indention start

makefile

@@ -11,7 +11,7 @@ $(TARGET): build/parser.o build/lexer.o build/main.o
 build/parser.cpp build/parser.hpp: src/parser.y | build
 	$(BISON) -d -o build/parser.cpp src/parser.y
 
-build/lexer.cpp build/lexer.hpp: src/lexer.l build/parser.hpp src/scanner.hpp | build
+build/lexer.cpp build/lexer.hpp: src/lexer.l build/parser.hpp src/scanner.hpp src/indent_helper.hpp| build
 	$(FLEX) --header-file=build/lexer.hpp -o build/lexer.cpp src/lexer.l
 
 build/%.o: build/%.cpp

src/indent_helper.hpp

@@ -0,0 +1,65 @@
+#pragma once
+#include <stack>
+
+class IndentHelper
+{
+    std::stack<int> indent_stack;
+    int pending_dedents = 0;
+
+public:
+    IndentHelper() { indent_stack.push(0); }
+
+    int processLine(int spaces)
+    {
+        int current = indent_stack.top();
+
+        if (spaces > current)
+        {
+            indent_stack.push(spaces);
+            return 1;
+        }
+
+        if (spaces == current)
+        {
+            return 0;
+        }
+
+        pending_dedents = 0;
+        while (indent_stack.top() > spaces)
+        {
+            indent_stack.pop();
+            pending_dedents++;
+        }
+
+        if (indent_stack.top() != spaces)
+        {
+            throw std::runtime_error("Indentation error");
+        }
+
+        return -1;
+    }
+
+    bool hasPendingDedents() const
+    {
+        return pending_dedents > 0;
+    }
+
+    void consumePendingDedent()
+    {
+        if (pending_dedents > 0)
+        {
+            pending_dedents--;
+        }
+    }
+
+    int flushAllDedents()
+    {
+        int total_dedents = 0;
+        while (indent_stack.size() > 1)
+        {
+            indent_stack.pop();
+            total_dedents++;
+        }
+        return total_dedents;
+    }
+};

src/lexer.l

@@ -5,28 +5,64 @@
 
 %{
 #include "scanner.hpp"
+#include "indent_helper.hpp"
+
+IndentHelper indent_helper;
+bool at_line_start = true;
 %}
 
 %%
+
+^[ \t]* {
+    if (!at_line_start) {
+        REJECT;
+    }
+    at_line_start = false;
+
+    int spaces = 0;
+    for (int i = 0; i < yyleng; i++)
+        spaces += (yytext[i] == '\t') ? 8 : 1;
+
+    int result = indent_helper.processLine(spaces);
+    if (result > 0) return yy::parser::token::TOK_INDENT;
+    if (result < 0) return yy::parser::token::TOK_DEDENT;
+}
+
+\n {
+    at_line_start = true;
+    return yy::parser::token::TOK_NEWLINE;
+}
+
 def      { return yy::parser::token::TOK_KW_DEF; }
 if       { return yy::parser::token::TOK_KW_IF; }
 return   { return yy::parser::token::TOK_KW_RETURN; }
 for      { return yy::parser::token::TOK_KW_FOR; }
-in              { return yy::parser::token::TOK_KW_IF;}
-\n              { return yy::parser::token::TOK_NEWLINE; }
+in       { return yy::parser::token::TOK_KW_IN; }
+
 \(       { return yy::parser::token::TOK_L_PAREN; }
 \)       { return yy::parser::token::TOK_R_PAREN; }
 \:       { return yy::parser::token::TOK_COLON; }
 \=       { return yy::parser::token::TOK_ASSIGN; }
 \+       { return yy::parser::token::TOK_PLUS; }
 \-       { return yy::parser::token::TOK_MINUS; }
-\"              { yymore(); BEGIN(STRING); }
-[0-9]+          { yylval->as<std::string>() = yytext; return yy::parser::token::TOK_NUMBER; }
-[_a-zA-Z]+      { yylval->as<std::string>() = yytext; return yy::parser::token::TOK_IDENTIFIER; }
-<<EOF>>         { return yy::parser::token::TOK_EOF; }
-.               { /* the last rule, ignore anything else (there should be nothing) */}
 
+\"       { yymore(); BEGIN(STRING); }
 <STRING>\"\" { yymore(); }
 <STRING>\"  { BEGIN(INITIAL); yylval->as<std::string>() = yytext; return yy::parser::token::TOK_STRING; }
 <STRING>.   { yymore(); }
-%%
+
+[0-9]+      { yylval->as<std::string>() = yytext; return yy::parser::token::TOK_NUMBER; }
+[A-Za-z_][A-Za-z0-9_]*  { yylval->as<std::string>() = yytext; return yy::parser::token::TOK_IDENTIFIER; }
+
+
+<<EOF>> {
+    if (indent_helper.hasPendingDedents()) {
+        indent_helper.consumeDedent();
+        return yy::parser::token::TOK_DEDENT;
+    }
+    int d;
+    if ((d = indent_helper.flushDedent()) != 0) return yy::parser::token::TOK_DEDENT;
+    return yy::parser::token::TOK_EOF;
+}
+
+.           { /* ignore other characters */ }

src/parser.y

@@ -29,7 +29,7 @@
 %token <int> NUMBER
 %token ASSIGN PLUS MINUS L_PAREN R_PAREN COLON
 %token KW_DEF KW_IF KW_RETURN KW_FOR KW_IN
-%token EOF NEWLINE INDENT_UP INDENT_DOWN
+%token EOF NEWLINE INDENT DEDENT
 
