| 1 |
#!/usr/bin/env python3 |
| 2 |
# ################################################################ |
| 3 |
# Copyright (c) Meta Platforms, Inc. and affiliates. |
| 4 |
# All rights reserved. |
| 5 |
# |
| 6 |
# This source code is licensed under both the BSD-style license (found in the |
| 7 |
# LICENSE file in the root directory of this source tree) and the GPLv2 (found |
| 8 |
# in the COPYING file in the root directory of this source tree). |
| 9 |
# You may select, at your option, one of the above-listed licenses. |
| 10 |
# ########################################################################## |
| 11 |
|
| 12 |
import argparse |
| 13 |
import contextlib |
| 14 |
import os |
| 15 |
import re |
| 16 |
import shutil |
| 17 |
import sys |
| 18 |
from typing import Optional |
| 19 |
|
| 20 |
|
| 21 |
INCLUDED_SUBDIRS = ["common", "compress", "decompress"] |
| 22 |
|
| 23 |
SKIPPED_FILES = [ |
| 24 |
"common/mem.h", |
| 25 |
"common/zstd_deps.h", |
| 26 |
"common/pool.c", |
| 27 |
"common/pool.h", |
| 28 |
"common/threading.c", |
| 29 |
"common/threading.h", |
| 30 |
"common/zstd_trace.h", |
| 31 |
"compress/zstdmt_compress.h", |
| 32 |
"compress/zstdmt_compress.c", |
| 33 |
] |
| 34 |
|
| 35 |
XXHASH_FILES = [ |
| 36 |
"common/xxhash.c", |
| 37 |
"common/xxhash.h", |
| 38 |
] |
| 39 |
|
| 40 |
|
| 41 |
class FileLines(object): |
| 42 |
def __init__(self, filename): |
| 43 |
self.filename = filename |
| 44 |
with open(self.filename, "r") as f: |
| 45 |
self.lines = f.readlines() |
| 46 |
|
| 47 |
def write(self): |
| 48 |
with open(self.filename, "w") as f: |
| 49 |
f.write("".join(self.lines)) |
| 50 |
|
| 51 |
|
| 52 |
class PartialPreprocessor(object): |
| 53 |
""" |
| 54 |
Looks for simple ifdefs and ifndefs and replaces them. |
| 55 |
Handles && and ||. |
| 56 |
Has fancy logic to handle translating elifs to ifs. |
| 57 |
Only looks for macros in the first part of the expression with no |
| 58 |
parens. |
| 59 |
Does not handle multi-line macros (only looks in first line). |
| 60 |
""" |
| 61 |
def __init__(self, defs: [(str, Optional[str])], replaces: [(str, str)], undefs: [str]): |
| 62 |
MACRO_GROUP = r"(?P<macro>[a-zA-Z_][a-zA-Z_0-9]*)" |
| 63 |
ELIF_GROUP = r"(?P<elif>el)?" |
| 64 |
OP_GROUP = r"(?P<op>&&|\|\|)?" |
| 65 |
|
| 66 |
self._defs = {macro:value for macro, value in defs} |
| 67 |
self._replaces = {macro:value for macro, value in replaces} |
| 68 |
self._defs.update(self._replaces) |
| 69 |
self._undefs = set(undefs) |
| 70 |
|
| 71 |
self._define = re.compile(r"\s*#\s*define") |
| 72 |
self._if = re.compile(r"\s*#\s*if") |
| 73 |
self._elif = re.compile(r"\s*#\s*(?P<elif>el)if") |
| 74 |
self._else = re.compile(r"\s*#\s*(?P<else>else)") |
| 75 |
self._endif = re.compile(r"\s*#\s*endif") |
| 76 |
|
| 77 |
self._ifdef = re.compile(fr"\s*#\s*if(?P<not>n)?def {MACRO_GROUP}\s*") |
| 78 |
self._if_defined = re.compile( |
| 79 |
fr"\s*#\s*{ELIF_GROUP}if\s+(?P<not>!)?\s*defined\s*\(\s*{MACRO_GROUP}\s*\)\s*{OP_GROUP}" |
| 80 |
) |
| 81 |
self._if_defined_value = re.compile( |
| 82 |
fr"\s*#\s*{ELIF_GROUP}if\s+defined\s*\(\s*{MACRO_GROUP}\s*\)\s*" |
| 83 |
fr"(?P<op>&&)\s*" |
| 84 |
fr"(?P<openp>\()?\s*" |
| 85 |
fr"(?P<macro2>[a-zA-Z_][a-zA-Z_0-9]*)\s*" |
| 86 |
fr"(?P<cmp>[=><!]+)\s*" |
| 87 |
fr"(?P<value>[0-9]*)\s*" |
| 88 |
fr"(?P<closep>\))?\s*" |
| 89 |
) |
| 90 |
self._if_true = re.compile( |
| 91 |
fr"\s*#\s*{ELIF_GROUP}if\s+{MACRO_GROUP}\s*{OP_GROUP}" |
| 92 |
) |
| 93 |
|
| 94 |
self._c_comment = re.compile(r"/\*.*?\*/") |
| 95 |
self._cpp_comment = re.compile(r"//") |
