937907d721b8e4e9628b6d13bb59423e4406a5f1,src/python/pants/backend/python/tasks/pytest_run.py,PytestRun,_do_run_tests,#PytestRun#Any#Any#,348

Before Change

      args.extend(test_args)
      args.extend(sources)

      try:
        // The pytest runner we use accepts a --pdb argument that will launch an interactive pdb
        // session on any test failure.  In order to support use of this pass-through flag we must
        // turn off stdin buffering that otherwise occurs.  Setting the PYTHONUNBUFFERED env var to
        // any value achieves this in python2.7.  We"ll need a different solution when we support
        // running pants under CPython 3 which does not unbuffer stdin using this trick.
        env = {
          "PYTHONUNBUFFERED": "1",
        }
        // If profiling a test run, this will enable profiling on the test code itself.
        // Note that tests may run in a different cwd, so it"s best to set PANTS_PROFILE
        // to an absolute path to make it easy to find the subprocess profiles later.
        if "PANTS_PROFILE" in os.environ:
          env["PEX_PROFILE"] = "{0}.subprocess.{1:.6f}".format(os.environ["PANTS_PROFILE"],
                                                               time.time())
        with environment_as(**env):
          rc = self._pex_run(pex, workunit, args=args, setsid=True)
          return PythonTestResult.rc(rc)
      except Exception:
        self.context.log.error("Failed to run test!")
        self.context.log.info(traceback.format_exc())
        return PythonTestResult.exception()

  def _pex_run(self, pex, workunit, args, setsid=False):
    return pex.run(args=args, setsid=setsid,
                   stdout=workunit.output("stdout"), stderr=workunit.output("stderr"))

After Change

      // The user might have already specified the resultlog option. In such case, reuse it.
      resultlogs = [arg.split("=", 1)[-1] for arg in args if arg.startswith("--resultlog=")]

      if resultlogs:
        return run_and_analyze(resultlogs[-1])
      else:
        with temporary_file_path() as resultlog_path:
          args.append("--resultlog={0}".format(resultlog_path))
          return run_and_analyze(resultlog_path)

  def _pex_run(self, pex, workunit, args, setsid=False):
    return pex.run(args=args, setsid=setsid,
                   stdout=workunit.output("stdout"), stderr=workunit.output("stderr"))

Instances