[Python-checkins] cpython (3.3): Back out patch for #1159051, which caused backwards compatibility problems.
georg.brandl
python-checkins at python.org
Sun May 12 12:47:28 CEST 2013
http://hg.python.org/cpython/rev/9c2831fe84e9
changeset: 83742:9c2831fe84e9
branch: 3.3
user: Georg Brandl <georg at python.org>
date: Sun May 12 11:57:26 2013 +0200
summary:
Back out patch for #1159051, which caused backwards compatibility problems.
files:
Lib/gzip.py | 81 ++++++++++++++++--------------
Lib/test/test_bz2.py | 14 -----
Lib/test/test_gzip.py | 14 -----
Lib/test/test_lzma.py | 14 -----
Misc/NEWS | 3 -
5 files changed, 44 insertions(+), 82 deletions(-)
diff --git a/Lib/gzip.py b/Lib/gzip.py
--- a/Lib/gzip.py
+++ b/Lib/gzip.py
@@ -65,6 +65,9 @@
# or unsigned.
output.write(struct.pack("<L", value))
+def read32(input):
+ return struct.unpack("<I", input.read(4))[0]
+
class _PaddedFile:
"""Minimal read-only file object that prepends a string to the contents
of an actual file. Shouldn't be used outside of gzip.py, as it lacks
@@ -278,32 +281,28 @@
self.crc = zlib.crc32(b"") & 0xffffffff
self.size = 0
- def _read_exact(self, n):
- data = self.fileobj.read(n)
- while len(data) < n:
- b = self.fileobj.read(n - len(data))
- if not b:
- raise EOFError("Compressed file ended before the "
- "end-of-stream marker was reached")
- data += b
- return data
-
def _read_gzip_header(self):
magic = self.fileobj.read(2)
if magic == b'':
- return False
+ raise EOFError("Reached EOF")
if magic != b'\037\213':
raise IOError('Not a gzipped file')
- method, flag, self.mtime = struct.unpack("<BBIxx", self._read_exact(8))
+ method = ord( self.fileobj.read(1) )
if method != 8:
raise IOError('Unknown compression method')
+ flag = ord( self.fileobj.read(1) )
+ self.mtime = read32(self.fileobj)
+ # extraflag = self.fileobj.read(1)
+ # os = self.fileobj.read(1)
+ self.fileobj.read(2)
if flag & FEXTRA:
# Read & discard the extra field, if present
- extra_len, = struct.unpack("<H", self._read_exact(2))
- self._read_exact(extra_len)
+ xlen = ord(self.fileobj.read(1))
+ xlen = xlen + 256*ord(self.fileobj.read(1))
+ self.fileobj.read(xlen)
if flag & FNAME:
# Read and discard a null-terminated string containing the filename
while True:
@@ -317,13 +316,12 @@
if not s or s==b'\000':
break
if flag & FHCRC:
- self._read_exact(2) # Read & discard the 16-bit header CRC
+ self.fileobj.read(2) # Read & discard the 16-bit header CRC
unused = self.fileobj.unused()
if unused:
uncompress = self.decompress.decompress(unused)
self._add_read_data(uncompress)
- return True
def write(self,data):
self._check_closed()
@@ -357,16 +355,20 @@
readsize = 1024
if size < 0: # get the whole thing
- while self._read(readsize):
- readsize = min(self.max_read_chunk, readsize * 2)
- size = self.extrasize
+ try:
+ while True:
+ self._read(readsize)
+ readsize = min(self.max_read_chunk, readsize * 2)
+ except EOFError:
+ size = self.extrasize
else: # just get some more of it
- while size > self.extrasize:
- if not self._read(readsize):
- if size > self.extrasize:
- size = self.extrasize
- break
- readsize = min(self.max_read_chunk, readsize * 2)
+ try:
+ while size > self.extrasize:
+ self._read(readsize)
+ readsize = min(self.max_read_chunk, readsize * 2)
+ except EOFError:
+ if size > self.extrasize:
+ size = self.extrasize
offset = self.offset - self.extrastart
chunk = self.extrabuf[offset: offset + size]
@@ -384,9 +386,12 @@
if self.extrasize <= 0 and self.fileobj is None:
return b''
- # For certain input data, a single call to _read() may not return
- # any data. In this case, retry until we get some data or reach EOF.
- while self.extrasize <= 0 and self._read():
+ try:
+ # For certain input data, a single call to _read() may not return
+ # any data. In this case, retry until we get some data or reach EOF.
+ while self.extrasize <= 0:
+ self._read()
+ except EOFError:
pass
if size < 0 or size > self.extrasize:
size = self.extrasize
@@ -409,9 +414,12 @@
if self.extrasize == 0:
if self.fileobj is None:
return b''
- # Ensure that we don't return b"" if we haven't reached EOF.
- # 1024 is the same buffering heuristic used in read()
- while self.extrasize == 0 and self._read(max(n, 1024)):
+ try:
+ # Ensure that we don't return b"" if we haven't reached EOF.
