cpython/Lib/test/test_tarfile.py

1550 lines
52 KiB
Python
Raw Normal View History

# -*- coding: iso-8859-15 -*-
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
import sys
import os
import shutil
import StringIO
from hashlib import md5
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
import errno
import unittest
import tarfile
from test import test_support
# Check for our compression modules.
try:
import gzip
gzip.GzipFile
except (ImportError, AttributeError):
gzip = None
try:
import bz2
except ImportError:
bz2 = None
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
def md5sum(data):
return md5(data).hexdigest()
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
TEMPDIR = os.path.abspath(test_support.TESTFN)
tarname = test_support.findfile("testtar.tar")
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
gzipname = os.path.join(TEMPDIR, "testtar.tar.gz")
bz2name = os.path.join(TEMPDIR, "testtar.tar.bz2")
tmpname = os.path.join(TEMPDIR, "tmp.tar")
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
md5_regtype = "65f477c818ad9e15f7feab0c6d37742f"
md5_sparse = "a54fbc4ca4f4399a90e1b27164012fc6"
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
class ReadTest(unittest.TestCase):
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
tarname = tarname
mode = "r:"
def setUp(self):
self.tar = tarfile.open(self.tarname, mode=self.mode, encoding="iso8859-1")
def tearDown(self):
self.tar.close()
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
class UstarReadTest(ReadTest):
def test_fileobj_regular_file(self):
tarinfo = self.tar.getmember("ustar/regtype")
fobj = self.tar.extractfile(tarinfo)
data = fobj.read()
2009-06-30 19:57:08 -03:00
self.assertTrue((len(data), md5sum(data)) == (tarinfo.size, md5_regtype),
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
"regular file extraction failed")
def test_fileobj_readlines(self):
self.tar.extract("ustar/regtype", TEMPDIR)
tarinfo = self.tar.getmember("ustar/regtype")
fobj1 = open(os.path.join(TEMPDIR, "ustar/regtype"), "rU")
fobj2 = self.tar.extractfile(tarinfo)
lines1 = fobj1.readlines()
lines2 = fobj2.readlines()
2009-06-30 19:57:08 -03:00
self.assertTrue(lines1 == lines2,
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
"fileobj.readlines() failed")
2009-06-30 19:57:08 -03:00
self.assertTrue(len(lines2) == 114,
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
"fileobj.readlines() failed")
self.assertTrue(lines2[83] ==
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
"I will gladly admit that Python is not the fastest running scripting language.\n",
"fileobj.readlines() failed")
def test_fileobj_iter(self):
self.tar.extract("ustar/regtype", TEMPDIR)
tarinfo = self.tar.getmember("ustar/regtype")
fobj1 = open(os.path.join(TEMPDIR, "ustar/regtype"), "rU")
fobj2 = self.tar.extractfile(tarinfo)
lines1 = fobj1.readlines()
lines2 = [line for line in fobj2]
2009-06-30 19:57:08 -03:00
self.assertTrue(lines1 == lines2,
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
"fileobj.__iter__() failed")
def test_fileobj_seek(self):
self.tar.extract("ustar/regtype", TEMPDIR)
fobj = open(os.path.join(TEMPDIR, "ustar/regtype"), "rb")
data = fobj.read()
fobj.close()
tarinfo = self.tar.getmember("ustar/regtype")
fobj = self.tar.extractfile(tarinfo)
text = fobj.read()
fobj.seek(0)
2009-06-30 19:57:08 -03:00
self.assertTrue(0 == fobj.tell(),
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
"seek() to file's start failed")
fobj.seek(2048, 0)
2009-06-30 19:57:08 -03:00
self.assertTrue(2048 == fobj.tell(),
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
"seek() to absolute position failed")
fobj.seek(-1024, 1)
2009-06-30 19:57:08 -03:00
self.assertTrue(1024 == fobj.tell(),
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
"seek() to negative relative position failed")
fobj.seek(1024, 1)
2009-06-30 19:57:08 -03:00
self.assertTrue(2048 == fobj.tell(),
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
"seek() to positive relative position failed")
s = fobj.read(10)
2009-06-30 19:57:08 -03:00
self.assertTrue(s == data[2048:2058],
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
"read() after seek failed")
fobj.seek(0, 2)
2009-06-30 19:57:08 -03:00
self.assertTrue(tarinfo.size == fobj.tell(),
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
"seek() to file's end failed")
2009-06-30 19:57:08 -03:00
self.assertTrue(fobj.read() == "",
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
"read() at file's end did not return empty string")
fobj.seek(-tarinfo.size, 2)
2009-06-30 19:57:08 -03:00
self.assertTrue(0 == fobj.tell(),
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
"relative seek() to file's start failed")
fobj.seek(512)
s1 = fobj.readlines()
fobj.seek(512)
s2 = fobj.readlines()
2009-06-30 19:57:08 -03:00
self.assertTrue(s1 == s2,
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
"readlines() after seek failed")
fobj.seek(0)
2009-06-30 19:57:08 -03:00
self.assertTrue(len(fobj.readline()) == fobj.tell(),
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
"tell() after readline() failed")
fobj.seek(512)
2009-06-30 19:57:08 -03:00
self.assertTrue(len(fobj.readline()) + 512 == fobj.tell(),
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
"tell() after seek() and readline() failed")
fobj.seek(0)
line = fobj.readline()
2009-06-30 19:57:08 -03:00
self.assertTrue(fobj.read() == data[len(line):],
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
"read() after readline() failed")
fobj.close()
# Test if symbolic and hard links are resolved by extractfile(). The
# test link members each point to a regular member whose data is
# supposed to be exported.
def _test_fileobj_link(self, lnktype, regtype):
a = self.tar.extractfile(lnktype)
b = self.tar.extractfile(regtype)
self.assertEqual(a.name, b.name)
def test_fileobj_link1(self):
self._test_fileobj_link("ustar/lnktype", "ustar/regtype")
def test_fileobj_link2(self):
self._test_fileobj_link("./ustar/linktest2/lnktype", "ustar/linktest1/regtype")
def test_fileobj_symlink1(self):
self._test_fileobj_link("ustar/symtype", "ustar/regtype")
def test_fileobj_symlink2(self):
self._test_fileobj_link("./ustar/linktest2/symtype", "ustar/linktest1/regtype")
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
class CommonReadTest(ReadTest):
def test_empty_tarfile(self):
# Test for issue6123: Allow opening empty archives.
# This test checks if tarfile.open() is able to open an empty tar
# archive successfully. Note that an empty tar archive is not the
# same as an empty file!
tarfile.open(tmpname, self.mode.replace("r", "w")).close()
try:
tar = tarfile.open(tmpname, self.mode)
tar.getnames()
except tarfile.ReadError:
self.fail("tarfile.open() failed on empty archive")
self.assertListEqual(tar.getmembers(), [])
def test_null_tarfile(self):
