# Copyright (C) 2010 Google Inc. All rights reserved. # # Redistribution and use in source and binary forms, with or without # modification, are permitted provided that the following conditions are # met: # # * Redistributions of source code must retain the above copyright # notice, this list of conditions and the following disclaimer. # * Redistributions in binary form must reproduce the above # copyright notice, this list of conditions and the following disclaimer # in the documentation and/or other materials provided with the # distribution. # * Neither the name of Google Inc. nor the names of its # contributors may be used to endorse or promote products derived from # this software without specific prior written permission. # # THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS # "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT # LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR # A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT # OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, # SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT # LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, # DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY # THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT # (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE # OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. import cPickle import logging from webkitpy.layout_tests.models import test_expectations _log = logging.getLogger(__name__) def is_reftest_failure(failure_list): failure_types = [type(f) for f in failure_list] return set((FailureReftestMismatch, FailureReftestMismatchDidNotOccur, FailureReftestNoImagesGenerated)).intersection(failure_types) # FIXME: This is backwards. Each TestFailure subclass should know what # test_expectation type it corresponds too. Then this method just # collects them all from the failure list and returns the worst one. def determine_result_type(failure_list): """Takes a set of test_failures and returns which result type best fits the list of failures. "Best fits" means we use the worst type of failure. Returns: one of the test_expectations result types - PASS, FAIL, CRASH, etc.""" if not failure_list or len(failure_list) == 0: return test_expectations.PASS failure_types = [type(f) for f in failure_list] if FailureCrash in failure_types: return test_expectations.CRASH elif FailureTimeout in failure_types: return test_expectations.TIMEOUT elif FailureEarlyExit in failure_types: return test_expectations.SKIP elif (FailureMissingResult in failure_types or FailureMissingImage in failure_types or FailureMissingImageHash in failure_types or FailureMissingAudio in failure_types): return test_expectations.MISSING else: is_text_failure = FailureTextMismatch in failure_types is_image_failure = (FailureImageHashIncorrect in failure_types or FailureImageHashMismatch in failure_types) is_audio_failure = (FailureAudioMismatch in failure_types) if is_text_failure and is_image_failure: return test_expectations.IMAGE_PLUS_TEXT elif is_text_failure: return test_expectations.TEXT elif is_image_failure or is_reftest_failure(failure_list): return test_expectations.IMAGE elif is_audio_failure: return test_expectations.AUDIO else: raise ValueError("unclassifiable set of failures: " + str(failure_types)) class TestFailure(object): """Abstract base class that defines the failure interface.""" @staticmethod def loads(s): """Creates a TestFailure object from the specified string.""" return cPickle.loads(s) def message(self): """Returns a string describing the failure in more detail.""" raise NotImplementedError def __eq__(self, other): return self.__class__.__name__ == other.__class__.__name__ def __ne__(self, other): return self.__class__.__name__ != other.__class__.__name__ def __hash__(self): return hash(self.__class__.__name__) def dumps(self): """Returns the string/JSON representation of a TestFailure.""" return cPickle.dumps(self) def driver_needs_restart(self): """Returns True if we should kill DumpRenderTree/WebKitTestRunner before the next test.""" return False def write_failure(self, writer, driver_output, expected_driver_output, port): assert isinstance(self, (FailureTimeout, FailureReftestNoImagesGenerated)) class FailureText(TestFailure): def write_failure(self, writer, driver_output, expected_driver_output, port): writer.write_text_files(driver_output.text, expected_driver_output.text) writer.create_text_diff_and_write_result(driver_output.text, expected_driver_output.text) class FailureAudio(TestFailure): def write_failure(self, writer, driver_output, expected_driver_output, port): writer.write_audio_files(driver_output.audio, expected_driver_output.audio) writer.create_audio_diff_and_write_result(driver_output.audio, expected_driver_output.audio) class FailureTimeout(TestFailure): def __init__(self, is_reftest=False): super(FailureTimeout, self).