More C++ stuff. Don't try to compile the Node stuff yet.

src/document.hpp

+#include <vector>
+#include "node.hpp"
+
+namespace Sass {
+  struct Document {
+    using std::vector;
+    
+    char* source;
+    unsigned int position;
+    vector<Node> statements;
+  };
+}
\ No newline at end of file

src/node.hpp

+#include <vector>
+#include "token.hpp"
+
+namespace Sass {
+  struct Node {
+    
\ No newline at end of file

src/prelexer.cpp

+#include <cctype>
 #include "prelexer.hpp"
 
 namespace Sass {
@@ -7,11 +8,118 @@ namespace Sass {
     char *epsilon(char *src) {
       return src;
     }
-    
     // Matches the empty string.
     char *empty(char *src) {
       return *src ? 0 : src;
     }
     
+    // Match a single character satisfying the ctype predicates.
+    char* space(char* src) { return std::isspace(*src) ? src+1 : 0; }
+    char* alpha(char* src) { return std::isalpha(*src) ? src+1 : 0; }
+    char* digit(char* src) { return std::isdigit(*src) ? src+1 : 0; }
+    char* xdigit(char* src) { return std::isxdigit(*src) ? src+1 : 0; }
+    char* alnum(char* src) { return std::isalnum(*src) ? src+1 : 0; }
+    char* punct(char* src) { return std::ispunct(*src) ? src+1 : 0; }
+    // Match multiple ctype characters.
+    char* spaces(char* src) { return one_plus<space>(src); }
+    char* alphas(char* src) { return one_plus<alpha>(src); }
+    char* digits(char* src) { return one_plus<digit>(src); }
+    char* xdigits(char* src) { return one_plus<xdigit>(src); }
+    char* alnums(char* src) { return one_plus<alnum>(src); }
+    char* puncts(char* src) { return one_plus<punct>(src); }
+    
+    // Match a line comment.
+    extern const char slash_slash[] = "//";
+    char* line_comment(char* src) { return to_endl<slash_slash>(src); }
+    // Match a block comment.
+    extern const char slash_star[] = "/*";
+    extern const char star_slash[] = "*/";
+    char* block_comment(char* src) {
+      return delimited_by<slash_star, star_slash, false>(src);
+    }
+    // Match double- and single-quoted strings.
+    char* double_quoted_string(char* src) {
+      return delimited_by<'"', '"', true>(src);
+    }
+    char* single_quoted_string(char* src) {
+      return delimited_by<'\'', '\'', true>(src);
+    }
+    char* string_constant(char* src) {
+      return alternatives<double_quoted_string, single_quoted_string>(src);
+    }
+    // Match interpolants.
+    extern const char hash_lbrace[] = "#{";
+    extern const char rbrace[] = "}";
+    char* interpolant(char* src) {
+      return delimited_by<hash_lbrace, rbrace, false>(src);
+    }
+    
+    // Match CSS identifiers.
+    char* identifier(char* src) {
+      return sequence< optional< exactly<'-'> >,
+                       alternatives< alpha, exactly<'_'> >,
+                       zero_plus< alternatives< alnum,
+                                                exactly<'-'>,
+                                                exactly<'_'> > > >(src);
+    }
+    // Match CSS '@' keywords.
+    char* at_keyword(char* src) {
+      return sequence<exactly<'@'>, identifier>(src);
+    }
+    char* name(char* src) {
+      return one_plus< alternatives< alnum,
+                                     exactly<'-'>,
+                                     exactly<'_'> > >(src);
+    }
+    // Match CSS id names.
+    char* id_name(char* src) {
+      return sequence<exactly<'#'>, name>(src);
+    }
+    // Match CSS class names.
+    char* class_name(char* src) {
+      return sequence<exactly<'.'>, identifier>(src);
+    }
+    // Match CSS numeric constants.
+    extern const char sign[] = "-+";
+    char* unsigned_number(char* src) {
+      return alternatives<sequence< zero_plus<digits>,
+                                    exactly<'.'>,
+                                    one_plus<digits> >,
+                          digits>(src);
+    }
+    char* number(char* src) {
+      return sequence< optional< class_char<sign> >, unsigned_number>(src);
+    }
+    char* percentage(char* src) {
+      return sequence< number, exactly<'%'> >(src);
+    }
+    char* dimension(char* src) {
+      return sequence<number, identifier>(src);
+    }
+    // Match CSS uri specifiers.
+    extern const char url_call[] = "url(";
+    char* uri(char* src) {
+      return sequence< exactly<url_call>,
+                       optional<spaces>,
+                       string_constant,
+                       optional<spaces>,
+                       exactly<')'> >(src);
+    }
+    // Match CSS function call openers.
+    char* function(char* src) {
+      return sequence< identifier, exactly<'('> >(src);
+    }
+    // Match CSS attribute-matching operators.
+    extern const char tilde_equal[]  = "~=";
+    extern const char pipe_equal[]   = "|=";
+    extern const char caret_equal[]  = "^=";
+    extern const char dollar_equal[] = "$=";
+    extern const char star_equal[]   = "*=";
+    char* exact_match(char* src) { return exactly<'='>(src); }
+    char* class_match(char* src) { return exactly<tilde_equal>(src); }
+    char* dash_match(char* src) { return exactly<pipe_equal>(src); }
+    char* prefix_match(char* src) { return exactly<caret_equal>(src); }
+    char* suffix_match(char* src) { return exactly<dollar_equal>(src); }
+    char* substring_match(char* src) { return exactly<star_equal>(src); }
   }
 }
\ No newline at end of file