# Test for issue6123: Allow opening empty archives.
# This test guarantees that tarfile.open() does not treat an empty
# file as an empty tar archive.
open(tmpname, "wb").close()
self.assertRaises(tarfile.ReadError, tarfile.open, tmpname, self.mode)
self.assertRaises(tarfile.ReadError, tarfile.open, tmpname)
def test_ignore_zeros(self):
# Test TarFile's ignore_zeros option.
if self.mode.endswith(":gz"):
_open = gzip.GzipFile
elif self.mode.endswith(":bz2"):
_open = bz2.BZ2File
else:
_open = open
for char in ('\0', 'a'):
# Test if EOFHeaderError ('\0') and InvalidHeaderError ('a')
# are ignored correctly.
fobj = _open(tmpname, "wb")
fobj.write(char * 1024)
fobj.write(tarfile.TarInfo("foo").tobuf())
fobj.close()
tar = tarfile.open(tmpname, mode="r", ignore_zeros=True)
self.assertListEqual(tar.getnames(), ["foo"],
"ignore_zeros=True should have skipped the %r-blocks" % char)
tar.close()
class MiscReadTest(CommonReadTest):
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
def test_no_name_argument(self):
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
fobj = open(self.tarname, "rb")
tar = tarfile.open(fileobj=fobj, mode=self.mode)
self.assertEqual(tar.name, os.path.abspath(fobj.name))
def test_no_name_attribute(self):
data = open(self.tarname, "rb").read()
fobj = StringIO.StringIO(data)
self.assertRaises(AttributeError, getattr, fobj, "name")
tar = tarfile.open(fileobj=fobj, mode=self.mode)
self.assertEqual(tar.name, None)
def test_empty_name_attribute(self):
data = open(self.tarname, "rb").read()
fobj = StringIO.StringIO(data)
fobj.name = ""
tar = tarfile.open(fileobj=fobj, mode=self.mode)
self.assertEqual(tar.name, None)
def test_fileobj_with_offset(self):
# Skip the first member and store values from the second member
# of the testtar.
tar = tarfile.open(self.tarname, mode=self.mode)
tar.next()
t = tar.next()
name = t.name
offset = t.offset
data = tar.extractfile(t).read()
tar.close()
# Open the testtar and seek to the offset of the second member.
if self.mode.endswith(":gz"):
_open = gzip.GzipFile
elif self.mode.endswith(":bz2"):
_open = bz2.BZ2File
else:
_open = open
fobj = _open(self.tarname, "rb")
fobj.seek(offset)
# Test if the tarfile starts with the second member.
tar = tar.open(self.tarname, mode="r:", fileobj=fobj)
t = tar.next()
self.assertEqual(t.name, name)
# Read to the end of fileobj and test if seeking back to the
# beginning works.
tar.getmembers()
self.assertEqual(tar.extractfile(t).read(), data,
"seek back did not work")
tar.close()
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
def test_fail_comp(self):
# For Gzip and Bz2 Tests: fail with a ReadError on an uncompressed file.
if self.mode == "r:":
return
self.assertRaises(tarfile.ReadError, tarfile.open, tarname, self.mode)
fobj = open(tarname, "rb")
self.assertRaises(tarfile.ReadError, tarfile.open, fileobj=fobj, mode=self.mode)
def test_v7_dirtype(self):
# Test old style dirtype member (bug #1336623):
# Old V7 tars create directory members using an AREGTYPE
# header with a "/" appended to the filename field.
tarinfo = self.tar.getmember("misc/dirtype-old-v7")
2009-06-30 19:57:08 -03:00
self.assertTrue(tarinfo.type == tarfile.DIRTYPE,
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
"v7 dirtype failed")
def test_xstar_type(self):
# The xstar format stores extra atime and ctime fields inside the
# space reserved for the prefix field. The prefix field must be
# ignored in this case, otherwise it will mess up the name.
try:
self.tar.getmember("misc/regtype-xstar")
except KeyError:
self.fail("failed to find misc/regtype-xstar (mangled prefix?)")
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
def test_check_members(self):
for tarinfo in self.tar:
2009-06-30 19:57:08 -03:00
self.assertTrue(int(tarinfo.mtime) == 07606136617,
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
"wrong mtime for %s" % tarinfo.name)
if not tarinfo.name.startswith("ustar/"):
continue
2009-06-30 19:57:08 -03:00
self.assertTrue(tarinfo.uname == "tarfile",
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
"wrong uname for %s" % tarinfo.name)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
def test_find_members(self):
2009-06-30 19:57:08 -03:00
self.assertTrue(self.tar.getmembers()[-1].name == "misc/eof",
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
"could not find all members")
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
def test_extract_hardlink(self):
# Test hardlink extraction (e.g. bug #857297).
tar = tarfile.open(tarname, errorlevel=1, encoding="iso8859-1")
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
tar.extract("ustar/regtype", TEMPDIR)
try:
tar.extract("ustar/lnktype", TEMPDIR)
except EnvironmentError, e:
if e.errno == errno.ENOENT:
self.fail("hardlink not extracted properly")
data = open(os.path.join(TEMPDIR, "ustar/lnktype"), "rb").read()
self.assertEqual(md5sum(data), md5_regtype)
try:
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
tar.extract("ustar/symtype", TEMPDIR)
except EnvironmentError, e:
if e.errno == errno.ENOENT:
self.fail("symlink not extracted properly")
data = open(os.path.join(TEMPDIR, "ustar/symtype"), "rb").read()
self.assertEqual(md5sum(data), md5_regtype)
def test_extractall(self):
# Test if extractall() correctly restores directory permissions
# and times (see issue1735).
tar = tarfile.open(tarname, encoding="iso8859-1")
directories = [t for t in tar if t.isdir()]
tar.extractall(TEMPDIR, directories)
for tarinfo in directories:
path = os.path.join(TEMPDIR, tarinfo.name)
if sys.platform != "win32":
# Win32 has no support for fine grained permissions.
self.assertEqual(tarinfo.mode & 0777, os.stat(path).st_mode & 0777)
self.assertEqual(tarinfo.mtime, os.path.getmtime(path))
tar.close()
def test_init_close_fobj(self):
# Issue #7341: Close the internal file object in the TarFile
# constructor in case of an error. For the test we rely on
# the fact that opening an empty file raises a ReadError.
empty = os.path.join(TEMPDIR, "empty")
open(empty, "wb").write("")
try:
tar = object.__new__(tarfile.TarFile)
try:
tar.__init__(empty)
except tarfile.ReadError:
self.assertTrue(tar.fileobj.closed)
else:
self.fail("ReadError not raised")
finally:
os.remove(empty)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
class StreamReadTest(CommonReadTest):
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
mode="r|"
def test_fileobj_regular_file(self):
tarinfo = self.tar.next() # get "regtype" (can't use getmember)
fobj = self.tar.extractfile(tarinfo)
data = fobj.read()
2009-06-30 19:57:08 -03:00
self.assertTrue((len(data), md5sum(data)) == (tarinfo.size, md5_regtype),
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
"regular file extraction failed")
def test_provoke_stream_error(self):
tarinfos = self.tar.getmembers()
f = self.tar.extractfile(tarinfos[0]) # read the first member
self.assertRaises(tarfile.StreamError, f.read)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
def test_compare_members(self):
tar1 = tarfile.open(tarname, encoding="iso8859-1")
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
tar2 = self.tar
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
while True:
t1 = tar1.next()
t2 = tar2.next()
if t1 is None:
break
2009-06-30 19:57:08 -03:00
self.assertTrue(t2 is not None, "stream.next() failed.")
