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
138
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
|
#!/usr/bin/env python3
"""Produce a report of all assertions in the MongoDB server codebase.
Parses .cpp files for assertions and verifies assertion codes are distinct.
Optionally replaces zero codes in source code with new distinct values.
"""
import bisect
import os.path
import sys
from collections import defaultdict, namedtuple
from optparse import OptionParser
from functools import reduce
from pathlib import Path
# Get relative imports to work when the package is not installed on the PYTHONPATH.
if __name__ == "__main__" and __package__ is None:
sys.path.append(os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
try:
import regex as re
except ImportError:
print("*** Run 'pip3 install --user regex' to speed up error code checking")
import re # type: ignore
ASSERT_NAMES = ["uassert", "massert", "fassert", "fassertFailed"]
MINIMUM_CODE = 10000
# This limit is intended to be increased by 1000 when we get close.
MAXIMUM_CODE = 51999
FIRST_BACKPORTED_CODE = 3873100
# pylint: disable=invalid-name
codes = [] # type: ignore
# pylint: enable=invalid-name
# Each AssertLocation identifies the C++ source location of an assertion
AssertLocation = namedtuple("AssertLocation", ['sourceFile', 'byteOffset', 'lines', 'code'])
list_files = False # pylint: disable=invalid-name
_CODE_PATTERNS = [
re.compile(p + r'\s*(?P<code>\d+)', re.MULTILINE) for p in [
# All the asserts and their optional variant suffixes
r"(?:f|i|m|msg|t|u)(?:assert)"
r"(?:ed)?"
r"(?:Failed)?"
r"(?:WithStatus)?"
r"(?:NoTrace)?"
r"(?:StatusOK)?"
r"(?:WithContext)?"
r"\s*\(",
# DBException and AssertionException constructors
r"(?:DB|Assertion)Exception\s*[({]",
# Calls to all LOGV2* variants
r"LOGV2(?:\w*)?\s*\(",
# Forwards a dynamic code to LOGV2
r"logAndBackoff\(",
# Error coersions
r"ErrorCodes::Error\s*[({]",
]
]
_DIR_EXCLUDE_RE = re.compile(r'(\..*'
r'|pcre-.*'
r'|32bit.*'
r'|mongodb-.*'
r'|debian.*'
r'|mongo-cxx-driver.*'
r'|.*gotools.*'
r'|.*mozjs.*'
r')')
_FILE_INCLUDE_RE = re.compile(r'.*\.(cpp|c|h|py|idl)')
def get_all_source_files(prefix='.'):
"""Return source files."""
def walk(path):
for fx in path.iterdir():
if fx.is_dir():
if fx.is_symlink() and fx.parent.name != "modules":
continue
if _DIR_EXCLUDE_RE.fullmatch(fx.name):
continue
for child in walk(fx):
yield child
elif fx.is_file() and _FILE_INCLUDE_RE.fullmatch(fx.name):
yield fx
for child in walk(Path(prefix)):
yield str(child)
def foreach_source_file(callback, src_root):
"""Invoke a callback on the text of each source file."""
for source_file in get_all_source_files(prefix=src_root):
with open(source_file, 'r', encoding='utf-8') as fh:
text = fh.read()
callback(source_file, text)
def parse_source_files(callback, src_root):
"""Walk MongoDB sourcefiles and invoke a callback for each AssertLocation found."""
def scan_for_codes(source_file, text):
for pat in _CODE_PATTERNS:
for match in pat.finditer(text):
# Note that this will include the text of the full match but will report the
# position of the beginning of the code portion rather than the beginning of the
# match. This is to position editors on the spot that needs to change.
loc = AssertLocation(source_file, match.start('code'), match.group(0),
match.group('code'))
callback(loc)
foreach_source_file(scan_for_codes, src_root)
def get_line_and_column_for_position(loc, _file_cache=None):
"""Convert an absolute position in a file into a line number."""
if _file_cache is None:
_file_cache = {}
if loc.sourceFile not in _file_cache:
with open(loc.sourceFile) as fh:
text = fh.read()
line_offsets = [0]
for line in text.splitlines(True):
line_offsets.append(line_offsets[-1] + len(line))
_file_cache[loc.sourceFile] = line_offsets
# These are both 1-based, but line is handled by starting the list with 0.
line = bisect.bisect(_file_cache[loc.sourceFile], loc.byteOffset)
column = loc.byteOffset - _file_cache[loc.sourceFile][line - 1] + 1
return (line, column)
def is_terminated(lines):
"""Determine if assert is terminated, from .cpp/.h source lines as text."""
code_block = " ".join(lines)
return ';' in code_block or code_block.count('(') - code_block.count(')') <= 0
def get_next_code():
"""Find next unused assertion code.
