1#!/usr/bin/env python3 2# SPDX-License-Identifier: GPL-2.0 3# Copyright Thomas Gleixner <[email protected]> 4 5from argparse import ArgumentParser 6from ply import lex, yacc 7import locale 8import traceback 9import sys 10import git 11import re 12import os 13 14class ParserException(Exception): 15 def __init__(self, tok, txt): 16 self.tok = tok 17 self.txt = txt 18 19class SPDXException(Exception): 20 def __init__(self, el, txt): 21 self.el = el 22 self.txt = txt 23 24class SPDXdata(object): 25 def __init__(self): 26 self.license_files = 0 27 self.exception_files = 0 28 self.licenses = [ ] 29 self.exceptions = { } 30 31class dirinfo(object): 32 def __init__(self): 33 self.missing = 0 34 self.total = 0 35 self.files = [] 36 37 def update(self, fname, basedir, miss): 38 self.total += 1 39 self.missing += miss 40 if miss: 41 fname = './' + fname 42 bdir = os.path.dirname(fname) 43 if bdir == basedir.rstrip('/'): 44 self.files.append(fname) 45 46# Read the spdx data from the LICENSES directory 47def read_spdxdata(repo): 48 49 # The subdirectories of LICENSES in the kernel source 50 # Note: exceptions needs to be parsed as last directory. 51 license_dirs = [ "preferred", "dual", "deprecated", "exceptions" ] 52 lictree = repo.head.commit.tree['LICENSES'] 53 54 spdx = SPDXdata() 55 56 for d in license_dirs: 57 for el in lictree[d].traverse(): 58 if not os.path.isfile(el.path): 59 continue 60 61 exception = None 62 for l in open(el.path, encoding="utf-8").readlines(): 63 if l.startswith('Valid-License-Identifier:'): 64 lid = l.split(':')[1].strip().upper() 65 if lid in spdx.licenses: 66 raise SPDXException(el, 'Duplicate License Identifier: %s' %lid) 67 else: 68 spdx.licenses.append(lid) 69 70 elif l.startswith('SPDX-Exception-Identifier:'): 71 exception = l.split(':')[1].strip().upper() 72 spdx.exceptions[exception] = [] 73 74 elif l.startswith('SPDX-Licenses:'): 75 for lic in l.split(':')[1].upper().strip().replace(' ', '').replace('\t', '').split(','): 76 if not lic in spdx.licenses: 77 raise SPDXException(None, 'Exception %s missing license %s' %(exception, lic)) 78 spdx.exceptions[exception].append(lic) 79 80 elif l.startswith("License-Text:"): 81 if exception: 82 if not len(spdx.exceptions[exception]): 83 raise SPDXException(el, 'Exception %s is missing SPDX-Licenses' %exception) 84 spdx.exception_files += 1 85 else: 86 spdx.license_files += 1 87 break 88 return spdx 89 90class id_parser(object): 91 92 reserved = [ 'AND', 'OR', 'WITH' ] 93 tokens = [ 'LPAR', 'RPAR', 'ID', 'EXC' ] + reserved 94 95 precedence = ( ('nonassoc', 'AND', 'OR'), ) 96 97 t_ignore = ' \t' 98 99 def __init__(self, spdx): 100 self.spdx = spdx 101 self.lasttok = None 102 self.lastid = None 103 self.lexer = lex.lex(module = self, reflags = re.UNICODE) 104 # Initialize the parser. No debug file and no parser rules stored on disk 105 # The rules are small enough to be generated on the fly 106 self.parser = yacc.yacc(module = self, write_tables = False, debug = False) 107 self.lines_checked = 0 108 self.checked = 0 109 self.spdx_valid = 0 110 self.spdx_errors = 0 111 self.spdx_dirs = {} 112 self.dirdepth = -1 113 self.basedir = '.' 114 self.curline = 0 115 self.deepest = 0 116 117 def set_dirinfo(self, basedir, dirdepth): 118 if dirdepth >= 0: 119 self.basedir = basedir 120 bdir = basedir.lstrip('./').rstrip('/') 121 if bdir != '': 122 parts = bdir.split('/') 123 else: 124 parts = [] 125 self.dirdepth = dirdepth + len(parts) 126 127 # Validate License and Exception IDs 128 def validate(self, tok): 129 id = tok.value.upper() 130 if tok.type == 'ID': 131 if not id in self.spdx.licenses: 132 raise ParserException(tok, 'Invalid License ID') 133 self.lastid = id 134 elif tok.type == 'EXC': 135 if id not in self.spdx.exceptions: 136 raise ParserException(tok, 'Invalid Exception ID') 137 if