This project has been created as part of the 42 curriculum by mhnatovs.

Pipex

Description

Pipex is a 42 school project that drills into the depths of UNIX mechanisms. The program mimics the behavior of the shell pipe command: < file1 cmd1 | cmd2 > file2. It handles the redirection of standard input and output, executes system commands via child processes, and manages the synchronization between them using pipes.

Features

• Redirection: Handles input from a file and output to another file.
• Piping: Implements a pipe to pass the output of the first command as the input to the second.
• Path Resolution: Searches the PATH environment variable to find and execute binaries.
• Error Handling: Robust management of file permissions, non-existent commands, and memory allocation.

Instructions

Prerequisites

• Linux or macOS
• cc compiler
• make

Compilation

To compile the project, run:

make

This will produce the pipex executable.

Usage: The program should be executed as follows:

• file1: Input file (source).
• cmd1: First command with its parameters.
• cmd2: Second command with its parameters.
• file2: Output file (destination).

Example: This is equivalent to the shell command: < infile ls -l | wc -l > outfile

Technical Concepts

• pipe():Creating a unidirectional data channel for inter-process communication.
• fork(): Creating child processes to run multiple commands simultaneously.
• dup2(): Redirecting file descriptors (STDIN/STDOUT) to the pipe or files.
• execve(): Replacing the current process image with a new process (executing the command).
• Environment Variables: Parsing the char **envp to locate command binaries.

AI Usage

AI was utilized during development for:

• Learning support:

Visualizing the flow of file descriptors ($0, 1, 2$) during a fork() and dup2() call. Explaining the necessity of closing unused pipe ends to avoid hanging processes (EOF behavior). Understanding how waitpid() synchronizes the parent process with its children.

• Debugging assistance:

Identifying "zombie processes" and ensuring all children are properly reaped. Debugging edge cases like /dev/urandom as input or commands without absolute paths.

What I did NOT use AI for:

The logic of the pipe-to-child-process execution flow. Handling the environment path parsing and command split logic.

Team

mhnatovs (https://github.com/mhnatovs) — Single Contributor / Developer