if t2.islnk() or t2.issym():
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
self.assertRaises(tarfile.StreamError, tar2.extractfile, t2)
continue
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
v1 = tar1.extractfile(t1)
v2 = tar2.extractfile(t2)
if v1 is None:
continue
2009-06-30 19:57:08 -03:00
self.assertTrue(v2 is not None, "stream.extractfile() failed")
self.assertTrue(v1.read() == v2.read(), "stream extraction failed")
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
tar1.close()
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
class DetectReadTest(unittest.TestCase):
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
def _testfunc_file(self, name, mode):
try:
tarfile.open(name, mode)
except tarfile.ReadError:
self.fail()
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
def _testfunc_fileobj(self, name, mode):
try:
tarfile.open(name, mode, fileobj=open(name, "rb"))
except tarfile.ReadError:
self.fail()
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
def _test_modes(self, testfunc):
testfunc(tarname, "r")
testfunc(tarname, "r:")
testfunc(tarname, "r:*")
testfunc(tarname, "r|")
testfunc(tarname, "r|*")
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
if gzip:
self.assertRaises(tarfile.ReadError, tarfile.open, tarname, mode="r:gz")
self.assertRaises(tarfile.ReadError, tarfile.open, tarname, mode="r|gz")
self.assertRaises(tarfile.ReadError, tarfile.open, gzipname, mode="r:")
self.assertRaises(tarfile.ReadError, tarfile.open, gzipname, mode="r|")
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
testfunc(gzipname, "r")
testfunc(gzipname, "r:*")
testfunc(gzipname, "r:gz")
testfunc(gzipname, "r|*")
testfunc(gzipname, "r|gz")
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
if bz2:
self.assertRaises(tarfile.ReadError, tarfile.open, tarname, mode="r:bz2")
self.assertRaises(tarfile.ReadError, tarfile.open, tarname, mode="r|bz2")
self.assertRaises(tarfile.ReadError, tarfile.open, bz2name, mode="r:")
self.assertRaises(tarfile.ReadError, tarfile.open, bz2name, mode="r|")
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
testfunc(bz2name, "r")
testfunc(bz2name, "r:*")
testfunc(bz2name, "r:bz2")
testfunc(bz2name, "r|*")
testfunc(bz2name, "r|bz2")
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
def test_detect_file(self):
self._test_modes(self._testfunc_file)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
def test_detect_fileobj(self):
self._test_modes(self._testfunc_fileobj)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
class MemberReadTest(ReadTest):
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
def _test_member(self, tarinfo, chksum=None, **kwargs):
if chksum is not None:
2009-06-30 19:57:08 -03:00
self.assertTrue(md5sum(self.tar.extractfile(tarinfo).read()) == chksum,
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
"wrong md5sum for %s" % tarinfo.name)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
kwargs["mtime"] = 07606136617
kwargs["uid"] = 1000
kwargs["gid"] = 100
if "old-v7" not in tarinfo.name:
# V7 tar can't handle alphabetic owners.
kwargs["uname"] = "tarfile"
kwargs["gname"] = "tarfile"
for k, v in kwargs.iteritems():
2009-06-30 19:57:08 -03:00
self.assertTrue(getattr(tarinfo, k) == v,
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
"wrong value in %s field of %s" % (k, tarinfo.name))
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
def test_find_regtype(self):
tarinfo = self.tar.getmember("ustar/regtype")
self._test_member(tarinfo, size=7011, chksum=md5_regtype)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
def test_find_conttype(self):
tarinfo = self.tar.getmember("ustar/conttype")
self._test_member(tarinfo, size=7011, chksum=md5_regtype)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
def test_find_dirtype(self):
tarinfo = self.tar.getmember("ustar/dirtype")
self._test_member(tarinfo, size=0)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
def test_find_dirtype_with_size(self):
tarinfo = self.tar.getmember("ustar/dirtype-with-size")
self._test_member(tarinfo, size=255)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
def test_find_lnktype(self):
tarinfo = self.tar.getmember("ustar/lnktype")
self._test_member(tarinfo, size=0, linkname="ustar/regtype")
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
def test_find_symtype(self):
tarinfo = self.tar.getmember("ustar/symtype")
self._test_member(tarinfo, size=0, linkname="regtype")
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
def test_find_blktype(self):
tarinfo = self.tar.getmember("ustar/blktype")
self._test_member(tarinfo, size=0, devmajor=3, devminor=0)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
def test_find_chrtype(self):
tarinfo = self.tar.getmember("ustar/chrtype")
self._test_member(tarinfo, size=0, devmajor=1, devminor=3)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
def test_find_fifotype(self):
tarinfo = self.tar.getmember("ustar/fifotype")
self._test_member(tarinfo, size=0)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
def test_find_sparse(self):
tarinfo = self.tar.getmember("ustar/sparse")
self._test_member(tarinfo, size=86016, chksum=md5_sparse)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
def test_find_umlauts(self):
tarinfo = self.tar.getmember("ustar/umlauts-<2D><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>")
self._test_member(tarinfo, size=7011, chksum=md5_regtype)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
def test_find_ustar_longname(self):
name = "ustar/" + "12345/" * 39 + "1234567/longname"
2010-01-23 19:04:36 -04:00
self.assertIn(name, self.tar.getnames())
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
def test_find_regtype_oldv7(self):
tarinfo = self.tar.getmember("misc/regtype-old-v7")
self._test_member(tarinfo, size=7011, chksum=md5_regtype)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
def test_find_pax_umlauts(self):
self.tar = tarfile.open(self.tarname, mode=self.mode, encoding="iso8859-1")
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
tarinfo = self.tar.getmember("pax/umlauts-<2D><><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>")
self._test_member(tarinfo, size=7011, chksum=md5_regtype)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
class LongnameTest(ReadTest):
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
def test_read_longname(self):