src/prelexer.hpp

@@ -2,58 +2,80 @@ namespace Sass {
   namespace Prelexer {
 
     typedef int (*ctype_predicate)(int);
-    typedef char *(*prelexer)(char *);
+    typedef char* (*prelexer)(char*);
 
     // Match a single character literal.
     template <char pre>
-    char *exactly(char *src) {
+    char* exactly(char* src) {
       return *src == pre ? src + 1 : 0;
     }
   
     // Match a string constant.
-    template <const char *prefix>
-    char *exactly(char *src) {
-      char *pre = prefix;
+    template <const char* prefix>
+    char* exactly(char* src) {
+      const char* pre = prefix;
       while (*pre && *src == *pre) ++src, ++pre;
       return *pre ? 0 : src;
     }
 
     // Match a single character that satifies the supplied ctype predicate.
     template <ctype_predicate pred>
-    char *class_char(char *src) {
+    char* class_char(char* src) {
       return pred(*src) ? src + 1 : 0;
     }
 
     // Match a single character that is a member of the supplied class.
-    template <const char *char_class>
-    char *class_char(char *src) {
-      char *cc = char_class;
+    template <const char* char_class>
+    char* class_char(char* src) {
+      const char* cc = char_class;
       while (*cc && *src != *cc) ++cc;
       return *cc ? src + 1 : 0;
     }
 
     // Match a sequence of characters that all satisfy the supplied ctype predicate.
     template <ctype_predicate pred>
-    char *class_chars(char *src) {
-      char *p = src;
+    char* class_chars(char* src) {
+      char* p = src;
       while (pred(*p)) ++p;
       return p == src ? 0 : p;
     }
 
     // Match a sequence of characters that are all members of the supplied class.
-    template <const char *char_class>
-    char *class_chars(char *src) {
-      char *p = src;
+    template <const char* char_class>
+    char* class_chars(char* src) {
+      char* p = src;
       while (class_char<char_class>(p)) ++p;
       return p == src ? 0 : p;
     }
     
+    // Match a sequence of characters up to the next newline.
+    template <const char* prefix>
+    char* to_endl(char* src) {
+      if (!(src = exactly<prefix>(src))) return 0;
+      while (*src && *src != '\n') ++src;
+      return src;
+    }
+    
+    // Match a sequence of characters delimited by the supplied chars.
+    template <char beg, char end, bool esc>
+    char* delimited_by(char* src) {
+      src = exactly<beg>(src);
+      if (!src) return 0;
+      char* stop;
+      while (1) {
+        if (!*src) return 0;
+        stop = exactly<end>(src);
+        if (stop && (!esc || *(src - 1) != '\\')) return stop;
+        src = stop ? stop : src + 1;
+      }
+    }
+    
     // Match a sequence of characters delimited by the supplied strings.
-    template <const char *beg, const char *end, bool esc>
-    char *delimited_by(char *src) {
+    template <const char* beg, const char* end, bool esc>
+    char* delimited_by(char* src) {
       src = exactly<beg>(src);
       if (!src) return 0;
-      char *stop;
+      char* stop;
       while (1) {
         if (!*src) return 0;
         stop = exactly<end>(src);
@@ -63,37 +85,37 @@ namespace Sass {
     }
     
     // Matches zero characters (always succeeds without consuming input).
-    char *epsilon(char *);
+    char* epsilon(char*);
     