self.lastid not in self.spdx.exceptions[id]: 138 raise ParserException(tok, 'Exception not valid for license %s' %self.lastid) 139 self.lastid = None 140 elif tok.type != 'WITH': 141 self.lastid = None 142 143 # Lexer functions 144 def t_RPAR(self, tok): 145 r'\)' 146 self.lasttok = tok.type 147 return tok 148 149 def t_LPAR(self, tok): 150 r'\(' 151 self.lasttok = tok.type 152 return tok 153 154 def t_ID(self, tok): 155 r'[A-Za-z.0-9\-+]+' 156 157 if self.lasttok == 'EXC': 158 print(tok) 159 raise ParserException(tok, 'Missing parentheses') 160 161 tok.value = tok.value.strip() 162 val = tok.value.upper() 163 164 if val in self.reserved: 165 tok.type = val 166 elif self.lasttok == 'WITH': 167 tok.type = 'EXC' 168 169 self.lasttok = tok.type 170 self.validate(tok) 171 return tok 172 173 def t_error(self, tok): 174 raise ParserException(tok, 'Invalid token') 175 176 def p_expr(self, p): 177 '''expr : ID 178 | ID WITH EXC 179 | expr AND expr 180 | expr OR expr 181 | LPAR expr RPAR''' 182 pass 183 184 def p_error(self, p): 185 if not p: 186 raise ParserException(None, 'Unfinished license expression') 187 else: 188 raise ParserException(p, 'Syntax error') 189 190 def parse(self, expr): 191 self.lasttok = None 192 self.lastid = None 193 self.parser.parse(expr, lexer = self.lexer) 194 195 def parse_lines(self, fd, maxlines, fname): 196 self.checked += 1 197 self.curline = 0 198 fail = 1 199 try: 200 for line in fd: 201 line = line.decode(locale.getpreferredencoding(False), errors='ignore') 202 self.curline += 1 203 if self.curline > maxlines: 204 break 205 self.lines_checked += 1 206 if line.find("SPDX-License-Identifier:") < 0: 207 continue 208 expr = line.split(':')[1].strip() 209 # Remove trailing comment closure 210 if line.strip().endswith('*/'): 211 expr = expr.rstrip('*/').strip() 212 # Remove trailing xml comment closure 213 if line.strip().endswith('-->'): 214 expr = expr.rstrip('-->').strip() 215 # Special case for SH magic boot code files 216 if line.startswith('LIST \"'): 217 expr = expr.rstrip('\"').strip() 218 self.parse(expr) 219 self.spdx_valid += 1 220 # 221 # Should we check for more SPDX ids in the same file and 222 # complain if there are any? 223 # 224 fail = 0 225 break 226 227 except ParserException as pe: 228 if pe.tok: 229 col = line.find(expr) + pe.tok.lexpos 230 tok = pe.tok.value 231 sys.stdout.write('%s: %d:%d %s: %s\n' %(fname, self.curline, col, pe.txt, tok)) 232 else: 233 sys.stdout.write('%s: %d:0 %s\n' %(fname, self.curline, pe.txt)) 234 self.spdx_errors += 1 235 236 if fname == '-': 237 return 238 239 base = os.path.dirname(fname) 240 if self.dirdepth > 0: 241 parts = base.split('/') 242 i = 0 243 base = '.' 244 while i < self.dirdepth and i < len(parts) and len(parts[i]): 245 base += '/' + parts[i] 246 i += 1 247 elif self.dirdepth == 0: 248 base = self.basedir 249 else: 250 base = './' + base.rstrip('/') 251 base += '/' 252 253 di = self.spdx_dirs.get(base, dirinfo()) 254 di.update(fname, base, fail) 255 self.spdx_dirs[base] = di 256 257def scan_git_tree(tree, basedir, dirdepth): 258 parser.set_dirinfo(basedir, dirdepth) 259 for el in tree.traverse(): 260 # Exclude stuff which would make pointless noise 261 # FIXME: Put this somewhere more sensible 262 if el.path.startswith("LICENSES"): 263 continue 264 if el.path.find("license-rules.rst") >= 0: 265 continue 266 if not os.path.isfile(el.path): 267 continue 268 with open(el.path, 'rb') as fd: 269 parser.parse_lines(fd, args.maxlines, el.path) 270 271def scan_git_subtree(tree, path, dirdepth): 272 for p in path.strip('/').split('/'): 273 tree = tree[p] 274 scan_git_tree(tree, path.strip('/'), dirdepth) 275 276if __name__ == '__main__': 277 278 ap = ArgumentParser(description='SPDX expression checker') 279 ap.add_argument('path', nargs='*', help='Check path or file. If not given full git tree scan. For stdin use "-"') 280 ap.add_argument('-d', '--dirs', action='store_true', 281 help='Show [sub]directory statistics.') 