| 96 |
|
| 97 |
def _log(self, *args, **kwargs): |
| 98 |
print(*args, **kwargs) |
| 99 |
|
| 100 |
def _strip_comments(self, line): |
| 101 |
# First strip c-style comments (may include //) |
| 102 |
while True: |
| 103 |
m = self._c_comment.search(line) |
| 104 |
if m is None: |
| 105 |
break |
| 106 |
line = line[:m.start()] + line[m.end():] |
| 107 |
|
| 108 |
# Then strip cpp-style comments |
| 109 |
m = self._cpp_comment.search(line) |
| 110 |
if m is not None: |
| 111 |
line = line[:m.start()] |
| 112 |
|
| 113 |
return line |
| 114 |
|
| 115 |
def _fixup_indentation(self, macro, replace: [str]): |
| 116 |
if len(replace) == 0: |
| 117 |
return replace |
| 118 |
if len(replace) == 1 and self._define.match(replace[0]) is None: |
| 119 |
# If there is only one line, only replace defines |
| 120 |
return replace |
| 121 |
|
| 122 |
|
| 123 |
all_pound = True |
| 124 |
for line in replace: |
| 125 |
if not line.startswith('#'): |
| 126 |
all_pound = False |
| 127 |
if all_pound: |
| 128 |
replace = [line[1:] for line in replace] |
| 129 |
|
| 130 |
min_spaces = len(replace[0]) |
| 131 |
for line in replace: |
| 132 |
spaces = 0 |
| 133 |
for i, c in enumerate(line): |
| 134 |
if c != ' ': |
| 135 |
# Non-preprocessor line ==> skip the fixup |
| 136 |
if not all_pound and c != '#': |
| 137 |
return replace |
| 138 |
spaces = i |
| 139 |
break |
| 140 |
min_spaces = min(min_spaces, spaces) |
| 141 |
|
| 142 |
replace = [line[min_spaces:] for line in replace] |
| 143 |
|
| 144 |
if all_pound: |
| 145 |
replace = ["#" + line for line in replace] |
| 146 |
|
| 147 |
return replace |
| 148 |
|
| 149 |
def _handle_if_block(self, macro, idx, is_true, prepend): |
| 150 |
""" |
| 151 |
Remove the #if or #elif block starting on this line. |
| 152 |
""" |
| 153 |
REMOVE_ONE = 0 |
| 154 |
KEEP_ONE = 1 |
| 155 |
REMOVE_REST = 2 |
| 156 |
|
| 157 |
if is_true: |
| 158 |
state = KEEP_ONE |
| 159 |
else: |
| 160 |
state = REMOVE_ONE |
| 161 |
|
| 162 |
line = self._inlines[idx] |
| 163 |
is_if = self._if.match(line) is not None |
| 164 |
assert is_if or self._elif.match(line) is not None |
| 165 |
depth = 0 |
| 166 |
|
| 167 |
start_idx = idx |
| 168 |
|
| 169 |
idx += 1 |
| 170 |
replace = prepend |
| 171 |
finished = False |
| 172 |
while idx < len(self._inlines): |
| 173 |
line = self._inlines[idx] |
| 174 |
# Nested if statement |
| 175 |
if self._if.match(line): |
| 176 |
depth += 1 |
| 177 |
idx += 1 |
| 178 |
continue |
| 179 |
# We're inside a nested statement |
| 180 |
if depth > 0: |
| 181 |
if self._endif.match(line): |
| 182 |
depth -= 1 |
| 183 |
idx += 1 |
| 184 |
continue |
| 185 |
|
| 186 |
# We're at the original depth |
| 187 |
|
| 188 |
# Looking only for an endif. |
| 189 |
# We've found a true statement, but haven't |
| 190 |
# completely elided the if block, so we just |
| 191 |
# remove the remainder. |
| 192 |
if state == REMOVE_REST: |
| 193 |
if self._endif.match(line): |
| 194 |
if is_if: |
| 195 |
# Remove the endif because we took the first if |
| 196 |
idx += 1 |
| 197 |
finished = True |
| 198 |
break |
| 199 |
idx += 1 |
| 200 |
continue |
| 201 |
|
| 202 |
if state == KEEP_ONE: |
| 203 |
m = self._elif.match(line) |
| 204 |
if self._endif.match(line): |
| 205 |
replace += self._inlines[start_idx + 1:idx] |
| 206 |
idx += 1 |
| 207 |
finished = True |
| 208 |
break |
| 209 |
if self._elif.match(line) or self._else.match(line): |
| 210 |
replace += self._inlines[start_idx + 1:idx] |
| 211 |
state = REMOVE_REST |
| 212 |
idx += 1 |
| 213 |
