| 1 |
import io |
| 2 |
import logging |
| 3 |
import os |
| 4 |
import os.path |
| 5 |
import re |
| 6 |
import sys |
| 7 |
|
| 8 |
from c_common import fsutil |
| 9 |
from c_common.logging import VERBOSITY, Printer |
| 10 |
from c_common.scriptutil import ( |
| 11 |
add_verbosity_cli, |
| 12 |
add_traceback_cli, |
| 13 |
add_sepval_cli, |
| 14 |
add_progress_cli, |
| 15 |
add_files_cli, |
| 16 |
add_commands_cli, |
| 17 |
process_args_by_key, |
| 18 |
configure_logger, |
| 19 |
get_prog, |
| 20 |
filter_filenames, |
| 21 |
) |
| 22 |
from c_parser.info import KIND |
| 23 |
from .match import filter_forward |
| 24 |
from . import ( |
| 25 |
analyze as _analyze, |
| 26 |
datafiles as _datafiles, |
| 27 |
check_all as _check_all, |
| 28 |
) |
| 29 |
|
| 30 |
|
| 31 |
KINDS = [ |
| 32 |
KIND.TYPEDEF, |
| 33 |
KIND.STRUCT, |
| 34 |
KIND.UNION, |
| 35 |
KIND.ENUM, |
| 36 |
KIND.FUNCTION, |
| 37 |
KIND.VARIABLE, |
| 38 |
KIND.STATEMENT, |
| 39 |
] |
| 40 |
|
| 41 |
logger = logging.getLogger(__name__) |
| 42 |
|
| 43 |
|
| 44 |
####################################### |
| 45 |
# table helpers |
| 46 |
|
| 47 |
TABLE_SECTIONS = { |
| 48 |
'types': ( |
| 49 |
['kind', 'name', 'data', 'file'], |
| 50 |
KIND.is_type_decl, |
| 51 |
(lambda v: (v.kind.value, v.filename or '', v.name)), |
| 52 |
), |
| 53 |
'typedefs': 'types', |
| 54 |
'structs': 'types', |
| 55 |
'unions': 'types', |
| 56 |
'enums': 'types', |
| 57 |
'functions': ( |
| 58 |
['name', 'data', 'file'], |
| 59 |
(lambda kind: kind is KIND.FUNCTION), |
| 60 |
(lambda v: (v.filename or '', v.name)), |
| 61 |
), |
| 62 |
'variables': ( |
| 63 |
['name', 'parent', 'data', 'file'], |
| 64 |
(lambda kind: kind is KIND.VARIABLE), |
| 65 |
(lambda v: (v.filename or '', str(v.parent) if v.parent else '', v.name)), |
| 66 |
), |
| 67 |
'statements': ( |
| 68 |
['file', 'parent', 'data'], |
| 69 |
(lambda kind: kind is KIND.STATEMENT), |
| 70 |
(lambda v: (v.filename or '', str(v.parent) if v.parent else '', v.name)), |
| 71 |
), |
| 72 |
KIND.TYPEDEF: 'typedefs', |
| 73 |
KIND.STRUCT: 'structs', |
| 74 |
KIND.UNION: 'unions', |
| 75 |
KIND.ENUM: 'enums', |
| 76 |
KIND.FUNCTION: 'functions', |
| 77 |
KIND.VARIABLE: 'variables', |
| 78 |
KIND.STATEMENT: 'statements', |
| 79 |
} |
| 80 |
|
| 81 |
|
| 82 |
def _render_table(items, columns, relroot=None): |
| 83 |
# XXX improve this |
| 84 |
header = '\t'.join(columns) |
| 85 |
div = '--------------------' |
| 86 |
yield header |
| 87 |
yield div |
| 88 |
total = 0 |
| 89 |
for item in items: |
| 90 |
rowdata = item.render_rowdata(columns) |
| 91 |
row = [rowdata[c] for c in columns] |
| 92 |
if relroot and 'file' in columns: |
| 93 |
index = columns.index('file') |
| 94 |
row[index] = os.path.relpath(row[index], relroot) |
| 95 |
yield '\t'.join(row) |
| 96 |
total += 1 |
| 97 |
yield div |
| 98 |
yield f'total: {total}' |
| 99 |
|
| 100 |
|
| 101 |
def build_section(name, groupitems, *, relroot=None): |
