun/javadoc
langtools/test/tools/javadoc
Two regression tests fail when the patch is
On 09/19/2013 11:02 AM, Jonathan Gibbons wrote:
On 09/19/2013 10:00 AM, Michael Simacek wrote:
Hi,
I thought about improving performance of the default doclet
implementation a bit.
According to profiler results, most of the CPU time is spent in
constructing the member map in VisibleMemberMap.
On 09/19/2013 10:00 AM, Michael Simacek wrote:
Hi,
I thought about improving performance of the default doclet implementation a
bit.
According to profiler results, most of the CPU time is spent in constructing
the member map in VisibleMemberMap.java.
So I've rewritten part of the VisbleMemberM
On 09/19/2013 10:00 AM, Michael Simacek wrote:
Is there any chance of this patch (attached) being accepted into OpenJDK?
I've never made any contribution to OpenJDK before, so I would like to ask for
code review and guidance through the contribution process.
Michael
For reference, the general
On 09/19/2013 10:00 AM, Michael Simacek wrote:
Hi,
I thought about improving performance of the default doclet implementation a
bit.
According to profiler results, most of the CPU time is spent in constructing
the member map in VisibleMemberMap.java.
So I've rewritten part of the VisbleMemberM
Hi,
I thought about improving performance of the default doclet implementation a
bit.
According to profiler results, most of the CPU time is spent in constructing
the member map in VisibleMemberMap.java.
So I've rewritten part of the VisbleMemberMap and reduced the javadoc run time
to 60-70% of