+ while self.extrasize == 0:
+ # 1024 is the same buffering heuristic used in read()
+ self._read(max(n, 1024))
+ except EOFError:
pass
offset = self.offset - self.extrastart
remaining = self.extrasize
@@ -424,14 +432,13 @@
def _read(self, size=1024):
if self.fileobj is None:
- return False
+ raise EOFError("Reached EOF")
if self._new_member:
# If the _new_member flag is set, we have to
# jump to the next member, if there is one.
self._init_read()
- if not self._read_gzip_header():
- return False
+ self._read_gzip_header()
self.decompress = zlib.decompressobj(-zlib.MAX_WBITS)
self._new_member = False
@@ -448,7 +455,7 @@
self.fileobj.prepend(self.decompress.unused_data, True)
self._read_eof()
self._add_read_data( uncompress )
- return False
+ raise EOFError('Reached EOF')
uncompress = self.decompress.decompress(buf)
self._add_read_data( uncompress )
@@ -464,7 +471,6 @@
# a new member on the next call
self._read_eof()
self._new_member = True
- return True
def _add_read_data(self, data):
self.crc = zlib.crc32(data, self.crc) & 0xffffffff
@@ -479,7 +485,8 @@
# We check the that the computed CRC and size of the
# uncompressed data matches the stored values. Note that the size
# stored is the true file size mod 2**32.
- crc32, isize = struct.unpack("<II", self._read_exact(8))
+ crc32 = read32(self.fileobj)
+ isize = read32(self.fileobj) # may exceed 2GB
if crc32 != self.crc:
raise IOError("CRC check failed %s != %s" % (hex(crc32),
hex(self.crc)))
diff --git a/Lib/test/test_bz2.py b/Lib/test/test_bz2.py
--- a/Lib/test/test_bz2.py
+++ b/Lib/test/test_bz2.py
@@ -577,20 +577,6 @@
bz2f.seek(-150, 1)
self.assertEqual(bz2f.read(), self.TEXT[500-150:])
- def test_read_truncated(self):
- # Drop the eos_magic field (6 bytes) and CRC (4 bytes).
- truncated = self.DATA[:-10]
- with BZ2File(BytesIO(truncated)) as f:
- self.assertRaises(EOFError, f.read)
- with BZ2File(BytesIO(truncated)) as f:
- self.assertEqual(f.read(len(self.TEXT)), self.TEXT)
- self.assertRaises(EOFError, f.read, 1)
- # Incomplete 4-byte file header, and block header of at least 146 bits.
- for i in range(22):
- with BZ2File(BytesIO(truncated[:i])) as f:
- self.assertRaises(EOFError, f.read, 1)
-
-
class BZ2CompressorTest(BaseTest):
def testCompress(self):
bz2c = BZ2Compressor()
diff --git a/Lib/test/test_gzip.py b/Lib/test/test_gzip.py
old mode 100644
new mode 100755
--- a/Lib/test/test_gzip.py
+++ b/Lib/test/test_gzip.py
@@ -389,20 +389,6 @@
datac = gzip.compress(data)
self.assertEqual(gzip.decompress(datac), data)
- def test_read_truncated(self):
- data = data1*50
- # Drop the CRC (4 bytes) and file size (4 bytes).
- truncated = gzip.compress(data)[:-8]
- with gzip.GzipFile(fileobj=io.BytesIO(truncated)) as f:
- self.assertRaises(EOFError, f.read)
- with gzip.GzipFile(fileobj=io.BytesIO(truncated)) as f:
- self.assertEqual(f.read(len(data)), data)
- self.assertRaises(EOFError, f.read, 1)
- # Incomplete 10-byte header.
- for i in range(2, 10):
- with gzip.GzipFile(fileobj=io.BytesIO(truncated[:i])) as f:
- self.assertRaises(EOFError, f.read, 1)
-
def test_read_with_extra(self):
# Gzip data with an extra field
gzdata = (b'\x1f\x8b\x08\x04\xb2\x17cQ\x02\xff'
diff --git a/Lib/test/test_lzma.py b/Lib/test/test_lzma.py
--- a/Lib/test/test_lzma.py
+++ b/Lib/test/test_lzma.py
@@ -669,20 +669,6 @@
with LZMAFile(BytesIO(COMPRESSED_XZ[:128])) as f:
self.assertRaises(EOFError, f.read)
- def test_read_truncated(self):
- # Drop stream footer: CRC (4 bytes), index size (4 bytes),
- # flags (2 bytes) and magic number (2 bytes).
- truncated = COMPRESSED_XZ[:-12]
- with LZMAFile(BytesIO(truncated)) as f:
- self.assertRaises(EOFError, f.read)
- with LZMAFile(BytesIO(truncated)) as f:
- self.assertEqual(f.read(len(INPUT)), INPUT)
- self.assertRaises(EOFError, f.read, 1)
- # Incomplete 12-byte header.
- for i in range(12):
- with LZMAFile(BytesIO(truncated[:i])) as f:
- self.assertRaises(EOFError, f.read, 1)
-
def test_read_bad_args(self):
f = LZMAFile(BytesIO(COMPRESSED_XZ))
f.close()
diff --git a/Misc/NEWS b/Misc/NEWS
--- a/Misc/NEWS
+++ b/Misc/NEWS
@@ -628,9 +628,6 @@
current directory on Unix and no longer searches a relative file path with
a directory part in PATH directories. Patch by Thomas Kluyver.
-- Issue #1159051: GzipFile now raises EOFError when reading a corrupted file
- with truncated header or footer.
-
- Issue #16993: shutil.which() now preserves the case of the path and extension
on Windows.
--
Repository URL: http://hg.python.org/cpython
More information about the Python-checkins
mailing list