check-c-globals.py 12.5 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137

from collections import namedtuple
import glob
import os.path
import re
import shutil
import sys
import subprocess


VERBOSITY = 2

C_GLOBALS_DIR = os.path.abspath(os.path.dirname(__file__))
TOOLS_DIR = os.path.dirname(C_GLOBALS_DIR)
ROOT_DIR = os.path.dirname(TOOLS_DIR)
GLOBALS_FILE = os.path.join(C_GLOBALS_DIR, 'ignored-globals.txt')

SOURCE_DIRS = ['Include', 'Objects', 'Modules', 'Parser', 'Python']

CAPI_REGEX = re.compile(r'^ *PyAPI_DATA\([^)]*\) \W*(_?Py\w+(?:, \w+)*\w).*;.*$')


IGNORED_VARS = {
        '_DYNAMIC',
        '_GLOBAL_OFFSET_TABLE_',
        '__JCR_LIST__',
        '__JCR_END__',
        '__TMC_END__',
        '__bss_start',
        '__data_start',
        '__dso_handle',
        '_edata',
        '_end',
        }


def find_capi_vars(root):
    capi_vars = {}
    for dirname in SOURCE_DIRS:
        for filename in glob.glob(os.path.join(ROOT_DIR, dirname, '**/*.[hc]'),
                                  recursive=True):
            with open(filename) as file:
                for name in _find_capi_vars(file):
                    if name in capi_vars:
                        assert not filename.endswith('.c')
                        assert capi_vars[name].endswith('.c')
                    capi_vars[name] = filename
    return capi_vars


def _find_capi_vars(lines):
    for line in lines:
        if not line.startswith('PyAPI_DATA'):
            continue
        assert '{' not in line
        match = CAPI_REGEX.match(line)
        assert match
        names, = match.groups()
        for name in names.split(', '):
            yield name


def _read_global_names(filename):
    # These variables are shared between all interpreters in the process.
    with open(filename) as file:
        return {line.partition('#')[0].strip()
                for line in file
                if line.strip() and not line.startswith('#')}


def _is_global_var(name, globalnames):
    if _is_autogen_var(name):
        return True
    if _is_type_var(name):
        return True
    if _is_module(name):
        return True
    if _is_exception(name):
        return True
    if _is_compiler(name):
        return True
    return name in globalnames


def _is_autogen_var(name):
    return (
        name.startswith('PyId_') or
        '.' in name or
        # Objects/typeobject.c
        name.startswith('op_id.') or
        name.startswith('rop_id.') or
        # Python/graminit.c
        name.startswith('arcs_') or
        name.startswith('states_')
        )


def _is_type_var(name):
    if name.endswith(('Type', '_Type', '_type')):  # XXX Always a static type?
        return True
    if name.endswith('_desc'):  # for structseq types
        return True
    return (
        name.startswith('doc_') or
        name.endswith(('_doc', '__doc__', '_docstring')) or
        name.endswith('_methods') or
        name.endswith('_fields') or
        name.endswith(('_memberlist', '_members')) or
        name.endswith('_slots') or
        name.endswith(('_getset', '_getsets', '_getsetlist')) or
        name.endswith('_as_mapping') or
        name.endswith('_as_number') or
        name.endswith('_as_sequence') or
        name.endswith('_as_buffer') or
        name.endswith('_as_async')
        )


def _is_module(name):
    if name.endswith(('_functions', 'Methods', '_Methods')):
        return True
    if name == 'module_def':
        return True
    if name == 'initialized':
        return True
    return name.endswith(('module', '_Module'))


def _is_exception(name):
    # Other vars are enumerated in globals-core.txt.
    if not name.startswith(('PyExc_', '_PyExc_')):
        return False
    return name.endswith(('Error', 'Warning'))


def _is_compiler(name):
    return (
138
        # Python/Python-ast.c
139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446
        name.endswith('_type') or
        name.endswith('_singleton') or
        name.endswith('_attributes')
        )


class Var(namedtuple('Var', 'name kind scope capi filename')):

