Issue python#14366: Support lzma compression in zip files.

Patch by Serhiy Storchaka.

Author: loewis

Doc/library/zipfile.rst

@@ -97,12 +97,20 @@ The module defines the following items:
 
    .. versionadded:: 3.3
 
+.. data:: ZIP_LZMA
+
+   The numeric constant for the LZMA compression method.  This requires the
+   lzma module.
+
+   .. versionadded:: 3.3
+
    .. note::
 
       The ZIP file format specification has included support for bzip2 compression
-      since 2001. However, some tools (including older Python releases) do not
-      support it, and may either refuse to process the ZIP file altogether, or
-      fail to extract individual files.
+      since 2001, and for LZMA compression since 2006. However, some tools
+      (including older Python releases) do not support these compression
+      methods, and may either refuse to process the ZIP file altogether,
+      or fail to extract individual files.
 
 
 .. seealso::
@@ -133,11 +141,11 @@ ZipFile Objects
    adding a ZIP archive to another file (such as :file:`python.exe`).  If
    *mode* is ``a`` and the file does not exist at all, it is created.
    *compression* is the ZIP compression method to use when writing the archive,
-   and should be :const:`ZIP_STORED`, :const:`ZIP_DEFLATED`; or
-   :const:`ZIP_DEFLATED`; unrecognized
-   values will cause :exc:`RuntimeError` to be raised.  If :const:`ZIP_DEFLATED` or
-   :const:`ZIP_BZIP2` is specified but the corresponded module
-   (:mod:`zlib` or :mod:`bz2`) is not available, :exc:`RuntimeError`
+   and should be :const:`ZIP_STORED`, :const:`ZIP_DEFLATED`,
+   :const:`ZIP_BZIP2` or :const:`ZIP_LZMA`; unrecognized
+   values will cause :exc:`RuntimeError` to be raised.  If :const:`ZIP_DEFLATED`,
+   :const:`ZIP_BZIP2` or :const:`ZIP_LZMA` is specified but the corresponded module
+   (:mod:`zlib`, :mod:`bz2` or :mod:`lzma`) is not available, :exc:`RuntimeError`
    is also raised. The default is :const:`ZIP_STORED`.  If *allowZip64* is
    ``True`` zipfile will create ZIP files that use the ZIP64 extensions when
    the zipfile is larger than 2 GB. If it is  false (the default) :mod:`zipfile`
@@ -161,7 +169,7 @@ ZipFile Objects
       Added the ability to use :class:`ZipFile` as a context manager.
 
    .. versionchanged:: 3.3
-      Added support for :mod:`bzip2` compression.
+      Added support for :mod:`bzip2` and :mod:`lzma` compression.
 
 
 .. method:: ZipFile.close()

Lib/test/support.py

@@ -45,6 +45,11 @@
 except ImportError:
     bz2 = None
 
+try:
+    import lzma
+except ImportError:
+    lzma = None
+
 __all__ = [
     "Error", "TestFailed", "ResourceDenied", "import_module",
     "verbose", "use_resources", "max_memuse", "record_original_stdout",
@@ -62,7 +67,7 @@
     "get_attribute", "swap_item", "swap_attr", "requires_IEEE_754",
     "TestHandler", "Matcher", "can_symlink", "skip_unless_symlink",
     "import_fresh_module", "requires_zlib", "PIPE_MAX_SIZE", "failfast",
-    "anticipate_failure", "run_with_tz", "requires_bz2"
+    "anticipate_failure", "run_with_tz", "requires_bz2", "requires_lzma"
     ]
 
 class Error(Exception):
@@ -513,6 +518,8 @@ def _is_ipv6_enabled():
 
 requires_bz2 = unittest.skipUnless(bz2, 'requires bz2')
 
+requires_lzma = unittest.skipUnless(lzma, 'requires lzma')
+
 is_jython = sys.platform.startswith('java')
 
 # Filename used for testing

Lib/test/test_zipfile.py

@@ -13,7 +13,7 @@
 from random import randint, random
 from unittest import skipUnless
 
-from test.support import TESTFN, run_unittest, findfile, unlink, requires_zlib, requires_bz2
+from test.support import TESTFN, run_unittest, findfile, unlink, requires_zlib, requires_bz2, requires_lzma
 
 TESTFN2 = TESTFN + "2"
 TESTFNDIR = TESTFN + "d"
@@ -361,6 +361,55 @@ def test_low_compression_bzip2(self):
                 self.assertEqual(openobj.read(1), b'1')
                 self.assertEqual(openobj.read(1), b'2')
 