| 102 |
info = TABLE_SECTIONS[name] |
| 103 |
while type(info) is not tuple: |
| 104 |
if name in KINDS: |
| 105 |
name = info |
| 106 |
info = TABLE_SECTIONS[info] |
| 107 |
|
| 108 |
columns, match_kind, sortkey = info |
| 109 |
items = (v for v in groupitems if match_kind(v.kind)) |
| 110 |
items = sorted(items, key=sortkey) |
| 111 |
def render(): |
| 112 |
yield '' |
| 113 |
yield f'{name}:' |
| 114 |
yield '' |
| 115 |
for line in _render_table(items, columns, relroot): |
| 116 |
yield line |
| 117 |
return items, render |
| 118 |
|
| 119 |
|
| 120 |
####################################### |
| 121 |
# the checks |
| 122 |
|
| 123 |
CHECKS = { |
| 124 |
#'globals': _check_globals, |
| 125 |
} |
| 126 |
|
| 127 |
|
| 128 |
def add_checks_cli(parser, checks=None, *, add_flags=None): |
| 129 |
default = False |
| 130 |
if not checks: |
| 131 |
checks = list(CHECKS) |
| 132 |
default = True |
| 133 |
elif isinstance(checks, str): |
| 134 |
checks = [checks] |
| 135 |
if (add_flags is None and len(checks) > 1) or default: |
| 136 |
add_flags = True |
| 137 |
|
| 138 |
process_checks = add_sepval_cli(parser, '--check', 'checks', checks) |
| 139 |
if add_flags: |
| 140 |
for check in checks: |
| 141 |
parser.add_argument(f'--{check}', dest='checks', |
| 142 |
action='append_const', const=check) |
| 143 |
return [ |
| 144 |
process_checks, |
| 145 |
] |
| 146 |
|
| 147 |
|
| 148 |
def _get_check_handlers(fmt, printer, verbosity=VERBOSITY): |
| 149 |
div = None |
| 150 |
def handle_after(): |
| 151 |
pass |
| 152 |
if not fmt: |
| 153 |
div = '' |
| 154 |
def handle_failure(failure, data): |
| 155 |
data = repr(data) |
| 156 |
if verbosity >= 3: |
| 157 |
logger.info(f'failure: {failure}') |
| 158 |
logger.info(f'data: {data}') |
| 159 |
else: |
| 160 |
logger.warn(f'failure: {failure} (data: {data})') |
| 161 |
elif fmt == 'raw': |
| 162 |
def handle_failure(failure, data): |
| 163 |
print(f'{failure!r} {data!r}') |
| 164 |
elif fmt == 'brief': |
| 165 |
def handle_failure(failure, data): |
| 166 |
parent = data.parent or '' |
| 167 |
funcname = parent if isinstance(parent, str) else parent.name |
| 168 |
name = f'({funcname}).{data.name}' if funcname else data.name |
| 169 |
failure = failure.split('\t')[0] |
| 170 |
print(f'{data.filename}:{name} - {failure}') |
| 171 |
elif fmt == 'summary': |
| 172 |
def handle_failure(failure, data): |
| 173 |
print(_fmt_one_summary(data, failure)) |
| 174 |
elif fmt == 'full': |
| 175 |
div = '' |
| 176 |
def handle_failure(failure, data): |
| 177 |
name = data.shortkey if data.kind is KIND.VARIABLE else data.name |
| 178 |
parent = data.parent or '' |
| 179 |
funcname = parent if isinstance(parent, str) else parent.name |
| 180 |
known = 'yes' if data.is_known else '*** NO ***' |
| 181 |
print(f'{data.kind.value} {name!r} failed ({failure})') |
| 182 |
print(f' file: {data.filename}') |
| 183 |
print(f' func: {funcname or "-"}') |
| 184 |
