OLD | NEW |
---|---|
1 # Copyright 2013 The Chromium Authors. All rights reserved. | 1 # Copyright 2013 The Chromium Authors. All rights reserved. |
2 # Use of this source code is governed by a BSD-style license that can be | 2 # Use of this source code is governed by a BSD-style license that can be |
3 # found in the LICENSE file. | 3 # found in the LICENSE file. |
4 | 4 |
5 import difflib | 5 import difflib |
6 import hashlib | 6 import hashlib |
7 import os | 7 import os |
8 import re | 8 import re |
9 import sys | 9 import sys |
10 | 10 |
11 from util import build_utils | |
12 | |
13 if build_utils.COLORAMA_ROOT not in sys.path: | |
14 sys.path.append(build_utils.COLORAMA_ROOT) | |
15 import colorama | |
16 | |
17 | 11 |
18 # When set and a difference is detected, a diff of what changed is printed. | 12 # When set and a difference is detected, a diff of what changed is printed. |
19 _PRINT_MD5_DIFFS = int(os.environ.get('PRINT_MD5_DIFFS', 0)) | 13 _PRINT_MD5_DIFFS = int(os.environ.get('PRINT_MD5_DIFFS', 0)) |
20 | 14 |
21 # Used to strip off temp dir prefix. | 15 # Used to strip off temp dir prefix. |
22 _TEMP_DIR_PATTERN = re.compile(r'^/tmp/.*?/') | 16 _TEMP_DIR_PATTERN = re.compile(r'^/tmp/.*?/') |
23 | 17 |
24 | 18 |
25 def CallAndRecordIfStale( | 19 def CallAndRecordIfStale( |
26 function, record_path=None, input_paths=None, input_strings=None, | 20 function, record_path=None, input_paths=None, input_strings=None, |
27 force=False): | 21 output_paths=None, force=False): |
28 """Calls function if the md5sum of the input paths/strings has changed. | 22 """Calls function if outputs are stale. |
29 | 23 |
30 The md5sum of the inputs is compared with the one stored in record_path. If | 24 Outputs are considered stale if: |
31 this has changed (or the record doesn't exist), function will be called and | 25 - any output_paths are missing, or |
32 the new md5sum will be recorded. | 26 - the contents of any file within input_paths has changed, or |
27 - the contents of input_strings has changed. | |
33 | 28 |
34 If force is True, the function will be called regardless of whether the | 29 To debug which files are out-of-date, set the environment variable: |
35 md5sum is out of date. | 30 PRINT_MD5_DIFFS=1 |
31 | |
32 Args: | |
33 function: The function to call. | |
34 record_path: Path to record metadata. | |
35 Defaults to output_paths[0] + '.md5.stamp' | |
36 input_paths: List of paths to calcualte an md5 sum on. | |
37 input_strings: List of strings to record verbatim. | |
38 output_paths: List of output paths. | |
39 force: When True, function is always called. | |
36 """ | 40 """ |
37 if not input_paths: | 41 input_paths = input_paths or [] |
38 input_paths = [] | 42 input_strings = input_strings or [] |
39 if not input_strings: | 43 output_paths = output_paths or [] |
40 input_strings = [] | 44 record_path = record_path or output_paths[0] + '.md5.stamp' |
jbudorick
2015/09/24 15:16:48
This looks like it doesn't work with the default p
agrieve
2015/09/24 15:40:44
Right. I want an exception to be thrown in this ca
jbudorick
2015/09/24 15:42:57
Yes, an assert makes it clear that it's intentiona
| |
41 md5_checker = _Md5Checker( | 45 md5_checker = _Md5Checker( |
42 record_path=record_path, | 46 record_path=record_path, |
43 input_paths=input_paths, | 47 input_paths=input_paths, |
44 input_strings=input_strings) | 48 input_strings=input_strings) |
45 | 49 |
50 missing_outputs = [x for x in output_paths if not os.path.exists(x)] | |
46 is_stale = md5_checker.old_digest != md5_checker.new_digest | 51 is_stale = md5_checker.old_digest != md5_checker.new_digest |
47 if force or is_stale: | 52 |
48 if is_stale and _PRINT_MD5_DIFFS: | 53 if force or missing_outputs or is_stale: |
49 print '%sDifference found in %s:%s' % ( | 54 if _PRINT_MD5_DIFFS: |
50 colorama.Fore.YELLOW, record_path, colorama.Fore.RESET) | 55 print '=' * 80 |
