Implement your own shell that runs on top of the regular command-line interpreter for Linux

For this assignment you will implement your own shell that runs on top of the regular command-line interpreter for Linux. Your shell should read lines of user input into a 1024 byte buffer, then parse and execute the commands by forking/creating new processes.

For each command, your shell should call fork() followed by execvp(). Following each command, your shell should wait for its child process to complete, and then reprint the command prompt. The user should be able to specify the command to execute by giving a path to the executable file (e.g. /bin/ls) or by using path expansion to locate the executable file (i.e.searching each directory in the PATH environment variable). (Note that the execvp() function perform this processing automatically; you do not need to program it yourself.)

If your shell encounters an error while reading a line of input it should report the error and exit. If your shell encounters EOF while reading a line of input, it should exit gracefully without reporting an error.

Ensure that you do not overflow your 1024 byte buffer when fetching the line of input (functions that do not accept the size of your buffer are not able to prevent overflows whereas functions that do accept a size generally do; be sure to check the manpage of any function you use carefully). You do not need to report an error if the user's input line is larger than the 1024 byte buffer; just use the truncated input as the command.

Before your shell forks a new process to call execvp(), it should parse the input string and separate it into a collection of substrings representing the executable file and any command-line arguments. If the user entered an empty line, report an error and fetch a new line of input. Your code must handle up to four command-line arguments (in addition to the name of the executable file itself). If the user enters a command with more than four arguments, report an error and fetch a new line of input.

You should store pointers to the substrings in an array (similar to the “argv” array passed to main()) and pass this array of arguments to execvp(). Note that the number of command-line arguments is variable; this is indicated in the array by including a NULL pointer in the array after the last substring. (This means that if the user specifies N substrings, your array must hold N + 1 pointers where the last pointer is NULL.) If the user enters the exit command, your shell should terminate (returning to the regular shell).

Here is a sample execution:

/bin/ls -l /dev/null

crw-rw-rw- 1 root wheel 3, 2 Feb 21 15:26 /dev/null

echo these are some args

these are some args

echo this is too many args

Too many arguments

Empty command

Powered by