Parsing pt3

[YvanoffP]

Create token process with enum and identification of each parts. See the ressource image for more

[YvanoffP]

List of tokens

• [ ] literal
• [ ] dollar
• [ ] exec [./]
• [ ] pipe [|]
• [ ] redirection [> >> < <<]

[YvanoffP]

The goal would be to create a realloc tab that will be called at each new word met. The principle of this will be to add one more cell to the tab to hold the new arg found.

Work with indexes and an offset to size the words we are about to malloc with substr() func.

Nodes will be as followed : char *arg; // this will be where the args stand t_node next; // next node

Any succession of chars will be counted as a string. here's a list of the splitters :

• ' '
• " "
• =
• [space]

a separator or a redirect will be threated as a new node holding only the sep/redir we found

if we meet a $, it will be parsed with the entire word attached to it

pay attention to the $> ./exec_me case in this case, we would store the following : arg[x] = "./exec_me" and arg[x+1] = "case"

[Bima42]

At this moment, Chained list are created, we cut arguments words by words like we tell before Then, "" ' ' and space are ok

We need :

• $
• =
• execution ./
• redirection < > << >>
• token (?)

[Bima42]

So now, parsing of $ is ok. But we have to change quotes parsing :

• ' ' : everything inside is duplicate, $ and other special characters doesnt matter
• " " : $ have to check if variable already exist, but for the rest it's same as ' '
• \ is not interpreted here so we just print \
• ; not interpreted too

TO DO :

• execution ./
• redirection < > << >>
• token (?)