lexer

3 files changed, 69 insertions, 0 deletions

diff --git a/include/lexer/lexer.h b/include/lexer/lexer.h
new file mode 100644
index 0000000..f24e92e
--- /dev/null
+++ b/include/lexer/lexer.h
@@ -0,0 +1,27 @@
+#ifndef LEXER_H
+#define LEXER_H
+
+#include "token.h"
+
+#include <vector>
+#include <string>
+
+class Lexer {
+public:
+    explicit Lexer (const std::string& src);
+
+    std::vector<Token> tokenise();
+
+private:
+    int line;
+    int column;
+    int position;
+    std::string src;
+    std::vector<Token> tokens;
+
+    char advance();
+    void skipWhitespace();
+    void skipComment();
+};
+
+#endif
diff --git a/include/lexer/token.h b/include/lexer/token.h
new file mode 100644
index 0000000..54ac116
--- /dev/null
+++ b/include/lexer/token.h
@@ -0,0 +1,17 @@
+#ifndef TOKEN_H
+#define TOKEN_H
+
+#include "token_type.h"
+
+#include <string>
+
+struct Token {
+    TokenType type;
+    std::string value;
+    int line;
+    int column;
+
+    Token(TokenType t, const std::string& val, int line, int col) : type{t}, value{val}, line{line}, column{col} {};
+};
+
+#endif
diff --git a/include/lexer/token_type.h b/include/lexer/token_type.h
new file mode 100644
index 0000000..f83c6d6
--- /dev/null
+++ b/include/lexer/token_type.h
@@ -0,0 +1,25 @@
+#ifndef TOKEN_TYPE_H
+#define TOKEN_TYPE_H
+
+enum class TokenType {
+    INT,
+    STRING,
+
+    NUMBER,
+    IDENTIFIER,
+
+    PLUS,
+    MINUS,
+    MULTIPLY,
+    DIVIDE,
+
+    ASSIGN,
+
+    SEMICOLON,
+
+    END_OF_FILE,
+    INVALID
+
+};
+
+#endif