Skip to content

C
cpython
Commit 02512fb1 authored by Antoine Pitrou's avatar Antoine Pitrou
Browse files
Options

Issue #16408: Fix file descriptors not being closed in error conditions in the zipfile module. 

Patch by Serhiy Storchaka.
parent 115bc79d
Hide whitespace changes
Inline Side-by-side
Showing with 207 additions and 210 deletions
Lib/zipfile.py
View file @ 02512fb1
...@@ -718,30 +718,34 @@ class ZipFile(object): ...@@ -718,30 +718,34 @@ class ZipFile(object):
self.fp = file self.fp = file
self.filename = getattr(file, 'name', None) self.filename = getattr(file, 'name', None)
if key == 'r': try:
self._GetContents() if key == 'r':
elif key == 'w':
# set the modified flag so central directory gets written
# even if no files are added to the archive
self._didModify = True
elif key == 'a':
try:
# See if file is a zip file
self._RealGetContents() self._RealGetContents()
# seek to start of directory and overwrite elif key == 'w':
self.fp.seek(self.start_dir, 0)
except BadZipfile:
# file is not a zip file, just append
self.fp.seek(0, 2)
# set the modified flag so central directory gets written # set the modified flag so central directory gets written
# even if no files are added to the archive # even if no files are added to the archive
self._didModify = True self._didModify = True
else: elif key == 'a':
try:
# See if file is a zip file
self._RealGetContents()
# seek to start of directory and overwrite
self.fp.seek(self.start_dir, 0)
except BadZipfile:
# file is not a zip file, just append
self.fp.seek(0, 2)
# set the modified flag so central directory gets written
# even if no files are added to the archive
self._didModify = True
else:
raise RuntimeError('Mode must be "r", "w" or "a"')
except:
fp = self.fp
self.fp = None
if not self._filePassed: if not self._filePassed:
self.fp.close() fp.close()
self.fp = None raise
raise RuntimeError, 'Mode must be "r", "w" or "a"'
def __enter__(self): def __enter__(self):
return self return self
...@@ -749,17 +753,6 @@ class ZipFile(object): ...@@ -749,17 +753,6 @@ class ZipFile(object):
def __exit__(self, type, value, traceback): def __exit__(self, type, value, traceback):
self.close() self.close()
def _GetContents(self):
"""Read the directory, making sure we close the file if the format
is bad."""
try:
self._RealGetContents()
except BadZipfile:
if not self._filePassed:
self.fp.close()
self.fp = None
raise
def _RealGetContents(self): def _RealGetContents(self):
"""Read in the table of contents for the ZIP file.""" """Read in the table of contents for the ZIP file."""
fp = self.fp fp = self.fp
...@@ -853,9 +846,9 @@ class ZipFile(object): ...@@ -853,9 +846,9 @@ class ZipFile(object):
try: try:
# Read by chunks, to avoid an OverflowError or a # Read by chunks, to avoid an OverflowError or a
# MemoryError with very large embedded files. # MemoryError with very large embedded files.
f = self.open(zinfo.filename, "r") with self.open(zinfo.filename, "r") as f:
while f.read(chunk_size): # Check CRC-32 while f.read(chunk_size): # Check CRC-32
pass pass
except BadZipfile: except BadZipfile:
return zinfo.filename return zinfo.filename
...@@ -909,60 +902,65 @@ class ZipFile(object): ...@@ -909,60 +902,65 @@ class ZipFile(object):
zef_file = open(self.filename, 'rb') zef_file = open(self.filename, 'rb')
should_close = True should_close = True
# Make sure we have an info object try:
if isinstance(name, ZipInfo): # Make sure we have an info object
# 'name' is already an info object if isinstance(name, ZipInfo):
zinfo = name # 'name' is already an info object
else: zinfo = name
# Get info object for name
zinfo = self.getinfo(name)
zef_file.seek(zinfo.header_offset, 0)
# Skip the file header:
fheader = zef_file.read(sizeFileHeader)
if fheader[0:4] != stringFileHeader:
raise BadZipfile, "Bad magic number for file header"
fheader = struct.unpack(structFileHeader, fheader)
fname = zef_file.read(fheader[_FH_FILENAME_LENGTH])
if fheader[_FH_EXTRA_FIELD_LENGTH]:
zef_file.read(fheader[_FH_EXTRA_FIELD_LENGTH])
if fname != zinfo.orig_filename:
raise BadZipfile, \
'File name in directory "%s" and header "%s" differ.' % (
zinfo.orig_filename, fname)
# check for encrypted flag & handle password
is_encrypted = zinfo.flag_bits & 0x1
zd = None
if is_encrypted:
if not pwd:
pwd = self.pwd
if not pwd:
raise RuntimeError, "File %s is encrypted, " \
"password required for extraction" % name
zd = _ZipDecrypter(pwd)