     // Matches the empty string.
-    char *empty(char *);
+    char* empty(char*);
     
     // Succeeds of the supplied matcher fails, and vice versa.
     template <prelexer mx>
-    char *negate(char *src) {
+    char* negate(char* src) {
       return mx(src) ? 0 : src;
     }
     
     // Tries the matchers in sequence and returns the first match (or none)
     template <prelexer mx1, prelexer mx2>
-    char *alternatives(char *src) {
-      char *rslt;
+    char* alternatives(char* src) {
+      char* rslt;
       (rslt = mx1(src)) || (rslt = mx2(src));
       return rslt;
     }
     
     // Same as above, but with 3 arguments.
     template <prelexer mx1, prelexer mx2, prelexer mx3>
-    char *alternatives(char *src) {
-      char *rslt;
+    char* alternatives(char* src) {
+      char* rslt;
       (rslt = mx1(src)) || (rslt = mx2(src)) || (rslt = mx3(src));
       return rslt;
     }
     
     // Same as above, but with 4 arguments.
     template <prelexer mx1, prelexer mx2, prelexer mx3, prelexer mx4>
-    char *alternatives(char *src) {
-      char *rslt;
+    char* alternatives(char* src) {
+      char* rslt;
       (rslt = mx1(src)) || (rslt = mx2(src)) ||
       (rslt = mx3(src)) || (rslt = mx4(src));
       return rslt;
@@ -101,37 +123,114 @@ namespace Sass {
     
     // Tries the matchers in sequence and succeeds if they all succeed.
     template <prelexer mx1, prelexer mx2>
-    char *sequence(char *src) {
-      char *rslt = src;
+    char* sequence(char* src) {
+      char* rslt = src;
       (rslt = mx1(rslt)) && (rslt = mx2(rslt));
       return rslt;
     }
     
     // Same as above, but with 3 arguments.
     template <prelexer mx1, prelexer mx2, prelexer mx3>
-    char *sequence(char *src) {
-      char *rslt = src;
+    char* sequence(char* src) {
+      char* rslt = src;
       (rslt = mx1(rslt)) && (rslt = mx2(rslt)) && (rslt = mx3(rslt));
       return rslt;
     }
     
     // Same as above, but with 4 arguments.
     template <prelexer mx1, prelexer mx2, prelexer mx3, prelexer mx4>
-    char *sequence(char *src) {
-      char *rslt = src;
+    char* sequence(char* src) {
+      char* rslt = src;
       (rslt = mx1(rslt)) && (rslt = mx2(rslt)) &&
       (rslt = mx3(rslt)) && (rslt = mx4(rslt));
       return rslt;
     }
     
+    // Same as above, but with 5 arguments.
+    template <prelexer mx1, prelexer mx2,
+              prelexer mx3, prelexer mx4,
+              prelexer mx5>
+    char* sequence(char* src) {
+      char* rslt = src;
+      (rslt = mx1(rslt)) && (rslt = mx2(rslt)) &&
+      (rslt = mx3(rslt)) && (rslt = mx4(rslt)) &&
+      (rslt = mx5(rslt));
+      return rslt;
+    }
+    
     // Match a pattern or not. Always succeeds.
     template <prelexer mx>
-    char *optional(char *src) {
-      char *p = mx(src);
+    char* optional(char* src) {
+      char* p = mx(src);
       return p ? p : src;
     }
     
+    // Match zero or more of the supplied pattern
+    template <prelexer mx>
+    char* zero_plus(char* src) {
+      char* p = mx(src);
+      while (p) src = p, p = mx(src);
+      return src;
+    }
     
