File Utilities and the find Command

Working with Files

Often, you may copy files from one directory to another. Use the cp command to perform this task. To copy the file /usr/X11R6/lib/X11/xinit/Xclients to the Xclients.sample file in the current directory (such as your home directory), type the following:

cp /usr/X11R6/lib/X11/xinit/Xclients Xclients.sample

If you want to copy a file to the current directory and retain the same name, use a period (.) as the second argument of the cp command. Thus, the following command copies the XF86Config file from the /etc/X11 directory to the current directory (denoted by a single period):

cp /etc/X11/XF86Config .

The cp command makes a new copy of a file and leaves the original intact.

Secret

If you want to copy the entire contents of a directory—including all subdirectories and their contents—to another directory, use the command cp -ar sourcedir destdir (this copies everything under sourcedir directory to destdir). For example, to copy all files from the /etc/X11 directory to the current directory, type the following command:

cp -ar /etc/X11 .

Another GNU utility command, mv, moves a file to a new location. The original copy is gone, and a new copy appears at the specified destination. You can use mv to rename a file. If you want to change the name of the today.list to old.list, use the mv command, as follows:

mv today.list old.list

On the other hand, if you want to move the today.list file to a subdirectory named saved, use this command:

mv today.list saved

An interesting feature of mv is the fact that you can use it to move entire directories, with all their subdirectories and files, to a new location. If you have a directory named data that contains many files and subdirectories, you can move that entire directory structure to old_data by using the following command:

mv data old_data

Another common file operation is deleting a file. Use the rm command to delete a file named old.list, for example, by typing the following command:

rm old.list

Manipulating Directories

To organize files in your home directory, you have to create new directories. Use the mkdir command to create a directory. For example, to create a directory named images in the current directory, type the following:

mkdir images

After you create the directory, you can type the cd images command to change to that directory.

You can create an entire directory tree by using the -p option of the mkdir command. For example, suppose that your system has a /usr/src directory and you want to create the directory tree /usr/src/book/java/examples/applets. You can create this directory hierarchy by typing the following command:

mkdir -p /usr/src/book/java/examples/applets

When you no longer need a directory, use the rmdir command to delete it. You can delete a directory only when the directory is empty.

To remove an empty directory tree, you can use the –p option, like this:

rmdir -p /usr/src/book/java/examples/applets

This command removes the empty parent directories of applets. The command stops when it encounters a directory that’s not empty.

Copying Disks with dd

The dd command is useful for copying binary data to a floppy disk (or to other devices, such as tape). For example, if you want to make a Red Hat boot disk on a Linux system, you can use the dd command to prepare the boot disk before you install Red Hat Linux from this book’s companion CD-ROMs. Perform the following steps to prepare the boot disk:

1. Log in as root and mount the first CD from this book’s companion CD-ROMs on /mnt/cdrom (type mount /mnt/cdrom). Change to the directory with the boot images as follows:

   cd /mnt/cdrom/images

2. Place a formatted floppy disk in the floppy drive, then type the following command to copy the bootdisk.img file to the floppy disk:

   dd if=bootdisk.img of=/dev/fd0 bs=1440k

If you want to copy another boot image to the floppy, replace bootdisk.img with that file’s name.

Viewing Disk Usage Information

Two programs in the GNU file utilities—df and du—enable you to check disk-space usage. These commands are simple to use. The df command shows you a summary of disk-space usage for all mounted devices, as shown in the following example:

df
Filesystem           1K-blocks      Used Available Use% Mounted on
/dev/hda5              7392428   3412688   3604224  49% /
/dev/hda3               101107      8865     87021  10% /boot
none                    127328         0    127328   0% /dev/shm
/dev/cdrom               75322     75322         0 100% /mnt/cdrom

The output is a table that shows the device, the total kilobytes of storage, how much is in use, how much is available, the percentage being used, and the mount point. For example, on my system, the /dev/hda5 device (a disk partition) is mounted on the Linux file system’s root directory; it has about 7.4GB of space, of which 3.4GB (or 49 percent) is being used, and 3.6GB is available. Similarly, you can see from the last line that the CD-ROM has about 75MB of storage in use.

