make string length function use our helper code point function

Author: wonja

Makefile

@@ -35,7 +35,8 @@ SOURCES = \
 	source_map.cpp \
 	to_c.cpp \
 	to_string.cpp \
-	units.cpp
+	units.cpp \
+	utf8_string.cpp
 
 OBJECTS = $(SOURCES:.cpp=.o)
 

functions.cpp

@@ -7,6 +7,7 @@
 #include "to_string.hpp"
 #include "inspect.hpp"
 #include "eval.hpp"
+#include "utf8_string.hpp"
 
 #include <cmath>
 #include <cctype>
@@ -695,7 +696,6 @@ namespace Sass {
     {
       String_Constant* s = ARG("$string", String_Constant);
       string str = s->value();
-      size_t len = 0;
       size_t length_of_s = str.size();
       size_t i = 0;
 
@@ -704,24 +704,8 @@ namespace Sass {
         --length_of_s;
       }
 
-      while (i < length_of_s) {
-        unsigned char c = static_cast<unsigned char>(str[i]);
-        if (c < 128) {
-          // it's a single-byte character
-          ++len;
-          ++i;
-        }
-        // it's a multi bit sequence and presumably it's a leading bit
-        else {
-          ++i; // go to the next byte
-          // see if it's still part of the sequence
-          while ((i < length_of_s) && ((static_cast<unsigned char>(str[i]) & 0b11000000) == 0b10000000)) {
-            ++i;
-          }
-          // when it's not [aka a new leading bit], increment and move on
-          ++len;
-        }
-      }
+      size_t len = UTF_8::code_point_count(str, i, length_of_s);
+
       return new (ctx.mem) Number(path, position, len);
     }
 

utf8_string.cpp

@@ -0,0 +1,53 @@
+#ifndef SASS_UTF8_STRING
+#define SASS_UTF8_STRING
+
+#include <string>
+
+namespace Sass {
+  namespace UTF_8 {
+    using std::string;
+    // class utf8_string {
+    //   string s_;
+    // public:
+    //   utf8_string(const string &s): s_(s) {}
+    //   utf8_string(const char* c): s_(string(c)) {}
+
+    //   char operator[](size_t i);
+    //   size_t length();
+    //   size_t byte_to_char(size_t i);
+    // };
+
+    // function that will count the number of code points (utf-8 characters) from the given beginning to the given end
+    size_t code_point_count(const string& str, size_t start, size_t end) {
+      size_t len = 0;
+      size_t i = start;
+
+      while (i < end) {
+        unsigned char c = static_cast<unsigned char>(str[i]);
+        if (c < 128) {
+          // it's a single-byte character
+          ++len;
+          ++i;
+        }
+        // it's a multi bit sequence and presumably it's a leading bit
+        else {
+          ++i; // go to the next byte
+          // see if it's still part of the sequence
+          while ((i < end) && ((static_cast<unsigned char>(str[i]) & 0b11000000) == 0b10000000)) {
+            ++i;
+          }
+          // when it's not [aka a new leading bit], increment and move on
+          ++len;
+        }
+      }
+      return len;
+    }
+
+    size_t code_point_count(const string& str) {
+      return code_point_count(str, 0, str.length());
+    }
+
+  }
+}
+
+#endif 

utf8_string.hpp

@@ -0,0 +1,26 @@
+#ifndef SASS_UTF8_STRING
+#define SASS_UTF8_STRING
+
+#include <string>
+
+namespace Sass {
+  namespace UTF_8 {
+    // class utf8_string {
+    //   string s_;
+    // public:
+    //   utf8_string(const string &s): s_(s) {}
+    //   utf8_string(const char* c): s_(string(c)) {}
+
+    //   char operator[](size_t i);
+    //   size_t length();
+    //   size_t byte_to_char(size_t i);
+    // };
+
+    // function that will count the number of code points (utf-8 characters) from the beginning to the given end
+    size_t code_point_count(const string& str, size_t start, size_t end);
+    size_t code_point_count(const string& str);
+
+  }
+}
+
+#endif