    @classmethod
    def parse_nm(cls, line, expected, ignored, capi_vars, globalnames):
        _, _, line = line.partition(' ')  # strip off the address
        line = line.strip()
        kind, _, line = line.partition(' ')
        if kind in ignored or ():
            return None
        elif kind not in expected or ():
            raise RuntimeError('unsupported NM type {!r}'.format(kind))

        name, _, filename = line.partition('\t')
        name = name.strip()
        if _is_autogen_var(name):
            return None
        if _is_global_var(name, globalnames):
            scope = 'global'
        else:
            scope = None
        capi = (name in capi_vars or ())
        if filename:
            filename = os.path.relpath(filename.partition(':')[0])
        return cls(name, kind, scope, capi, filename or '~???~')

    @property
    def external(self):
        return self.kind.isupper()


def find_vars(root, globals_filename=GLOBALS_FILE):
    python = os.path.join(root, 'python')
    if not os.path.exists(python):
        raise RuntimeError('python binary missing (need to build it first?)')
    capi_vars = find_capi_vars(root)
    globalnames = _read_global_names(globals_filename)

    nm = shutil.which('nm')
    if nm is None:
        # XXX Use dumpbin.exe /SYMBOLS on Windows.
        raise NotImplementedError
    else:
        yield from (var
                    for var in _find_var_symbols(python, nm, capi_vars,
                                                 globalnames)
                    if var.name not in IGNORED_VARS)


NM_FUNCS = set('Tt')
NM_PUBLIC_VARS = set('BD')
NM_PRIVATE_VARS = set('bd')
NM_VARS = NM_PUBLIC_VARS | NM_PRIVATE_VARS
NM_DATA = set('Rr')
NM_OTHER = set('ACGgiINpSsuUVvWw-?')
NM_IGNORED = NM_FUNCS | NM_DATA | NM_OTHER


def _find_var_symbols(python, nm, capi_vars, globalnames):
    args = [nm,
            '--line-numbers',
            python]
    out = subprocess.check_output(args)
    for line in out.decode('utf-8').splitlines():
        var = Var.parse_nm(line, NM_VARS, NM_IGNORED, capi_vars, globalnames)
        if var is None:
            continue
        yield var


#######################################

class Filter(namedtuple('Filter', 'name op value action')):

    @classmethod
    def parse(cls, raw):
        action = '+'
        if raw.startswith(('+', '-')):
            action = raw[0]
            raw = raw[1:]
        # XXX Support < and >?
        name, op, value = raw.partition('=')
        return cls(name, op, value, action)

    def check(self, var):
        value = getattr(var, self.name, None)
        if not self.op:
            matched = bool(value)
        elif self.op == '=':
            matched = (value == self.value)
        else:
            raise NotImplementedError

        if self.action == '+':
            return matched
        elif self.action == '-':
            return not matched
        else:
            raise NotImplementedError


def filter_var(var, filters):
    for filter in filters:
        if not filter.check(var):
            return False
    return True


def make_sort_key(spec):
    columns = [(col.strip('_'), '_' if col.startswith('_') else '')
               for col in spec]
    def sort_key(var):
        return tuple(getattr(var, col).lstrip(prefix)
                     for col, prefix in columns)
    return sort_key


def make_groups(allvars, spec):
    group = spec
    groups = {}
    for var in allvars:
        value = getattr(var, group)
        key = '{}: {}'.format(group, value)
        try:
            groupvars = groups[key]
        except KeyError:
            groupvars = groups[key] = []
        groupvars.append(var)
    return groups


def format_groups(groups, columns, fmts, widths):
    for group in sorted(groups):
        groupvars = groups[group]
        yield '', 0
        yield '  # {}'.format(group), 0
        yield from format_vars(groupvars, columns, fmts, widths)


def format_vars(allvars, columns, fmts, widths):
    fmt = ' '.join(fmts[col] for col in columns)
    fmt = ' ' + fmt.replace(' ', '   ') + ' '  # for div margin
    header = fmt.replace(':', ':^').format(*(col.upper() for col in columns))
    yield header, 0
    div = ' '.join('-'*(widths[col]+2) for col in columns)
    yield div, 0
    for var in allvars:
        values = (getattr(var, col) for col in columns)
        row = fmt.format(*('X' if val is True else val or ''
                           for val in values))
        yield row, 1
    yield div, 0


#######################################

COLUMNS = 'name,external,capi,scope,filename'
COLUMN_NAMES = COLUMNS.split(',')

