648db22b |
1 | #!/usr/bin/env python3 |
2 | |
3 | # ################################################################ |
4 | # Copyright (c) Meta Platforms, Inc. and affiliates. |
5 | # All rights reserved. |
6 | # |
7 | # This source code is licensed under both the BSD-style license (found in the |
8 | # LICENSE file in the root directory of this source tree) and the GPLv2 (found |
9 | # in the COPYING file in the root directory of this source tree). |
10 | # You may select, at your option, one of the above-listed licenses. |
11 | # ################################################################ |
12 | |
13 | import enum |
14 | import glob |
15 | import os |
16 | import re |
17 | import sys |
18 | |
19 | ROOT = os.path.join(os.path.dirname(__file__), "..") |
20 | |
21 | RELDIRS = [ |
22 | "doc", |
23 | "examples", |
24 | "lib", |
25 | "programs", |
26 | "tests", |
27 | "contrib/linux-kernel", |
28 | ] |
29 | |
30 | REL_EXCLUDES = [ |
31 | "contrib/linux-kernel/test/include", |
32 | ] |
33 | |
34 | def to_abs(d): |
35 | return os.path.normpath(os.path.join(ROOT, d)) + "/" |
36 | |
37 | DIRS = [to_abs(d) for d in RELDIRS] |
38 | EXCLUDES = [to_abs(d) for d in REL_EXCLUDES] |
39 | |
40 | SUFFIXES = [ |
41 | ".c", |
42 | ".h", |
43 | "Makefile", |
44 | ".mk", |
45 | ".py", |
46 | ".S", |
47 | ] |
48 | |
49 | # License should certainly be in the first 10 KB. |
50 | MAX_BYTES = 10000 |
51 | MAX_LINES = 50 |
52 | |
53 | LICENSE_LINES = [ |
54 | "This source code is licensed under both the BSD-style license (found in the", |
55 | "LICENSE file in the root directory of this source tree) and the GPLv2 (found", |
56 | "in the COPYING file in the root directory of this source tree).", |
57 | "You may select, at your option, one of the above-listed licenses.", |
58 | ] |
59 | |
60 | COPYRIGHT_EXCEPTIONS = { |
61 | # From zstdmt |
62 | "threading.c", |
63 | "threading.h", |
64 | # From divsufsort |
65 | "divsufsort.c", |
66 | "divsufsort.h", |
67 | } |
68 | |
69 | LICENSE_EXCEPTIONS = { |
70 | # From divsufsort |
71 | "divsufsort.c", |
72 | "divsufsort.h", |
73 | # License is slightly different because it references GitHub |
74 | "linux_zstd.h", |
75 | } |
76 | |
77 | |
78 | def valid_copyright(lines): |
79 | YEAR_REGEX = re.compile("\d\d\d\d|present") |
80 | for line in lines: |
81 | line = line.strip() |
82 | if "Copyright" not in line: |
83 | continue |
84 | if "present" in line: |
85 | return (False, f"Copyright line '{line}' contains 'present'!") |
86 | if "Meta Platforms, Inc" not in line: |
87 | return (False, f"Copyright line '{line}' does not contain 'Meta Platforms, Inc'") |
88 | year = YEAR_REGEX.search(line) |
89 | if year is not None: |
90 | return (False, f"Copyright line '{line}' contains {year.group(0)}; it should be yearless") |
91 | if " (c) " not in line: |
92 | return (False, f"Copyright line '{line}' does not contain ' (c) '!") |
93 | return (True, "") |
94 | return (False, "Copyright not found!") |
95 | |
96 | |
97 | def valid_license(lines): |
98 | for b in range(len(lines)): |
99 | if LICENSE_LINES[0] not in lines[b]: |
100 | continue |
101 | for l in range(len(LICENSE_LINES)): |
102 | if LICENSE_LINES[l] not in lines[b + l]: |
103 | message = f"""Invalid license line found starting on line {b + l}! |
104 | Expected: '{LICENSE_LINES[l]}' |
105 | Actual: '{lines[b + l]}'""" |
106 | return (False, message) |
107 | return (True, "") |
108 | return (False, "License not found!") |
109 | |
110 | |
111 | def valid_file(filename): |
112 | with open(filename, "r") as f: |
113 | lines = f.readlines(MAX_BYTES) |
114 | lines = lines[:min(len(lines), MAX_LINES)] |
115 | |
116 | ok = True |
117 | if os.path.basename(filename) not in COPYRIGHT_EXCEPTIONS: |
118 | c_ok, c_msg = valid_copyright(lines) |
119 | if not c_ok: |
120 | print(f"{filename}: {c_msg}", file=sys.stderr) |
121 | ok = False |
122 | if os.path.basename(filename) not in LICENSE_EXCEPTIONS: |
123 | l_ok, l_msg = valid_license(lines) |
124 | if not l_ok: |
125 | print(f"{filename}: {l_msg}", file=sys.stderr) |
126 | ok = False |
127 | return ok |
128 | |
129 | |
130 | def exclude(filename): |
131 | for x in EXCLUDES: |
132 | if filename.startswith(x): |
133 | return True |
134 | return False |
135 | |
136 | def main(): |
137 | invalid_files = [] |
138 | for directory in DIRS: |
139 | for suffix in SUFFIXES: |
140 | files = set(glob.glob(f"{directory}/**/*{suffix}", recursive=True)) |
141 | for filename in files: |
142 | if exclude(filename): |
143 | continue |
144 | if not valid_file(filename): |
145 | invalid_files.append(filename) |
146 | if len(invalid_files) > 0: |
147 | print("Fail!", file=sys.stderr) |
148 | for f in invalid_files: |
149 | print(f) |
150 | return 1 |
151 | else: |
152 | print("Pass!", file=sys.stderr) |
153 | return 0 |
154 | |
155 | if __name__ == "__main__": |
156 | sys.exit(main()) |