feat: basic lexer with flex

1 files changed, 58 insertions, 0 deletions

diff --git a/sfinae.l b/sfinae.l
new file mode 100644
index 0000000..8453699
--- /dev/null
+++ b/sfinae.l
@@ -0,0 +1,58 @@
+%option noyywrap nounput noinput yylineno
+
+%{
+#include "lexer.h"
+
+#include <stdbool.h>
+#include <stdio.h>
+
+static bool failed = false;
+static char character = '\0';
+
+static char unescape(char escaped)
+{
+    switch (escaped)
+    {
+        case 'n': return '\n';
+        case 't': return '\t';
+        case '0': return '\0';
+        default: return escaped;
+    }
+}
+
+// complains, remembers, and hands back TOK_END so the scan stops where it went wrong
+static int fail(const char* message)
+{
+    (void)fprintf(stderr, "line %d: %s\n", yylineno, message);
+    failed = true;
+    return TOK_END;
+}
+%}
+
+%%
+
+"::"                    return TOK_COLONCOLON;
+"emit"                  return TOK_EMIT;
+"commit"                return TOK_COMMIT;
+"decltype"              return TOK_DECLTYPE;
+"difference"            return TOK_DIFFERENCE;
+[<>(),;]                return yytext[0];
+'\\.'                   character = unescape(yytext[2]); return TOK_CHAR;
+'.'                     character = yytext[1]; return TOK_CHAR;
+"//"[^\n]*              /* a comment */
+[ \t\r\n]+              /* blanks */
+[0-9]+                  return fail("numeric literals are for the weak, subtract two characters");
+[A-Za-z_][A-Za-z0-9_]*  return fail("that is not a notion this language recognises");
+.                       return fail("what is this character even");
+
+%%
+
+bool lexer_failed(void)
+{
+    return failed;
+}
+
+char lexer_character(void)
+{
+    return character;
+}