print(f' name: {data.name}') |
| 185 |
print(f' data: ...') |
| 186 |
print(f' type unknown: {known}') |
| 187 |
else: |
| 188 |
if fmt in FORMATS: |
| 189 |
raise NotImplementedError(fmt) |
| 190 |
raise ValueError(f'unsupported fmt {fmt!r}') |
| 191 |
return handle_failure, handle_after, div |
| 192 |
|
| 193 |
|
| 194 |
####################################### |
| 195 |
# the formats |
| 196 |
|
| 197 |
def fmt_raw(analysis): |
| 198 |
for item in analysis: |
| 199 |
yield from item.render('raw') |
| 200 |
|
| 201 |
|
| 202 |
def fmt_brief(analysis): |
| 203 |
# XXX Support sorting. |
| 204 |
items = sorted(analysis) |
| 205 |
for kind in KINDS: |
| 206 |
if kind is KIND.STATEMENT: |
| 207 |
continue |
| 208 |
for item in items: |
| 209 |
if item.kind is not kind: |
| 210 |
continue |
| 211 |
yield from item.render('brief') |
| 212 |
yield f' total: {len(items)}' |
| 213 |
|
| 214 |
|
| 215 |
def fmt_summary(analysis): |
| 216 |
# XXX Support sorting and grouping. |
| 217 |
items = list(analysis) |
| 218 |
total = len(items) |
| 219 |
|
| 220 |
def section(name): |
| 221 |
_, render = build_section(name, items) |
| 222 |
yield from render() |
| 223 |
|
| 224 |
yield from section('types') |
| 225 |
yield from section('functions') |
| 226 |
yield from section('variables') |
| 227 |
yield from section('statements') |
| 228 |
|
| 229 |
yield '' |
| 230 |
# yield f'grand total: {len(supported) + len(unsupported)}' |
| 231 |
yield f'grand total: {total}' |
| 232 |
|
| 233 |
|
| 234 |
def _fmt_one_summary(item, extra=None): |
| 235 |
parent = item.parent or '' |
| 236 |
funcname = parent if isinstance(parent, str) else parent.name |
| 237 |
if extra: |
| 238 |
return f'{item.filename:35}\t{funcname or "-":35}\t{item.name:40}\t{extra}' |
| 239 |
else: |
| 240 |
return f'{item.filename:35}\t{funcname or "-":35}\t{item.name}' |
| 241 |
|
| 242 |
|
| 243 |
def fmt_full(analysis): |
| 244 |
# XXX Support sorting. |
| 245 |
items = sorted(analysis, key=lambda v: v.key) |
| 246 |
yield '' |
| 247 |
for item in items: |
| 248 |
yield from item.render('full') |
| 249 |
yield '' |
| 250 |
yield f'total: {len(items)}' |
| 251 |
|
| 252 |
|
| 253 |
FORMATS = { |
| 254 |
'raw': fmt_raw, |
| 255 |
'brief': fmt_brief, |
| 256 |
'summary': fmt_summary, |
| 257 |
'full': fmt_full, |
| 258 |
} |
| 259 |
|
| 260 |
|
| 261 |
def add_output_cli(parser, *, default='summary'): |
| 262 |
parser.add_argument('--format', dest='fmt', default=default, choices=tuple(FORMATS)) |
| 263 |
|
| 264 |
def process_args(args, *, argv=None): |
| 265 |
pass |
| 266 |
return process_args |
| 267 |
|
| 268 |
|
| 269 |
####################################### |
| 270 |
# the commands |
| 271 |
|
| 272 |
def _cli_check(parser, checks=None, **kwargs): |
| 273 |
if isinstance(checks, str): |
| 274 |
checks = [checks] |
| 275 |
if checks is False: |
| 276 |
process_checks = None |
| 277 |
elif checks is None: |
| 278 |
process_checks = add_checks_cli(parser) |