# Test reading of longname (bug #1471427).
longname = self.subdir + "/" + "123/" * 125 + "longname"
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
try:
tarinfo = self.tar.getmember(longname)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
except KeyError:
self.fail("longname not found")
2009-06-30 19:57:08 -03:00
self.assertTrue(tarinfo.type != tarfile.DIRTYPE, "read longname as dirtype")
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
def test_read_longlink(self):
longname = self.subdir + "/" + "123/" * 125 + "longname"
longlink = self.subdir + "/" + "123/" * 125 + "longlink"
try:
tarinfo = self.tar.getmember(longlink)
except KeyError:
self.fail("longlink not found")
2009-06-30 19:57:08 -03:00
self.assertTrue(tarinfo.linkname == longname, "linkname wrong")
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
def test_truncated_longname(self):
longname = self.subdir + "/" + "123/" * 125 + "longname"
tarinfo = self.tar.getmember(longname)
offset = tarinfo.offset
self.tar.fileobj.seek(offset)
fobj = StringIO.StringIO(self.tar.fileobj.read(3 * 512))
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
self.assertRaises(tarfile.ReadError, tarfile.open, name="foo.tar", fileobj=fobj)
def test_header_offset(self):
# Test if the start offset of the TarInfo object includes
# the preceding extended header.
longname = self.subdir + "/" + "123/" * 125 + "longname"
offset = self.tar.getmember(longname).offset
fobj = open(tarname)
fobj.seek(offset)
tarinfo = tarfile.TarInfo.frombuf(fobj.read(512))
self.assertEqual(tarinfo.type, self.longnametype)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
class GNUReadTest(LongnameTest):
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
subdir = "gnu"
longnametype = tarfile.GNUTYPE_LONGNAME
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
def test_sparse_file(self):
tarinfo1 = self.tar.getmember("ustar/sparse")
fobj1 = self.tar.extractfile(tarinfo1)
tarinfo2 = self.tar.getmember("gnu/sparse")
fobj2 = self.tar.extractfile(tarinfo2)
2009-06-30 19:57:08 -03:00
self.assertTrue(fobj1.read() == fobj2.read(),
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
"sparse file extraction failed")
class PaxReadTest(LongnameTest):
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
subdir = "pax"
longnametype = tarfile.XHDTYPE
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
def test_pax_global_headers(self):
tar = tarfile.open(tarname, encoding="iso8859-1")
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
tarinfo = tar.getmember("pax/regtype1")
self.assertEqual(tarinfo.uname, "foo")
self.assertEqual(tarinfo.gname, "bar")
self.assertEqual(tarinfo.pax_headers.get("VENDOR.umlauts"), u"<EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>")
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
tarinfo = tar.getmember("pax/regtype2")
self.assertEqual(tarinfo.uname, "")
self.assertEqual(tarinfo.gname, "bar")
self.assertEqual(tarinfo.pax_headers.get("VENDOR.umlauts"), u"<EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>")
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
tarinfo = tar.getmember("pax/regtype3")
self.assertEqual(tarinfo.uname, "tarfile")
self.assertEqual(tarinfo.gname, "tarfile")
self.assertEqual(tarinfo.pax_headers.get("VENDOR.umlauts"), u"<EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD><EFBFBD>")
def test_pax_number_fields(self):
# All following number fields are read from the pax header.
tar = tarfile.open(tarname, encoding="iso8859-1")
tarinfo = tar.getmember("pax/regtype4")
self.assertEqual(tarinfo.size, 7011)
self.assertEqual(tarinfo.uid, 123)
self.assertEqual(tarinfo.gid, 123)
self.assertEqual(tarinfo.mtime, 1041808783.0)
self.assertEqual(type(tarinfo.mtime), float)
self.assertEqual(float(tarinfo.pax_headers["atime"]), 1041808783.0)
self.assertEqual(float(tarinfo.pax_headers["ctime"]), 1041808783.0)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
class WriteTestBase(unittest.TestCase):
# Put all write tests in here that are supposed to be tested
# in all possible mode combinations.
def test_fileobj_no_close(self):
fobj = StringIO.StringIO()
tar = tarfile.open(fileobj=fobj, mode=self.mode)
tar.addfile(tarfile.TarInfo("foo"))
tar.close()
2009-06-30 19:57:08 -03:00
self.assertTrue(fobj.closed is False, "external fileobjs must never closed")
class WriteTest(WriteTestBase):
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
mode = "w:"
def test_100_char_name(self):
# The name field in a tar header stores strings of at most 100 chars.
# If a string is shorter than 100 chars it has to be padded with '\0',
# which implies that a string of exactly 100 chars is stored without
# a trailing '\0'.
name = "0123456789" * 10
tar = tarfile.open(tmpname, self.mode)
t = tarfile.TarInfo(name)
tar.addfile(t)
tar.close()
tar = tarfile.open(tmpname)
2009-06-30 19:57:08 -03:00
self.assertTrue(tar.getnames()[0] == name,
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
"failed to store 100 char filename")
tar.close()
def test_tar_size(self):
# Test for bug #1013882.
tar = tarfile.open(tmpname, self.mode)
path = os.path.join(TEMPDIR, "file")
fobj = open(path, "wb")
fobj.write("aaa")
fobj.close()
tar.add(path)
tar.close()
2009-06-30 19:57:08 -03:00
self.assertTrue(os.path.getsize(tmpname) > 0,
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
"tarfile is empty")
# The test_*_size tests test for bug #1167128.
def test_file_size(self):
tar = tarfile.open(tmpname, self.mode)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
path = os.path.join(TEMPDIR, "file")
fobj = open(path, "wb")
fobj.close()
tarinfo = tar.gettarinfo(path)
self.assertEqual(tarinfo.size, 0)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
fobj = open(path, "wb")
fobj.write("aaa")
fobj.close()
tarinfo = tar.gettarinfo(path)
self.assertEqual(tarinfo.size, 3)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
tar.close()
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
def test_directory_size(self):
path = os.path.join(TEMPDIR, "directory")
os.mkdir(path)
try:
tar = tarfile.open(tmpname, self.mode)
tarinfo = tar.gettarinfo(path)
self.assertEqual(tarinfo.size, 0)
finally:
os.rmdir(path)
def test_link_size(self):
if hasattr(os, "link"):
link = os.path.join(TEMPDIR, "link")
target = os.path.join(TEMPDIR, "link_target")
fobj = open(target, "wb")
fobj.write("aaa")
fobj.close()
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
os.link(target, link)
try:
tar = tarfile.open(tmpname, self.mode)
# Record the link target in the inodes list.
tar.gettarinfo(target)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
tarinfo = tar.gettarinfo(link)
self.assertEqual(tarinfo.size, 0)
finally:
os.remove(target)
os.remove(link)
def test_symlink_size(self):
if hasattr(os, "symlink"):
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
path = os.path.join(TEMPDIR, "symlink")
os.symlink("link_target", path)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
try:
tar = tarfile.open(tmpname, self.mode)
tarinfo = tar.gettarinfo(path)
self.assertEqual(tarinfo.size, 0)
finally:
os.remove(path)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
def test_add_self(self):