continue |
| 214 |
|
| 215 |
if state == REMOVE_ONE: |
| 216 |
m = self._elif.match(line) |
| 217 |
if m is not None: |
| 218 |
if is_if: |
| 219 |
idx += 1 |
| 220 |
b = m.start('elif') |
| 221 |
e = m.end('elif') |
| 222 |
assert e - b == 2 |
| 223 |
replace.append(line[:b] + line[e:]) |
| 224 |
finished = True |
| 225 |
break |
| 226 |
m = self._else.match(line) |
| 227 |
if m is not None: |
| 228 |
if is_if: |
| 229 |
idx += 1 |
| 230 |
while self._endif.match(self._inlines[idx]) is None: |
| 231 |
replace.append(self._inlines[idx]) |
| 232 |
idx += 1 |
| 233 |
idx += 1 |
| 234 |
finished = True |
| 235 |
break |
| 236 |
if self._endif.match(line): |
| 237 |
if is_if: |
| 238 |
# Remove the endif because no other elifs |
| 239 |
idx += 1 |
| 240 |
finished = True |
| 241 |
break |
| 242 |
idx += 1 |
| 243 |
continue |
| 244 |
if not finished: |
| 245 |
raise RuntimeError("Unterminated if block!") |
| 246 |
|
| 247 |
replace = self._fixup_indentation(macro, replace) |
| 248 |
|
| 249 |
self._log(f"\tHardwiring {macro}") |
| 250 |
if start_idx > 0: |
| 251 |
self._log(f"\t\t {self._inlines[start_idx - 1][:-1]}") |
| 252 |
for x in range(start_idx, idx): |
| 253 |
self._log(f"\t\t- {self._inlines[x][:-1]}") |
| 254 |
for line in replace: |
| 255 |
self._log(f"\t\t+ {line[:-1]}") |
| 256 |
if idx < len(self._inlines): |
| 257 |
self._log(f"\t\t {self._inlines[idx][:-1]}") |
| 258 |
|
| 259 |
return idx, replace |
| 260 |
|
| 261 |
def _preprocess_once(self): |
| 262 |
outlines = [] |
| 263 |
idx = 0 |
| 264 |
changed = False |
| 265 |
while idx < len(self._inlines): |
| 266 |
line = self._inlines[idx] |
| 267 |
sline = self._strip_comments(line) |
| 268 |
m = self._ifdef.fullmatch(sline) |
| 269 |
if_true = False |
| 270 |
if m is None: |
| 271 |
m = self._if_defined_value.fullmatch(sline) |
| 272 |
if m is None: |
| 273 |
m = self._if_defined.match(sline) |
| 274 |
if m is None: |
| 275 |
m = self._if_true.match(sline) |
| 276 |
if_true = (m is not None) |
| 277 |
if m is None: |
| 278 |
outlines.append(line) |
| 279 |
idx += 1 |
| 280 |
continue |
| 281 |
|
| 282 |
groups = m.groupdict() |
| 283 |
macro = groups['macro'] |
| 284 |
op = groups.get('op') |
| 285 |
|
| 286 |
if not (macro in self._defs or macro in self._undefs): |
| 287 |
outlines.append(line) |
| 288 |
idx += 1 |
| 289 |
continue |
| 290 |
|
| 291 |
defined = macro in self._defs |
| 292 |
|
| 293 |
# Needed variables set: |
| 294 |
# resolved: Is the statement fully resolved? |
| 295 |
# is_true: If resolved, is the statement true? |
| 296 |
ifdef = False |
| 297 |
if if_true: |
| 298 |
if not defined: |
| 299 |
outlines.append(line) |
| 300 |
idx += 1 |
| 301 |
continue |
| 302 |
|
| 303 |
defined_value = self._defs[macro] |
| 304 |
is_int = True |
| 305 |
try: |
| 306 |
defined_value = int(defined_value) |
| 307 |
except TypeError: |
| 308 |
is_int = False |
| 309 |
except ValueError: |
| 310 |
is_int = False |
| 311 |
|
| 312 |
resolved = is_int |
| 313 |
is_true = (defined_value != 0) |
| 314 |
|
| 315 |
if resolved and op is not None: |
| 316 |
if op == '&&': |
| 317 |
resolved = not is_true |
| 318 |
else: |
| 319 |
assert op == '||' |
| 320 |
resolved = is_true |
| 321 |
|
| 322 |
else: |
| 323 |
ifdef = groups.get('not') is None |
| 324 |
elseif = groups.get('elif') is not None |
| 325 |
|
| 326 |
macro2 = groups.get('macro2') |
| 327 |
cmp = groups.get('cmp') |
| 328 |
value = groups.get('value') |
| 329 |
openp = groups.get('openp') |
| 330 |
closep = groups.get('closep') |
| 331 |
|
| 332 |
is_true = (ifdef == defined) |