| 279 |
elif len(checks) == 1 and type(checks) is not dict and re.match(r'^<.*>$', checks[0]): |
| 280 |
check = checks[0][1:-1] |
| 281 |
def process_checks(args, *, argv=None): |
| 282 |
args.checks = [check] |
| 283 |
else: |
| 284 |
process_checks = add_checks_cli(parser, checks=checks) |
| 285 |
process_progress = add_progress_cli(parser) |
| 286 |
process_output = add_output_cli(parser, default=None) |
| 287 |
process_files = add_files_cli(parser, **kwargs) |
| 288 |
return [ |
| 289 |
process_checks, |
| 290 |
process_progress, |
| 291 |
process_output, |
| 292 |
process_files, |
| 293 |
] |
| 294 |
|
| 295 |
|
| 296 |
def cmd_check(filenames, *, |
| 297 |
checks=None, |
| 298 |
ignored=None, |
| 299 |
fmt=None, |
| 300 |
failfast=False, |
| 301 |
iter_filenames=None, |
| 302 |
relroot=fsutil.USE_CWD, |
| 303 |
track_progress=None, |
| 304 |
verbosity=VERBOSITY, |
| 305 |
_analyze=_analyze, |
| 306 |
_CHECKS=CHECKS, |
| 307 |
**kwargs |
| 308 |
): |
| 309 |
if not checks: |
| 310 |
checks = _CHECKS |
| 311 |
elif isinstance(checks, str): |
| 312 |
checks = [checks] |
| 313 |
checks = [_CHECKS[c] if isinstance(c, str) else c |
| 314 |
for c in checks] |
| 315 |
printer = Printer(verbosity) |
| 316 |
(handle_failure, handle_after, div |
| 317 |
) = _get_check_handlers(fmt, printer, verbosity) |
| 318 |
|
| 319 |
filenames, relroot = fsutil.fix_filenames(filenames, relroot=relroot) |
| 320 |
filenames = filter_filenames(filenames, iter_filenames, relroot) |
| 321 |
if track_progress: |
| 322 |
filenames = track_progress(filenames) |
| 323 |
|
| 324 |
logger.info('analyzing files...') |
| 325 |
analyzed = _analyze(filenames, **kwargs) |
| 326 |
analyzed.fix_filenames(relroot, normalize=False) |
| 327 |
decls = filter_forward(analyzed, markpublic=True) |
| 328 |
|
| 329 |
logger.info('checking analysis results...') |
| 330 |
failed = [] |
| 331 |
for data, failure in _check_all(decls, checks, failfast=failfast): |
| 332 |
if data is None: |
| 333 |
printer.info('stopping after one failure') |
| 334 |
break |
| 335 |
if div is not None and len(failed) > 0: |
| 336 |
printer.info(div) |
| 337 |
failed.append(data) |
| 338 |
handle_failure(failure, data) |
| 339 |
handle_after() |
| 340 |
|
| 341 |
printer.info('-------------------------') |
| 342 |
logger.info(f'total failures: {len(failed)}') |
| 343 |
logger.info('done checking') |
| 344 |
|
| 345 |
if fmt == 'summary': |
| 346 |
print('Categorized by storage:') |
| 347 |
print() |
| 348 |
from .match import group_by_storage |
| 349 |
grouped = group_by_storage(failed, ignore_non_match=False) |
| 350 |
for group, decls in grouped.items(): |
| 351 |
print() |
| 352 |
print(group) |
| 353 |
for decl in decls: |
| 354 |
print(' ', _fmt_one_summary(decl)) |
| 355 |
print(f'subtotal: {len(decls)}') |
| 356 |
|
| 357 |
if len(failed) > 0: |
| 358 |
sys.exit(len(failed)) |
| 359 |
|
| 360 |
|
| 361 |
def _cli_analyze(parser, **kwargs): |