# Test for #1257255.
dstname = os.path.abspath(tmpname)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
tar = tarfile.open(tmpname, self.mode)
2009-06-30 19:57:08 -03:00
self.assertTrue(tar.name == dstname, "archive name must be absolute")
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
tar.add(dstname)
2009-06-30 19:57:08 -03:00
self.assertTrue(tar.getnames() == [], "added the archive to itself")
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
cwd = os.getcwd()
os.chdir(TEMPDIR)
tar.add(dstname)
os.chdir(cwd)
2009-06-30 19:57:08 -03:00
self.assertTrue(tar.getnames() == [], "added the archive to itself")
def test_exclude(self):
tempdir = os.path.join(TEMPDIR, "exclude")
os.mkdir(tempdir)
try:
for name in ("foo", "bar", "baz"):
name = os.path.join(tempdir, name)
open(name, "wb").close()
exclude = os.path.isfile
tar = tarfile.open(tmpname, self.mode, encoding="iso8859-1")
with test_support.check_warnings(("use the filter argument",
DeprecationWarning)):
tar.add(tempdir, arcname="empty_dir", exclude=exclude)
tar.close()
tar = tarfile.open(tmpname, "r")
self.assertEqual(len(tar.getmembers()), 1)
self.assertEqual(tar.getnames()[0], "empty_dir")
finally:
shutil.rmtree(tempdir)
def test_filter(self):
tempdir = os.path.join(TEMPDIR, "filter")
os.mkdir(tempdir)
try:
for name in ("foo", "bar", "baz"):
name = os.path.join(tempdir, name)
open(name, "wb").close()
def filter(tarinfo):
if os.path.basename(tarinfo.name) == "bar":
return
tarinfo.uid = 123
tarinfo.uname = "foo"
return tarinfo
tar = tarfile.open(tmpname, self.mode, encoding="iso8859-1")
tar.add(tempdir, arcname="empty_dir", filter=filter)
tar.close()
tar = tarfile.open(tmpname, "r")
for tarinfo in tar:
self.assertEqual(tarinfo.uid, 123)
self.assertEqual(tarinfo.uname, "foo")
self.assertEqual(len(tar.getmembers()), 3)
tar.close()
finally:
shutil.rmtree(tempdir)
# Guarantee that stored pathnames are not modified. Don't
# remove ./ or ../ or double slashes. Still make absolute
# pathnames relative.
# For details see bug #6054.
def _test_pathname(self, path, cmp_path=None, dir=False):
# Create a tarfile with an empty member named path
# and compare the stored name with the original.
foo = os.path.join(TEMPDIR, "foo")
if not dir:
open(foo, "w").close()
else:
os.mkdir(foo)
tar = tarfile.open(tmpname, self.mode)
tar.add(foo, arcname=path)
tar.close()
tar = tarfile.open(tmpname, "r")
t = tar.next()
tar.close()
if not dir:
os.remove(foo)
else:
os.rmdir(foo)
self.assertEqual(t.name, cmp_path or path.replace(os.sep, "/"))
def test_pathnames(self):
self._test_pathname("foo")
self._test_pathname(os.path.join("foo", ".", "bar"))
self._test_pathname(os.path.join("foo", "..", "bar"))
self._test_pathname(os.path.join(".", "foo"))
self._test_pathname(os.path.join(".", "foo", "."))
self._test_pathname(os.path.join(".", "foo", ".", "bar"))
self._test_pathname(os.path.join(".", "foo", "..", "bar"))
self._test_pathname(os.path.join(".", "foo", "..", "bar"))
self._test_pathname(os.path.join("..", "foo"))
self._test_pathname(os.path.join("..", "foo", ".."))
self._test_pathname(os.path.join("..", "foo", ".", "bar"))
self._test_pathname(os.path.join("..", "foo", "..", "bar"))
self._test_pathname("foo" + os.sep + os.sep + "bar")
self._test_pathname("foo" + os.sep + os.sep, "foo", dir=True)
def test_abs_pathnames(self):
if sys.platform == "win32":
self._test_pathname("C:\\foo", "foo")
else:
self._test_pathname("/foo", "foo")
self._test_pathname("///foo", "foo")
def test_cwd(self):
# Test adding the current working directory.
cwd = os.getcwd()
os.chdir(TEMPDIR)
try:
open("foo", "w").close()
tar = tarfile.open(tmpname, self.mode)
tar.add(".")
tar.close()
tar = tarfile.open(tmpname, "r")
for t in tar:
self.assert_(t.name == "." or t.name.startswith("./"))
tar.close()
finally:
os.chdir(cwd)
class StreamWriteTest(WriteTestBase):
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
mode = "w|"
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
def test_stream_padding(self):
# Test for bug #1543303.
tar = tarfile.open(tmpname, self.mode)
tar.close()
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
if self.mode.endswith("gz"):
fobj = gzip.GzipFile(tmpname)
data = fobj.read()
fobj.close()
elif self.mode.endswith("bz2"):
dec = bz2.BZ2Decompressor()
data = open(tmpname, "rb").read()
data = dec.decompress(data)
2009-06-30 19:57:08 -03:00
self.assertTrue(len(dec.unused_data) == 0,
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
"found trailing data")
else:
fobj = open(tmpname, "rb")
data = fobj.read()
fobj.close()
2009-06-30 19:57:08 -03:00
self.assertTrue(data.count("\0") == tarfile.RECORDSIZE,
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
"incorrect zero padding")
def test_file_mode(self):
# Test for issue #8464: Create files with correct
# permissions.
if sys.platform == "win32" or not hasattr(os, "umask"):
return
if os.path.exists(tmpname):
os.remove(tmpname)
original_umask = os.umask(0022)
try:
tar = tarfile.open(tmpname, self.mode)
tar.close()
mode = os.stat(tmpname).st_mode & 0777
self.assertEqual(mode, 0644, "wrong file permissions")
finally:
os.umask(original_umask)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
class GNUWriteTest(unittest.TestCase):
# This testcase checks for correct creation of GNU Longname
# and Longlink extended headers (cp. bug #812325).
def _length(self, s):
blocks, remainder = divmod(len(s) + 1, 512)
if remainder:
blocks += 1
return blocks * 512
def _calc_size(self, name, link=None):
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
# Initial tar header
count = 512
if len(name) > tarfile.LENGTH_NAME:
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
# GNU longname extended header + longname
count += 512
count += self._length(name)
if link is not None and len(link) > tarfile.LENGTH_LINK:
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
# GNU longlink extended header + longlink
count += 512
count += self._length(link)
return count
def _test(self, name, link=None):
tarinfo = tarfile.TarInfo(name)
if link:
tarinfo.linkname = link
tarinfo.type = tarfile.LNKTYPE
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
tar = tarfile.open(tmpname, "w")
tar.format = tarfile.GNU_FORMAT
tar.addfile(tarinfo)
v1 = self._calc_size(name, link)
v2 = tar.offset
2009-06-30 19:57:08 -03:00
self.assertTrue(v1 == v2, "GNU longname/longlink creation failed")
tar.close()
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
tar = tarfile.open(tmpname)
member = tar.next()
self.assertIsNotNone(member,
"unable to read longname member")
self.assertEqual(tarinfo.name, member.name,
"unable to read longname member")
self.assertEqual(tarinfo.linkname, member.linkname,
"unable to read longname member")
def test_longname_1023(self):
self._test(("longnam/" * 127) + "longnam")
def test_longname_1024(self):
self._test(("longnam/" * 127) + "longname")
def test_longname_1025(self):
self._test(("longnam/" * 127) + "longname_")
def test_longlink_1023(self):
self._test("name", ("longlnk/" * 127) + "longlnk")
def test_longlink_1024(self):
self._test("name", ("longlnk/" * 127) + "longlink")
def test_longlink_1025(self):
self._test("name", ("longlnk/" * 127) + "longlink_")
def test_longnamelink_1023(self):
self._test(("longnam/" * 127) + "longnam",
("longlnk/" * 127) + "longlnk")
def test_longnamelink_1024(self):
self._test(("longnam/" * 127) + "longname",
("longlnk/" * 127) + "longlink")
def test_longnamelink_1025(self):
self._test(("longnam/" * 127) + "longname_",
("longlnk/" * 127) + "longlink_")
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
class HardlinkTest(unittest.TestCase):