 
 %%

test/lexer/2.out

@@ -4,28 +4,28 @@ L_PAREN
 R_PAREN
 COLON
 NEWLINE
-INDENT_UP
+INDENT
 KW_IF
 IDENTIFIER condition
 COLON
 NEWLINE
-INDENT_UP
+INDENT
 KW_RETURN
 IDENTIFIER value
 NEWLINE
-INDENT_DOWN
+DEDENT
 KW_FOR
 IDENTIFIER item
 KW_IN
 IDENTIFIER list
 COLON
 NEWLINE
-INDENT_UP
+INDENT
 IDENTIFIER print
 L_PAREN
 IDENTIFIER item
 R_PAREN
 NEWLINE
-INDENT_DOWN
-INDENT_DOWN
+DEDENT
+DEDENT
 EOF

test/lexer/3.in

@@ -1,3 +1,12 @@
-x == y != z < > <= >=
-a + b - c * d / e % f
-( ) [ ] { } , : . ;
+def outer():
+    def inner1():
+        if cond1:
+            action1()
+        if cond2:
+            action2()
+    def inner2():
+        for i in range(3):
+            if cond3:
+                action3()
+            action4()
+    return result

test/lexer/3.out

@@ -0,0 +1,73 @@
+KW_DEF
+IDENTIFIER outer
+L_PAREN
+R_PAREN
+COLON
+NEWLINE
+INDENT
+KW_DEF
+IDENTIFIER inner1
+L_PAREN
+R_PAREN
+COLON
+NEWLINE
+INDENT
+KW_IF
+IDENTIFIER cond1
+COLON
+NEWLINE
+INDENT
+IDENTIFIER action1
+L_PAREN
+R_PAREN
+NEWLINE
+DEDENT
+KW_IF
+IDENTIFIER cond2
+COLON
+NEWLINE
+INDENT
+IDENTIFIER action2
+L_PAREN
+R_PAREN
+NEWLINE
+DEDENT
+DEDENT
+KW_DEF
+IDENTIFIER inner2
+L_PAREN
+R_PAREN
+COLON
+NEWLINE
+INDENT
+KW_FOR
+IDENTIFIER i
+KW_IN
+IDENTIFIER range
+L_PAREN
+NUMBER 3
+R_PAREN
+COLON
+NEWLINE
+INDENT
+KW_IF
+IDENTIFIER cond3
+COLON
+NEWLINE
+INDENT
+IDENTIFIER action3
+L_PAREN
+R_PAREN
+NEWLINE
+DEDENT
+IDENTIFIER action4
+L_PAREN
+R_PAREN
+NEWLINE
+DEDENT
+DEDENT
+KW_RETURN
+IDENTIFIER result
+NEWLINE
+DEDENT
+EOF

test/lexer/4.in

@@ -1,3 +1,3 @@
-42 3.14 0xFF 0b1010
-"hello" 'world' """multiline"""
-'test' "escaped\"quote"
+x == y != z < > <= >=
+a + b - c * d / e % f
+( ) [ ] { } , : . ;

test/lexer/4.out

@@ -0,0 +1,34 @@
+IDENTIFIER x
+EQ
+IDENTIFIER y
+NE
+IDENTIFIER z
+LT
+GT
+LTE
+GTE
+NEWLINE
+IDENTIFIER a
+PLUS
+IDENTIFIER b
+MINUS
+IDENTIFIER c
+MULTIPLY
+IDENTIFIER d
+DIVIDE
+IDENTIFIER e
+MODULO
+IDENTIFIER f
+NEWLINE
+L_PAREN
+R_PAREN
+L_BRACKET
+R_BRACKET
+L_BRACE
+R_BRACE
+COMMA
+COLON
+DOT
+SEMICOLON
+NEWLINE
+EOF

test/lexer/5.in

@@ -1,8 +1,3 @@
-def calculate(a, b=10):
-    result = (a + b) * 2
-    if result > 100:
-        print("Large result:", result)
-    return result
-
-# This is a comment
-x = calculate(5, 15)
+42 3.14 0xFF 0b1010
+"hello" 'world' """multiline"""
+'test' "escaped\"quote"

test/lexer/6.in

@@ -1,5 +1,8 @@
-_private = "valid"
-var123 = 456
-CamelCase = True
-snake_case = False
-very_long_variable_name_123 = "test"
+def calculate(a, b=10):
+    result = (a + b) * 2
+    if result > 100:
+        print("Large result:", result)
+    return result
+
+# This is a comment
+x = calculate(5, 15)

test/lexer/7.in

@@ -0,0 +1,5 @@
+_private = "valid"
+var123 = 456
+CamelCase = True
+snake_case = False
+very_long_variable_name_123 = "test"