Suggestions for the expand_tokenlist function

[multitudes]

Since you will be working on this function again:

oid expand_tokenlist(t_data *data, t_list *tokenlist)
{

    // flags to ensure correct expansions are done depending on other tokens in the list eg VAR=~"string"$EXP:~
    while (tokenlist)
    {
        if (!get_curr_token(tokenlist))
            print_error_status("minishell: system error: missing token\n", 1);
        debug("Token to check for expansion - token type: %d, lexeme: %s", get_token_type(tokenlist), get_token_lexeme(tokenlist));
        if (get_token_type(tokenlist) == S_QUOTED_STRING)
            extract_string(get_curr_token(tokenlist));
        else if (get_token_type(tokenlist) == QUOTED_STRING)
            expand_string(data, get_curr_token(tokenlist));
        else if (get_token_type(tokenlist) == VAR_EXPANSION || get_token_type(tokenlist) == DOLLAR)
            expand_dollar(data->env_arr, get_curr_token(tokenlist));
        if (get_token_type(tokenlist) == TILDE \
                || get_token_type(tokenlist) == PATHNAME) // specify condition
            expand_path(data->env_arr, get_curr_token(tokenlist));
        else if (get_token_type(tokenlist)  == DOLLAR_QUESTION)
            expand_exit_status(data, get_curr_token(tokenlist));
        else if (get_token_type(tokenlist) == GLOBBING)
            expand_globbing(tokenlist);
        else
            debug("Token type not expanded");
        // Merge tokens
        tokenlist = tokenlist->next;
    }

Some suggestions:

• better name for extract_string() and expand_string. I mean how would somebody looking at the code understand what they are doing? Maybe something like expand_single_quotes and expand_double_quotes?
• I thought about the tilde in the scanner. There are so many edge cases for the tilde. I will have a look in the scanner if I can assign every string containing tilde to be a path. But the best is to check at this point, after checking for single quotes, after checking for double quotes (and if you assign a special token type after the expansion of double quotes and single quotes then you will not need a flag for the quotes) so after checking for dollar expansions, then do a strchr() for a tilde, and if it is not a type coming from the expansion of single code or double quotes, then pass it to be expanded together according to the flags, first tilde? has a equal sign been there?
• then pls do a folder for expansion, but you can already move the call to the function inside of the executer.
• pls start with the redirections asap because the analyser might need to do some extra work and we dont know yet until we start with the redirection and heredoc. Thank you!

PS after looking at the function recognizing if PATHNAME, I did a lot of checks trying to get a mostly correct check for paths, but if I put tilde in here then this -w-+}[]=w{~or^%ld- becomes a pathname which makes me cringe a bit...

[multitudes]

THis would be the function in scanner:

bool    str_is_pathname(const char *str)
{
    if (((ft_strchr(str, '/') && (!ft_strchr(str, '~'))) || peek(str, ".", \
    false) || peek(str, "./", FUZZY) || peek(str, "../", FUZZY) || peek(str, \
    "~/", FUZZY) || peek(str, "~+/", FUZZY) || peek(str, "~+", EXACT) || \
    peek(str, "~-/", FUZZY) || peek(str, "~-", EXACT)) && !ft_strchr(str, '*'))
    {
        while (*str)
        {
            if (!char_is_in_pathname(*str))
                return (false);
            str++;
        }
        return (true);
    }
    return (false);
}

Mayeb if you have sggestions to improve this? here I try to get the pathname type whjih includes all of the valid tilde cases (I tried)