51 print md5_checker.DescribeDifference() | 56 print 'Difference found in %s:' % record_path |
57 if missing_outputs: | |
58 print 'Outputs do not exist:\n' + '\n'.join(missing_outputs) | |
59 elif force: | |
60 print 'force=True' | |
61 else: | |
62 print md5_checker.DescribeDifference() | |
63 print '=' * 80 | |
52 function() | 64 function() |
53 md5_checker.Write() | 65 md5_checker.Write() |
54 | 66 |
55 | 67 |
56 def _UpdateMd5ForFile(md5, path, block_size=2**16): | 68 def _UpdateMd5ForFile(md5, path, block_size=2**16): |
57 with open(path, 'rb') as infile: | 69 with open(path, 'rb') as infile: |
58 while True: | 70 while True: |
59 data = infile.read(block_size) | 71 data = infile.read(block_size) |
60 if not data: | 72 if not data: |
61 break | 73 break |
(...skipping 37 matching lines...) Expand 10 before | Expand all | Expand 10 after Loading... | |
99 extended_info = [] | 111 extended_info = [] |
100 outer_md5 = hashlib.md5() | 112 outer_md5 = hashlib.md5() |
101 for i in sorted(input_paths): | 113 for i in sorted(input_paths): |
102 inner_md5 = hashlib.md5() | 114 inner_md5 = hashlib.md5() |
103 _UpdateMd5ForPath(inner_md5, i) | 115 _UpdateMd5ForPath(inner_md5, i) |
104 i = _TrimPathPrefix(i) | 116 i = _TrimPathPrefix(i) |
105 extended_info.append(i + '=' + inner_md5.hexdigest()) | 117 extended_info.append(i + '=' + inner_md5.hexdigest()) |
106 # Include the digest in the overall diff, but not the path | 118 # Include the digest in the overall diff, but not the path |
107 outer_md5.update(inner_md5.hexdigest()) | 119 outer_md5.update(inner_md5.hexdigest()) |
108 | 120 |
109 for s in input_strings: | 121 for s in map(str, input_strings): |
jbudorick
2015/09/24 15:16:48
nit: we tend to avoid map, filter, etc in favor of
agrieve
2015/09/24 15:40:44
Agree that reads better :) Done.
| |
110 outer_md5.update(s) | 122 outer_md5.update(s) |
111 extended_info.append(s) | 123 extended_info.append(s) |
112 | 124 |
113 self.new_digest = outer_md5.hexdigest() | 125 self.new_digest = outer_md5.hexdigest() |
114 self.new_extended_info = extended_info | 126 self.new_extended_info = extended_info |
115 | 127 |
116 self.old_digest = '' | 128 self.old_digest = '' |
117 self.old_extended_info = [] | 129 self.old_extended_info = [] |
118 if os.path.exists(self.record_path): | 130 if os.path.exists(self.record_path): |
119 with open(self.record_path, 'r') as old_record: | 131 with open(self.record_path, 'r') as old_record: |
120 self.old_extended_info = [line.strip() for line in old_record] | 132 self.old_extended_info = [line.strip() for line in old_record] |
121 self.old_digest = self.old_extended_info.pop(0) | 133 if self.old_extended_info: |
134 self.old_digest = self.old_extended_info.pop(0) | |
122 | 135 |
123 def Write(self): | 136 def Write(self): |
124 with open(self.record_path, 'w') as new_record: | 137 with open(self.record_path, 'w') as new_record: |
125 new_record.write(self.new_digest) | 138 new_record.write(self.new_digest) |
126 new_record.write('\n' + '\n'.join(self.new_extended_info) + '\n') | 139 new_record.write('\n' + '\n'.join(self.new_extended_info) + '\n') |
127 | 140 |
128 def DescribeDifference(self): | 141 def DescribeDifference(self): |
129 if self.old_digest == self.new_digest: | 142 if self.old_digest == self.new_digest: |
130 return "There's no difference." | 143 return "There's no difference." |
131 if not self.old_digest: | 144 if not self.old_digest: |
132 return 'Previous stamp file not found.' | 145 return 'Previous stamp file not found.' |
133 if not self.old_extended_info: | 146 if not self.old_extended_info: |
134 return 'Previous stamp file lacks extended info.' | 147 return 'Previous stamp file lacks extended info.' |
135 diff = difflib.unified_diff(self.old_extended_info, self.new_extended_info) | 148 diff = difflib.unified_diff(self.old_extended_info, self.new_extended_info) |
136 return '\n'.join(diff) | 149 return '\n'.join(diff) |
OLD | NEW |