2 # SPDX-License-Identifier: GPL-2.0
3 # Copyright Thomas Gleixner <tglx@linutronix.de>
5 from argparse import ArgumentParser
6 from ply import lex, yacc
14 class ParserException(Exception):
15 def __init__(self, tok, txt):
19 class SPDXException(Exception):
20 def __init__(self, el, txt):
24 class SPDXdata(object):
26 self.license_files = 0
27 self.exception_files = 0
31 # Read the spdx data from the LICENSES directory
32 def read_spdxdata(repo):
34 # The subdirectories of LICENSES in the kernel source
35 license_dirs = [ "preferred", "other", "exceptions" ]
36 lictree = repo.head.commit.tree['LICENSES']
40 for d in license_dirs:
41 for el in lictree[d].traverse():
42 if not os.path.isfile(el.path):
46 for l in open(el.path).readlines():
47 if l.startswith('Valid-License-Identifier:'):
48 lid = l.split(':')[1].strip().upper()
49 if lid in spdx.licenses:
50 raise SPDXException(el, 'Duplicate License Identifier: %s' %lid)
52 spdx.licenses.append(lid)
54 elif l.startswith('SPDX-Exception-Identifier:'):
55 exception = l.split(':')[1].strip().upper()
56 spdx.exceptions[exception] = []
58 elif l.startswith('SPDX-Licenses:'):
59 for lic in l.split(':')[1].upper().strip().replace(' ', '').replace('\t', '').split(','):
60 if not lic in spdx.licenses:
61 raise SPDXException(None, 'Exception %s missing license %s' %(ex, lic))
62 spdx.exceptions[exception].append(lic)
64 elif l.startswith("License-Text:"):
66 if not len(spdx.exceptions[exception]):
67 raise SPDXException(el, 'Exception %s is missing SPDX-Licenses' %excid)
68 spdx.exception_files += 1
70 spdx.license_files += 1
74 class id_parser(object):
76 reserved = [ 'AND', 'OR', 'WITH' ]
77 tokens = [ 'LPAR', 'RPAR', 'ID', 'EXC' ] + reserved
79 precedence = ( ('nonassoc', 'AND', 'OR'), )
83 def __init__(self, spdx):
87 self.lexer = lex.lex(module = self, reflags = re.UNICODE)
88 # Initialize the parser. No debug file and no parser rules stored on disk
89 # The rules are small enough to be generated on the fly
90 self.parser = yacc.yacc(module = self, write_tables = False, debug = False)
91 self.lines_checked = 0
98 # Validate License and Exception IDs
99 def validate(self, tok):
100 id = tok.value.upper()
102 if not id in self.spdx.licenses:
103 raise ParserException(tok, 'Invalid License ID')
105 elif tok.type == 'EXC':
106 if id not in self.spdx.exceptions:
107 raise ParserException(tok, 'Invalid Exception ID')
108 if self.lastid not in self.spdx.exceptions[id]:
109 raise ParserException(tok, 'Exception not valid for license %s' %self.lastid)
111 elif tok.type != 'WITH':
115 def t_RPAR(self, tok):
117 self.lasttok = tok.type
120 def t_LPAR(self, tok):
122 self.lasttok = tok.type
128 if self.lasttok == 'EXC':
130 raise ParserException(tok, 'Missing parentheses')
132 tok.value = tok.value.strip()
133 val = tok.value.upper()
135 if val in self.reserved:
137 elif self.lasttok == 'WITH':
140 self.lasttok = tok.type
144 def t_error(self, tok):
145 raise ParserException(tok, 'Invalid token')
155 def p_error(self, p):
157 raise ParserException(None, 'Unfinished license expression')
159 raise ParserException(p, 'Syntax error')
161 def parse(self, expr):
164 self.parser.parse(expr, lexer = self.lexer)
166 def parse_lines(self, fd, maxlines, fname):
171 line = line.decode(locale.getpreferredencoding(False), errors='ignore')
173 if self.curline > maxlines:
175 self.lines_checked += 1
176 if line.find("SPDX-License-Identifier:") < 0:
178 expr = line.split(':')[1].replace('*/', '').strip()
182 # Should we check for more SPDX ids in the same file and
183 # complain if there are any?
187 except ParserException as pe:
189 col = line.find(expr) + pe.tok.lexpos
191 sys.stdout.write('%s: %d:%d %s: %s\n' %(fname, self.curline, col, pe.txt, tok))
193 sys.stdout.write('%s: %d:0 %s\n' %(fname, self.curline, col, pe.txt))
194 self.spdx_errors += 1
196 def scan_git_tree(tree):
197 for el in tree.traverse():
198 # Exclude stuff which would make pointless noise
199 # FIXME: Put this somewhere more sensible
200 if el.path.startswith("LICENSES"):
202 if el.path.find("license-rules.rst") >= 0:
204 if not os.path.isfile(el.path):
206 with open(el.path, 'rb') as fd:
207 parser.parse_lines(fd, args.maxlines, el.path)
209 def scan_git_subtree(tree, path):
210 for p in path.strip('/').split('/'):
214 if __name__ == '__main__':
216 ap = ArgumentParser(description='SPDX expression checker')
217 ap.add_argument('path', nargs='*', help='Check path or file. If not given full git tree scan. For stdin use "-"')
218 ap.add_argument('-m', '--maxlines', type=int, default=15,
219 help='Maximum number of lines to scan in a file. Default 15')
220 ap.add_argument('-v', '--verbose', action='store_true', help='Verbose statistics output')
221 args = ap.parse_args()
223 # Sanity check path arguments
224 if '-' in args.path and len(args.path) > 1:
225 sys.stderr.write('stdin input "-" must be the only path argument\n')
229 # Use git to get the valid license expressions
230 repo = git.Repo(os.getcwd())
233 # Initialize SPDX data
234 spdx = read_spdxdata(repo)
236 # Initilize the parser
237 parser = id_parser(spdx)
239 except SPDXException as se:
241 sys.stderr.write('%s: %s\n' %(se.el.path, se.txt))
243 sys.stderr.write('%s\n' %se.txt)
246 except Exception as ex:
247 sys.stderr.write('FAIL: %s\n' %ex)
248 sys.stderr.write('%s\n' %traceback.format_exc())
252 if len(args.path) and args.path[0] == '-':
253 parser.parse_lines(sys.stdin, args.maxlines, '-')
257 if os.path.isfile(p):
258 parser.parse_lines(open(p), args.maxlines, p)
259 elif os.path.isdir(p):
260 scan_git_subtree(repo.head.reference.commit.tree, p)
262 sys.stderr.write('path %s does not exist\n' %p)
266 scan_git_tree(repo.head.commit.tree)
269 sys.stderr.write('\n')
270 sys.stderr.write('License files: %12d\n' %spdx.license_files)
271 sys.stderr.write('Exception files: %12d\n' %spdx.exception_files)
272 sys.stderr.write('License IDs %12d\n' %len(spdx.licenses))
273 sys.stderr.write('Exception IDs %12d\n' %len(spdx.exceptions))
274 sys.stderr.write('\n')
275 sys.stderr.write('Files checked: %12d\n' %parser.checked)
276 sys.stderr.write('Lines checked: %12d\n' %parser.lines_checked)
277 sys.stderr.write('Files with SPDX: %12d\n' %parser.spdx_valid)
278 sys.stderr.write('Files with errors: %12d\n' %parser.spdx_errors)
282 except Exception as ex:
283 sys.stderr.write('FAIL: %s\n' %ex)
284 sys.stderr.write('%s\n' %traceback.format_exc())