|
| 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 | +from __future__ import print_function |
| 18 | + |
| 19 | +import argparse |
| 20 | +import datetime |
| 21 | +import difflib |
| 22 | +import glob |
| 23 | +import os |
| 24 | +import re |
| 25 | +import sys |
| 26 | + |
| 27 | +parser = argparse.ArgumentParser() |
| 28 | +parser.add_argument( |
| 29 | + "filenames", |
| 30 | + help="list of files to check, all files if unspecified", |
| 31 | + nargs='*') |
| 32 | + |
| 33 | +rootdir = os.path.dirname(__file__) + "/../../" |
| 34 | +rootdir = os.path.abspath(rootdir) |
| 35 | +parser.add_argument( |
| 36 | + "--rootdir", default=rootdir, help="root directory to examine") |
| 37 | + |
| 38 | +default_boilerplate_dir = os.path.join(rootdir, "hack/boilerplate") |
| 39 | +parser.add_argument( |
| 40 | + "--boilerplate-dir", default=default_boilerplate_dir) |
| 41 | + |
| 42 | +parser.add_argument( |
| 43 | + "-v", "--verbose", |
| 44 | + help="give verbose output regarding why a file does not pass", |
| 45 | + action="store_true") |
| 46 | + |
| 47 | +args = parser.parse_args() |
| 48 | + |
| 49 | +verbose_out = sys.stderr if args.verbose else open("/dev/null", "w") |
| 50 | + |
| 51 | +def get_refs(): |
| 52 | + refs = {} |
| 53 | + |
| 54 | + for path in glob.glob(os.path.join(args.boilerplate_dir, "boilerplate.*.txt")): |
| 55 | + extension = os.path.basename(path).split(".")[1] |
| 56 | + |
| 57 | + ref_file = open(path, 'r') |
| 58 | + ref = ref_file.read().splitlines() |
| 59 | + ref_file.close() |
| 60 | + refs[extension] = ref |
| 61 | + |
| 62 | + return refs |
| 63 | + |
| 64 | +def is_generated_file(filename, data, regexs): |
| 65 | + for d in skipped_ungenerated_files: |
| 66 | + if d in filename: |
| 67 | + return False |
| 68 | + |
| 69 | + p = regexs["generated"] |
| 70 | + return p.search(data) |
| 71 | + |
| 72 | +def file_passes(filename, refs, regexs): |
| 73 | + try: |
| 74 | + f = open(filename, 'r') |
| 75 | + except Exception as exc: |
| 76 | + print("Unable to open %s: %s" % (filename, exc), file=verbose_out) |
| 77 | + return False |
| 78 | + |
| 79 | + data = f.read() |
| 80 | + f.close() |
| 81 | + |
| 82 | + # determine if the file is automatically generated |
| 83 | + generated = is_generated_file(filename, data, regexs) |
| 84 | + |
| 85 | + basename = os.path.basename(filename) |
| 86 | + extension = file_extension(filename) |
| 87 | + if generated: |
| 88 | + if extension == "go": |
| 89 | + extension = "generatego" |
| 90 | + elif extension == "bzl": |
| 91 | + extension = "generatebzl" |
| 92 | + |
| 93 | + if extension != "": |
| 94 | + ref = refs[extension] |
| 95 | + else: |
| 96 | + ref = refs[basename] |
| 97 | + |
| 98 | + # remove extra content from the top of files |
| 99 | + if extension == "go" or extension == "generatego": |
| 100 | + p = regexs["go_build_constraints"] |
| 101 | + (data, found) = p.subn("", data, 1) |
| 102 | + elif extension in ["sh", "py"]: |
| 103 | + p = regexs["shebang"] |
| 104 | + (data, found) = p.subn("", data, 1) |
| 105 | + |
| 106 | + data = data.splitlines() |
| 107 | + |
| 108 | + # if our test file is smaller than the reference it surely fails! |
| 109 | + if len(ref) > len(data): |
| 110 | + print('File %s smaller than reference (%d < %d)' % |
| 111 | + (filename, len(data), len(ref)), |
| 112 | + file=verbose_out) |
| 113 | + return False |
| 114 | + |
| 115 | + # trim our file to the same number of lines as the reference file |
| 116 | + data = data[:len(ref)] |
| 117 | + |
| 118 | + p = regexs["year"] |
| 119 | + for d in data: |
| 120 | + if p.search(d): |
| 121 | + if generated: |
| 122 | + print('File %s has the YEAR field, but it should not be in generated file' % |
| 123 | + filename, file=verbose_out) |
| 124 | + else: |
| 125 | + print('File %s has the YEAR field, but missing the year of date' % |
| 126 | + filename, file=verbose_out) |
| 127 | + return False |
| 128 | + |
| 129 | + if not generated: |
| 130 | + # Replace all occurrences of the regex "2014|2015|2016|2017|2018" with "YEAR" |
| 131 | + p = regexs["date"] |
| 132 | + for i, d in enumerate(data): |
| 133 | + (data[i], found) = p.subn('YEAR', d) |
