feat(lexer): comments done + operator WIP

This commit is contained in:
matteo 2026-01-23 15:56:21 +01:00
parent 82f35cfa08
commit 3cdba0ad9c
3 changed files with 60 additions and 9 deletions

View file

@ -133,13 +133,14 @@ struct token *peek_token(struct lexer_context *ctx)
i++;
}
struct token *tok = new_token(stream, i);
struct token *tok = new_token(stream, i, ctx->only_digits);
// if token is comment, we don't want it
if (tok->type == TOKEN_COMMENT)
{
// drop current stream
get_next_stream(ctx);
// Find next newline or EOF.
go_end_of_line(ctx);
free_token(&tok);
tok = peek_token(ctx);
}
@ -170,6 +171,9 @@ struct token *pop_token(struct lexer_context *ctx)
if (!update_lexing_mode(stream, i, &lexing_mode)
&& lexing_mode == LEXER_NORMAL)
{
// TODO call here a function
// it must check if is a spe char or an operator
// and sets i accordingly.
if (is_special_char(stream[i]))
{
if (i == 0) // where we create spe_char token
@ -195,7 +199,7 @@ struct token *pop_token(struct lexer_context *ctx)
// (this should never happen)
if (ctx->current_token == NULL)
{
ctx->current_token = new_token(stream, i);
ctx->current_token = new_token(stream, i, ctx->only_digits);
}
save_state(stream, i, ctx);

View file

@ -132,12 +132,30 @@ static void set_token_word(struct token *tok, char *begin, ssize_t size)
}
}
struct token *new_token(char *begin, ssize_t size)
/* @brief: Sets the token to an IO number
* Also allocates the data and fills it.
*/
static void set_token_ION(struct token *tok, char *begin, ssize_t size)
{
if (tok->type == TOKEN_NULL && size != 0)
{
tok->type = TOKEN_IONUMBER;
tok->data = calloc(size + 1, sizeof(char));
if (tok->data == NULL)
return;
strncpy(tok->data, begin, size);
}
}
struct token *new_token(char *begin, ssize_t size, bool only_digits)
{
struct token *tok = calloc(1, sizeof(struct token));
if (tok == NULL)
return NULL;
if (only_digits)
set_token_ION(tok, begin, size);
set_token_spechar(tok, begin, size);
set_token_keyword(tok, begin, size);
set_token_word(tok, begin, size);
@ -186,6 +204,21 @@ void stream_init(struct lexer_context *ctx)
ctx->end_previous_token = trimed_stream;
}
void go_end_of_line(struct lexer_context *ctx)
{
if (ctx == NULL || ctx->end_previous_token == NULL)
return;
ssize_t i = 0;
while (ctx->end_previous_token[i] != '\n'
&& ctx->end_previous_token[i] != EOF)
{
i++;
}
ctx->end_previous_token += i;
ctx->remaining_chars -= i;
}
void get_next_stream(struct lexer_context *ctx)
{
ctx->remaining_chars = 0;

View file

@ -1,6 +1,7 @@
#ifndef LEXER_UTILS_H
#define LEXER_UTILS_H
#include <stdbool.h>
#include <stddef.h>
#include <sys/types.h>
@ -9,6 +10,10 @@ struct lexer_context
char *end_previous_token;
ssize_t remaining_chars;
// usefull to detect IO numbers.
// tells us if we only lexed digits in current token.
bool only_digits;
struct token *previous_token;
struct token *current_token;
};
@ -39,17 +44,20 @@ enum token_type
TOKEN_GRAVE,
TOKEN_SEMICOLON,
TOKEN_COMMENT,
TOKEN_PIPE,
TOKEN_AMPERSAND,
TOKEN_STAR,
TOKEN_BACKSLASH,
TOKEN_DOLLAR,
TOKEN_LEFT_PAREN,
TOKEN_RIGHT_PAREN,
TOKEN_LEFT_BRACKET,
TOKEN_RIGHT_BRACKET,
// redirections
TOKEN_LESS,
TOKEN_GREATER,
TOKEN_STAR,
TOKEN_PIPE,
TOKEN_AMPERSAND,
TOKEN_IONUMBER,
// Keywords
TOKEN_IF,
@ -73,7 +81,7 @@ struct token
*
* @return: NULL on error, a token otherwise.
*/
struct token *new_token(char *begin, ssize_t size);
struct token *new_token(char *begin, ssize_t size, bool only_digits);
/* @brief: frees the token given in argument
*/
@ -90,6 +98,12 @@ void free_token(struct token **tok);
*/
void stream_init(struct lexer_context *ctx);
/* @brief: finds the next '\n' or EOF character,
* starting at [ctx->end_previous_token],
* and updates the stream and remaining_chars accordingly.
*/
void go_end_of_line(struct lexer_context *ctx);
/*
* @brief: drops the current stream and asks IOB for a new one
*/