| 362 |
process_progress = add_progress_cli(parser) |
| 363 |
process_output = add_output_cli(parser) |
| 364 |
process_files = add_files_cli(parser, **kwargs) |
| 365 |
return [ |
| 366 |
process_progress, |
| 367 |
process_output, |
| 368 |
process_files, |
| 369 |
] |
| 370 |
|
| 371 |
|
| 372 |
# XXX Support filtering by kind. |
| 373 |
def cmd_analyze(filenames, *, |
| 374 |
fmt=None, |
| 375 |
iter_filenames=None, |
| 376 |
relroot=fsutil.USE_CWD, |
| 377 |
track_progress=None, |
| 378 |
verbosity=None, |
| 379 |
_analyze=_analyze, |
| 380 |
formats=FORMATS, |
| 381 |
**kwargs |
| 382 |
): |
| 383 |
verbosity = verbosity if verbosity is not None else 3 |
| 384 |
|
| 385 |
try: |
| 386 |
do_fmt = formats[fmt] |
| 387 |
except KeyError: |
| 388 |
raise ValueError(f'unsupported fmt {fmt!r}') |
| 389 |
|
| 390 |
filenames, relroot = fsutil.fix_filenames(filenames, relroot=relroot) |
| 391 |
filenames = filter_filenames(filenames, iter_filenames, relroot) |
| 392 |
if track_progress: |
| 393 |
filenames = track_progress(filenames) |
| 394 |
|
| 395 |
logger.info('analyzing files...') |
| 396 |
analyzed = _analyze(filenames, **kwargs) |
| 397 |
analyzed.fix_filenames(relroot, normalize=False) |
| 398 |
decls = filter_forward(analyzed, markpublic=True) |
| 399 |
|
| 400 |
for line in do_fmt(decls): |
| 401 |
print(line) |
| 402 |
|
| 403 |
|
| 404 |
def _cli_data(parser, filenames=None, known=None): |
| 405 |
ArgumentParser = type(parser) |
| 406 |
common = ArgumentParser(add_help=False) |
| 407 |
# These flags will get processed by the top-level parse_args(). |
| 408 |
add_verbosity_cli(common) |
| 409 |
add_traceback_cli(common) |
| 410 |
|
| 411 |
subs = parser.add_subparsers(dest='datacmd') |
| 412 |
|
| 413 |
sub = subs.add_parser('show', parents=[common]) |
| 414 |
if known is None: |
| 415 |
sub.add_argument('--known', required=True) |
| 416 |
if filenames is None: |
| 417 |
sub.add_argument('filenames', metavar='FILE', nargs='+') |
| 418 |
|
| 419 |
sub = subs.add_parser('dump', parents=[common]) |
| 420 |
if known is None: |
| 421 |
sub.add_argument('--known') |
| 422 |
sub.add_argument('--show', action='store_true') |
| 423 |
process_progress = add_progress_cli(sub) |
| 424 |
|
| 425 |
sub = subs.add_parser('check', parents=[common]) |
| 426 |
if known is None: |
| 427 |
sub.add_argument('--known', required=True) |
| 428 |
|
| 429 |
def process_args(args, *, argv): |
| 430 |
if args.datacmd == 'dump': |
| 431 |
process_progress(args, argv) |
| 432 |
return process_args |
| 433 |
|
| 434 |
|
| 435 |
def cmd_data(datacmd, filenames, known=None, *, |
| 436 |
_analyze=_analyze, |
| 437 |
formats=FORMATS, |
| 438 |
extracolumns=None, |
| 439 |
relroot=fsutil.USE_CWD, |
| 440 |
track_progress=None, |
| 441 |
**kwargs |
| 442 |
): |
| 443 |
kwargs.pop('verbosity', None) |
| 444 |
usestdout = kwargs.pop('show', None) |
| 445 |
if datacmd == 'show': |
| 446 |
do_fmt = formats['summary'] |
