Started implementing the lexer

This commit is contained in:
2023-06-17 00:45:00 +01:00
parent f4b826a299
commit cad6b938b3
9 changed files with 887 additions and 0 deletions

154
src/lexer/lexer_states.c Normal file
View File

@@ -0,0 +1,154 @@
#include "lexer_states.h"
#include "aliases.h"
#include <ctype.h>
// clang-format off
typedef enum {
// GENERAL STATES
LEXER_STATE_START,
LEXER_STATE_SUCCESS,
LEXER_STATE_ERROR,
LEXER_STATE_VALUE,
LEXER_STATE_KEYWORD,
// COLLECTION STATES
LEXER_STATE_OBJECT,
LEXER_STATE_ARRAY,
// OBJECT STATES
LEXER_STATE_KEY_START,
LEXER_STATE_KEY_END,
// NUMBER STATES
LEXER_STATE_NUMBER_START,
LEXER_STATE_DECIMAL,
LEXER_STATE_NUMBER,
LEXER_STATE_FRACTION,
LEXER_STATE_EXPONENT,
LEXER_STATE_EXP_SIGN,
LEXER_STATE_POWER,
LEXER_STATE_NUMBER_END,
// STRING STATES
LEXER_STATE_STRING,
LEXER_STATE_ESCAPE_SEQUENCE,
LEXER_STATE_UNICODE_HEX,
COUNT_LEXER_STATES,
} lexer_state_t;
// clang-format on
struct lexer {
lexer_state_t current;
u64 line;
u64 column;
};
void lexer_state_machine(lexer_t *lexer, char input);
lexer_state_t handle_lexer_start(char input);
lexer_state_t handle_object(char input);
lexer_state_t handle_character(char input);
lexer_state_t handle_escape_sequence(char input);
void lexer_state_machine(lexer_t *lexer, char input) {
switch (lexer->current) {
case LEXER_STATE_START:
lexer->current = handle_lexer_start(input);
break;
case LEXER_STATE_SUCCESS:
break;
case LEXER_STATE_ERROR:
break;
case LEXER_STATE_VALUE:
break;
case LEXER_STATE_KEYWORD:
break;
case LEXER_STATE_OBJECT:
lexer->current = handle_object(input);
break;
case LEXER_STATE_ARRAY:
lexer->current = LEXER_STATE_VALUE;
break;
case LEXER_STATE_KEY_START:
lexer->current = LEXER_STATE_STRING;
break;
case LEXER_STATE_KEY_END:
break;
case LEXER_STATE_NUMBER_START:
break;
case LEXER_STATE_DECIMAL:
break;
case LEXER_STATE_NUMBER:
break;
case LEXER_STATE_FRACTION:
break;
case LEXER_STATE_EXPONENT:
break;
case LEXER_STATE_EXP_SIGN:
break;
case LEXER_STATE_POWER:
break;
case LEXER_STATE_NUMBER_END:
break;
case LEXER_STATE_STRING:
lexer->current = handle_character(input);
break;
case LEXER_STATE_ESCAPE_SEQUENCE:
lexer->current = handle_escape_sequence(input);
break;
case LEXER_STATE_UNICODE_HEX:
break;
default:
lexer->current = LEXER_STATE_ERROR;
}
}
lexer_state_t handle_lexer_start(char input) {
if (isspace(input)) {
return LEXER_STATE_START;
}
switch (input) {
case '{':
return LEXER_STATE_OBJECT;
case '[':
return LEXER_STATE_ARRAY;
}
return LEXER_STATE_ERROR;
}
lexer_state_t handle_object(char input) {
if (isspace(input)) {
return LEXER_STATE_OBJECT;
} else if (input == '"') {
return LEXER_STATE_KEY_START;
}
return LEXER_STATE_ERROR;
}
lexer_state_t handle_character(char input) {
switch (input) {
case '\\':
return LEXER_STATE_ESCAPE_SEQUENCE;
case '"':
return LEXER_STATE_STRING; // WHAT TO RETURN HERE?
}
return LEXER_STATE_STRING;
}
lexer_state_t handle_escape_sequence(char input) {
switch (input) {
case '"':
case '/':
case '\\':
case 'b':
case 'f':
case 'n':
case 'r':
case 't':
return LEXER_STATE_STRING;
case 'u':
return LEXER_STATE_UNICODE_HEX;
}
return LEXER_STATE_ERROR;
}

3
src/main.c Normal file
View File

@@ -0,0 +1,3 @@
#include <stdlib.h>
int main(int argc, char *argv[]) { return EXIT_SUCCESS; }