cmdgroup parsing: parse cmdlist surrounded by parentheses

Makefile

@@ -30,8 +30,9 @@ srcs = \
 	src/executing/here_doc/strip_newline.c \
 	src/ft_errno.c \
 	src/get_command.c \
-	src/parser/cmdgroup/cmdgroup_builder.c \
 	src/parser/cmdgroup/cmdgroup.c \
+	src/parser/cmdgroup/cmdgroup_builder.c \
+	src/parser/cmdgroup/paren.c \
 	src/parser/cmdlist/cmdlist.c \
 	src/parser/cmdlist/cmdlist_builder.c \
 	src/parser/cmdlist/cmdlist_debug.c \

src/parser/cmdgroup/cmdgroup.c

@@ -6,7 +6,7 @@
 /*   By: khais <marvin@42.fr>                       +#+  +:+       +#+        */
 /*                                                +#+#+#+#+#+   +#+           */
 /*   Created: 2025/03/11 15:18:02 by khais             #+#    #+#             */
-/*   Updated: 2025/03/20 12:27:57 by khais            ###   ########.fr       */
+/*   Updated: 2025/03/20 14:46:25 by khais            ###   ########.fr       */
 /*                                                                            */
 /* ************************************************************************** */
 
@@ -14,14 +14,25 @@
 #include "cmdgroup_builder.h"
 #include <stdlib.h>
 #include "libft.h"
+#include "paren.h"
 #include "../../treedrawing.h"
 #include "../../postprocess/redirections/redirection_list.h"
 #include "libft/libft.h"
 
+/*
+** Recursively parse a cmdgroup
+**
+** This is responsible for parsing one optional set of parentheses surrounding a
+** cmdlist.
+**
+** If parentheses are present, also parse redirections, which may only occur
+** after the closing parenthesis.
+*/
 t_cmdgroup	*cmdgroup_from_wordlist(t_wordlist *wordlist)
 {
 	t_cmdgroup_builder	builder;
 	t_wordlist			*words;
+	t_paren				paren;
 
 	words = wordlist_copy(wordlist);
 	if (words == NULL)
@@ -30,7 +41,11 @@ t_cmdgroup	*cmdgroup_from_wordlist(t_wordlist *wordlist)
 		return (wordlist_destroy(words), NULL);
 	while (builder.error == false && builder.current_word != NULL)
 	{
+		paren = match_paren(builder.current_word->word);
+		if (paren == PAREN_INVALID)
 			cmdgroup_builder_next_word(&builder, &words);
+		else
+			cmdgroup_builder_paren(&builder, &words, paren);
 	}
 	if (builder.current_wordlist != NULL)
 		cmdgroup_builder_delimit(&builder, &words);

src/parser/cmdgroup/cmdgroup_builder.c

@@ -6,7 +6,7 @@
 /*   By: khais <marvin@42.fr>                       +#+  +:+       +#+        */
 /*                                                +#+#+#+#+#+   +#+           */
 /*   Created: 2025/03/20 12:32:14 by khais             #+#    #+#             */
-/*   Updated: 2025/03/20 12:49:37 by khais            ###   ########.fr       */
+/*   Updated: 2025/03/20 17:19:55 by khais            ###   ########.fr       */
 /*                                                                            */
 /* ************************************************************************** */
 
@@ -51,3 +51,23 @@ void	cmdgroup_builder_delimit(t_cmdgroup_builder *builder,
 	builder->current_wordlist = NULL;
 	(void)words;
 }
+
+/*
+** discard the current word
+*/
+void	cmdgroup_builder_discard_word(t_cmdgroup_builder *builder,
+			t_wordlist **words)
+{
+	worddesc_destroy(builder->current_word);
+	builder->current_word = wordlist_pop(words);
+}
+
+/*
+** Handle encountering a parenthesis token
+*/
+void	cmdgroup_builder_paren(t_cmdgroup_builder *builder,
+			t_wordlist **words, t_paren paren)
+{
+	cmdgroup_builder_discard_word(builder, words);
+	(void)paren;
+}

src/parser/cmdgroup/cmdgroup_builder.h

@@ -6,7 +6,7 @@
 /*   By: khais <marvin@42.fr>                       +#+  +:+       +#+        */
 /*                                                +#+#+#+#+#+   +#+           */
 /*   Created: 2025/03/20 12:27:20 by khais             #+#    #+#             */
-/*   Updated: 2025/03/20 12:50:09 by khais            ###   ########.fr       */
+/*   Updated: 2025/03/20 17:12:52 by khais            ###   ########.fr       */
 /*                                                                            */
 /* ************************************************************************** */
 
@@ -15,6 +15,7 @@
 
 # include "../wordlist/wordlist.h"
 # include "cmdgroup.h"
+# include "paren.h"
 
 typedef struct s_cmdgroup_builder
 {
@@ -34,6 +35,12 @@ typedef struct s_cmdgroup_builder
 	** cmdgroup that is being constructed
 	*/
 	t_cmdgroup	*cmdgroup;
+	/*
+	** count of parenthesis tokens
+	**
+	** +1 for each opening parenthesis, -1 for each closing
+	*/
+	int			paren;
 }	t_cmdgroup_builder;
 
 t_cmdgroup_builder	*setup_cmdgroup_builder(t_cmdgroup_builder *builder,
@@ -42,5 +49,7 @@ void				cmdgroup_builder_next_word(t_cmdgroup_builder *builder,
 						t_wordlist **words);
 void				cmdgroup_builder_delimit(t_cmdgroup_builder *builder,
 						t_wordlist **words);
+void				cmdgroup_builder_paren(t_cmdgroup_builder *builder,
+						t_wordlist **words, t_paren paren);
 
