1#!/usr/bin/python
2
3'''
4Copyright 2013 Google Inc.
5
6Use of this source code is governed by a BSD-style license that can be
7found in the LICENSE file.
8'''
9
10'''
11Gathers diffs between 2 JSON expectations files, or between actual and
12expected results within a single JSON actual-results file,
13and generates an old-vs-new diff dictionary.
14
15TODO(epoger): Fix indentation in this file (2-space indents, not 4-space).
16'''
17
18# System-level imports
19import argparse
20import json
21import os
22import sys
23import urllib2
24
25# Imports from within Skia
26#
27# We need to add the 'gm' directory, so that we can import gm_json.py within
28# that directory.  That script allows us to parse the actual-results.json file
29# written out by the GM tool.
30# Make sure that the 'gm' dir is in the PYTHONPATH, but add it at the *end*
31# so any dirs that are already in the PYTHONPATH will be preferred.
32#
33# This assumes that the 'gm' directory has been checked out as a sibling of
34# the 'tools' directory containing this script, which will be the case if
35# 'trunk' was checked out as a single unit.
36GM_DIRECTORY = os.path.realpath(
37    os.path.join(os.path.dirname(os.path.dirname(__file__)), 'gm'))
38if GM_DIRECTORY not in sys.path:
39    sys.path.append(GM_DIRECTORY)
40import gm_json
41
42
43# Object that generates diffs between two JSON gm result files.
44class GMDiffer(object):
45
46    def __init__(self):
47        pass
48
49    def _GetFileContentsAsString(self, filepath):
50        """Returns the full contents of a file, as a single string.
51        If the filename looks like a URL, download its contents.
52        If the filename is None, return None."""
53        if filepath is None:
54            return None
55        elif filepath.startswith('http:') or filepath.startswith('https:'):
56            return urllib2.urlopen(filepath).read()
57        else:
58            return open(filepath, 'r').read()
59
60    def _GetExpectedResults(self, contents):
61        """Returns the dictionary of expected results from a JSON string,
62        in this form:
63
64        {
65          'test1' : 14760033689012826769,
66          'test2' : 9151974350149210736,
67          ...
68        }
69
70        We make these simplifying assumptions:
71        1. Each test has either 0 or 1 allowed results.
72        2. All expectations are of type JSONKEY_HASHTYPE_BITMAP_64BITMD5.
73
74        Any tests which violate those assumptions will cause an exception to
75        be raised.
76
77        Any tests for which we have no expectations will be left out of the
78        returned dictionary.
79        """
80        result_dict = {}
81        json_dict = gm_json.LoadFromString(contents)
82        all_expectations = json_dict[gm_json.JSONKEY_EXPECTEDRESULTS]
83
84        # Prevent https://code.google.com/p/skia/issues/detail?id=1588
85        # ('svndiff.py: 'NoneType' object has no attribute 'keys'')
86        if not all_expectations:
87            return result_dict
88
89        for test_name in all_expectations.keys():
90            test_expectations = all_expectations[test_name]
91            allowed_digests = test_expectations[
92                gm_json.JSONKEY_EXPECTEDRESULTS_ALLOWEDDIGESTS]
93            if allowed_digests:
94                num_allowed_digests = len(allowed_digests)
95                if num_allowed_digests > 1:
96                    raise ValueError(
97                        'test %s has %d allowed digests' % (
98                            test_name, num_allowed_digests))
99                digest_pair = allowed_digests[0]
100                if digest_pair[0] != gm_json.JSONKEY_HASHTYPE_BITMAP_64BITMD5:
101                    raise ValueError(
102                        'test %s has unsupported hashtype %s' % (
103                            test_name, digest_pair[0]))
104                result_dict[test_name] = digest_pair[1]
105        return result_dict
106
107    def _GetActualResults(self, contents):
108        """Returns the dictionary of actual results from a JSON string,
109        in this form:
110
111        {
112          'test1' : 14760033689012826769,
113          'test2' : 9151974350149210736,
114          ...
115        }
116
117        We make these simplifying assumptions:
118        1. All results are of type JSONKEY_HASHTYPE_BITMAP_64BITMD5.
119
120        Any tests which violate those assumptions will cause an exception to
121        be raised.
122
123        Any tests for which we have no actual results will be left out of the
124        returned dictionary.
125        """
126        result_dict = {}
127        json_dict = gm_json.LoadFromString(contents)
128        all_result_types = json_dict[gm_json.JSONKEY_ACTUALRESULTS]
129        for result_type in all_result_types.keys():
130            results_of_this_type = all_result_types[result_type]
131            if results_of_this_type:
132                for test_name in results_of_this_type.keys():
133                    digest_pair = results_of_this_type[test_name]
134                    if digest_pair[0] != gm_json.JSONKEY_HASHTYPE_BITMAP_64BITMD5:
135                        raise ValueError(
136                            'test %s has unsupported hashtype %s' % (
137                                test_name, digest_pair[0]))
138                    result_dict[test_name] = digest_pair[1]
139        return result_dict
140
141    def _DictionaryDiff(self, old_dict, new_dict):
142        """Generate a dictionary showing the diffs between old_dict and new_dict.
143        Any entries which are identical across them will be left out."""
144        diff_dict = {}
145        all_keys = set(old_dict.keys() + new_dict.keys())
146        for key in all_keys:
147            if old_dict.get(key) != new_dict.get(key):
148                new_entry = {}
149                new_entry['old'] = old_dict.get(key)
150                new_entry['new'] = new_dict.get(key)
151                diff_dict[key] = new_entry
152        return diff_dict
153
154    def GenerateDiffDict(self, oldfile, newfile=None):
155        """Generate a dictionary showing the diffs:
156        old = expectations within oldfile
157        new = expectations within newfile
158
159        If newfile is not specified, then 'new' is the actual results within
160        oldfile.
161        """
162        return self.GenerateDiffDictFromStrings(self._GetFileContentsAsString(oldfile),
163                                                self._GetFileContentsAsString(newfile))
164
165    def GenerateDiffDictFromStrings(self, oldjson, newjson=None):
166        """Generate a dictionary showing the diffs:
167        old = expectations within oldjson
168        new = expectations within newjson
169
170        If newfile is not specified, then 'new' is the actual results within
171        oldfile.
172        """
173        old_results = self._GetExpectedResults(oldjson)
174        if newjson:
175            new_results = self._GetExpectedResults(newjson)
176        else:
177            new_results = self._GetActualResults(oldjson)
178        return self._DictionaryDiff(old_results, new_results)
179
180
181def _Main():
182    parser = argparse.ArgumentParser()
183    parser.add_argument(
184        'old',
185        help='Path to JSON file whose expectations to display on ' +
186        'the "old" side of the diff. This can be a filepath on ' +
187        'local storage, or a URL.')
188    parser.add_argument(
189        'new', nargs='?',
190        help='Path to JSON file whose expectations to display on ' +
191        'the "new" side of the diff; if not specified, uses the ' +
192        'ACTUAL results from the "old" JSON file. This can be a ' +
193        'filepath on local storage, or a URL.')
194    args = parser.parse_args()
195    differ = GMDiffer()
196    diffs = differ.GenerateDiffDict(oldfile=args.old, newfile=args.new)
197    json.dump(diffs, sys.stdout, sort_keys=True, indent=2)
198
199
200if __name__ == '__main__':
201    _Main()
202