1#!/usr/bin/env python3
2
3# Copyright 2015 The Kubernetes Authors.
4#
5# Licensed under the Apache License, Version 2.0 (the "License");
6# you may not use this file except in compliance with the License.
7# You may obtain a copy of the License at
8#
9# http://www.apache.org/licenses/LICENSE-2.0
10#
11# Unless required by applicable law or agreed to in writing, software
12# distributed under the License is distributed on an "AS IS" BASIS,
13# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
14# See the License for the specific language governing permissions and
15# limitations under the License.
16
17# This file is copied from https://github.com/kubernetes/kubernetes/blob/04c2b1fbdc1289c9a72eda87cf7072346e60d241/hack/boilerplate/boilerplate.py
18
19from __future__ import print_function
20
21import argparse
22import datetime
23import difflib
24import glob
25import os
26import re
27import sys
28
29parser = argparse.ArgumentParser()
30parser.add_argument(
31 "filenames",
32 help="list of files to check, all files if unspecified",
33 nargs='*')
34
35rootdir = os.path.dirname(__file__) + "/../../"
36rootdir = os.path.abspath(rootdir)
37parser.add_argument(
38 "--rootdir", default=rootdir, help="root directory to examine")
39
40default_boilerplate_dir = os.path.join(rootdir, "hack/boilerplate")
41parser.add_argument(
42 "--boilerplate-dir", default=default_boilerplate_dir)
43
44parser.add_argument(
45 "-v", "--verbose",
46 help="give verbose output regarding why a file does not pass",
47 action="store_true")
48
49args = parser.parse_args()
50
51verbose_out = sys.stderr if args.verbose else open("/dev/null", "w")
52
53
54def get_refs():
55 refs = {}
56
57 for path in glob.glob(os.path.join(args.boilerplate_dir, "boilerplate.*.txt")):
58 extension = os.path.basename(path).split(".")[1]
59
60 ref_file = open(path, 'r')
61 ref = ref_file.read().splitlines()
62 ref_file.close()
63 refs[extension] = ref
64
65 return refs
66
67
68def is_generated_file(filename, data, regexs):
69 for d in skipped_ungenerated_files:
70 if d in filename:
71 return False
72
73 p = regexs["generated"]
74 return p.search(data)
75
76
77def file_passes(filename, refs, regexs):
78 try:
79 f = open(filename, 'r')
80 except Exception as exc:
81 print("Unable to open %s: %s" % (filename, exc), file=verbose_out)
82 return False
83
84 data = f.read()
85 f.close()
86
87 # determine if the file is automatically generated
88 generated = is_generated_file(filename, data, regexs)
89
90 basename = os.path.basename(filename)
91 extension = file_extension(filename)
92 if generated:
93 if extension == "go":
94 extension = "generatego"
95 elif extension == "bzl":
96 extension = "generatebzl"
97
98 if extension != "":
99 ref = refs[extension]
100 else:
101 ref = refs[basename]
102
103 # remove extra content from the top of files
104 if extension == "go" or extension == "generatego":
105 p = regexs["go_build_constraints"]
106 (data, found) = p.subn("", data, 1)
107 elif extension in ["sh", "py"]:
108 p = regexs["shebang"]
109 (data, found) = p.subn("", data, 1)
110
111 data = data.splitlines()
112
113 # if our test file is smaller than the reference it surely fails!
