<html>
  <head>
    <meta content="text/html; charset=ISO-8859-1"
      http-equiv="Content-Type">
  </head>
  <body bgcolor="#FFFFFF" text="#000000">
    <div class="moz-cite-prefix">Hi,<br>
      <br>
      On 03/06/2015 12:37 PM, Laurent Bourgès wrote:<br>
    </div>
    <blockquote
cite="mid:CAKjRUT79+xn02-Fi68stcfD_THodT9jzghKOyTjx3nTQs8FoRg@mail.gmail.com"
      type="cite">
      <div dir="ltr">
        <div>Phil,<br>
          <br>
        </div>
        Thanks for your feedback !<br>
        <div>
          <div>
            <div class="gmail_extra"><br>
              <div class="gmail_quote">2015-03-06 20:06 GMT+01:00 Phil
                Race <span dir="ltr"><<a moz-do-not-send="true"
                    href="mailto:philip.race@oracle.com" target="_blank">philip.race@oracle.com</a>></span>:<br>
                <blockquote class="gmail_quote" style="margin:0px 0px
                  0px 0.8ex;border-left:1px solid
                  rgb(204,204,204);padding-left:1ex">Hi,<br>
                  <br>
                  you placed the test in the java.awt.geom package.<br>
                  <br>
                    25 package java.awt.geom;<br>
                  <br>
                  and are accessing internals of that package.<br>
                  <br>
                  In jigsaw/modular mode that won't even compile.<br>
                </blockquote>
                <div><br>
                </div>
                <div>Ok it is annoying: <br>
                  as all Path2D fields are package protected, I designed
                  the test using direct access to any fields ...<br>
                  <br>
                </div>
                <blockquote class="gmail_quote" style="margin:0px 0px
                  0px 0.8ex;border-left:1px solid
                  rgb(204,204,204);padding-left:1ex">
                  <br>
                  So the test should go in the anonymous package and
                  avoid accessing internals.<br>
                  It should be possible to use just public API  to
                  verify the arrays  of a shape<br>
                  being cloned are trimmed .<br>
                </blockquote>
                <div><br>
                </div>
                <div>No, it is not possible to use Shape API to access
                  arrays nor fields (numTypes ...): <br>
                  only getPathIterator() could give me data but it won't
                  tell me if the underlying arrays or fields are
                  correct.<br>
                </div>
              </div>
            </div>
          </div>
        </div>
      </div>
    </blockquote>
    <br>
    That is true .. <br>
    <br>
    Well, if you need it to be in java.awt.geom, I think even today
    you'll find it won't work<br>
    unless you jump through some jtreg hoops to install it on the
    bootclasspath.<br>
    I think its something like "@run main/othervm -Xbootclasspath/a:. "<br>
    And later in the modular JDK it will need to be modified again.<br>
    <br>
    I'd say either update the test to work with jtreg today - and test
    it to be sure,<br>
    or provide the test without an @test tag, or with an @ignore tag, so
    people can<br>
    still manually verify it but the harness won't run it.<br>
    <br>
    <blockquote
cite="mid:CAKjRUT79+xn02-Fi68stcfD_THodT9jzghKOyTjx3nTQs8FoRg@mail.gmail.com"
      type="cite">
      <div dir="ltr">
        <div>
          <div>
            <div class="gmail_extra">
              <div class="gmail_quote">
                <div><br>
                </div>
                <div>Maybe I could use introspection to
                  getDeclaredField(name) and setAccessible(true) to get
                  internal data.<br>
                </div>
              </div>
            </div>
          </div>
        </div>
      </div>
    </blockquote>
    <br>
    That won't work either. So maybe this is a "noreg-hard" or
    "noreg-cleanup" bug.<br>
    We add those labels to the JBS/JIRA bug when something isn't
    testable.<br>
    <br>
    <br>
    <blockquote
cite="mid:CAKjRUT79+xn02-Fi68stcfD_THodT9jzghKOyTjx3nTQs8FoRg@mail.gmail.com"
      type="cite">
      <div dir="ltr">
        <div>
          <div>
            <div class="gmail_extra">
              <div class="gmail_quote">
                <div><br>
                </div>
                <div>Any idea or utility class I could use </div>
              </div>
            </div>
          </div>
        </div>
      </div>
    </blockquote>
    <br>
    <blockquote
cite="mid:CAKjRUT79+xn02-Fi68stcfD_THodT9jzghKOyTjx3nTQs8FoRg@mail.gmail.com"
      type="cite">
      <div dir="ltr">
        <div>
          <div>
            <div class="gmail_extra">
              <div class="gmail_quote">
                <blockquote class="gmail_quote" style="margin:0px 0px
                  0px 0.8ex;border-left:1px solid
                  rgb(204,204,204);padding-left:1ex">
                  <br>
                  Why is it necessary to explicitly add the call to
                  super(); ?<br>
                  <br>
                  223             super();<br>
                </blockquote>
                <div><br>
                </div>
                <div>I agree it is not necessary but it explicitely says
                  that I use the empty constructor:<br>
                  <span style="font-family:monospace,monospace"><br>
                        /**<br>
                         * Constructs a new empty {@code Path2D} object.<br>
                         * It is assumed that the package sibling
                    subclass that is<br>
                         * defaulting to this constructor will fill in
                    all values.<br>
                         *<br>
                         * @since 1.6<br>
                         */<br>
                        /* private protected */<br>
                        Path2D() {<br>
                        }</span><br>
                </div>
              </div>
            </div>
          </div>
        </div>
      </div>
    </blockquote>
    <br>
    If we all did this, all of the time, there'd be a lot of extra lines
    in the code, that the compiler<br>
    would fill in for us anyway.<br>
    <br>
    -phil.<br>
    <br>
    <blockquote
cite="mid:CAKjRUT79+xn02-Fi68stcfD_THodT9jzghKOyTjx3nTQs8FoRg@mail.gmail.com"
      type="cite">
      <div dir="ltr">
        <div>
          <div>
            <div class="gmail_extra">
              <div class="gmail_quote">
                <div> <br>
                </div>
                <div>Laurent<br>
                </div>
                <br>
              </div>
            </div>
          </div>
        </div>
      </div>
    </blockquote>
    <br>
  </body>
</html>