__init__() self.is_reftest = is_reftest def message(self): return "test timed out" def driver_needs_restart(self): return True class FailureCrash(TestFailure): def __init__(self, is_reftest=False, process_name='DumpRenderTree', pid=None): super(FailureCrash, self).__init__() self.process_name = process_name self.pid = pid self.is_reftest = is_reftest def message(self): if self.pid: return "%s crashed [pid=%d]" % (self.process_name, self.pid) return self.process_name + " crashed" def driver_needs_restart(self): return True def write_failure(self, writer, driver_output, expected_driver_output, port): crashed_driver_output = expected_driver_output if self.is_reftest else driver_output writer.write_crash_log(crashed_driver_output.crash_log) class FailureMissingResult(FailureText): def message(self): return "-expected.txt was missing" class FailureTextMismatch(FailureText): def message(self): return "text diff" class FailureMissingImageHash(TestFailure): def message(self): return "-expected.png was missing an embedded checksum" def write_failure(self, writer, driver_output, expected_driver_output, port): writer.write_image_files(driver_output.image, expected_driver_output.image) class FailureMissingImage(TestFailure): def message(self): return "-expected.png was missing" def write_failure(self, writer, driver_output, expected_driver_output, port): writer.write_image_files(driver_output.image, expected_image=None) class FailureImageHashMismatch(TestFailure): def __init__(self, diff_percent=0): super(FailureImageHashMismatch, self).__init__() self.diff_percent = diff_percent def message(self): return "image diff" def write_failure(self, writer, driver_output, expected_driver_output, port): writer.write_image_files(driver_output.image, expected_driver_output.image) writer.write_image_diff_files(driver_output.image_diff) class FailureImageHashIncorrect(TestFailure): def message(self): return "-expected.png embedded checksum is incorrect" class FailureReftestMismatch(TestFailure): def __init__(self, reference_filename=None): super(FailureReftestMismatch, self).__init__() self.reference_filename = reference_filename self.diff_percent = None def message(self): return "reference mismatch" def write_failure(self, writer, driver_output, expected_driver_output, port): writer.write_image_files(driver_output.image, expected_driver_output.image) # FIXME: This work should be done earlier in the pipeline (e.g., when we compare images for non-ref tests). # FIXME: We should always have 2 images here. if driver_output.image and expected_driver_output.image: diff_image, diff_percent, err_str = port.diff_image(expected_driver_output.image, driver_output.image, tolerance=0) if diff_image: writer.write_image_diff_files(diff_image) self.diff_percent = diff_percent else: _log.warn('ref test mismatch did not produce an image diff.') writer.write_reftest(self.reference_filename) class FailureReftestMismatchDidNotOccur(TestFailure): def __init__(self, reference_filename=None): super(FailureReftestMismatchDidNotOccur, self).__init__() self.reference_filename = reference_filename def message(self): return "reference mismatch didn't happen" def write_failure(self, writer, driver_output, expected_driver_output, port): writer.write_image_files(driver_output.image, expected_image=None) writer.write_reftest(self.reference_filename) class FailureReftestNoImagesGenerated(TestFailure): def __init__(self, reference_filename=None): super(FailureReftestNoImagesGenerated, self).__init__() self.reference_filename = reference_filename def message(self): return "reference didn't generate pixel results." class FailureMissingAudio(FailureAudio): def message(self): return "expected audio result was missing" class FailureAudioMismatch(FailureAudio): def message(self): return "audio mismatch" class FailureEarlyExit(TestFailure): def message(self): return "skipped due to early exit" # Convenient collection of all failure classes for anything that might # need to enumerate over them all. ALL_FAILURE_CLASSES = (FailureTimeout, FailureCrash, FailureMissingResult, FailureTextMismatch, FailureMissingImageHash, FailureMissingImage, FailureImageHashMismatch, FailureImageHashIncorrect, FailureReftestMismatch, FailureReftestMismatchDidNotOccur, FailureReftestNoImagesGenerated, FailureMissingAudio, FailureAudioMismatch, FailureEarlyExit)