diff --git a/Doc/library/gzip.rst b/Doc/library/gzip.rst index 1f644286c9c..a8305936157 100644 --- a/Doc/library/gzip.rst +++ b/Doc/library/gzip.rst @@ -72,6 +72,9 @@ The module defines the following items: .. versionchanged:: 3.1 Support for the :keyword:`with` statement was added. + .. versionchanged:: 3.2 + Support for zero-padded files was added. + .. function:: open(filename, mode='rb', compresslevel=9) diff --git a/Lib/gzip.py b/Lib/gzip.py index 66fc88daa0d..ef6befcb7d3 100644 --- a/Lib/gzip.py +++ b/Lib/gzip.py @@ -348,6 +348,15 @@ class GzipFile(io.BufferedIOBase): elif isize != (self.size & 0xffffffff): raise IOError("Incorrect length of data produced") + # Gzip files can be padded with zeroes and still have archives. + # Consume all zero bytes and set the file position to the first + # non-zero byte. See http://www.gzip.org/#faq8 + c = b"\x00" + while c == b"\x00": + c = self.fileobj.read(1) + if c: + self.fileobj.seek(-1, 1) + @property def closed(self): return self.fileobj is None diff --git a/Lib/test/test_gzip.py b/Lib/test/test_gzip.py index 320adfda8a0..429ada0dcd4 100644 --- a/Lib/test/test_gzip.py +++ b/Lib/test/test_gzip.py @@ -253,6 +253,18 @@ class TestGzip(unittest.TestCase): else: self.fail("1/0 didn't raise an exception") + def test_zero_padded_file(self): + with gzip.GzipFile(self.filename, "wb") as f: + f.write(data1 * 50) + + # Pad the file with zeroes + with open(self.filename, "ab") as f: + f.write(b"\x00" * 50) + + with gzip.GzipFile(self.filename, "rb") as f: + d = f.read() + self.assertEqual(d, data1 * 50, "Incorrect data in file") + def test_main(verbose=None): support.run_unittest(TestGzip) diff --git a/Misc/ACKS b/Misc/ACKS index 72e2096c22c..f0c667a57c8 100644 --- a/Misc/ACKS +++ b/Misc/ACKS @@ -164,6 +164,7 @@ Simon Cross Drew Csillag John Cugini Tom Culliton +Brian Curtin Lisandro Dalcin Andrew Dalke Lars Damerow diff --git a/Misc/NEWS b/Misc/NEWS index 3e4073d43d0..a577e65a239 100644 --- a/Misc/NEWS +++ b/Misc/NEWS @@ -213,6 +213,9 @@ C-API Library ------- +- Issue #2846: Add support for gzip.GzipFile reading zero-padded files. + Patch by Brian Curtin. + - Issue #7681: Use floor division in appropiate places in the wave module. - Issue #5372: Drop the reuse of .o files in Distutils' ccompiler (since