The other command, du, is useful for finding out how much space a directory is using. For example, type the following command to view the contents of all the subdirectories in the /var/log directory (this directory contains various error logs):

du /var/log
4       /var/log/vbox
12528   /var/log/cups
4       /var/log/samba
4       /var/log/news/OLD
8       /var/log/news
60      /var/log/httpd
4       /var/log/squid
36      /var/log/gdm
14764   /var/log

Each directory name is preceded by a number—that number denotes the number of kilobytes of disk space that directory uses. Thus, the /var/log directory, as a whole, uses 14764KB, or about 14.7MB, of disk space, whereas the /var/log/httpd subdirectory uses 60KB.

You can use the -h option with the du command to view the disk-space usage in human-readable format. For example, here’s what you get when you type du -h /var/log to view the disk space used by the /var/log directory and its contents:

du -h /var/log
4.0K    /var/log/vbox
13M     /var/log/cups
4.0K    /var/log/samba
4.0K    /var/log/news/OLD
8.0K    /var/log/news
60K     /var/log/httpd
4.0K    /var/log/squid
36K     /var/log/gdm
15M     /var/log

If you simply want the total disk space a directory uses (including all the files and subdirectories contained in that directory), use the -s option together with the -h option, as follows:

du -sh /var/log
15M     /var/log

Notice that the -s option causes du to print just the summary information for the /var/log directory.

Learning the find Command

The find command is very useful for locating files (and directories) that meet specified search criteria. The Linux version of the find command also comes from GNU, and it has more extensive options than the standard UNIX version. I show the syntax for the standard UNIX find command, however, because it syntax works in GNU find, and you can use the same format on other UNIX systems.

I must admit that when I began using UNIX many years ago (Berkeley UNIX in the early 1980s), I was confounded by the find command. I stayed with one basic syntax of find for a long time before graduating to more complex forms. The basic syntax I learned first was for finding a file anywhere in the file system.

Suppose that you want to find any file or directory with a name that starts with gnome. You can use find to perform this search, as follows:

find / -name "gnome*" -print

This command tells find to start looking at the root directory (/), to look for filenames that match gnome*, and to display the full pathname of any matching file.

You can use variations of this simple form of find to locate a file in any directory (as well as any subdirectories contained in the directory). If you forget where in your home directory you have stored all files named report* (names that start with the text string report), you can search for the files by using the following command:

find ~ -name "report*" -print

When you become comfortable with this syntax of find, you can use other options of find. For example, to find only specific types of files (such as directories), use the -type option. The following command displays all top-level directory names in your Linux system:

find / -type d -maxdepth 1 -print

To find all files that exceed 20,000KB (20MB) in size, you can use the following find command:

find / -size +20000k -print

Secret

You can also execute some command [i1]on a file after you find it. For example, suppose that you want to search for the files that exceed 20,000KB in size and then print a detailed directory listing of that file. You can use the -exec operator of find to execute the ls -l command on each file that matches the search criteria, like this:

find / -size +20000k -exec ls -l {} \;

The -exec operator enables you to run any command on the files that meet whatever search criteria you use with find. You can type the command to execute following the -exec operator, then list any arguments, and end the list with a semicolon (;). However, the shell treats semicolon as command-separator, so you must place a backslash (\) in front of the semicolon to stop the shell from interpreting it. That’s why every -exec operator ends with the characters \;. The curly brace pair {} is also has special meaning. The find command replaces the characters {} with the name of the file it has found—the file that satisfies whatever search criteria you have provided.

Another good use of find’s -exec operator is to find the files with permission settings that may be too dangerous and then change the permission by using the chmod command. For example, suppose you want to locate any file with a group-write permission (that means the group owner is allowed to alter the file) and take away that permission. Here is a find command that uses the -exec operator to accomplish this task:

find / -perm -20 -exec chmod g-w {} \;

You probably do not have to use the complex forms of find very often in a typical Linux system, but you can look up the rest of the find options by using the following command:

man find