From da96e479ed930ca56d7007cb218da7b4192c0ae5 Mon Sep 17 00:00:00 2001
From: Hye-Shik Chang <hyeshik@gmail.com>
Date: Tue, 5 Jun 2007 19:02:59 +0000
Subject: (Backport from r55770)

Bug #1728403: Fix a bug that CJKCodecs StreamReader hangs when it
reads a file that ends with incomplete sequence and sizehint argument
for .read() is specified.
---
 Lib/test/test_multibytecodec.py    | 13 ++++++++++++-
 Misc/NEWS                          |  4 ++++
 Modules/cjkcodecs/multibytecodec.c |  6 +++++-
 3 files changed, 21 insertions(+), 2 deletions(-)

diff --git a/Lib/test/test_multibytecodec.py b/Lib/test/test_multibytecodec.py
index 3fcc8ba81a..052f1681dd 100644
--- a/Lib/test/test_multibytecodec.py
+++ b/Lib/test/test_multibytecodec.py
@@ -7,7 +7,17 @@
 
 from test import test_support
 from test import test_multibytecodec_support
-import unittest, StringIO, codecs, sys
+from test.test_support import TESTFN
+import unittest, StringIO, codecs, sys, os
+
+class Test_StreamReader(unittest.TestCase):
+    def test_bug1728403(self):
+        try:
+            open(TESTFN, 'w').write('\xa1')
+            f = codecs.open(TESTFN, encoding='cp949')
+            self.assertRaises(UnicodeDecodeError, f.read, 2)
+        finally:
+            os.unlink(TESTFN)
 
 class Test_StreamWriter(unittest.TestCase):
     if len(u'\U00012345') == 2: # UCS2
@@ -99,6 +109,7 @@ class Test_ISO2022(unittest.TestCase):
 
 def test_main():
     suite = unittest.TestSuite()
+    suite.addTest(unittest.makeSuite(Test_StreamReader))
     suite.addTest(unittest.makeSuite(Test_StreamWriter))
     suite.addTest(unittest.makeSuite(Test_ISO2022))
     test_support.run_suite(suite)
diff --git a/Misc/NEWS b/Misc/NEWS
index e8ea8cc01a..002ea18d9c 100644
--- a/Misc/NEWS
+++ b/Misc/NEWS
@@ -26,6 +26,10 @@ Extension Modules
 Library
 -------
 
+- Bug #1728403: Fix a bug that CJKCodecs StreamReader hangs when it
+  reads a file that ends with incomplete sequence and sizehint argument
+  for .read() is specified.
+
 - HTML-escape the plain traceback in cgitb's HTML output, to prevent
   the traceback inadvertently or maliciously closing the comment and
   injecting HTML into the error page.
diff --git a/Modules/cjkcodecs/multibytecodec.c b/Modules/cjkcodecs/multibytecodec.c
index 7d2d15ebfb..b69eefff0a 100644
--- a/Modules/cjkcodecs/multibytecodec.c
+++ b/Modules/cjkcodecs/multibytecodec.c
@@ -705,6 +705,8 @@ mbstreamreader_iread(MultibyteStreamReaderObject *self,
 	cres = NULL;
 
 	for (;;) {
+		int endoffile;
+
 		if (sizehint < 0)
 			cres = PyObject_CallMethod(self->stream,
 					(char *)method, NULL);
@@ -721,6 +723,8 @@ mbstreamreader_iread(MultibyteStreamReaderObject *self,
 			goto errorexit;
 		}
 
+		endoffile = (PyString_GET_SIZE(cres) == 0);
+
 		if (self->pendingsize > 0) {
 			PyObject *ctr;
 			char *ctrdata;
@@ -772,7 +776,7 @@ mbstreamreader_iread(MultibyteStreamReaderObject *self,
 					goto errorexit;
 			}
 
-		if (rsize == 0 || sizehint < 0) { /* end of file */
+		if (endoffile || sizehint < 0) { /* end of file */
 			if (buf.inbuf < buf.inbuf_end &&
 			    multibytecodec_decerror(self->codec, &self->state,
 					&buf, self->errors, MBERR_TOOFEW))
-- 
cgit v1.2.1