-
+    // Match one or more of the supplied pattern
+    template <prelexer mx>
+    char* one_plus(char* src) {
+      char* p = mx(src);
+      if (!p) return 0;
+      while (p) src = p, p = mx(src);
+      return src;
+    }
+    
+    // Match a single character satisfying the ctype predicates.
+    char *space(char *src);
+    char *alpha(char *src);
+    char *digit(char *src);
+    char *xdigit(char *src);
+    char *alnum(char *src);
+    char *punct(char *src);
+    // Match multiple ctype characters.
+    char* spaces(char* src);
+    char* alphas(char* src);
+    char* digits(char* src);
+    char* xdigits(char* src);
+    char* alnums(char* src);
+    char* puncts(char* src);
+    
+    // Match a line comment.
+    char* line_comment(char* src);
+    // Match a block comment.
+    char* block_comment(char* src);
+    // Match double- and single-quoted strings.
+    char* double_quoted_string(char* src);
+    char* single_quoted_string(char* src);
+    char* string_constant(char* src);
+    // Match interpolants.
+    char* interpolant(char* src);
+
+    // Match a CSS identifier.
+    char* identifier(char* src);
+    // Match CSS '@' keywords.
+    char* at_keyword(char* src);
+    // Match CSS id names.
+    char* id_name(char* src);
+    // Match CSS class names.
+    char* class_name(char* src);
+    // Match CSS numeric constants.
+    char* unsigned_number(char* src);
+    char* number(char* src);
+    char* percentage(char* src);
+    char* dimension(char* src);
+    // Match CSS uri specifiers.
+    char* uri(char* src);
+    // Match CSS function call openers.
+    char* function(char* src);
+    // Match CSS attribute-matching operators.
+    char* exact_match(char* src);
+    char* class_match(char* src);
+    char* dash_match(char* src);
+    char* prefix_match(char* src);
+    char* suffix_match(char* src);
+    char* substring_match(char* src);
   }
 }

src/test.cpp

-#include <iostream>
-#include <string>
-#include "prelexer.hpp"
-
-using std::cout;
-using std::endl;
-using std::string;
-using namespace Sass::Prelexer;
-
-extern const char msg1[] = "Matched an 'a'!";
-extern const char msg2[] = "Failed to match a 'b'!";
-
-int main() {
-  
-  prelexer p = exactly<'a'>;
-  prelexer q = exactly<'b'>;
-  
-  if (p("abcd")) cout << msg1 << endl;
-  if (!q("abcd")) cout << msg2 << endl;
-
-  return 0;
-}
\ No newline at end of file

src/unit-test-prelexer.cpp

+#include <cstdio>
+#include "prelexer.hpp"
+
+using namespace Sass::Prelexer;
+
+void print_slice(const char *s, const char *t) {
+  if (t) {
+    printf("succeeded with %ld characters:\t", t - s);
+    while (s < t) putchar(*s++);
+    putchar('\n');
+  }
+  else {
+    printf("failed\n");
+  }
+}
+
+#define check_twice(matcher, inp1, inp2) \
+  (printf("Attempting to match %s\n", #matcher), \
+   print_slice(inp1, matcher(inp1)), \
+   print_slice(inp2, matcher(inp2)), \
+   putchar('\n'))
+
+char num1[]     = "123foogoo";
+char num2[]     = ".456e7";
+char num3[]     = "-23.45";
+char ident1[]   = "-webkit-box-sizing: border-box;";
+char interp1[]  = "#{$stuff + $more_stuff}";
+char atkwd1[]   = "@media screen blah blah { ... }";
+char idnm1[]    = "#-blah_hoo { color: red; }";
+char classnm1[] = ".hux-boo23 { color: blue; }";
+char percent1[] = "110%; color: red; }";
+char dim1[]     = "12px; color: blue; }";
+char uri1[]     = "url(  'www.bork.com/bork.png' )";
+char ident2[]   = "url ()// not a uri!";
+char func1[]    = "lighten(indigo, 20%)";
+char exact1[]   = "='blah'] { ... }";
+char inc1[]     = "~='foo'] { ... }";
+char dash1[]    = "|='bar'] { ... }";
+char pre1[]     = "^='hux'] { ... }";
+char suf1[]     = "$='baz'] { ... }";
+char sub1[]     = "*='bum'] { ... }";
+
+int main() {
+  
+  check_twice(identifier, ident1, num1);
+  check_twice(interpolant, interp1, idnm1);
+  check_twice(at_keyword, atkwd1, classnm1);
+  check_twice(id_name, idnm1, interp1);
+  check_twice(class_name, classnm1, num2);
+  check_twice(number, num1, ident1);
+  check_twice(number, num2, classnm1);
+  check_twice(number, num3, ident1);
+  check_twice(percentage, percent1, dim1);
+  check_twice(dimension, dim1, percent1);
+  check_twice(uri, uri1, ident2);
+  check_twice(function, func1, ident2);
+  check_twice(exact_match, exact1, inc1);
+  check_twice(class_match, inc1, pre1);
+  check_twice(dash_match, dash1, suf1);
+  check_twice(prefix_match, pre1, dash1);
+  check_twice(suffix_match, suf1, pre1);
+  check_twice(substring_match, sub1, suf1);
+
+  return 0;
+}
\ No newline at end of file