COLUMN_WIDTHS = {col: len(col)
                 for col in COLUMN_NAMES}
COLUMN_WIDTHS.update({
        'name': 50,
        'scope': 7,
        'filename': 40,
        })
COLUMN_FORMATS = {col: '{:%s}' % width
                  for col, width in COLUMN_WIDTHS.items()}
for col in COLUMN_FORMATS:
    if COLUMN_WIDTHS[col] == len(col):
        COLUMN_FORMATS[col] = COLUMN_FORMATS[col].replace(':', ':^')


def _parse_filters_arg(raw, error):
    filters = []
    for value in raw.split(','):
        value=value.strip()
        if not value:
            continue
        try:
            filter = Filter.parse(value)
            if filter.name not in COLUMN_NAMES:
                raise Exception('unsupported column {!r}'.format(filter.name))
        except Exception as e:
            error('bad filter {!r}: {}'.format(raw, e))
        filters.append(filter)
    return filters


def _parse_columns_arg(raw, error):
    columns = raw.split(',')
    for column in columns:
        if column not in COLUMN_NAMES:
            error('unsupported column {!r}'.format(column))
    return columns


def _parse_sort_arg(raw, error):
    sort = raw.split(',')
    for column in sort:
        if column.lstrip('_') not in COLUMN_NAMES:
            error('unsupported column {!r}'.format(column))
    return sort


def _parse_group_arg(raw, error):
    if not raw:
        return raw
    group = raw
    if group not in COLUMN_NAMES:
        error('unsupported column {!r}'.format(group))
    if group != 'filename':
        error('unsupported group {!r}'.format(group))
    return group


def parse_args(argv=None):
    if argv is None:
        argv = sys.argv[1:]

    import argparse
    parser = argparse.ArgumentParser()

    parser.add_argument('-v', '--verbose', action='count', default=0)
    parser.add_argument('-q', '--quiet', action='count', default=0)

    parser.add_argument('--filters', default='-scope',
                        help='[[-]<COLUMN>[=<GLOB>]] ...')

    parser.add_argument('--columns', default=COLUMNS,
                        help='a comma-separated list of columns to show')
    parser.add_argument('--sort', default='filename,_name',
                        help='a comma-separated list of columns to sort')
    parser.add_argument('--group',
                        help='group by the given column name (- to not group)')

    parser.add_argument('--rc-on-match', dest='rc', type=int)

    parser.add_argument('filename', nargs='?', default=GLOBALS_FILE)

    args = parser.parse_args(argv)

    verbose = vars(args).pop('verbose', 0)
    quiet = vars(args).pop('quiet', 0)
    args.verbosity = max(0, VERBOSITY + verbose - quiet)

    if args.sort.startswith('filename') and not args.group:
        args.group = 'filename'

    if args.rc is None:
        if '-scope=core' in args.filters or 'core' not in args.filters:
            args.rc = 0
        else:
            args.rc = 1

    args.filters = _parse_filters_arg(args.filters, parser.error)
    args.columns = _parse_columns_arg(args.columns, parser.error)
    args.sort = _parse_sort_arg(args.sort, parser.error)
    args.group = _parse_group_arg(args.group, parser.error)

    return args


def main(root=ROOT_DIR, filename=GLOBALS_FILE,
         filters=None, columns=COLUMN_NAMES, sort=None, group=None,
         verbosity=VERBOSITY, rc=1):

    log = lambda msg: ...
    if verbosity >= 2:
        log = lambda msg: print(msg)

    allvars = (var
               for var in find_vars(root, filename)
               if filter_var(var, filters))
    if sort:
        allvars = sorted(allvars, key=make_sort_key(sort))

    if group:
        try:
            columns.remove(group)
        except ValueError:
            pass
        grouped = make_groups(allvars, group)
        lines = format_groups(grouped, columns, COLUMN_FORMATS, COLUMN_WIDTHS)
    else:
        lines = format_vars(allvars, columns, COLUMN_FORMATS, COLUMN_WIDTHS)

    total = 0
    for line, count in lines:
        total += count
        log(line)
    log('\ntotal: {}'.format(total))

    if total and rc:
        print('ERROR: found unsafe globals', file=sys.stderr)
        return rc
    return 0


if __name__ == '__main__':
    args = parse_args()
    sys.exit(
            main(**vars(args)))