# The first 12 bytes in the cypher stream is an encryption header
# used to strengthen the algorithm. The first 11 bytes are
# completely random, while the 12th contains the MSB of the CRC,
# or the MSB of the file time depending on the header type
# and is used to check the correctness of the password.
bytes = zef_file.read(12)
h = map(zd, bytes[0:12])
if zinfo.flag_bits & 0x8:
# compare against the file type from extended local headers
check_byte = (zinfo._raw_time >> 8) & 0xff
else: else:
# compare against the CRC otherwise # Get info object for name
check_byte = (zinfo.CRC >> 24) & 0xff zinfo = self.getinfo(name)
if ord(h[11]) != check_byte:
raise RuntimeError("Bad password for file", name) zef_file.seek(zinfo.header_offset, 0)
return ZipExtFile(zef_file, mode, zinfo, zd, # Skip the file header:
close_fileobj=should_close) fheader = zef_file.read(sizeFileHeader)
if fheader[0:4] != stringFileHeader:
raise BadZipfile, "Bad magic number for file header"
fheader = struct.unpack(structFileHeader, fheader)
fname = zef_file.read(fheader[_FH_FILENAME_LENGTH])
if fheader[_FH_EXTRA_FIELD_LENGTH]:
zef_file.read(fheader[_FH_EXTRA_FIELD_LENGTH])
if fname != zinfo.orig_filename:
raise BadZipfile, \
'File name in directory "%s" and header "%s" differ.' % (
zinfo.orig_filename, fname)
# check for encrypted flag & handle password
is_encrypted = zinfo.flag_bits & 0x1
zd = None
if is_encrypted:
if not pwd:
pwd = self.pwd
if not pwd:
raise RuntimeError, "File %s is encrypted, " \
"password required for extraction" % name
zd = _ZipDecrypter(pwd)
# The first 12 bytes in the cypher stream is an encryption header
# used to strengthen the algorithm. The first 11 bytes are
# completely random, while the 12th contains the MSB of the CRC,
# or the MSB of the file time depending on the header type
# and is used to check the correctness of the password.
bytes = zef_file.read(12)
h = map(zd, bytes[0:12])
if zinfo.flag_bits & 0x8:
# compare against the file type from extended local headers
check_byte = (zinfo._raw_time >> 8) & 0xff
else:
# compare against the CRC otherwise
check_byte = (zinfo.CRC >> 24) & 0xff
if ord(h[11]) != check_byte:
raise RuntimeError("Bad password for file", name)
return ZipExtFile(zef_file, mode, zinfo, zd,
close_fileobj=should_close)
except:
if should_close:
zef_file.close()
raise
def extract(self, member, path=None, pwd=None): def extract(self, member, path=None, pwd=None):
"""Extract a member from the archive to the current working directory, """Extract a member from the archive to the current working directory,
...@@ -1019,11 +1017,9 @@ class ZipFile(object): ...@@ -1019,11 +1017,9 @@ class ZipFile(object):
os.mkdir(targetpath) os.mkdir(targetpath)
return targetpath return targetpath
source = self.open(member, pwd=pwd) with self.open(member, pwd=pwd) as source, \
target = file(targetpath, "wb") file(targetpath, "wb") as target:
shutil.copyfileobj(source, target) shutil.copyfileobj(source, target)
source.close()
target.close()
return targetpath return targetpath
...@@ -1184,102 +1180,104 @@ class ZipFile(object): ...@@ -1184,102 +1180,104 @@ class ZipFile(object):
if self.fp is None: if self.fp is None:
return return
if self.mode in ("w", "a") and self._didModify: # write ending records try:
count = 0 if self.mode in ("w", "a") and self._didModify: # write ending records
pos1 = self.fp.tell() count = 0
for zinfo in self.filelist: # write central directory pos1 = self.fp.tell()
count = count + 1 for zinfo in self.filelist: # write central directory
dt = zinfo.date_time count = count + 1
dosdate = (dt[0] - 1980) << 9 | dt[1] << 5 | dt[2] dt = zinfo.date_time
dostime = dt[3] << 11 | dt[4] << 5 | (dt[5] // 2) dosdate = (dt[0] - 1980) << 9 | dt[1] << 5 | dt[2]
extra = [] dostime = dt[3] << 11 | dt[4] << 5 | (dt[5] // 2)
if zinfo.file_size > ZIP64_LIMIT \ extra = []
or zinfo.compress_size > ZIP64_LIMIT: if zinfo.file_size > ZIP64_LIMIT \
