Issue #11461: Fix the incremental UTF-16 decoder. Original patch by

Amaury Forgeot d'Arc. Added tests for partial decoding of non-BMP
characters.

3 files changed, 47 insertions, 9 deletions

diff --git a/Lib/test/test_codecs.py b/Lib/test/test_codecs.py
index f2a1ae3..15cd687 100644
--- a/Lib/test/test_codecs.py
+++ b/Lib/test/test_codecs.py
@@ -330,7 +330,7 @@ class UTF32Test(ReadTest):
 
     def test_partial(self):
         self.check_partial(
-            "\x00\xff\u0100\uffff",
+            "\x00\xff\u0100\uffff\U00010000",
             [
                 "", # first byte of BOM read
                 "", # second byte of BOM read
@@ -352,6 +352,10 @@ class UTF32Test(ReadTest):
                 "\x00\xff\u0100",
                 "\x00\xff\u0100",
                 "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff\U00010000",
             ]
         )
 
@@ -386,7 +390,7 @@ class UTF32LETest(ReadTest):
 
     def test_partial(self):
         self.check_partial(
-            "\x00\xff\u0100\uffff",
+            "\x00\xff\u0100\uffff\U00010000",
             [
                 "",
                 "",
@@ -404,6 +408,10 @@ class UTF32LETest(ReadTest):
                 "\x00\xff\u0100",
                 "\x00\xff\u0100",
                 "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff\U00010000",
             ]
         )
 
@@ -426,7 +434,7 @@ class UTF32BETest(ReadTest):
 
     def test_partial(self):
         self.check_partial(
-            "\x00\xff\u0100\uffff",
+            "\x00\xff\u0100\uffff\U00010000",
             [
                 "",
                 "",
@@ -444,6 +452,10 @@ class UTF32BETest(ReadTest):
                 "\x00\xff\u0100",
                 "\x00\xff\u0100",
                 "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff\U00010000",
             ]
         )
 
@@ -494,7 +506,7 @@ class UTF16Test(ReadTest):
 
     def test_partial(self):
         self.check_partial(
-            "\x00\xff\u0100\uffff",
+            "\x00\xff\u0100\uffff\U00010000",
             [
                 "", # first byte of BOM read
                 "", # second byte of BOM read => byteorder known
@@ -506,6 +518,10 @@ class UTF16Test(ReadTest):
                 "\x00\xff\u0100",
                 "\x00\xff\u0100",
                 "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff\U00010000",
             ]
         )
 
@@ -543,7 +559,7 @@ class UTF16LETest(ReadTest):
 
     def test_partial(self):
         self.check_partial(
-            "\x00\xff\u0100\uffff",
+            "\x00\xff\u0100\uffff\U00010000",
             [
                 "",
                 "\x00",
@@ -553,6 +569,10 @@ class UTF16LETest(ReadTest):
                 "\x00\xff\u0100",
                 "\x00\xff\u0100",
                 "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff\U00010000",
             ]
         )
 
@@ -582,7 +602,7 @@ class UTF16BETest(ReadTest):
 
     def test_partial(self):
         self.check_partial(
-            "\x00\xff\u0100\uffff",
+            "\x00\xff\u0100\uffff\U00010000",
             [
                 "",
                 "\x00",
@@ -592,6 +612,10 @@ class UTF16BETest(ReadTest):
                 "\x00\xff\u0100",
                 "\x00\xff\u0100",
                 "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff\U00010000",
             ]
         )
 
@@ -621,7 +645,7 @@ class UTF8Test(ReadTest):
 
     def test_partial(self):
         self.check_partial(
-            "\x00\xff\u07ff\u0800\uffff",
+            "\x00\xff\u07ff\u0800\uffff\U00010000",
             [
                 "\x00",
                 "\x00",
@@ -634,6 +658,10 @@ class UTF8Test(ReadTest):
                 "\x00\xff\u07ff\u0800",
                 "\x00\xff\u07ff\u0800",
                 "\x00\xff\u07ff\u0800\uffff",
+                "\x00\xff\u07ff\u0800\uffff",
+                "\x00\xff\u07ff\u0800\uffff",
+                "\x00\xff\u07ff\u0800\uffff",
+                "\x00\xff\u07ff\u0800\uffff\U00010000",
             ]
         )
 
@@ -816,7 +844,7 @@ class UTF8SigTest(ReadTest):
 
     def test_partial(self):
         self.check_partial(
-            "\ufeff\x00\xff\u07ff\u0800\uffff",
+            "\ufeff\x00\xff\u07ff\u0800\uffff\U00010000",
             [
                 "",
                 "",
@@ -835,6 +863,10 @@ class UTF8SigTest(ReadTest):
                 "\ufeff\x00\xff\u07ff\u0800",
                 "\ufeff\x00\xff\u07ff\u0800",
                 "\ufeff\x00\xff\u07ff\u0800\uffff",
+                "\ufeff\x00\xff\u07ff\u0800\uffff",
+                "\ufeff\x00\xff\u07ff\u0800\uffff",
+                "\ufeff\x00\xff\u07ff\u0800\uffff",
+                "\ufeff\x00\xff\u07ff\u0800\uffff\U00010000",
             ]
         )
 
diff --git a/Misc/NEWS b/Misc/NEWS
index b40d4c7..8a5d14d 100644
--- a/Misc/NEWS
+++ b/Misc/NEWS
@@ -12,6 +12,9 @@ What's New in Python 3.3.1?
 Core and Builtins
 -----------------
 
+- Issue #11461: Fix the incremental UTF-16 decoder. Original patch by
+  Amaury Forgeot d'Arc.
+
 - Issue #16881: Fix Py_ARRAY_LENGTH macro for GCC < 3.1.
 
 - Issue #16856: Fix a segmentation fault from calling repr() on a dict with
diff --git a/Objects/unicodeobject.c b/Objects/unicodeobject.c
index 1522a16..b4fc004 100644
--- a/Objects/unicodeobject.c
+++ b/Objects/unicodeobject.c
@@ -5284,8 +5284,11 @@ PyUnicode_DecodeUTF16Stateful(const char *s,
             /* The remaining input chars are ignored if the callback
                chooses to skip the input */
         case 1:
+            q -= 2;
+            if (consumed)
+                goto End;
             errmsg = "unexpected end of data";
-            startinpos = ((const char *)q) - 2 - starts;
+            startinpos = ((const char *)q) - starts;
             endinpos = ((const char *)e) - starts;
             break;
         case 2: