1 files changed, 97 insertions, 7 deletions

diff --git a/Tests/CMakeLib/testUTF8.cxx b/Tests/CMakeLib/testUTF8.cxx
index c99c46d..a0bb5cd 100644
--- a/Tests/CMakeLib/testUTF8.cxx
+++ b/Tests/CMakeLib/testUTF8.cxx
@@ -13,6 +13,21 @@ static void test_utf8_char_print(test_utf8_char const c)
          static_cast<int>(d[3]));
 }
 
+static void byte_array_print(char const* s)
+{
+  unsigned char const* d = reinterpret_cast<unsigned char const*>(s);
+  bool started = false;
+  printf("[");
+  for (; *d; ++d) {
+    if (started) {
+      printf(",");
+    }
+    started = true;
+    printf("0x%02X", static_cast<int>(*d));
+  }
+  printf("]");
+}
+
 struct test_utf8_entry
 {
   int n;
@@ -21,17 +36,36 @@ struct test_utf8_entry
 };
 
 static test_utf8_entry const good_entry[] = {
-  { 1, "\x20\x00\x00\x00", 0x0020 },  /* Space.  */
-  { 2, "\xC2\xA9\x00\x00", 0x00A9 },  /* Copyright.  */
-  { 3, "\xE2\x80\x98\x00", 0x2018 },  /* Open-single-quote.  */
-  { 3, "\xE2\x80\x99\x00", 0x2019 },  /* Close-single-quote.  */
-  { 4, "\xF0\xA3\x8E\xB4", 0x233B4 }, /* Example from RFC 3629.  */
+  { 1, "\x20\x00\x00\x00", 0x0020 },   /* Space.  */
+  { 2, "\xC2\xA9\x00\x00", 0x00A9 },   /* Copyright.  */
+  { 3, "\xE2\x80\x98\x00", 0x2018 },   /* Open-single-quote.  */
+  { 3, "\xE2\x80\x99\x00", 0x2019 },   /* Close-single-quote.  */
+  { 4, "\xF0\xA3\x8E\xB4", 0x233B4 },  /* Example from RFC 3629.  */
+  { 3, "\xED\x80\x80\x00", 0xD000 },   /* Valid 0xED prefixed codepoint.  */
+  { 4, "\xF4\x8F\xBF\xBF", 0x10FFFF }, /* Highest valid RFC codepoint. */
   { 0, { 0, 0, 0, 0, 0 }, 0 }
 };
 
 static test_utf8_char const bad_chars[] = {
-  "\x80\x00\x00\x00", "\xC0\x00\x00\x00", "\xE0\x00\x00\x00",
-  "\xE0\x80\x80\x00", "\xF0\x80\x80\x80", { 0, 0, 0, 0, 0 }
+  "\x80\x00\x00\x00", /* Leading continuation byte. */
+  "\xC0\x80\x00\x00", /* Overlong encoding. */
+  "\xC1\x80\x00\x00", /* Overlong encoding. */
+  "\xC2\x00\x00\x00", /* Missing continuation byte. */
+  "\xE0\x00\x00\x00", /* Missing continuation bytes. */
+  "\xE0\x80\x80\x00", /* Overlong encoding. */
+  "\xF0\x80\x80\x80", /* Overlong encoding. */
+  "\xED\xA0\x80\x00", /* UTF-16 surrogate half. */
+  "\xED\xBF\xBF\x00", /* UTF-16 surrogate half. */
+  "\xF4\x90\x80\x80", /* Lowest out-of-range codepoint. */
+  "\xF5\x80\x80\x80", /* Prefix forces out-of-range codepoints. */
+  { 0, 0, 0, 0, 0 }
+};
+
+static char const* good_strings[] = { "", "ASCII", "\xC2\xA9 Kitware", 0 };
+
+static char const* bad_strings[] = {
+  "\xC0\x80", /* Modified UTF-8 for embedded 0-byte. */
+  0
 };
 
 static void report_good(bool passed, test_utf8_char const c)
@@ -87,6 +121,46 @@ static bool decode_bad(test_utf8_char const s)
   return true;
 }
 
+static void report_valid(bool passed, char const* s)
+{
+  printf("%s: validity good ", passed ? "pass" : "FAIL");
+  byte_array_print(s);
+  printf(" (%s) ", s);
+}
+
+static void report_invalid(bool passed, char const* s)
+{
+  printf("%s: validity bad  ", passed ? "pass" : "FAIL");
+  byte_array_print(s);
+  printf(" ");
+}
+
+static bool is_valid(const char* s)
+{
+  bool valid = cm_utf8_is_valid(s) != 0;
+  if (!valid) {
+    report_valid(false, s);
+    printf("expected valid, reported as invalid\n");
+    return false;
+  }
+  report_valid(true, s);
+  printf("valid as expected\n");
+  return true;
+}
+
+static bool is_invalid(const char* s)
+{
+  bool valid = cm_utf8_is_valid(s) != 0;
+  if (valid) {
+    report_invalid(false, s);
+    printf("expected invalid, reported as valid\n");
+    return false;
+  }
+  report_invalid(true, s);
+  printf("invalid as expected\n");
+  return true;
+}
+
 int testUTF8(int /*unused*/, char* /*unused*/ [])
 {
   int result = 0;
@@ -94,11 +168,27 @@ int testUTF8(int /*unused*/, char* /*unused*/ [])
     if (!decode_good(*e)) {
       result = 1;
     }
+    if (!is_valid(e->str)) {
+      result = 1;
+    }
   }
   for (test_utf8_char const* c = bad_chars; (*c)[0]; ++c) {
     if (!decode_bad(*c)) {
       result = 1;
     }
+    if (!is_invalid(*c)) {
+      result = 1;
+    }
+  }
+  for (char const** s = good_strings; *s; ++s) {
+    if (!is_valid(*s)) {
+      result = 1;
+    }
+  }
+  for (char const** s = bad_strings; *s; ++s) {
+    if (!is_invalid(*s)) {
+      result = 1;
+    }
   }
   return result;
 }