+    @requires_lzma
+    def test_lzma(self):
+        for f in (TESTFN2, TemporaryFile(), io.BytesIO()):
+            self.zip_test(f, zipfile.ZIP_LZMA)
+
+    @requires_lzma
+    def test_open_lzma(self):
+        for f in (TESTFN2, TemporaryFile(), io.BytesIO()):
+            self.zip_open_test(f, zipfile.ZIP_LZMA)
+
+    @requires_lzma
+    def test_random_open_lzma(self):
+        for f in (TESTFN2, TemporaryFile(), io.BytesIO()):
+            self.zip_random_open_test(f, zipfile.ZIP_LZMA)
+
+    @requires_lzma
+    def test_readline_read_lzma(self):
+        # Issue #7610: calls to readline() interleaved with calls to read().
+        for f in (TESTFN2, TemporaryFile(), io.BytesIO()):
+            self.zip_readline_read_test(f, zipfile.ZIP_LZMA)
+
+    @requires_lzma
+    def test_readline_lzma(self):
+        for f in (TESTFN2, TemporaryFile(), io.BytesIO()):
+            self.zip_readline_test(f, zipfile.ZIP_LZMA)
+
+    @requires_lzma
+    def test_readlines_lzma(self):
+        for f in (TESTFN2, TemporaryFile(), io.BytesIO()):
+            self.zip_readlines_test(f, zipfile.ZIP_LZMA)
+
+    @requires_lzma
+    def test_iterlines_lzma(self):
+        for f in (TESTFN2, TemporaryFile(), io.BytesIO()):
+            self.zip_iterlines_test(f, zipfile.ZIP_LZMA)
+
+    @requires_lzma
+    def test_low_compression_lzma(self):
+        """Check for cases where compressed data is larger than original."""
+        # Create the ZIP archive
+        with zipfile.ZipFile(TESTFN2, "w", zipfile.ZIP_LZMA) as zipfp:
+            zipfp.writestr("strfile", '12')
+
+        # Get an open object for strfile
+        with zipfile.ZipFile(TESTFN2, "r", zipfile.ZIP_LZMA) as zipfp:
+            with zipfp.open("strfile") as openobj:
+                self.assertEqual(openobj.read(1), b'1')
+                self.assertEqual(openobj.read(1), b'2')
+
     def test_absolute_arcnames(self):
         with zipfile.ZipFile(TESTFN2, "w", zipfile.ZIP_STORED) as zipfp:
             zipfp.write(TESTFN, "/absolute")
@@ -508,6 +557,13 @@ def test_writestr_compression_bzip2(self):
         info = zipfp.getinfo('b.txt')
         self.assertEqual(info.compress_type, zipfile.ZIP_BZIP2)
 
+    @requires_lzma
+    def test_writestr_compression_lzma(self):
+        zipfp = zipfile.ZipFile(TESTFN2, "w")
+        zipfp.writestr("b.txt", "hello world", compress_type=zipfile.ZIP_LZMA)
+        info = zipfp.getinfo('b.txt')
+        self.assertEqual(info.compress_type, zipfile.ZIP_LZMA)
+
     def zip_test_writestr_permissions(self, f, compression):
         # Make sure that writestr creates files with mode 0600,
         # when it is passed a name rather than a ZipInfo instance.
@@ -686,6 +742,11 @@ def test_bzip2(self):
         for f in (TESTFN2, TemporaryFile(), io.BytesIO()):
             self.zip_test(f, zipfile.ZIP_BZIP2)
 
+    @requires_lzma
+    def test_lzma(self):
+        for f in (TESTFN2, TemporaryFile(), io.BytesIO()):
+            self.zip_test(f, zipfile.ZIP_LZMA)
+
     def test_absolute_arcnames(self):
         with zipfile.ZipFile(TESTFN2, "w", zipfile.ZIP_STORED,
                              allowZip64=True) as zipfp:
@@ -826,6 +887,16 @@ class OtherTests(unittest.TestCase):
             b'\x00 \x80\x80\x81\x00\x00\x00\x00afilePK'
             b'\x05\x06\x00\x00\x00\x00\x01\x00\x01\x003\x00\x00\x00[\x00'
             b'\x00\x00\x00\x00'),
+        zipfile.ZIP_LZMA: (
+            b'PK\x03\x04\x14\x03\x00\x00\x0e\x00nu\x0c=FA'
+            b'KE\x1b\x00\x00\x00n\x00\x00\x00\x05\x00\x00\x00af'
+            b'ile\t\x04\x05\x00]\x00\x00\x00\x04\x004\x19I'
+            b'\xee\x8d\xe9\x17\x89:3`\tq!.8\x00PK'
+            b'\x01\x02\x14\x03\x14\x03\x00\x00\x0e\x00nu\x0c=FA'
+            b'KE\x1b\x00\x00\x00n\x00\x00\x00\x05\x00\x00\x00\x00\x00'
+            b'\x00\x00\x00\x00 \x80\x80\x81\x00\x00\x00\x00afil'
+            b'ePK\x05\x06\x00\x00\x00\x00\x01\x00\x01\x003\x00\x00'
+            b'\x00>\x00\x00\x00\x00\x00'),
     }
 
     def test_unsupported_version(self):
@@ -1104,6 +1175,10 @@ def test_testzip_with_bad_crc_deflated(self):
     def test_testzip_with_bad_crc_bzip2(self):
         self.check_testzip_with_bad_crc(zipfile.ZIP_BZIP2)
 
