macpeng / shell-practice

shell-practice

This Is About Unix Utilities

Build a few different UNIX utilities, simple versions of commonly used commands like cat, ls, etc. We'll call each of them a slightly different name to avoid confusion; for example, instead of cat, implementing wcat (i.e., "wisconsin" cat).

Objectives:

• Re-familiarize with the C/C++ programming language
• Re-familiarize with a shell / terminal / command-line of UNIX
• Learn (as a side effect) how to use a proper code editor such as emacs
• Learn a little about how UNIX utilities are implemented

Summary of what gets turned in:

• A bunch of single .cpp files for each of the utilities below: wcat.cpp, wgrep.cpp, wzip.cpp, and wunzip.cpp.
• Make sure use the open(), read(), write(), and close() calls for file, STDOUT, and STDIN I/O
• Each should compile successfully when compiled with the -Wall and -Werror flags.

See these tutorials for git instructions:

• Git Handbook
• GitHub / Git Cheatsheet

wcat

The program wcat is a simple program. Generally, it reads a file as specified by the user and prints its contents. A typical usage is as follows, in which the user wants to see the contents of main.cpp, and thus types:

prompt> ./wcat main.cpp
#include <stdio.h>
...

As shown, wcat reads the file main.cpp and prints out its contents. The "./" before the wcat above is a UNIX thing; it just tells the system which directory to find wcat in (in this case, in the "." (dot) directory, which means the current working directory).

To create the wcat binary wcat.cpp, and writing a little C++ code to implement this simplified version of cat. To compile this program, will do the following:

prompt> g++ -o wcat wcat.cpp -Wall -Werror
prompt> 

This will make a single executable binary called wcat which you can then run as above.

You'll need to learn how to use a few library routines from the C++ standard library (often called libc++) to implement the source code for this program, which we'll assume is in a file called wcat.cpp. All C++ code is automatically linked with the C++ library, which is full of useful functions you can call to implement the program.

For this project,required to use the following routines to do file input and output: open, read, write, and close.

On UNIX systems, the best way to read about such functions is to use what are called the man pages (short for manual). In our HTML/web-driven world, the man pages feel a bit antiquated, but they are useful and informative and generally quite easy to use.

To access the man page for open, for example, just type the following at your UNIX shell prompt:

prompt> man 2 open

Then, read! Reading man pages effectively takes practice; why not start learning now?

We will also give a simple overview here. The open function "opens" a file, which is a common way in UNIX systems to begin the process of file access. In this case, opening a file just gives you back a file descriptor, which is an identifier that you can then passed to other routines to read, write, etc.

Here is a typical usage of open:

int fileDescriptor = open("main.cpp", O_RDONLY);
if (fileDescriptor < 0) {
    cerr << "cannot open file" << endl;
    exit(1);
}

A couple of points here. First, note that open takes two arguments: the name of the file and the mode. The latter just indicates what we plan to do with the file. In this case, because we wish to read the file, we pass O_RDONLY as the second argument. Read the man pages to see what other options are available.

Second, note the critical checking of whether the open actually succeeded. This is not Java where an exception will be thrown when things goes wrong; rather, it is C++, and it is expected (in good programs, i.e., the only kind you'd want to write) that you always will check if the call succeeded. Reading the man page tells you the details of what is returned when an error is encountered; in this case, the macOS man page says:

If successful, open() returns a non-negative integer, termed a file
descriptor.  It returns -1 on failure.  The file pointer (used to mark
the current position within the file) is set to the beginning of the
file. 

Thus, as the code above does, please check that open does not return -1 before trying to use the file descriptor it returns.

Third, note that when the error case occurs, the program prints a message and then exits with error status of 1. In UNIX systems, it is traditional to return 0 upon success, and non-zero upon failure. Here, we will use 1 to indicate failure.

Side note: if open() does fail, there are many reasons possible as to why. You can use the functions perror() or strerror() to print out more about why the error occurred; learn about those on your own (using ... you guessed it ... the man pages!).

Once a file is open, there are many different ways to read from it. The one we're requiring here to you is read(), which is used to get input from file descriptors.

To print out file contents, use write(). For example, after reading in a line with read() into a variable buffer and keeping track of the number of bytes that you read in a variable bytesRead, you can just print out the buffer as follows:

write(fileDescriptor, buffer, bytesRead)

Finally, when you are done reading and printing, use close() to close the file (thus indicating you no longer need to read from it).

