v4.0.0

[dhershman1]

Should address #30

BREAKING CHANGES

• Converted from tap-out to tap-parser
  • This change means that the xml generated is now "Flat" (see below)
  • This removes the error count as there isn't an event for this anymore
  • Skips are handled WAY better and more naturally with TAP
  • This should flow better with proper tap output
• If a file extension is set in the command, it will be used for the generated file, if not .xml will be used
• Error counter removed with new parser
• The testcase name no longer has test number, the test number is now provided in the id attribute of testcase

New

• Todo support

Improved

• Optimized performance from both parsing and serializing
• Data is built more dynamically, instead of relying on certain keys to exist
• Better JUnit formatting
• Moved from xmlbuilder to xmlbuilder2 (In Progress)

Flat XML

Now <testsuits> containes only a singular <testsuite> element, within that all of the <testcase> elements now live. Here's an example:

v3 tap-junit output:

<testsuites tests="4" name="suite-name" failures="0" errors="0">
  <testsuite tests="3" failures="0" errors="0" name="1 === 1">
    <testcase name="#1 test is equal"/>
    <testcase name="#2 test skip extra # SKIP">
      <skipped/>
    </testcase>
    <testcase name="#3 should not be equal"/>
  </testsuite>
  <testsuite tests="1" failures="0" errors="0" name="2 === 2">
    <testcase name="#4 should be equal"/>
  </testsuite>
  <testsuite tests="0" failures="0" errors="0" name="SKIP skipped test"/>
</testsuites>

v4 tap-junit output:

<testsuites tests="4" name="Tap-Junit" failures="2">
  <testsuite tests="4" failures="2" skipped="1">
    <testcase id="1" name="test is equal"/>
    <testcase id="2" name="test skip extra # SKIP">
      <skipped/>
    </testcase>
    <testcase name="#3 should not be equal"/>
    <testcase id="3" name="should not be equal"/>
    <testcase id="4" name="should be equal"/>
  </testsuite>
</testsuites>

[dhershman1]

Still working out the kinks on some BAT file Comment based messages, as mentioned in #23

[dhershman1]

Also re writing how the serialization works in v4.0.0 with xmlbuilder2