| 134 | + if found != 0: |
| 135 | + break |
| 136 | + |
| 137 | + # if we don't match the reference at this point, fail |
| 138 | + if ref != data: |
| 139 | + print("Header in %s does not match reference, diff:" % |
| 140 | + filename, file=verbose_out) |
| 141 | + if args.verbose: |
| 142 | + print(file=verbose_out) |
| 143 | + for line in difflib.unified_diff(ref, data, 'reference', filename, lineterm=''): |
| 144 | + print(line, file=verbose_out) |
| 145 | + print(file=verbose_out) |
| 146 | + |
| 147 | + return False |
| 148 | + |
| 149 | + return True |
| 150 | + |
| 151 | +def file_extension(filename): |
| 152 | + return os.path.splitext(filename)[1].split(".")[-1].lower() |
| 153 | + |
| 154 | +skipped_dirs = ['_output', '.git', "hack/boilerplate/test"] |
| 155 | + |
| 156 | +# list all the files contain 'DO NOT EDIT', but are not generated |
| 157 | +skipped_ungenerated_files = [ |
| 158 | + 'hack/lib/swagger.sh', |
| 159 | + 'hack/boilerplate/boilerplate.py', |
| 160 | + # The generator injects `DO NOT EDIT` and thus needs to get excluded to not |
| 161 | + # get detected as false positive. |
| 162 | + 'hack/tools/prowjob-gen/generator.go', |
| 163 | + ] |
| 164 | + |
| 165 | +def normalize_files(files): |
| 166 | + newfiles = [] |
| 167 | + for pathname in files: |
| 168 | + if any(x in pathname for x in skipped_dirs): |
| 169 | + continue |
| 170 | + newfiles.append(pathname) |
| 171 | + for i, pathname in enumerate(newfiles): |
| 172 | + if not os.path.isabs(pathname): |
| 173 | + newfiles[i] = os.path.join(args.rootdir, pathname) |
| 174 | + return newfiles |
| 175 | + |
| 176 | +def get_files(extensions): |
| 177 | + files = [] |
| 178 | + if len(args.filenames) > 0: |
| 179 | + files = args.filenames |
| 180 | + else: |
| 181 | + for root, dirs, walkfiles in os.walk(args.rootdir): |
| 182 | + # don't visit certain dirs. This is just a performance improvement |
| 183 | + # as we would prune these later in normalize_files(). But doing it |
| 184 | + # cuts down the amount of filesystem walking we do and cuts down |
| 185 | + # the size of the file list |
| 186 | + for d in skipped_dirs: |
| 187 | + if d in dirs: |
| 188 | + dirs.remove(d) |
| 189 | + |
| 190 | + for name in walkfiles: |
| 191 | + pathname = os.path.join(root, name) |
| 192 | + files.append(pathname) |
| 193 | + |
| 194 | + files = normalize_files(files) |
| 195 | + outfiles = [] |
| 196 | + for pathname in files: |
| 197 | + basename = os.path.basename(pathname) |
| 198 | + extension = file_extension(pathname) |
| 199 | + if extension in extensions or basename in extensions: |
| 200 | + outfiles.append(pathname) |
| 201 | + return outfiles |
| 202 | + |
| 203 | +def get_dates(): |
| 204 | + years = datetime.datetime.now().year |
| 205 | + return '(%s)' % '|'.join((str(year) for year in range(2014, years+1))) |
| 206 | + |
| 207 | +def get_regexs(): |
| 208 | + regexs = {} |
| 209 | + # Search for "YEAR" which exists in the boilerplate, but shouldn't in the real thing |
| 210 | + regexs["year"] = re.compile('YEAR') |
| 211 | + # get_dates return 2014, 2015, 2016, 2017, or 2018 until the current year as a regex like: "(2014|2015|2016|2017|2018)"; |
| 212 | + # company holder names can be anything |
| 213 | + regexs["date"] = re.compile(get_dates()) |
| 214 | + # strip the following build constraints/tags: |
| 215 | + # //go:build |
| 216 | + # // +build \n\n |
| 217 | + regexs["go_build_constraints"] = re.compile( |
| 218 | + r"^(//(go:build| \+build).*\n)+\n", re.MULTILINE) |
| 219 | + # strip #!.* from scripts |
| 220 | + regexs["shebang"] = re.compile(r"^(#!.*\n)\n*", re.MULTILINE) |
| 221 | + # Search for generated files |
| 222 | + regexs["generated"] = re.compile('DO NOT EDIT') |
| 223 | + return regexs |
| 224 | + |
| 225 | +def main(): |
| 226 | + regexs = get_regexs() |
| 227 | + refs = get_refs() |
| 228 | + filenames = get_files(refs.keys()) |
| 229 | + |
| 230 | + for filename in filenames: |
| 231 | + if not file_passes(filename, refs, regexs): |
| 232 | + print(filename, file=sys.stdout) |
| 233 | + |
| 234 | + return 0 |
| 235 | + |
| 236 | +if __name__ == "__main__": |
| 237 | + sys.exit(main()) |
0 commit comments