Issue #11461: Fix the incremental UTF-16 decoder. Original patch by
Amaury Forgeot d'Arc. Added tests for partial decoding of non-BMP
characters.
diff --git a/Lib/test/test_codecs.py b/Lib/test/test_codecs.py
index f620b4c..5baf225 100644
--- a/Lib/test/test_codecs.py
+++ b/Lib/test/test_codecs.py
@@ -281,7 +281,7 @@
def test_partial(self):
self.check_partial(
- u"\x00\xff\u0100\uffff",
+ u"\x00\xff\u0100\uffff\U00010000",
[
u"", # first byte of BOM read
u"", # second byte of BOM read
@@ -303,6 +303,10 @@
u"\x00\xff\u0100",
u"\x00\xff\u0100",
u"\x00\xff\u0100\uffff",
+ u"\x00\xff\u0100\uffff",
+ u"\x00\xff\u0100\uffff",
+ u"\x00\xff\u0100\uffff",
+ u"\x00\xff\u0100\uffff\U00010000",
]
)
@@ -331,7 +335,7 @@
def test_partial(self):
self.check_partial(
- u"\x00\xff\u0100\uffff",
+ u"\x00\xff\u0100\uffff\U00010000",
[
u"",
u"",
@@ -349,6 +353,10 @@
u"\x00\xff\u0100",
u"\x00\xff\u0100",
u"\x00\xff\u0100\uffff",
+ u"\x00\xff\u0100\uffff",
+ u"\x00\xff\u0100\uffff",
+ u"\x00\xff\u0100\uffff",
+ u"\x00\xff\u0100\uffff\U00010000",
]
)
@@ -371,7 +379,7 @@
def test_partial(self):
self.check_partial(
- u"\x00\xff\u0100\uffff",
+ u"\x00\xff\u0100\uffff\U00010000",
[
u"",
u"",
@@ -389,6 +397,10 @@
u"\x00\xff\u0100",
u"\x00\xff\u0100",
u"\x00\xff\u0100\uffff",
+ u"\x00\xff\u0100\uffff",
+ u"\x00\xff\u0100\uffff",
+ u"\x00\xff\u0100\uffff",
+ u"\x00\xff\u0100\uffff\U00010000",
]
)
@@ -439,7 +451,7 @@
def test_partial(self):
self.check_partial(
- u"\x00\xff\u0100\uffff",
+ u"\x00\xff\u0100\uffff\U00010000",
[
u"", # first byte of BOM read
u"", # second byte of BOM read => byteorder known
@@ -451,6 +463,10 @@
u"\x00\xff\u0100",
u"\x00\xff\u0100",
u"\x00\xff\u0100\uffff",
+ u"\x00\xff\u0100\uffff",
+ u"\x00\xff\u0100\uffff",
+ u"\x00\xff\u0100\uffff",
+ u"\x00\xff\u0100\uffff\U00010000",
]
)
@@ -481,7 +497,7 @@
def test_partial(self):
self.check_partial(
- u"\x00\xff\u0100\uffff",
+ u"\x00\xff\u0100\uffff\U00010000",
[
u"",
u"\x00",
@@ -491,6 +507,10 @@
u"\x00\xff\u0100",
u"\x00\xff\u0100",
u"\x00\xff\u0100\uffff",
+ u"\x00\xff\u0100\uffff",
+ u"\x00\xff\u0100\uffff",
+ u"\x00\xff\u0100\uffff",
+ u"\x00\xff\u0100\uffff\U00010000",
]
)
@@ -514,7 +534,7 @@
def test_partial(self):
self.check_partial(
- u"\x00\xff\u0100\uffff",
+ u"\x00\xff\u0100\uffff\U00010000",
[
u"",
u"\x00",
@@ -524,6 +544,10 @@
u"\x00\xff\u0100",
u"\x00\xff\u0100",
u"\x00\xff\u0100\uffff",
+ u"\x00\xff\u0100\uffff",
+ u"\x00\xff\u0100\uffff",
+ u"\x00\xff\u0100\uffff",
+ u"\x00\xff\u0100\uffff\U00010000",
]
)
@@ -547,7 +571,7 @@
def test_partial(self):
self.check_partial(
- u"\x00\xff\u07ff\u0800\uffff",
+ u"\x00\xff\u07ff\u0800\uffff\U00010000",
[
u"\x00",
u"\x00",
@@ -560,6 +584,10 @@
u"\x00\xff\u07ff\u0800",
u"\x00\xff\u07ff\u0800",
u"\x00\xff\u07ff\u0800\uffff",
+ u"\x00\xff\u07ff\u0800\uffff",
+ u"\x00\xff\u07ff\u0800\uffff",
+ u"\x00\xff\u07ff\u0800\uffff",
+ u"\x00\xff\u07ff\u0800\uffff\U00010000",
]
)
@@ -619,7 +647,7 @@
def test_partial(self):
self.check_partial(
- u"\ufeff\x00\xff\u07ff\u0800\uffff",
+ u"\ufeff\x00\xff\u07ff\u0800\uffff\U00010000",
[
u"",
u"",
@@ -638,6 +666,10 @@
u"\ufeff\x00\xff\u07ff\u0800",
u"\ufeff\x00\xff\u07ff\u0800",
u"\ufeff\x00\xff\u07ff\u0800\uffff",
+ u"\ufeff\x00\xff\u07ff\u0800\uffff",
+ u"\ufeff\x00\xff\u07ff\u0800\uffff",
+ u"\ufeff\x00\xff\u07ff\u0800\uffff",
+ u"\ufeff\x00\xff\u07ff\u0800\uffff\U00010000",
]
)