| 333 |
resolved = True |
| 334 |
if op is not None: |
| 335 |
if op == '&&': |
| 336 |
resolved = not is_true |
| 337 |
else: |
| 338 |
assert op == '||' |
| 339 |
resolved = is_true |
| 340 |
|
| 341 |
if macro2 is not None and not resolved: |
| 342 |
assert ifdef and defined and op == '&&' and cmp is not None |
| 343 |
# If the statement is true, but we have a single value check, then |
| 344 |
# check the value. |
| 345 |
defined_value = self._defs[macro] |
| 346 |
are_ints = True |
| 347 |
try: |
| 348 |
defined_value = int(defined_value) |
| 349 |
value = int(value) |
| 350 |
except TypeError: |
| 351 |
are_ints = False |
| 352 |
except ValueError: |
| 353 |
are_ints = False |
| 354 |
if ( |
| 355 |
macro == macro2 and |
| 356 |
((openp is None) == (closep is None)) and |
| 357 |
are_ints |
| 358 |
): |
| 359 |
resolved = True |
| 360 |
if cmp == '<': |
| 361 |
is_true = defined_value < value |
| 362 |
elif cmp == '<=': |
| 363 |
is_true = defined_value <= value |
| 364 |
elif cmp == '==': |
| 365 |
is_true = defined_value == value |
| 366 |
elif cmp == '!=': |
| 367 |
is_true = defined_value != value |
| 368 |
elif cmp == '>=': |
| 369 |
is_true = defined_value >= value |
| 370 |
elif cmp == '>': |
| 371 |
is_true = defined_value > value |
| 372 |
else: |
| 373 |
resolved = False |
| 374 |
|
| 375 |
if op is not None and not resolved: |
| 376 |
# Remove the first op in the line + spaces |
| 377 |
if op == '&&': |
| 378 |
opre = op |
| 379 |
else: |
| 380 |
assert op == '||' |
| 381 |
opre = r'\|\|' |
| 382 |
needle = re.compile(fr"(?P<if>\s*#\s*(el)?if\s+).*?(?P<op>{opre}\s*)") |
| 383 |
match = needle.match(line) |
| 384 |
assert match is not None |
| 385 |
newline = line[:match.end('if')] + line[match.end('op'):] |
| 386 |
|
| 387 |
self._log(f"\tHardwiring partially resolved {macro}") |
| 388 |
self._log(f"\t\t- {line[:-1]}") |
| 389 |
self._log(f"\t\t+ {newline[:-1]}") |
| 390 |
|
| 391 |
outlines.append(newline) |
| 392 |
idx += 1 |
| 393 |
continue |
| 394 |
|
| 395 |
# Skip any statements we cannot fully compute |
| 396 |
if not resolved: |
| 397 |
outlines.append(line) |
| 398 |
idx += 1 |
| 399 |
continue |
| 400 |
|
| 401 |
prepend = [] |
| 402 |
if macro in self._replaces: |
| 403 |
assert not ifdef |
| 404 |
assert op is None |
| 405 |
value = self._replaces.pop(macro) |
| 406 |
prepend = [f"#define {macro} {value}\n"] |
| 407 |
|
| 408 |
idx, replace = self._handle_if_block(macro, idx, is_true, prepend) |
| 409 |
outlines += replace |
| 410 |
changed = True |
| 411 |
|
| 412 |
return changed, outlines |
| 413 |
|
| 414 |
def preprocess(self, filename): |
| 415 |
with open(filename, 'r') as f: |
| 416 |
self._inlines = f.readlines() |
| 417 |
changed = True |
| 418 |
iters = 0 |
| 419 |
while changed: |
| 420 |
iters += 1 |
| 421 |
changed, outlines = self._preprocess_once() |
| 422 |
self._inlines = outlines |
| 423 |
|
| 424 |
with open(filename, 'w') as f: |
| 425 |
f.write(''.join(self._inlines)) |
| 426 |
|
| 427 |
|
| 428 |
class Freestanding(object): |
| 429 |
def __init__( |
| 430 |
self, zstd_deps: str, mem: str, source_lib: str, output_lib: str, |
| 431 |
external_xxhash: bool, xxh64_state: Optional[str], |
| 432 |
xxh64_prefix: Optional[str], rewritten_includes: [(str, str)], |
| 433 |
defs: [(str, Optional[str])], replaces: [(str, str)], |
| 434 |
undefs: [str], excludes: [str], seds: [str], spdx: bool, |
| 435 |
): |
| 436 |
self._zstd_deps = zstd_deps |
| 437 |
self._mem = mem |
| 438 |
self._src_lib = source_lib |
| 439 |
