feat(lexer): finished

This commit is contained in:
Matteo Flebus 2026-01-12 21:31:15 +01:00
parent 97b7240286
commit 745649df1c
2 changed files with 57 additions and 19 deletions

View file

@ -1,15 +1,53 @@
#include "lexer.h" #include "lexer.h"
#include <ctype.h>
#include <stdbool.h> #include <stdbool.h>
#include <stddef.h> #include <stddef.h>
#include <stdlib.h> #include <stdlib.h>
#include <string.h> #include <string.h>
#include "io_backend/io_backend.h" #include "io_backend/io_backend.h"
#include "utils/string_utils/string_utils.h"
static char *end_last_token; static char *end_last_token;
static ssize_t remaining_chars; static ssize_t remaining_chars;
/* @brief: saves state for the next call the the lexer.
*
*/
static void save_state(char *stream, ssize_t i)
{
remaining_chars -= i;
end_last_token = stream + i;
return;
}
/* @return: true if a special character from the grammar was found,
* false otherwise.
*
*/
static bool is_special_char(char c)
{
return c == '\'' || c == '\n' || c == ';';
}
/* @return: true if a keyword from the grammar was found, false otherwise.
*
*/
static bool is_keyword(char *stream, ssize_t i)
{
if (i == 2)
{
return strcmp(stream, "if") == 0 || strcmp(stream, "fi") == 0;
}
if (i == 4)
{
return strcmp(stream, "then") || strcmp(stream, "else")
|| strcmp(stream, "elif");
}
return false;
}
char *new_token(char *begin, ssize_t size) char *new_token(char *begin, ssize_t size)
{ {
char *res = calloc(size + 1, sizeof(char)); char *res = calloc(size + 1, sizeof(char));
@ -32,6 +70,10 @@ char *stream_init(void)
stream = end_last_token; stream = end_last_token;
} }
char *trimed_stream = trim_blanks_left(stream);
remaining_chars -= trimed_stream - stream;
stream = trimed_stream;
return stream; return stream;
} }
@ -39,34 +81,29 @@ char *get_token(void)
{ {
char *stream = stream_init(); char *stream = stream_init();
bool inquotes = false;
ssize_t i = 0; ssize_t i = 0;
while (i < remaining_chars) while (i < remaining_chars)
{ {
switch (stream[i]) if (is_special_char(stream[i]))
{ {
case '\'': if (i == 0) // where we create spe_char token
inquotes = !inquotes; i++;
break; break;
}
case ' ' | '\n' | '\t': if (isblank(stream[i]))
if (inquotes) {
break; break;
else }
{ else if (is_keyword(stream, i))
// token creation {
// skip blank char i++;
// exit from loop
char *token = new_token(stream, i);
return token;
}
default:
break; break;
} }
i++; i++;
} }
remaining_chars -= i; save_state(stream, i);
return NULL;
return new_token(stream, i);
} }

View file

@ -30,6 +30,7 @@ char *new_token(char *begin, ssize_t size);
* If it is, it calls stream_read() from IO_backend, * If it is, it calls stream_read() from IO_backend,
* and sets [remaing_chars]. * and sets [remaing_chars].
* If not, it starts from the end of the last token. * If not, it starts from the end of the last token.
* Also trims left blanks before returning.
* *
* @return: char* stream from which we tokenise. * @return: char* stream from which we tokenise.
*/ */