- parserutils_charset_utf16_to_ucs4(): fixed surrogate handling.

- cscodec-utf8.c(run_test): Added more asserts.
- Added UTF-16 tester (based on the UTF-8 one).

svn path=/trunk/libparserutils/; revision=4472

1 files changed, 33 insertions, 0 deletions

diff --git a/test/data/cscodec-utf16/simple.dat b/test/data/cscodec-utf16/simple.dat
new file mode 100644
index 0000000..1e7d324
--- /dev/null
+++ b/test/data/cscodec-utf16/simple.dat
@@ -0,0 +1,33 @@
+# *** Simple test:
+#data decode STRICT
+&#x0040&#x4142
+#expected PARSERUTILS_OK
+&#x00000040&#x00004142
+#reset
+
+# *** Surrogate test:
+#data decode STRICT
+&#xD800&#xDF02
+#expected PARSERUTILS_OK
+&#x00010302
+#reset
+
+# *** Lonely high surrogate:
+# This is a bit strange that end status is ok.
+#data decode STRICT
+&#xD805
+#expected PARSERUTILS_OK
+#reset
+
+# With an extra code point, the status is different.
+#data decode STRICT
+&#xD805&#x4142
+#expected PARSERUTILS_INVALID
+#reset
+
+# *** Wrong low surrogate start:
+#data decode STRICT
+&#xDC05
+#expected PARSERUTILS_INVALID
+#reset
+