282 ap.add_argument('-D', '--depth', type=int, default=-1, 283 help='Directory depth for -d statistics. Default: unlimited') 284 ap.add_argument('-f', '--files', action='store_true', 285 help='Show files without SPDX.') 286 ap.add_argument('-m', '--maxlines', type=int, default=15, 287 help='Maximum number of lines to scan in a file. Default 15') 288 ap.add_argument('-v', '--verbose', action='store_true', help='Verbose statistics output') 289 args = ap.parse_args() 290 291 # Sanity check path arguments 292 if '-' in args.path and len(args.path) > 1: 293 sys.stderr.write('stdin input "-" must be the only path argument\n') 294 sys.exit(1) 295 296 try: 297 # Use git to get the valid license expressions 298 repo = git.Repo(os.getcwd()) 299 assert not repo.bare 300 301 # Initialize SPDX data 302 spdx = read_spdxdata(repo) 303 304 # Initialize the parser 305 parser = id_parser(spdx) 306 307 except SPDXException as se: 308 if se.el: 309 sys.stderr.write('%s: %s\n' %(se.el.path, se.txt)) 310 else: 311 sys.stderr.write('%s\n' %se.txt) 312 sys.exit(1) 313 314 except Exception as ex: 315 sys.stderr.write('FAIL: %s\n' %ex) 316 sys.stderr.write('%s\n' %traceback.format_exc()) 317 sys.exit(1) 318 319 try: 320 if len(args.path) and args.path[0] == '-': 321 stdin = os.fdopen(sys.stdin.fileno(), 'rb') 322 parser.parse_lines(stdin, args.maxlines, '-') 323 else: 324 if args.path: 325 for p in args.path: 326 if os.path.isfile(p): 327 parser.parse_lines(open(p, 'rb'), args.maxlines, p) 328 elif os.path.isdir(p): 329 scan_git_subtree(repo.head.reference.commit.tree, p, 330 args.depth) 331 else: 332 sys.stderr.write('path %s does not exist\n' %p) 333 sys.exit(1) 334 else: 335 # Full git tree scan 336 scan_git_tree(repo.head.commit.tree, '.', args.depth) 337 338 ndirs = len(parser.spdx_dirs) 339 dirsok = 0 340 if ndirs: 341 for di in parser.spdx_dirs.values(): 342 if not di.missing: 343 dirsok += 1 344 345 if args.verbose: 346 sys.stderr.write('\n') 347 sys.stderr.write('License files: %12d\n' %spdx.license_files) 348 sys.stderr.write('Exception files: %12d\n' %spdx.exception_files) 349 sys.stderr.write('License IDs %12d\n' %len(spdx.licenses)) 350 sys.stderr.write('Exception IDs %12d\n' %len(spdx.exceptions)) 351 sys.stderr.write('\n') 352 sys.stderr.write('Files checked: %12d\n' %parser.checked) 353 sys.stderr.write('Lines checked: %12d\n' %parser.lines_checked) 354 if parser.checked: 355 pc = int(100 * parser.spdx_valid / parser.checked) 356 sys.stderr.write('Files with SPDX: %12d %3d%%\n' %(parser.spdx_valid, pc)) 357 sys.stderr.write('Files with errors: %12d\n' %parser.spdx_errors) 358 if ndirs: 359 sys.stderr.write('\n') 360 sys.stderr.write('Directories accounted: %8d\n' %ndirs) 361 pc = int(100 * dirsok / ndirs) 362 sys.stderr.write('Directories complete: %8d %3d%%\n' %(dirsok, pc)) 363 364 if ndirs and ndirs != dirsok and args.dirs: 365 if args.verbose: 366 sys.stderr.write('\n') 367 sys.stderr.write('Incomplete directories: SPDX in Files\n') 368 for f in sorted(parser.spdx_dirs.keys()): 369 di = parser.spdx_dirs[f] 370 if di.missing: 371 valid = di.total - di.missing 372 pc = int(100 * valid / di.total) 373 sys.stderr.write(' %-80s: %5d of %5d %3d%%\n' %(f, valid, di.total, pc)) 374 375 if ndirs and ndirs != dirsok and args.files: 376 if args.verbose or args.dirs: 377 sys.stderr.write('\n') 378 sys.stderr.write('Files without SPDX:\n') 379 for f in sorted(parser.spdx_dirs.keys()): 380 di = parser.spdx_dirs[f] 381 for f in sorted(di.files): 382 sys.stderr.write(' %s\n' %f) 383 384 sys.exit(0) 385 386 except Exception as ex: 387 sys.stderr.write('FAIL: %s\n' %ex) 388 sys.stderr.write('%s\n' %traceback.format_exc()) 389 sys.exit(1) 390