# Test the creation of LNKTYPE (hardlink) members in an archive.
def setUp(self):
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
self.foo = os.path.join(TEMPDIR, "foo")
self.bar = os.path.join(TEMPDIR, "bar")
fobj = open(self.foo, "wb")
fobj.write("foo")
fobj.close()
os.link(self.foo, self.bar)
self.tar = tarfile.open(tmpname, "w")
self.tar.add(self.foo)
def tearDown(self):
self.tar.close()
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
os.remove(self.foo)
os.remove(self.bar)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
def test_add_twice(self):
# The same name will be added as a REGTYPE every
# time regardless of st_nlink.
tarinfo = self.tar.gettarinfo(self.foo)
2009-06-30 19:57:08 -03:00
self.assertTrue(tarinfo.type == tarfile.REGTYPE,
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
"add file as regular failed")
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
def test_add_hardlink(self):
tarinfo = self.tar.gettarinfo(self.bar)
2009-06-30 19:57:08 -03:00
self.assertTrue(tarinfo.type == tarfile.LNKTYPE,
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
"add file as hardlink failed")
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
def test_dereference_hardlink(self):
self.tar.dereference = True
tarinfo = self.tar.gettarinfo(self.bar)
2009-06-30 19:57:08 -03:00
self.assertTrue(tarinfo.type == tarfile.REGTYPE,
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
"dereferencing hardlink failed")
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
class PaxWriteTest(GNUWriteTest):
def _test(self, name, link=None):
# See GNUWriteTest.
tarinfo = tarfile.TarInfo(name)
if link:
tarinfo.linkname = link
tarinfo.type = tarfile.LNKTYPE
tar = tarfile.open(tmpname, "w", format=tarfile.PAX_FORMAT)
tar.addfile(tarinfo)
tar.close()
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
tar = tarfile.open(tmpname)
if link:
l = tar.getmembers()[0].linkname
2009-06-30 19:57:08 -03:00
self.assertTrue(link == l, "PAX longlink creation failed")
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
else:
n = tar.getmembers()[0].name
2009-06-30 19:57:08 -03:00
self.assertTrue(name == n, "PAX longname creation failed")
def test_pax_global_header(self):
pax_headers = {
u"foo": u"bar",
u"uid": u"0",
u"mtime": u"1.23",
u"test": u"<EFBFBD><EFBFBD><EFBFBD>",
u"<EFBFBD><EFBFBD><EFBFBD>": u"test"}
tar = tarfile.open(tmpname, "w", format=tarfile.PAX_FORMAT,
pax_headers=pax_headers)
tar.addfile(tarfile.TarInfo("test"))
tar.close()
# Test if the global header was written correctly.
tar = tarfile.open(tmpname, encoding="iso8859-1")
self.assertEqual(tar.pax_headers, pax_headers)
self.assertEqual(tar.getmembers()[0].pax_headers, pax_headers)
# Test if all the fields are unicode.
for key, val in tar.pax_headers.iteritems():
2009-06-30 19:57:08 -03:00
self.assertTrue(type(key) is unicode)
self.assertTrue(type(val) is unicode)
if key in tarfile.PAX_NUMBER_FIELDS:
try:
tarfile.PAX_NUMBER_FIELDS[key](val)
except (TypeError, ValueError):
self.fail("unable to convert pax header field")
def test_pax_extended_header(self):
# The fields from the pax header have priority over the
# TarInfo.
pax_headers = {u"path": u"foo", u"uid": u"123"}
tar = tarfile.open(tmpname, "w", format=tarfile.PAX_FORMAT, encoding="iso8859-1")
t = tarfile.TarInfo()
t.name = u"<EFBFBD><EFBFBD><EFBFBD>" # non-ASCII
t.uid = 8**8 # too large
t.pax_headers = pax_headers
tar.addfile(t)
tar.close()
tar = tarfile.open(tmpname, encoding="iso8859-1")
t = tar.getmembers()[0]
self.assertEqual(t.pax_headers, pax_headers)
self.assertEqual(t.name, "foo")
self.assertEqual(t.uid, 123)
class UstarUnicodeTest(unittest.TestCase):
# All *UnicodeTests FIXME
format = tarfile.USTAR_FORMAT
def test_iso8859_1_filename(self):
self._test_unicode_filename("iso8859-1")
def test_utf7_filename(self):
self._test_unicode_filename("utf7")
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
def test_utf8_filename(self):
self._test_unicode_filename("utf8")
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
def _test_unicode_filename(self, encoding):
tar = tarfile.open(tmpname, "w", format=self.format, encoding=encoding, errors="strict")
name = u"<EFBFBD><EFBFBD><EFBFBD>"
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
tar.addfile(tarfile.TarInfo(name))
tar.close()
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
tar = tarfile.open(tmpname, encoding=encoding)
2009-06-30 19:57:08 -03:00
self.assertTrue(type(tar.getnames()[0]) is not unicode)
self.assertEqual(tar.getmembers()[0].name, name.encode(encoding))
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
tar.close()
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
def test_unicode_filename_error(self):
tar = tarfile.open(tmpname, "w", format=self.format, encoding="ascii", errors="strict")
tarinfo = tarfile.TarInfo()
tarinfo.name = "<EFBFBD><EFBFBD><EFBFBD>"
if self.format == tarfile.PAX_FORMAT:
self.assertRaises(UnicodeError, tar.addfile, tarinfo)
else:
tar.addfile(tarinfo)
tarinfo.name = u"<EFBFBD><EFBFBD><EFBFBD>"
self.assertRaises(UnicodeError, tar.addfile, tarinfo)
tarinfo.name = "foo"
tarinfo.uname = u"<EFBFBD><EFBFBD><EFBFBD>"
self.assertRaises(UnicodeError, tar.addfile, tarinfo)
def test_unicode_argument(self):
tar = tarfile.open(tarname, "r", encoding="iso8859-1", errors="strict")
for t in tar:
2009-06-30 19:57:08 -03:00
self.assertTrue(type(t.name) is str)
self.assertTrue(type(t.linkname) is str)
self.assertTrue(type(t.uname) is str)
self.assertTrue(type(t.gname) is str)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
tar.close()
def test_uname_unicode(self):
for name in (u"<EFBFBD><EFBFBD><EFBFBD>", "<EFBFBD><EFBFBD><EFBFBD>"):
t = tarfile.TarInfo("foo")
t.uname = name
t.gname = name
fobj = StringIO.StringIO()
tar = tarfile.open("foo.tar", mode="w", fileobj=fobj, format=self.format, encoding="iso8859-1")
tar.addfile(t)
tar.close()
fobj.seek(0)
tar = tarfile.open("foo.tar", fileobj=fobj, encoding="iso8859-1")
t = tar.getmember("foo")
self.assertEqual(t.uname, "<EFBFBD><EFBFBD><EFBFBD>")
self.assertEqual(t.gname, "<EFBFBD><EFBFBD><EFBFBD>")
class GNUUnicodeTest(UstarUnicodeTest):
format = tarfile.GNU_FORMAT
class PaxUnicodeTest(UstarUnicodeTest):
2006-12-20 07:55:16 -04:00
format = tarfile.PAX_FORMAT
def _create_unicode_name(self, name):
tar = tarfile.open(tmpname, "w", format=self.format)
t = tarfile.TarInfo()
t.pax_headers["path"] = name
tar.addfile(t)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
tar.close()
def test_error_handlers(self):