extra.append(zinfo.file_size) or zinfo.compress_size > ZIP64_LIMIT:
extra.append(zinfo.compress_size) extra.append(zinfo.file_size)
file_size = 0xffffffff extra.append(zinfo.compress_size)
compress_size = 0xffffffff file_size = 0xffffffff
else: compress_size = 0xffffffff
file_size = zinfo.file_size else:
compress_size = zinfo.compress_size file_size = zinfo.file_size
compress_size = zinfo.compress_size
if zinfo.header_offset > ZIP64_LIMIT:
extra.append(zinfo.header_offset) if zinfo.header_offset > ZIP64_LIMIT:
header_offset = 0xffffffffL extra.append(zinfo.header_offset)
else: header_offset = 0xffffffffL
header_offset = zinfo.header_offset else:
header_offset = zinfo.header_offset
extra_data = zinfo.extra
if extra: extra_data = zinfo.extra
# Append a ZIP64 field to the extra's if extra:
extra_data = struct.pack( # Append a ZIP64 field to the extra's
'<HH' + 'Q'*len(extra), extra_data = struct.pack(
1, 8*len(extra), *extra) + extra_data '<HH' + 'Q'*len(extra),
1, 8*len(extra), *extra) + extra_data
extract_version = max(45, zinfo.extract_version)
create_version = max(45, zinfo.create_version) extract_version = max(45, zinfo.extract_version)
else: create_version = max(45, zinfo.create_version)
extract_version = zinfo.extract_version else:
create_version = zinfo.create_version extract_version = zinfo.extract_version
create_version = zinfo.create_version
try:
filename, flag_bits = zinfo._encodeFilenameFlags() try:
centdir = struct.pack(structCentralDir, filename, flag_bits = zinfo._encodeFilenameFlags()
stringCentralDir, create_version, centdir = struct.pack(structCentralDir,
zinfo.create_system, extract_version, zinfo.reserved, stringCentralDir, create_version,
flag_bits, zinfo.compress_type, dostime, dosdate, zinfo.create_system, extract_version, zinfo.reserved,
zinfo.CRC, compress_size, file_size, flag_bits, zinfo.compress_type, dostime, dosdate,
len(filename), len(extra_data), len(zinfo.comment), zinfo.CRC, compress_size, file_size,
0, zinfo.internal_attr, zinfo.external_attr, len(filename), len(extra_data), len(zinfo.comment),
header_offset) 0, zinfo.internal_attr, zinfo.external_attr,
except DeprecationWarning: header_offset)
print >>sys.stderr, (structCentralDir, except DeprecationWarning:
stringCentralDir, create_version, print >>sys.stderr, (structCentralDir,
zinfo.create_system, extract_version, zinfo.reserved, stringCentralDir, create_version,
zinfo.flag_bits, zinfo.compress_type, dostime, dosdate, zinfo.create_system, extract_version, zinfo.reserved,
zinfo.CRC, compress_size, file_size, zinfo.flag_bits, zinfo.compress_type, dostime, dosdate,
len(zinfo.filename), len(extra_data), len(zinfo.comment), zinfo.CRC, compress_size, file_size,
0, zinfo.internal_attr, zinfo.external_attr, len(zinfo.filename), len(extra_data), len(zinfo.comment),
header_offset) 0, zinfo.internal_attr, zinfo.external_attr,
raise header_offset)
self.fp.write(centdir) raise
self.fp.write(filename) self.fp.write(centdir)
self.fp.write(extra_data) self.fp.write(filename)
self.fp.write(zinfo.comment) self.fp.write(extra_data)
self.fp.write(zinfo.comment)
pos2 = self.fp.tell()
# Write end-of-zip-archive record pos2 = self.fp.tell()
centDirCount = count # Write end-of-zip-archive record
centDirSize = pos2 - pos1 centDirCount = count
centDirOffset = pos1 centDirSize = pos2 - pos1
if (centDirCount >= ZIP_FILECOUNT_LIMIT or centDirOffset = pos1
centDirOffset > ZIP64_LIMIT or if (centDirCount >= ZIP_FILECOUNT_LIMIT or
centDirSize > ZIP64_LIMIT): centDirOffset > ZIP64_LIMIT or
# Need to write the ZIP64 end-of-archive records centDirSize > ZIP64_LIMIT):
zip64endrec = struct.pack( # Need to write the ZIP64 end-of-archive records
structEndArchive64, stringEndArchive64, zip64endrec = struct.pack(
44, 45, 45, 0, 0, centDirCount, centDirCount, structEndArchive64, stringEndArchive64,
centDirSize, centDirOffset) 44, 45, 45, 0, 0, centDirCount, centDirCount,
self.fp.write(zip64endrec) centDirSize, centDirOffset)