Called by: SConstruct and main()
Since SConstruct calls us, codes[] must be global OR WE REPARSE EVERYTHING
"""
if not codes:
read_error_codes()
highest = reduce(lambda x, y: max(int(x), int(y)), (loc.code for loc in codes))
return highest + 1
def check_error_codes():
"""Check error codes as SConstruct expects a boolean response from this function."""
(_, errors) = read_error_codes()
return len(errors) == 0
def read_error_codes():
"""Define callback, call parse_source_files() with callback, save matches to global codes list."""
seen = {}
errors = []
dups = defaultdict(list)
skips = []
# define callback
def check_dups(assert_loc):
"""Check for duplicates."""
codes.append(assert_loc)
code = assert_loc.code
if not code in seen:
seen[code] = assert_loc
# on first occurrence of a specific excessively large code, add to skips, errors
if int(code) >= FIRST_BACKPORTED_CODE:
# Large codes are used in newer versions, so allow them in backports.
pass
elif int(code) > MAXIMUM_CODE:
skips.append(assert_loc)
errors.append(assert_loc)
elif int(code) > MAXIMUM_CODE - 20:
print("Approaching maximum error code. Consider raising the limit soon.")
else:
if not code in dups:
# on first duplicate, add original to dups, errors
dups[code].append(seen[code])
errors.append(seen[code])
dups[code].append(assert_loc)
errors.append(assert_loc)
parse_source_files(check_dups, 'src/mongo')
if "0" in seen:
code = "0"
bad = seen[code]
errors.append(bad)
line, col = get_line_and_column_for_position(bad)
print("ZERO_CODE:")
print(" %s:%d:%d:%s" % (bad.sourceFile, line, col, bad.lines))
for loc in skips:
line, col = get_line_and_column_for_position(loc)
print("EXCESSIVE SKIPPING OF ERROR CODES:")
print(" %s:%d:%d:%s" % (loc.sourceFile, line, col, loc.lines))
for code, locations in list(dups.items()):
print("DUPLICATE IDS: %s" % code)
for loc in locations:
line, col = get_line_and_column_for_position(loc)
print(" %s:%d:%d:%s" % (loc.sourceFile, line, col, loc.lines))
return (codes, errors)
def replace_bad_codes(errors, next_code): # pylint: disable=too-many-locals
"""Modify C++ source files to replace invalid assertion codes.
For now, we only modify zero codes.
Args:
errors: list of AssertLocation
next_code: int, next non-conflicting assertion code
"""
zero_errors = [e for e in errors if int(e.code) == 0]
skip_errors = [e for e in errors if int(e.code) != 0]
for loc in skip_errors:
line, col = get_line_and_column_for_position(loc)
print("SKIPPING NONZERO code=%s: %s:%d:%d" % (loc.code, loc.sourceFile, line, col))
# Dedupe, sort, and reverse so we don't have to update offsets as we go.
for assert_loc in reversed(sorted(set(zero_errors))):
(source_file, byte_offset, _, _) = assert_loc
line_num, _ = get_line_and_column_for_position(assert_loc)
print("UPDATING_FILE: %s:%s" % (source_file, line_num))
ln = line_num - 1
with open(source_file, 'r+') as fh:
print("LINE_%d_BEFORE:%s" % (line_num, fh.readlines()[ln].rstrip()))
fh.seek(0)
text = fh.read()
assert text[byte_offset] == '0'
fh.seek(0)
fh.write(text[:byte_offset])
fh.write(str(next_code))
fh.write(text[byte_offset + 1:])
fh.seek(0)
print("LINE_%d_AFTER :%s" % (line_num, fh.readlines()[ln].rstrip()))
next_code += 1
def main():
"""Main."""
parser = OptionParser(description=__doc__.strip())
parser.add_option("--fix", dest="replace", action="store_true", default=False,
help="Fix zero codes in source files [default: %default]")
parser.add_option("-q", "--quiet", dest="quiet", action="store_true", default=False,
help="Suppress output on success [default: %default]")
parser.add_option("--list-files", dest="list_files", action="store_true", default=False,
help="Print the name of each file as it is scanned [default: %default]")
(options, _) = parser.parse_args()
global list_files # pylint: disable=global-statement,invalid-name
list_files = options.list_files
(_, errors) = read_error_codes()
ok = len(errors) == 0
if ok and options.quiet:
return
next_code = get_next_code()
print("ok: %s" % ok)
print("next: %s" % next_code)
if ok:
sys.exit(0)
elif options.replace:
replace_bad_codes(errors, next_code)
else:
print(ERROR_HELP)
sys.exit(1)
ERROR_HELP = """
ERRORS DETECTED. To correct, run "buildscripts/errorcodes.py --fix" to replace zero codes.
Other errors require manual correction.
"""
if __name__ == "__main__":
main()
|