 #endif // CMDGROUP_BUILDER_H

src/parser/cmdgroup/paren.c

@@ -0,0 +1,23 @@
+/* ************************************************************************** */
+/*                                                                            */
+/*                                                        :::      ::::::::   */
+/*   paren.c                                            :+:      :+:    :+:   */
+/*                                                    +:+ +:+         +:+     */
+/*   By: khais <marvin@42.fr>                       +#+  +:+       +#+        */
+/*                                                +#+#+#+#+#+   +#+           */
+/*   Created: 2025/03/20 14:48:55 by khais             #+#    #+#             */
+/*   Updated: 2025/03/20 14:49:46 by khais            ###   ########.fr       */
+/*                                                                            */
+/* ************************************************************************** */
+
+#include "paren.h"
+#include "libft.h"
+
+t_paren	match_paren(char *word)
+{
+	if (ft_strcmp("(", word) == 0)
+		return (PAREN_OPEN);
+	if (ft_strcmp(")", word) == 0)
+		return (PAREN_CLOSE);
+	return (PAREN_INVALID);
+}

src/parser/cmdgroup/paren.h

@@ -0,0 +1,25 @@
+/* ************************************************************************** */
+/*                                                                            */
+/*                                                        :::      ::::::::   */
+/*   paren.h                                            :+:      :+:    :+:   */
+/*                                                    +:+ +:+         +:+     */
+/*   By: khais <marvin@42.fr>                       +#+  +:+       +#+        */
+/*                                                +#+#+#+#+#+   +#+           */
+/*   Created: 2025/03/20 14:45:48 by khais             #+#    #+#             */
+/*   Updated: 2025/03/20 14:48:00 by khais            ###   ########.fr       */
+/*                                                                            */
+/* ************************************************************************** */
+
+#ifndef PAREN_H
+# define PAREN_H
+
+typedef enum e_paren
+{
+	PAREN_INVALID,
+	PAREN_OPEN,
+	PAREN_CLOSE,
+}	t_paren;
+
+t_paren	match_paren(char *word);
+
+#endif // PAREN_H

test.sh

@@ -192,4 +192,75 @@ expecting <<EOF
      ╰─ (no redirections)
 EOF
 
+when_run <<EOF "a single command in parentheses is has the parenthesis implicitly removed"
+(echo hi)
+EOF
+# this is how it is handled in bash:
+# doing a `ps uf` on bash running `sleep 10`
+# \_ bash --norc
+#     \_ sleep 10
+# and `(sleep 10)`
+# \_ bash --norc
+#     \_ sleep 10
+# this is exactly the same process graph
+expecting <<EOF
+ ╰─ t_cmdgroup
+     ├─ t_cmdlist
+     │   ├─ num_cmds = 1
+     │   ╰─ cmd[0]
+     │       ├─ t_cmdlist_item
+     │       │   ╰─ t_pipeline
+     │       │       ├─ num_cmd = 1
+     │       │       ╰─ cmd[0]
+     │       │           ╰─ t_simple_cmd
+     │       │               ├─ words = [echo][hi]
+     │       │               ╰─ (no redirections)
+     │       ╰─ t_operator = END
+     ╰─ (no redirections)
+EOF
+
+when_run <<EOF "a single cmdlist in parenthesis has the parenthesis implicitly removed"
+(echo hi >> append | echo < infile bye && echo hello > outfile)
+EOF
+expecting <<EOF
+ ╰─ t_cmdgroup
+     ├─ t_cmdlist
+     │   ├─ num_cmds = 2
+     │   ├─ cmd[0]
+     │   │   ├─ t_cmdlist_item
+     │   │   │   ╰─ t_pipeline
+     │   │   │       ├─ num_cmd = 2
+     │   │   │       ├─ cmd[0]
+     │   │   │       │   ╰─ t_simple_cmd
+     │   │   │       │       ├─ words = [echo][hi]
+     │   │   │       │       ╰─ t_redir_list
+     │   │   │       │           ╰─ redirection[0]
+     │   │   │       │               ╰─ t_redirection
+     │   │   │       │                   ├─ t_redir_type = REDIR_APPEND
+     │   │   │       │                   ╰─ marker = [append]
+     │   │   │       ╰─ cmd[1]
+     │   │   │           ╰─ t_simple_cmd
+     │   │   │               ├─ words = [echo][bye]
+     │   │   │               ╰─ t_redir_list
+     │   │   │                   ╰─ redirection[0]
+     │   │   │                       ╰─ t_redirection
+     │   │   │                           ├─ t_redir_type = REDIR_INPUT
+     │   │   │                           ╰─ marker = [infile]
+     │   │   ╰─ t_operator = AND
+     │   ╰─ cmd[1]
+     │       ├─ t_cmdlist_item
+     │       │   ╰─ t_pipeline
+     │       │       ├─ num_cmd = 1
+     │       │       ╰─ cmd[0]
+     │       │           ╰─ t_simple_cmd
+     │       │               ├─ words = [echo][hello]
+     │       │               ╰─ t_redir_list
+     │       │                   ╰─ redirection[0]
+     │       │                       ╰─ t_redirection
+     │       │                           ├─ t_redir_type = REDIR_OUTPUT
+     │       │                           ╰─ marker = [outfile]
+     │       ╰─ t_operator = END
+     ╰─ (no redirections)
+EOF
+
 finalize