# Test if the unicode error handlers work correctly for characters
# that cannot be expressed in a given encoding.
self._create_unicode_name(u"<EFBFBD><EFBFBD><EFBFBD>")
for handler, name in (("utf-8", u"<EFBFBD><EFBFBD><EFBFBD>".encode("utf8")),
("replace", "???"), ("ignore", "")):
tar = tarfile.open(tmpname, format=self.format, encoding="ascii",
errors=handler)
self.assertEqual(tar.getnames()[0], name)
self.assertRaises(UnicodeError, tarfile.open, tmpname,
encoding="ascii", errors="strict")
def test_error_handler_utf8(self):
# Create a pathname that has one component representable using
# iso8859-1 and the other only in iso8859-15.
self._create_unicode_name(u"<EFBFBD><EFBFBD><EFBFBD>/<2F>")
tar = tarfile.open(tmpname, format=self.format, encoding="iso8859-1",
errors="utf-8")
self.assertEqual(tar.getnames()[0], "<EFBFBD><EFBFBD><EFBFBD>/" + u"<EFBFBD>".encode("utf8"))
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
class AppendTest(unittest.TestCase):
# Test append mode (cp. patch #1652681).
def setUp(self):
self.tarname = tmpname
if os.path.exists(self.tarname):
os.remove(self.tarname)
def _add_testfile(self, fileobj=None):
tar = tarfile.open(self.tarname, "a", fileobj=fileobj)
tar.addfile(tarfile.TarInfo("bar"))
tar.close()
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
def _create_testtar(self, mode="w:"):
src = tarfile.open(tarname, encoding="iso8859-1")
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
t = src.getmember("ustar/regtype")
t.name = "foo"
f = src.extractfile(t)
tar = tarfile.open(self.tarname, mode)
tar.addfile(t, f)
tar.close()
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
def _test(self, names=["bar"], fileobj=None):
tar = tarfile.open(self.tarname, fileobj=fileobj)
self.assertEqual(tar.getnames(), names)
def test_non_existing(self):
self._add_testfile()
self._test()
def test_empty(self):
tarfile.open(self.tarname, "w:").close()
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
self._add_testfile()
self._test()
def test_empty_fileobj(self):
fobj = StringIO.StringIO("\0" * 1024)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
self._add_testfile(fobj)
fobj.seek(0)
self._test(fileobj=fobj)
def test_fileobj(self):
self._create_testtar()
data = open(self.tarname).read()
fobj = StringIO.StringIO(data)
self._add_testfile(fobj)
fobj.seek(0)
self._test(names=["foo", "bar"], fileobj=fobj)
def test_existing(self):
self._create_testtar()
self._add_testfile()
self._test(names=["foo", "bar"])
def test_append_gz(self):
if gzip is None:
return
self._create_testtar("w:gz")
self.assertRaises(tarfile.ReadError, tarfile.open, tmpname, "a")
def test_append_bz2(self):
if bz2 is None:
return
self._create_testtar("w:bz2")
self.assertRaises(tarfile.ReadError, tarfile.open, tmpname, "a")
# Append mode is supposed to fail if the tarfile to append to
# does not end with a zero block.
def _test_error(self, data):
open(self.tarname, "wb").write(data)
self.assertRaises(tarfile.ReadError, self._add_testfile)
def test_null(self):
self._test_error("")
def test_incomplete(self):
self._test_error("\0" * 13)
def test_premature_eof(self):
data = tarfile.TarInfo("foo").tobuf()
self._test_error(data)
def test_trailing_garbage(self):
data = tarfile.TarInfo("foo").tobuf()
self._test_error(data + "\0" * 13)
def test_invalid(self):
self._test_error("a" * 512)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
class LimitsTest(unittest.TestCase):
def test_ustar_limits(self):
# 100 char name
tarinfo = tarfile.TarInfo("0123456789" * 10)
tarinfo.tobuf(tarfile.USTAR_FORMAT)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
# 101 char name that cannot be stored
tarinfo = tarfile.TarInfo("0123456789" * 10 + "0")
self.assertRaises(ValueError, tarinfo.tobuf, tarfile.USTAR_FORMAT)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
# 256 char name with a slash at pos 156
tarinfo = tarfile.TarInfo("123/" * 62 + "longname")
tarinfo.tobuf(tarfile.USTAR_FORMAT)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
# 256 char name that cannot be stored
tarinfo = tarfile.TarInfo("1234567/" * 31 + "longname")
self.assertRaises(ValueError, tarinfo.tobuf, tarfile.USTAR_FORMAT)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
# 512 char name
tarinfo = tarfile.TarInfo("123/" * 126 + "longname")
self.assertRaises(ValueError, tarinfo.tobuf, tarfile.USTAR_FORMAT)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
# 512 char linkname
tarinfo = tarfile.TarInfo("longlink")
tarinfo.linkname = "123/" * 126 + "longname"
self.assertRaises(ValueError, tarinfo.tobuf, tarfile.USTAR_FORMAT)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
# uid > 8 digits
tarinfo = tarfile.TarInfo("name")
tarinfo.uid = 010000000
self.assertRaises(ValueError, tarinfo.tobuf, tarfile.USTAR_FORMAT)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
def test_gnu_limits(self):
tarinfo = tarfile.TarInfo("123/" * 126 + "longname")
tarinfo.tobuf(tarfile.GNU_FORMAT)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
tarinfo = tarfile.TarInfo("longlink")
tarinfo.linkname = "123/" * 126 + "longname"
tarinfo.tobuf(tarfile.GNU_FORMAT)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
# uid >= 256 ** 7
tarinfo = tarfile.TarInfo("name")
tarinfo.uid = 04000000000000000000L
self.assertRaises(ValueError, tarinfo.tobuf, tarfile.GNU_FORMAT)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
def test_pax_limits(self):
tarinfo = tarfile.TarInfo("123/" * 126 + "longname")
tarinfo.tobuf(tarfile.PAX_FORMAT)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
tarinfo = tarfile.TarInfo("longlink")
tarinfo.linkname = "123/" * 126 + "longname"
tarinfo.tobuf(tarfile.PAX_FORMAT)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
tarinfo = tarfile.TarInfo("name")
tarinfo.uid = 04000000000000000000L
tarinfo.tobuf(tarfile.PAX_FORMAT)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
class ContextManagerTest(unittest.TestCase):
def test_basic(self):
with tarfile.open(tarname) as tar:
self.assertFalse(tar.closed, "closed inside runtime context")
self.assertTrue(tar.closed, "context manager failed")
def test_closed(self):