| 447 |
if isinstance(known, str): |
| 448 |
known, _ = _datafiles.get_known(known, extracolumns, relroot) |
| 449 |
for line in do_fmt(known): |
| 450 |
print(line) |
| 451 |
elif datacmd == 'dump': |
| 452 |
filenames, relroot = fsutil.fix_filenames(filenames, relroot=relroot) |
| 453 |
if track_progress: |
| 454 |
filenames = track_progress(filenames) |
| 455 |
analyzed = _analyze(filenames, **kwargs) |
| 456 |
analyzed.fix_filenames(relroot, normalize=False) |
| 457 |
if known is None or usestdout: |
| 458 |
outfile = io.StringIO() |
| 459 |
_datafiles.write_known(analyzed, outfile, extracolumns, |
| 460 |
relroot=relroot) |
| 461 |
print(outfile.getvalue()) |
| 462 |
else: |
| 463 |
_datafiles.write_known(analyzed, known, extracolumns, |
| 464 |
relroot=relroot) |
| 465 |
elif datacmd == 'check': |
| 466 |
raise NotImplementedError(datacmd) |
| 467 |
else: |
| 468 |
raise ValueError(f'unsupported data command {datacmd!r}') |
| 469 |
|
| 470 |
|
| 471 |
COMMANDS = { |
| 472 |
'check': ( |
| 473 |
'analyze and fail if the given C source/header files have any problems', |
| 474 |
[_cli_check], |
| 475 |
cmd_check, |
| 476 |
), |
| 477 |
'analyze': ( |
| 478 |
'report on the state of the given C source/header files', |
| 479 |
[_cli_analyze], |
| 480 |
cmd_analyze, |
| 481 |
), |
| 482 |
'data': ( |
| 483 |
'check/manage local data (e.g. known types, ignored vars, caches)', |
| 484 |
[_cli_data], |
| 485 |
cmd_data, |
| 486 |
), |
| 487 |
} |
| 488 |
|
| 489 |
|
| 490 |
####################################### |
| 491 |
# the script |
| 492 |
|
| 493 |
def parse_args(argv=sys.argv[1:], prog=sys.argv[0], *, subset=None): |
| 494 |
import argparse |
| 495 |
parser = argparse.ArgumentParser( |
| 496 |
prog=prog or get_prog(), |
| 497 |
) |
| 498 |
|
| 499 |
processors = add_commands_cli( |
| 500 |
parser, |
| 501 |
commands={k: v[1] for k, v in COMMANDS.items()}, |
| 502 |
commonspecs=[ |
| 503 |
add_verbosity_cli, |
| 504 |
add_traceback_cli, |
| 505 |
], |
| 506 |
subset=subset, |
| 507 |
) |
| 508 |
|
| 509 |
args = parser.parse_args(argv) |
| 510 |
ns = vars(args) |
| 511 |
|
| 512 |
cmd = ns.pop('cmd') |
| 513 |
|
| 514 |
verbosity, traceback_cm = process_args_by_key( |
| 515 |
args, |
| 516 |
argv, |
| 517 |
processors[cmd], |
| 518 |
['verbosity', 'traceback_cm'], |
| 519 |
) |
| 520 |
# "verbosity" is sent to the commands, so we put it back. |
| 521 |
args.verbosity = verbosity |
| 522 |
|
| 523 |
return cmd, ns, verbosity, traceback_cm |
| 524 |
|
| 525 |
|
| 526 |
def main(cmd, cmd_kwargs): |
| 527 |
try: |
| 528 |
run_cmd = COMMANDS[cmd][0] |
| 529 |
except KeyError: |
| 530 |
raise ValueError(f'unsupported cmd {cmd!r}') |
| 531 |
run_cmd(**cmd_kwargs) |
| 532 |
|
| 533 |
|
| 534 |
if __name__ == '__main__': |
| 535 |
cmd, cmd_kwargs, verbosity, traceback_cm = parse_args() |
| 536 |
configure_logger(verbosity) |
| 537 |
with traceback_cm: |
| 538 |
main(cmd, cmd_kwargs) |