Details

• Your program wcat can be invoked with one or more files on the command line; it should just print out each file in turn.
• In all non-error cases, wcat should exit with status code 0, usually by returning a 0 from main() (or by calling exit(0)).
• If no files are specified on the command line, wcat should just exit and return 0. Note that this is slightly different than the behavior of normal UNIX cat (if you'd like to, figure out the difference).
• If the program tries to open() a file and fails, it should print the exact message "wcat: cannot open file" (followed by a newline) and exit with status code 1. If multiple files are specified on the command line, the files should be printed out in order until the end of the file list is reached or an error opening a file is reached (at which point the error message is printed and wcat exits).

wgrep

The second utility you will build is called wgrep, a variant of the UNIX tool grep. This tool looks through a file, line by line, trying to find a user-specified search term in the line. If a line has the word within it, the line is printed out, otherwise it is not.

Here is how a user would look for the term foo in the file bar.txt:

prompt> ./wgrep foo bar.txt
this line has foo in it
so does this foolish line; do you see where?
even this line, which has barfood in it, will be printed.

Details

• Your program wgrep is always passed a search term and zero or more files to grep through (thus, more than one is possible). It should go through each line and see if the search term is in it; if so, the line should be printed, and if not, the line should be skipped.
• The matching is case sensitive. Thus, if searching for foo, lines with Foo will not match.
• Lines can be arbitrarily long (that is, you may see many many characters before you encounter a newline character, \n). wgrep should work as expected even with very long lines. For this, you might want to write a function that reads from a file descriptor and buffers data until you reach a newline character.
• If wgrep is passed no command-line arguments, it should print "wgrep: searchterm [file ...]" (followed by a newline) and exit with status 1.
• If wgrep encounters a file that it cannot open, it should print "wgrep: cannot open file" (followed by a newline) and exit with status 1.
• In all other cases, wgrep should exit with return code 0.
• If a search term, but no file, is specified, wgrep should work, but instead of reading from a file, wgrep should read from standard input. Doing so is easy, because the file descriptor STDIN_FILENO is already open; you can use read() to read from it.
• For simplicity, if passed the empty string as a search string, wgrep can either match NO lines or match ALL lines, both are acceptable.

wzip and wunzip

The next tools you will build come in a pair, because one (wzip) is a file compression tool, and the other (wunzip) is a file decompression tool.

The type of compression used here is a simple form of compression called run-length encoding (RLE). RLE is quite simple: when you encounter n characters of the same type in a row, the compression tool (wzip) will turn that into the number n and a single instance of the character.

Thus, if we had a file with the following contents:

aaaaaaaaaabbbb

the tool would turn it (logically) into:

10a4b

However, the exact format of the compressed file is quite important; here, you will write out a 4-byte integer in binary format followed by the single character in ASCII. Thus, a compressed file will consist of some number of 5-byte entries, each of which is comprised of a 4-byte integer (the run length) and the single character.

To write out an integer in binary format (not ASCII), you should use write(). Read the man page for more details. For wzip, all output should be written to standard output (the STDOUT_FILENO file descriptor, which, is already open when the program starts running).

Note that typical usage of the wzip tool would thus use shell redirection in order to write the compressed output to a file. For example, to compress the file file.txt into a (hopefully smaller) file.z, you would type:

prompt> ./wzip file.txt > file.z

The "greater than" sign is a UNIX shell redirection; in this case, it ensures that the output from wzip is written to the file file.z (instead of being printed to the screen). You'll learn more about how this works a little later in the course.

The wunzip tool simply does the reverse of the wzip tool, taking in a compressed file and writing (to standard output again) the uncompressed results. For example, to see the contents of file.txt, you would type:

prompt> ./wunzip file.z

wunzip should read in the compressed file (using read()) and print out the uncompressed output to standard output using write().

Details

• Correct invocation should pass one or more files via the command line to the program; if no files are specified, the program should exit with return code 1 and print "wzip: file1 [file2 ...]" (followed by a newline) or "wunzip: file1 [file2 ...]" (followed by a newline) for wzip and wunzip respectively.
• The format of the compressed file must match the description above exactly (a 4-byte integer followed by a character for each run).
• Do note that if multiple files are passed to *wzip, they are compressed into a single compressed output, and when unzipped, will turn into a single uncompressed stream of text (thus, the information that multiple files were originally input into wzip is lost). The same thing holds for wunzip.