From 39079946a257522ecb66cb068c872f5d1fe70ea3 Mon Sep 17 00:00:00 2001 From: Nadeem Vawda Date: Mon, 5 Nov 2012 00:37:42 +0100 Subject: Issue #16350: Fix zlib decompressor handling of unused_data with multiple calls to decompress() after EOF. Patch by Serhiy Storchaka. --- Lib/test/test_zlib.py | 13 +++++++++++++ Misc/NEWS | 4 ++++ Modules/zlibmodule.c | 29 +++++++++++++++++++++++------ 3 files changed, 40 insertions(+), 6 deletions(-) diff --git a/Lib/test/test_zlib.py b/Lib/test/test_zlib.py index 60081e2..c17b4d0 100644 --- a/Lib/test/test_zlib.py +++ b/Lib/test/test_zlib.py @@ -434,6 +434,19 @@ class CompressObjectTestCase(BaseCompressTestCase, unittest.TestCase): y += dco.flush() self.assertEqual(y, b'foo') + def test_decompress_unused_data(self): + # Repeated calls to decompress() after EOF should accumulate data in + # dco.unused_data, instead of just storing the arg to the last call. + x = zlib.compress(HAMLET_SCENE) + HAMLET_SCENE + for step in 1, 2, 100: + dco = zlib.decompressobj() + data = b''.join(dco.decompress(x[i : i + step]) + for i in range(0, len(x), step)) + data += dco.flush() + + self.assertEqual(data, HAMLET_SCENE) + self.assertEqual(dco.unused_data, HAMLET_SCENE) + if hasattr(zlib.compressobj(), "copy"): def test_compresscopy(self): # Test copying a compression object diff --git a/Misc/NEWS b/Misc/NEWS index bb1d6f7..a7f0521 100644 --- a/Misc/NEWS +++ b/Misc/NEWS @@ -156,6 +156,10 @@ Core and Builtins Library ------- +- Issue #16350: zlib.Decompress.decompress() now accumulates data from + successive calls after EOF in unused_data, instead of only saving the argument + to the last call. Patch by Serhiy Storchaka. + - Issue #12759: sre_parse now raises a proper error when the name of the group is missing. Initial patch by Serhiy Storchaka. diff --git a/Modules/zlibmodule.c b/Modules/zlibmodule.c index a1e605b..7734ba6 100644 --- a/Modules/zlibmodule.c +++ b/Modules/zlibmodule.c @@ -610,12 +610,29 @@ PyZlib_objdecompress(compobject *self, PyObject *args) preserved. */ if (err == Z_STREAM_END) { - Py_XDECREF(self->unused_data); /* Free original empty string */ - self->unused_data = PyBytes_FromStringAndSize( - (char *)self->zst.next_in, self->zst.avail_in); - if (self->unused_data == NULL) { - Py_DECREF(RetVal); - goto error; + if (self->zst.avail_in > 0) { + /* Append the leftover data to the existing value of unused_data. */ + Py_ssize_t old_size = PyBytes_GET_SIZE(self->unused_data); + Py_ssize_t new_size = old_size + self->zst.avail_in; + PyObject *new_data; + if (new_size <= old_size) { /* Check for overflow. */ + PyErr_NoMemory(); + Py_DECREF(RetVal); + RetVal = NULL; + goto error; + } + new_data = PyBytes_FromStringAndSize(NULL, new_size); + if (new_data == NULL) { + Py_DECREF(RetVal); + RetVal = NULL; + goto error; + } + Py_MEMCPY(PyBytes_AS_STRING(new_data), + PyBytes_AS_STRING(self->unused_data), old_size); + Py_MEMCPY(PyBytes_AS_STRING(new_data) + old_size, + self->zst.next_in, self->zst.avail_in); + Py_DECREF(self->unused_data); + self->unused_data = new_data; } /* We will only get Z_BUF_ERROR if the output buffer was full but there wasn't more output when we tried again, so it is -- cgit v0.12