114 if len(ref) > len(data):
115 print('File %s smaller than reference (%d < %d)' %
116 (filename, len(data), len(ref)),
117 file=verbose_out)
118 return False
119
120 # trim our file to the same number of lines as the reference file
121 data = data[:len(ref)]
122
123 p = regexs["year"]
124 for d in data:
125 if p.search(d):
126 if generated:
127 print('File %s has the YEAR field, but it should not be in generated file' %
128 filename, file=verbose_out)
129 else:
130 print('File %s has the YEAR field, but missing the year of date' %
131 filename, file=verbose_out)
132 return False
133
134 if not generated:
135 # Replace all occurrences of the regex "2014|2015|2016|2017|2018" with "YEAR"
136 p = regexs["date"]
137 for i, d in enumerate(data):
138 (data[i], found) = p.subn('YEAR', d)
139 if found != 0:
140 break
141
142 # if we don't match the reference at this point, fail
143 if ref != data:
144 print("Header in %s does not match reference, diff:" %
145 filename, file=verbose_out)
146 if args.verbose:
147 print(file=verbose_out)
148 for line in difflib.unified_diff(ref, data, 'reference', filename, lineterm=''):
149 print(line, file=verbose_out)
150 print(file=verbose_out)
151 return False
152
153 return True
154
155
156def file_extension(filename):
157 return os.path.splitext(filename)[1].split(".")[-1].lower()
158
159
160skipped_dirs = [
161 'cluster/env.sh',
162 '.git',
163 '_gopath',
164 'hack/boilerplate/test',
165 '_output',
166 'staging/src/k8s.io/kubectl/pkg/generated/bindata.go',
167 'test/e2e/generated/bindata.go',
168 'third_party',
169 'vendor',
170 '.venv',
171]
172
173# list all the files contain 'DO NOT EDIT', but are not generated
174skipped_ungenerated_files = [
175 'hack/lib/swagger.sh', 'hack/boilerplate/boilerplate.py']
176
177
178def normalize_files(files):
179 newfiles = []
180 for pathname in files:
181 if any(x in pathname for x in skipped_dirs):
182 continue
183 newfiles.append(pathname)
184 for i, pathname in enumerate(newfiles):
185 if not os.path.isabs(pathname):
186 newfiles[i] = os.path.join(args.rootdir, pathname)
187 return newfiles
188
189
190def get_files(extensions):
191 files = []
192 if len(args.filenames) > 0:
193 files = args.filenames
194 else:
195 for root, dirs, walkfiles in os.walk(args.rootdir):
196 # don't visit certain dirs. This is just a performance improvement
197 # as we would prune these later in normalize_files(). But doing it
198 # cuts down the amount of filesystem walking we do and cuts down
199 # the size of the file list
200 for d in skipped_dirs:
201 if d in dirs:
202 dirs.remove(d)
203
204 for name in walkfiles:
205 pathname = os.path.join(root, name)
206 files.append(pathname)
207
208 files = normalize_files(files)
209 outfiles = []
210 for pathname in files:
211 basename = os.path.basename(pathname)
212 extension = file_extension(pathname)
213 if extension in extensions or basename in extensions:
214 outfiles.append(pathname)
215 return outfiles
216
217
218def get_dates():
219 years = datetime.datetime.now().year
220 return '(%s)' % '|'.join((str(year) for year in range(2014, years+1)))
221
222
223def get_regexs():
224 regexs = {}
225 # Search for "YEAR" which exists in the boilerplate, but shouldn't in the real thing
226 regexs["year"] = re.compile('YEAR')
227 # get_dates return 2014, 2015, 2016, 2017, or 2018 until the current year as a regex like: "(2014|2015|2016|2017|2018)";
228 # company holder names can be anything
229 regexs["date"] = re.compile(get_dates())
230 # strip the following build constraints/tags:
231 # //go:build
232 # // +build \n\n
233 regexs["go_build_constraints"] = re.compile(
234 r"^(//(go:build| \+build).*\n)+\n", re.MULTILINE)
235 # strip #!.* from scripts
236 regexs["shebang"] = re.compile(r"^(#!.*\n)\n*", re.MULTILINE)
237 # Search for generated files
238 regexs["generated"] = re.compile('DO NOT EDIT')
239 return regexs
240
241
242def main():
243 regexs = get_regexs()
244 refs = get_refs()
245 filenames = get_files(refs.keys())
246
247 for filename in filenames:
248 if not file_passes(filename, refs, regexs):
249 print(filename, file=sys.stdout)
250
251 print("Verified %d file headers match boilerplate" % (len(filenames),), file=sys.stderr)
252
253 return 0
254
255
256if __name__ == "__main__":
257 sys.exit(main())
View as plain text