Bug 175721

[ BUILD # : 200910281401 ]
[ JDK VERSION : 1.6.* ]

The code coverage report is only meaningful if the weaknesses in the coverage can be extracted.  This isn't easy with
the current code coverage report however because the table only sorts its columns alphabetically.  What is required is
to be able to sort table rows by measures of significance, such as percentage (treated as a number so that 10% and 100%
are correctly ordered) and number of lines *not* covered.

It is interesting that whilst the report details the number of lines covered, it is actually the number of lines not
covered that is actually the more interesting measure because that indicates what is left to do!

I suggest columns for %age coverage, lines not covered and total number of lines, each of which should be ordered
numerically.  The report should order using on of these measures by default.  My preference would be to use the number
of lines not covered.