cm_utf8: add an is_valid function

3 files changed, 101 insertions, 0 deletions

diff --git a/Source/cm_utf8.c b/Source/cm_utf8.c
index d41d097..62e7e8c 100644
--- a/Source/cm_utf8.c
+++ b/Source/cm_utf8.c
@@ -2,6 +2,8 @@
    file Copyright.txt or https://cmake.org/licensing for details.  */
 #include "cm_utf8.h"
 
+#include <string.h>
+
 /*
   RFC 3629
   07-bit: 0xxxxxxx
@@ -85,3 +87,20 @@ const char* cm_utf8_decode_character(const char* first, const char* last,
     return first;
   }
 }
+
+int cm_utf8_is_valid(const char* s)
+{
+  if (!s) {
+    return 0;
+  }
+
+  const char* last = s + strlen(s);
+  const char* pos = s;
+  unsigned int pc;
+
+  while (pos != last && (pos = cm_utf8_decode_character(pos, last, &pc))) {
+    /* Nothing to do. */
+  }
+
+  return pos == last;
+}
diff --git a/Source/cm_utf8.h b/Source/cm_utf8.h
index fcb43e0..27dc559 100644
--- a/Source/cm_utf8.h
+++ b/Source/cm_utf8.h
@@ -13,6 +13,10 @@ extern "C" {
 const char* cm_utf8_decode_character(const char* first, const char* last,
                                      unsigned int* pc);
 
+/** Returns whether a C string is a sequence of valid UTF-8 encoded Unicode
+    codepoints.  Returns non-zero on success. */
+int cm_utf8_is_valid(const char* s);
+
 #ifdef __cplusplus
 } /* extern "C" */
 #endif
diff --git a/Tests/CMakeLib/testUTF8.cxx b/Tests/CMakeLib/testUTF8.cxx
index 7f52c82..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;
@@ -46,6 +61,13 @@ static test_utf8_char const bad_chars[] = {
   { 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)
 {
   printf("%s: decoding good ", passed ? "pass" : "FAIL");
@@ -99,6 +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;
@@ -106,11 +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;
 }