+    @requires_lzma
+    def test_testzip_with_bad_crc_lzma(self):
+        self.check_testzip_with_bad_crc(zipfile.ZIP_LZMA)
+
     def check_read_with_bad_crc(self, compression):
         """Tests that files with bad CRCs raise a BadZipFile exception when read."""
         zipdata = self.zips_with_bad_crc[compression]
@@ -1136,6 +1211,10 @@ def test_read_with_bad_crc_deflated(self):
     def test_read_with_bad_crc_bzip2(self):
         self.check_read_with_bad_crc(zipfile.ZIP_BZIP2)
 
+    @requires_lzma
+    def test_read_with_bad_crc_lzma(self):
+        self.check_read_with_bad_crc(zipfile.ZIP_LZMA)
+
     def check_read_return_size(self, compression):
         # Issue #9837: ZipExtFile.read() shouldn't return more bytes
         # than requested.
@@ -1160,6 +1239,10 @@ def test_read_return_size_deflated(self):
     def test_read_return_size_bzip2(self):
         self.check_read_return_size(zipfile.ZIP_BZIP2)
 
+    @requires_lzma
+    def test_read_return_size_lzma(self):
+        self.check_read_return_size(zipfile.ZIP_LZMA)
+
     def test_empty_zipfile(self):
         # Check that creating a file in 'w' or 'a' mode and closing without
         # adding any files to the archives creates a valid empty ZIP file
@@ -1306,6 +1389,11 @@ def test_bzip2(self):
         for f in (TESTFN2, TemporaryFile(), io.BytesIO()):
             self.zip_test(f, zipfile.ZIP_BZIP2)
 
+    @requires_lzma
+    def test_lzma(self):
+        for f in (TESTFN2, TemporaryFile(), io.BytesIO()):
+            self.zip_test(f, zipfile.ZIP_LZMA)
+
     def zip_open_test(self, f, compression):
         self.make_test_archive(f, compression)
 
@@ -1351,6 +1439,11 @@ def test_open_bzip2(self):
         for f in (TESTFN2, TemporaryFile(), io.BytesIO()):
             self.zip_open_test(f, zipfile.ZIP_BZIP2)
 
+    @requires_lzma
+    def test_open_lzma(self):
+        for f in (TESTFN2, TemporaryFile(), io.BytesIO()):
+            self.zip_open_test(f, zipfile.ZIP_LZMA)
+
     def zip_random_open_test(self, f, compression):
         self.make_test_archive(f, compression)
 
@@ -1384,6 +1477,11 @@ def test_random_open_bzip2(self):
         for f in (TESTFN2, TemporaryFile(), io.BytesIO()):
             self.zip_random_open_test(f, zipfile.ZIP_BZIP2)
 
+    @requires_lzma
+    def test_random_open_lzma(self):
+        for f in (TESTFN2, TemporaryFile(), io.BytesIO()):
+            self.zip_random_open_test(f, zipfile.ZIP_LZMA)
+
 
 @requires_zlib
 class TestsWithMultipleOpens(unittest.TestCase):
@@ -1628,6 +1726,31 @@ def test_iterlines_bzip2(self):
         for f in (TESTFN2, TemporaryFile(), io.BytesIO()):
             self.iterlines_test(f, zipfile.ZIP_BZIP2)
 
+    @requires_lzma
+    def test_read_lzma(self):
+        for f in (TESTFN2, TemporaryFile(), io.BytesIO()):
+            self.read_test(f, zipfile.ZIP_LZMA)
+
+    @requires_lzma
+    def test_readline_read_lzma(self):
+        for f in (TESTFN2, TemporaryFile(), io.BytesIO()):
+            self.readline_read_test(f, zipfile.ZIP_LZMA)
+
+    @requires_lzma
+    def test_readline_lzma(self):
+        for f in (TESTFN2, TemporaryFile(), io.BytesIO()):
+            self.readline_test(f, zipfile.ZIP_LZMA)
+
+    @requires_lzma
+    def test_readlines_lzma(self):
+        for f in (TESTFN2, TemporaryFile(), io.BytesIO()):
+            self.readlines_test(f, zipfile.ZIP_LZMA)
+
+    @requires_lzma
+    def test_iterlines_lzma(self):
+        for f in (TESTFN2, TemporaryFile(), io.BytesIO()):
+            self.iterlines_test(f, zipfile.ZIP_LZMA)
+
     def tearDown(self):
         for sep, fn in self.arcfiles.items():
             os.remove(fn)