self._dst_lib = output_lib |
| 440 |
self._external_xxhash = external_xxhash |
| 441 |
self._xxh64_state = xxh64_state |
| 442 |
self._xxh64_prefix = xxh64_prefix |
| 443 |
self._rewritten_includes = rewritten_includes |
| 444 |
self._defs = defs |
| 445 |
self._replaces = replaces |
| 446 |
self._undefs = undefs |
| 447 |
self._excludes = excludes |
| 448 |
self._seds = seds |
| 449 |
self._spdx = spdx |
| 450 |
|
| 451 |
def _dst_lib_file_paths(self): |
| 452 |
""" |
| 453 |
Yields all the file paths in the dst_lib. |
| 454 |
""" |
| 455 |
for root, dirname, filenames in os.walk(self._dst_lib): |
| 456 |
for filename in filenames: |
| 457 |
filepath = os.path.join(root, filename) |
| 458 |
yield filepath |
| 459 |
|
| 460 |
def _log(self, *args, **kwargs): |
| 461 |
print(*args, **kwargs) |
| 462 |
|
| 463 |
def _copy_file(self, lib_path): |
| 464 |
suffixes = [".c", ".h", ".S"] |
| 465 |
if not any((lib_path.endswith(suffix) for suffix in suffixes)): |
| 466 |
return |
| 467 |
if lib_path in SKIPPED_FILES: |
| 468 |
self._log(f"\tSkipping file: {lib_path}") |
| 469 |
return |
| 470 |
if self._external_xxhash and lib_path in XXHASH_FILES: |
| 471 |
self._log(f"\tSkipping xxhash file: {lib_path}") |
| 472 |
return |
| 473 |
|
| 474 |
src_path = os.path.join(self._src_lib, lib_path) |
| 475 |
dst_path = os.path.join(self._dst_lib, lib_path) |
| 476 |
self._log(f"\tCopying: {src_path} -> {dst_path}") |
| 477 |
shutil.copyfile(src_path, dst_path) |
| 478 |
|
| 479 |
def _copy_source_lib(self): |
| 480 |
self._log("Copying source library into output library") |
| 481 |
|
| 482 |
assert os.path.exists(self._src_lib) |
| 483 |
os.makedirs(self._dst_lib, exist_ok=True) |
| 484 |
self._copy_file("zstd.h") |
| 485 |
self._copy_file("zstd_errors.h") |
| 486 |
for subdir in INCLUDED_SUBDIRS: |
| 487 |
src_dir = os.path.join(self._src_lib, subdir) |
| 488 |
dst_dir = os.path.join(self._dst_lib, subdir) |
| 489 |
|
| 490 |
assert os.path.exists(src_dir) |
| 491 |
os.makedirs(dst_dir, exist_ok=True) |
| 492 |
|
| 493 |
for filename in os.listdir(src_dir): |
| 494 |
lib_path = os.path.join(subdir, filename) |
| 495 |
self._copy_file(lib_path) |
| 496 |
|
| 497 |
def _copy_zstd_deps(self): |
| 498 |
dst_zstd_deps = os.path.join(self._dst_lib, "common", "zstd_deps.h") |
| 499 |
self._log(f"Copying zstd_deps: {self._zstd_deps} -> {dst_zstd_deps}") |
| 500 |
shutil.copyfile(self._zstd_deps, dst_zstd_deps) |
| 501 |
|
| 502 |
def _copy_mem(self): |
| 503 |
dst_mem = os.path.join(self._dst_lib, "common", "mem.h") |
| 504 |
self._log(f"Copying mem: {self._mem} -> {dst_mem}") |
| 505 |
shutil.copyfile(self._mem, dst_mem) |
| 506 |
|
| 507 |
def _hardwire_preprocessor(self, name: str, value: Optional[str] = None, undef=False): |
| 508 |
""" |
| 509 |
If value=None then hardwire that it is defined, but not what the value is. |
| 510 |
If undef=True then value must be None. |
| 511 |
If value='' then the macro is defined to '' exactly. |
| 512 |
""" |
| 513 |
assert not (undef and value is not None) |
| 514 |
for filepath in self._dst_lib_file_paths(): |
| 515 |
file = FileLines(filepath) |
| 516 |
|
| 517 |
def _hardwire_defines(self): |
| 518 |
self._log("Hardwiring macros") |
| 519 |
partial_preprocessor = PartialPreprocessor(self._defs, self._replaces, self._undefs) |
| 520 |
for filepath in self._dst_lib_file_paths(): |
| 521 |
partial_preprocessor.preprocess(filepath) |
| 522 |
|
| 523 |
def _remove_excludes(self): |
| 524 |
self._log("Removing excluded sections") |
| 525 |
for exclude in self._excludes: |