# The __enter__() method is supposed to raise IOError
# if the TarFile object is already closed.
tar = tarfile.open(tarname)
tar.close()
with self.assertRaises(IOError):
with tar:
pass
def test_exception(self):
# Test if the IOError exception is passed through properly.
with self.assertRaises(Exception) as exc:
with tarfile.open(tarname) as tar:
raise IOError
self.assertIsInstance(exc.exception, IOError,
"wrong exception raised in context manager")
self.assertTrue(tar.closed, "context manager failed")
def test_no_eof(self):
# __exit__() must not write end-of-archive blocks if an
# exception was raised.
try:
with tarfile.open(tmpname, "w") as tar:
raise Exception
except:
pass
self.assertEqual(os.path.getsize(tmpname), 0,
"context manager wrote an end-of-archive block")
self.assertTrue(tar.closed, "context manager failed")
def test_eof(self):
# __exit__() must write end-of-archive blocks, i.e. call
# TarFile.close() if there was no error.
with tarfile.open(tmpname, "w"):
pass
self.assertNotEqual(os.path.getsize(tmpname), 0,
"context manager wrote no end-of-archive block")
def test_fileobj(self):
# Test that __exit__() did not close the external file
# object.
fobj = open(tmpname, "wb")
try:
with tarfile.open(fileobj=fobj, mode="w") as tar:
raise Exception
except:
pass
self.assertFalse(fobj.closed, "external file object was closed")
self.assertTrue(tar.closed, "context manager failed")
fobj.close()
class LinkEmulationTest(ReadTest):
# Test for issue #8741 regression. On platforms that do not support
# symbolic or hard links tarfile tries to extract these types of members as
# the regular files they point to.
def _test_link_extraction(self, name):
self.tar.extract(name, TEMPDIR)
data = open(os.path.join(TEMPDIR, name), "rb").read()
self.assertEqual(md5sum(data), md5_regtype)
def test_hardlink_extraction1(self):
self._test_link_extraction("ustar/lnktype")
def test_hardlink_extraction2(self):
self._test_link_extraction("./ustar/linktest2/lnktype")
def test_symlink_extraction1(self):
self._test_link_extraction("ustar/symtype")
def test_symlink_extraction2(self):
self._test_link_extraction("./ustar/linktest2/symtype")
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
class GzipMiscReadTest(MiscReadTest):
tarname = gzipname
mode = "r:gz"
class GzipUstarReadTest(UstarReadTest):
tarname = gzipname
mode = "r:gz"
class GzipStreamReadTest(StreamReadTest):
tarname = gzipname
mode = "r|gz"
class GzipWriteTest(WriteTest):
mode = "w:gz"
class GzipStreamWriteTest(StreamWriteTest):
mode = "w|gz"
class Bz2MiscReadTest(MiscReadTest):
tarname = bz2name
mode = "r:bz2"
class Bz2UstarReadTest(UstarReadTest):
tarname = bz2name
mode = "r:bz2"
class Bz2StreamReadTest(StreamReadTest):
tarname = bz2name
mode = "r|bz2"
class Bz2WriteTest(WriteTest):
mode = "w:bz2"
class Bz2StreamWriteTest(StreamWriteTest):
mode = "w|bz2"
class Bz2PartialReadTest(unittest.TestCase):
# Issue5068: The _BZ2Proxy.read() method loops forever
# on an empty or partial bzipped file.
def _test_partial_input(self, mode):
class MyStringIO(StringIO.StringIO):
hit_eof = False
def read(self, n):
if self.hit_eof:
raise AssertionError("infinite loop detected in tarfile.open()")
self.hit_eof = self.pos == self.len
return StringIO.StringIO.read(self, n)
def seek(self, *args):
self.hit_eof = False
return StringIO.StringIO.seek(self, *args)
data = bz2.compress(tarfile.TarInfo("foo").tobuf())
for x in range(len(data) + 1):
try:
tarfile.open(fileobj=MyStringIO(data[:x]), mode=mode)
except tarfile.ReadError:
pass # we have no interest in ReadErrors
def test_partial_input(self):
self._test_partial_input("r")
def test_partial_input_bz2(self):
self._test_partial_input("r:bz2")
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
def test_main():
os.makedirs(TEMPDIR)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
tests = [
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
UstarReadTest,
MiscReadTest,
StreamReadTest,
DetectReadTest,
MemberReadTest,
GNUReadTest,
PaxReadTest,
WriteTest,
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
StreamWriteTest,
GNUWriteTest,
PaxWriteTest,
UstarUnicodeTest,
GNUUnicodeTest,
PaxUnicodeTest,
AppendTest,
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
LimitsTest,
ContextManagerTest,
]
if hasattr(os, "link"):
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
tests.append(HardlinkTest)
else:
tests.append(LinkEmulationTest)
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
fobj = open(tarname, "rb")
data = fobj.read()
fobj.close()
if gzip:
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
# Create testtar.tar.gz and add gzip-specific tests.
tar = gzip.open(gzipname, "wb")
tar.write(data)
tar.close()
tests += [
GzipMiscReadTest,
GzipUstarReadTest,
GzipStreamReadTest,
GzipWriteTest,
GzipStreamWriteTest,
]
if bz2:
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
# Create testtar.tar.bz2 and add bz2-specific tests.
tar = bz2.BZ2File(bz2name, "wb")
tar.write(data)
tar.close()
tests += [
Bz2MiscReadTest,
Bz2UstarReadTest,
Bz2StreamReadTest,
Bz2WriteTest,
Bz2StreamWriteTest,
Bz2PartialReadTest,
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
]
try:
test_support.run_unittest(*tests)
finally:
This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'.
2007-03-13 07:47:19 -03:00
if os.path.exists(TEMPDIR):
shutil.rmtree(TEMPDIR)
if __name__ == "__main__":
test_main()