self.fp.write(zip64endrec)
zip64locrec = struct.pack(
structEndArchive64Locator, zip64locrec = struct.pack(
stringEndArchive64Locator, 0, pos2, 1) structEndArchive64Locator,
self.fp.write(zip64locrec) stringEndArchive64Locator, 0, pos2, 1)
centDirCount = min(centDirCount, 0xFFFF) self.fp.write(zip64locrec)
centDirSize = min(centDirSize, 0xFFFFFFFF) centDirCount = min(centDirCount, 0xFFFF)
centDirOffset = min(centDirOffset, 0xFFFFFFFF) centDirSize = min(centDirSize, 0xFFFFFFFF)
centDirOffset = min(centDirOffset, 0xFFFFFFFF)
endrec = struct.pack(structEndArchive, stringEndArchive,
0, 0, centDirCount, centDirCount, endrec = struct.pack(structEndArchive, stringEndArchive,
centDirSize, centDirOffset, len(self._comment)) 0, 0, centDirCount, centDirCount,
self.fp.write(endrec) centDirSize, centDirOffset, len(self._comment))
self.fp.write(self._comment) self.fp.write(endrec)
self.fp.flush() self.fp.write(self._comment)
self.fp.flush()
if not self._filePassed: finally:
self.fp.close() fp = self.fp
self.fp = None self.fp = None
if not self._filePassed:
fp.close()
class PyZipFile(ZipFile): class PyZipFile(ZipFile):
...@@ -1401,16 +1399,15 @@ def main(args = None): ...@@ -1401,16 +1399,15 @@ def main(args = None):
if len(args) != 2: if len(args) != 2:
print USAGE print USAGE
sys.exit(1) sys.exit(1)
zf = ZipFile(args[1], 'r') with ZipFile(args[1], 'r') as zf:
zf.printdir() zf.printdir()
zf.close()
elif args[0] == '-t': elif args[0] == '-t':
if len(args) != 2: if len(args) != 2:
print USAGE print USAGE
sys.exit(1) sys.exit(1)
zf = ZipFile(args[1], 'r') with ZipFile(args[1], 'r') as zf:
badfile = zf.testzip() badfile = zf.testzip()
if badfile: if badfile:
print("The following enclosed file is corrupted: {!r}".format(badfile)) print("The following enclosed file is corrupted: {!r}".format(badfile))
print "Done testing" print "Done testing"
...@@ -1420,20 +1417,19 @@ def main(args = None): ...@@ -1420,20 +1417,19 @@ def main(args = None):
print USAGE print USAGE
sys.exit(1) sys.exit(1)
zf = ZipFile(args[1], 'r') with ZipFile(args[1], 'r') as zf:
out = args[2] out = args[2]
for path in zf.namelist(): for path in zf.namelist():
if path.startswith('./'): if path.startswith('./'):
tgt = os.path.join(out, path[2:]) tgt = os.path.join(out, path[2:])
else: else:
tgt = os.path.join(out, path) tgt = os.path.join(out, path)
tgtdir = os.path.dirname(tgt) tgtdir = os.path.dirname(tgt)
if not os.path.exists(tgtdir): if not os.path.exists(tgtdir):
os.makedirs(tgtdir) os.makedirs(tgtdir)
with open(tgt, 'wb') as fp: with open(tgt, 'wb') as fp:
fp.write(zf.read(path)) fp.write(zf.read(path))
zf.close()
elif args[0] == '-c': elif args[0] == '-c':
if len(args) < 3: if len(args) < 3:
...@@ -1449,11 +1445,9 @@ def main(args = None): ...@@ -1449,11 +1445,9 @@ def main(args = None):
os.path.join(path, nm), os.path.join(zippath, nm)) os.path.join(path, nm), os.path.join(zippath, nm))
# else: ignore # else: ignore
zf = ZipFile(args[1], 'w', allowZip64=True) with ZipFile(args[1], 'w', allowZip64=True) as zf:
for src in args[2:]: for src in args[2:]:
addToZip(zf, src, os.path.basename(src)) addToZip(zf, src, os.path.basename(src))
zf.close()
if __name__ == "__main__": if __name__ == "__main__":
main() main()
Misc/NEWS
View file @ 02512fb1
...@@ -148,6 +148,9 @@ Core and Builtins ...@@ -148,6 +148,9 @@ Core and Builtins
Library Library
------- -------
- Issue #16408: Fix file descriptors not being closed in error conditions
in the zipfile module. Patch by Serhiy Storchaka.
- Issue #16327: The subprocess module no longer leaks file descriptors - Issue #16327: The subprocess module no longer leaks file descriptors
used for stdin/stdout/stderr pipes to the child when fork() fails. used for stdin/stdout/stderr pipes to the child when fork() fails.
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment
GitLab Nexedi Edition | About GitLab | About Nexedi | 沪ICP备2021021310号-2 | 沪ICP备2021021310号-7