| 526 |
self._log(f"\tRemoving excluded sections for: {exclude}") |
| 527 |
begin_re = re.compile(f"BEGIN {exclude}") |
| 528 |
end_re = re.compile(f"END {exclude}") |
| 529 |
for filepath in self._dst_lib_file_paths(): |
| 530 |
file = FileLines(filepath) |
| 531 |
outlines = [] |
| 532 |
skipped = [] |
| 533 |
emit = True |
| 534 |
for line in file.lines: |
| 535 |
if emit and begin_re.search(line) is not None: |
| 536 |
assert end_re.search(line) is None |
| 537 |
emit = False |
| 538 |
if emit: |
| 539 |
outlines.append(line) |
| 540 |
else: |
| 541 |
skipped.append(line) |
| 542 |
if end_re.search(line) is not None: |
| 543 |
assert begin_re.search(line) is None |
| 544 |
self._log(f"\t\tRemoving excluded section: {exclude}") |
| 545 |
for s in skipped: |
| 546 |
self._log(f"\t\t\t- {s}") |
| 547 |
emit = True |
| 548 |
skipped = [] |
| 549 |
if not emit: |
| 550 |
raise RuntimeError("Excluded section unfinished!") |
| 551 |
file.lines = outlines |
| 552 |
file.write() |
| 553 |
|
| 554 |
def _rewrite_include(self, original, rewritten): |
| 555 |
self._log(f"\tRewriting include: {original} -> {rewritten}") |
| 556 |
regex = re.compile(f"\\s*#\\s*include\\s*(?P<include>{original})") |
| 557 |
for filepath in self._dst_lib_file_paths(): |
| 558 |
file = FileLines(filepath) |
| 559 |
for i, line in enumerate(file.lines): |
| 560 |
match = regex.match(line) |
| 561 |
if match is None: |
| 562 |
continue |
| 563 |
s = match.start('include') |
| 564 |
e = match.end('include') |
| 565 |
file.lines[i] = line[:s] + rewritten + line[e:] |
| 566 |
file.write() |
| 567 |
|
| 568 |
def _rewrite_includes(self): |
| 569 |
self._log("Rewriting includes") |
| 570 |
for original, rewritten in self._rewritten_includes: |
| 571 |
self._rewrite_include(original, rewritten) |
| 572 |
|
| 573 |
def _replace_xxh64_prefix(self): |
| 574 |
if self._xxh64_prefix is None: |
| 575 |
return |
| 576 |
self._log(f"Replacing XXH64 prefix with {self._xxh64_prefix}") |
| 577 |
replacements = [] |
| 578 |
if self._xxh64_state is not None: |
| 579 |
replacements.append( |
| 580 |
(re.compile(r"([^\w]|^)(?P<orig>XXH64_state_t)([^\w]|$)"), self._xxh64_state) |
| 581 |
) |
| 582 |
if self._xxh64_prefix is not None: |
| 583 |
replacements.append( |
| 584 |
(re.compile(r"([^\w]|^)(?P<orig>XXH64)[\(_]"), self._xxh64_prefix) |
| 585 |
) |
| 586 |
for filepath in self._dst_lib_file_paths(): |
| 587 |
file = FileLines(filepath) |
| 588 |
for i, line in enumerate(file.lines): |
| 589 |
modified = False |
| 590 |
for regex, replacement in replacements: |
| 591 |
match = regex.search(line) |
| 592 |
while match is not None: |
| 593 |
modified = True |
| 594 |
b = match.start('orig') |
| 595 |
e = match.end('orig') |
| 596 |
line = line[:b] + replacement + line[e:] |
| 597 |
match = regex.search(line) |
| 598 |
if modified: |
| 599 |
self._log(f"\t- {file.lines[i][:-1]}") |
| 600 |
self._log(f"\t+ {line[:-1]}") |
| 601 |
file.lines[i] = line |
| 602 |
file.write() |
| 603 |
|
| 604 |
def _parse_sed(self, sed): |
| 605 |
assert sed[0] == 's' |
| 606 |
delim = sed[1] |
| 607 |
match = re.fullmatch(f's{delim}(.+){delim}(.*){delim}(.*)', sed) |
| 608 |
assert match is not None |
| 609 |
regex = re.compile(match.group(1)) |
| 610 |
format_str = match.group(2) |
| 611 |
is_global = match.group(3) == 'g' |
| 612 |
return regex, format_str, is_global |
| 613 |
|
| 614 |
def _process_sed(self, sed): |
| 615 |
self._log(f"Processing sed: {sed}") |
| 616 |
regex, format_str, is_global = self._parse_sed(sed) |
| 617 |
|
| 618 |
for filepath in self._dst_lib_file_paths(): |
| 619 |
file = FileLines(filepath) |
| 620 |
for i, line in enumerate(file.lines): |
| 621 |
modified = False |
| 622 |
while True: |
| 623 |
match = regex.search(line) |
| 624 |
if match is None: |
| 625 |
break |
| 626 |
replacement = format_str.format(match.groups(''), match.groupdict('')) |
| 627 |
b = match.start() |
| 628 |
e = match.end() |
| 629 |
line = line[:b] + replacement + line[e:] |
| 630 |
modified = True |
| 631 |
if not is_global: |
| 632 |
break |
| 633 |
if modified: |
| 634 |
self._log(f"\t- {file.lines[i][:-1]}") |
| 635 |
self._log(f"\t+ {line[:-1]}") |
| 636 |
file.lines[i] = line |
| 637 |
file.write() |
| 638 |
|
| 639 |
def _process_seds(self): |
| 640 |
self._log("Processing seds") |
| 641 |
for sed in self._seds: |
| 642 |
self._process_sed(sed) |
| 643 |
|
| 644 |
def _process_spdx(self): |
| 645 |
if not self._spdx: |
| 646 |
return |
| 647 |
self._log("Processing spdx") |
| 648 |
SPDX_C = "// SPDX-License-Identifier: GPL-2.0+ OR BSD-3-Clause\n" |
| 649 |
SPDX_H_S = "/* SPDX-License-Identifier: GPL-2.0+ OR BSD-3-Clause */\n" |
| 650 |
for filepath in self._dst_lib_file_paths(): |
| 651 |
file = FileLines(filepath) |
| 652 |
if file.lines[0] == SPDX_C or file.lines[0] == SPDX_H_S: |
| 653 |
continue |
| 654 |
for line in file.lines: |
| 655 |
if "SPDX-License-Identifier" in line: |
| 656 |
raise RuntimeError(f"Unexpected SPDX license identifier: {file.filename} {repr(line)}") |
| 657 |
if file.filename.endswith(".c"): |
| 658 |
file.lines.insert(0, SPDX_C) |
| 659 |
elif file.filename.endswith(".h") or file.filename.endswith(".S"): |
| 660 |
file.lines.insert(0, SPDX_H_S) |
| 661 |
else: |
| 662 |
raise RuntimeError(f"Unexpected file extension: {file.filename}") |
| 663 |
file.write() |
| 664 |
|
| 665 |
|
| 666 |
|
| 667 |
def go(self): |
| 668 |
self._copy_source_lib() |
| 669 |
self._copy_zstd_deps() |
| 670 |
self._copy_mem() |
| 671 |
self._hardwire_defines() |
| 672 |
self._remove_excludes() |
| 673 |
self._rewrite_includes() |
| 674 |
self._replace_xxh64_prefix() |
| 675 |
self._process_seds() |
| 676 |
self._process_spdx() |
| 677 |
|
| 678 |
|
| 679 |
def parse_optional_pair(defines: [str]) -> [(str, Optional[str])]: |
| 680 |
output = [] |
| 681 |
for define in defines: |
| 682 |
parsed = define.split('=') |
| 683 |
if len(parsed) == 1: |
| 684 |
output.append((parsed[0], None)) |
| 685 |
elif len(parsed) == 2: |
| 686 |
output.append((parsed[0], parsed[1])) |
| 687 |
else: |
| 688 |
raise RuntimeError(f"Bad define: {define}") |
| 689 |
return output |
| 690 |
|
| 691 |
|
| 692 |
def parse_pair(rewritten_includes: [str]) -> [(str, str)]: |
| 693 |
output = [] |
| 694 |
for rewritten_include in rewritten_includes: |
| 695 |
parsed = rewritten_include.split('=') |
| 696 |
if len(parsed) == 2: |
| 697 |
output.append((parsed[0], parsed[1])) |
| 698 |
else: |
| 699 |
raise RuntimeError(f"Bad rewritten include: {rewritten_include}") |
| 700 |
return output |
| 701 |
|
| 702 |
|
| 703 |
|
| 704 |
def main(name, args): |
| 705 |
parser = argparse.ArgumentParser(prog=name) |
| 706 |
parser.add_argument("--zstd-deps", default="zstd_deps.h", help="Zstd dependencies file") |
| 707 |
parser.add_argument("--mem", default="mem.h", help="Memory module") |
| 708 |
parser.add_argument("--source-lib", default="../../lib", help="Location of the zstd library") |
| 709 |
parser.add_argument("--output-lib", default="./freestanding_lib", help="Where to output the freestanding zstd library") |
| 710 |
parser.add_argument("--xxhash", default=None, help="Alternate external xxhash include e.g. --xxhash='<xxhash.h>'. If set xxhash is not included.") |
| 711 |
parser.add_argument("--xxh64-state", default=None, help="Alternate XXH64 state type (excluding _) e.g. --xxh64-state='struct xxh64_state'") |
| 712 |
parser.add_argument("--xxh64-prefix", default=None, help="Alternate XXH64 function prefix (excluding _) e.g. --xxh64-prefix=xxh64") |
| 713 |
parser.add_argument("--rewrite-include", default=[], dest="rewritten_includes", action="append", help="Rewrite an include REGEX=NEW (e.g. '<stddef\\.h>=<linux/types.h>')") |
| 714 |
parser.add_argument("--sed", default=[], dest="seds", action="append", help="Apply a sed replacement. Format: `s/REGEX/FORMAT/[g]`. REGEX is a Python regex. FORMAT is a Python format string formatted by the regex dict.") |
| 715 |
parser.add_argument("--spdx", action="store_true", help="Add SPDX License Identifiers") |
| 716 |
parser.add_argument("-D", "--define", default=[], dest="defs", action="append", help="Pre-define this macro (can be passed multiple times)") |
| 717 |
parser.add_argument("-U", "--undefine", default=[], dest="undefs", action="append", help="Pre-undefine this macro (can be passed multiple times)") |
| 718 |
parser.add_argument("-R", "--replace", default=[], dest="replaces", action="append", help="Pre-define this macro and replace the first ifndef block with its definition") |
| 719 |
parser.add_argument("-E", "--exclude", default=[], dest="excludes", action="append", help="Exclude all lines between 'BEGIN <EXCLUDE>' and 'END <EXCLUDE>'") |
| 720 |
args = parser.parse_args(args) |
| 721 |
|
| 722 |
# Always remove threading |
| 723 |
if "ZSTD_MULTITHREAD" not in args.undefs: |
| 724 |
args.undefs.append("ZSTD_MULTITHREAD") |
| 725 |
|
| 726 |
args.defs = parse_optional_pair(args.defs) |
| 727 |
for name, _ in args.defs: |
| 728 |
if name in args.undefs: |
| 729 |
raise RuntimeError(f"{name} is both defined and undefined!") |
| 730 |
|
| 731 |
# Always set tracing to 0 |
| 732 |
if "ZSTD_NO_TRACE" not in (arg[0] for arg in args.defs): |
| 733 |
args.defs.append(("ZSTD_NO_TRACE", None)) |
| 734 |
args.defs.append(("ZSTD_TRACE", "0")) |
| 735 |
|
| 736 |
args.replaces = parse_pair(args.replaces) |
| 737 |
for name, _ in args.replaces: |
| 738 |
if name in args.undefs or name in args.defs: |
| 739 |
raise RuntimeError(f"{name} is both replaced and (un)defined!") |
| 740 |
|
| 741 |
args.rewritten_includes = parse_pair(args.rewritten_includes) |
| 742 |
|
| 743 |
external_xxhash = False |
| 744 |
if args.xxhash is not None: |
| 745 |
external_xxhash = True |
| 746 |
args.rewritten_includes.append(('"(\\.\\./common/)?xxhash.h"', args.xxhash)) |
| 747 |
|
| 748 |
if args.xxh64_prefix is not None: |
| 749 |
if not external_xxhash: |
| 750 |
raise RuntimeError("--xxh64-prefix may only be used with --xxhash provided") |
| 751 |
|
| 752 |
if args.xxh64_state is not None: |
| 753 |
if not external_xxhash: |
| 754 |
raise RuntimeError("--xxh64-state may only be used with --xxhash provided") |
| 755 |
|
| 756 |
Freestanding( |
| 757 |
args.zstd_deps, |
| 758 |
args.mem, |
| 759 |
args.source_lib, |
| 760 |
args.output_lib, |
| 761 |
external_xxhash, |
| 762 |
args.xxh64_state, |
| 763 |
args.xxh64_prefix, |
| 764 |
args.rewritten_includes, |
| 765 |
args.defs, |
| 766 |
args.replaces, |
| 767 |
args.undefs, |
| 768 |
args.excludes, |
| 769 |
args.seds, |
| 770 |
args.spdx, |
| 771 |
).go() |
| 772 |
|
| 773 |
if __name__ == "__main__": |
| 774 |
main(sys.argv[0], sys.argv[1:]) |