Organizing Your Hard Drive By Ken Johnson, Chicago Computer Society In last month's column we looked at files and file names. This month we'll discuss how files are organized on a disk. Your computer's hard disk can store many files, but finding and using specific files (whether programs or data) can be difficult without some sort of organization. Luckily DOS provides such organization via a "subdirectory structure." A physical disk is organized into a logical structure consisting of a root directory (designated by a backslash "\") and various named subdirectories which branch off the root. Graphically, the directory structure looks like an upside-down tree. This tree- structured directory system allows you to categorize your files, placing like files into the same directory. Most users divide their hard disk into several directories grouped around similar types of applications or data: C:\ (ROOT) 旼컴컴컴컴쩡컴컴컨컴컴컴컫컴컴컴컴컴� 123R23 WP51 UTIL DOS 旼컴컴켸 � � � 旼컴좔컴커 旼컴좔컴� BIN NORTON � � � � BUDGET TAXES MEMOS ARTICLES Thus, a directory structure grows like a tree. The root directory contains all the main "branches" and each branch may contain its own branches. The "leaves" of the tree are the files: C:\ (ROOT) 旼 � � � � � � WP51 � � 컫컴컴컨 � � MEMOS � MEMO5-4.92 MKTPLAN.FEB 2DO.LST CLIENTS.LTR The file MEMO5-4.92 is in the MEMOS subdirectory, which is in the WP51 subdirectory, which is in the root directory of the C: drive. The sequence of directories leading from the root to the file is called the path. As we discussed in last month's column, to find a file DOS needs to know the path to the file, also that file's name and extension. This is called the filespec. The filespec for the file above is: C:\WP51\MEMOS\MEMO5-4.92 where: "C:\" indicates the root directory of the C: drive, "WP51\MEMOS\" indicates the MEMOS subdirectory of the WP51 subdirectory, and "MEMO5-4.92" is the file name and extension. Organization of files into a logical subdirectory structure is very important. Most application programs will create their own subdirectories and install their files there. However, often you will need to decide where to store your data files. A general rule is to store your data in a separate subdirectory from the program that created that data. For example, don't store your WordPerfect documents with the WordPerfect programs. The basic reason for storing data separate from programs is that you care much more about the data than the programs. By keeping the data separate, it is much easier to backup those important files and more likely that you will do it. Data subdirectories tend to be smaller, and often you can copy all the files onto a single high density floppy using COPY or XCOPY. This can easily be automated in a batch file, letting you make backup copies of data files without having to really think about it. The Root Directory When you boot up your computer, you are logged into the root directory of the hard drive C: (or the A: drive if you've started with a DOS disk in the A: drive). The root directory doesn't really have a name, rather the backslash character "\" stands for the root directory. From the root, you use the CD (change directory) to move to the subdirectories. One sure sign of a disorganized hard drive is a root directory full of files. The root directory need only contain a few necessary files ("necessary" because DOS will look for them in the root of the boot drive), including: AUTOEXEC.BAT The AUTOmatically EXECute "startup" BATch file containing commands to execute whenever the computer starts up. CONFIG.SYS The file specifying CONFIGuration of your SYStem, which loads device drivers and configures memory. The root will also contain the two hidden DOS files (IBMBIO.COM & IBMDOS.COM, or IO.SYS & MSDOS.SYS), and a hidden file containing the volume label of the disk. Many people keep COMMAND.COM, the command interpreter, in the root directory but this is not required. The SHELL= statement in CONFIG.SYS can tell DOS where to find COMMAND.COM if not in the root directory. (The SHELL statement in CONFIG.SYS tells DOS the name of the command interpreter (usually COMMAND.COM), its location, and the size of the DOS environment. SHELL was discussed in the January 1992 and the September 1991 Hard-Copy Beginner's Columns.) There is a limit on the number of files that can exist in the root directory of any disk. The maximums are 112 files on a Double Density, Double Sided 5� or 3� floppy disk, 224 files on a high density 5� or 3� disk, and 512 files on most hard disks. However, there are no limits on the number of files in a subdirectory. Subdirectories are just files that keep track of other files, and like any file can grow if there is disk space. Subdirectories Most of DOS' work is done in subdirectories. As mentioned above, there is no limit on the number of files in a subdirectory, nor is there a limit to the number of subdirectories you can have on a hard disk. There is however a limit of the number of subdirectory levels you can have, based on the length of the subdirectory names. DOS limits a directory path to 63 characters, measured from the root to the bottom level. If you use 8-character subdirectory names, this gives you a limit of about seven levels. In practice, you won't have to worry about this limitation since most users seldom have subdirectories more than 3-4 levels deep. The most important thing to remember about the directory structure is that you can only be in one directory at a time. This is called the "default directory," meaning that unless you tell DOS otherwise, it will perform operations against files in that directory. You can tell DOS to look in other directories by using the PATH statement, and move to other subdirectories with the CD command. We'll look at these commands below. There are a couple of different ways to view subdirectories with the DIR command. If you have DOS 5.0, you can use the /AD switch to show only directories. Another method is to use "DIR *.", that will show all files without extensions. Since most subdirectories don't have extensions (though there is no reason they can't) and most files do have extensions, this command will usually show all the subdirectories one level down from where you are. For example, here is the output from "DIR /AD" on the root directory of my D: drive: Volume in drive D is STACVOL_DSK Directory of D:\ 123R23 12-01-91 2:45p ARTICLES 04-12-92 6:54p CSERVE 12-08-91 7:15p EXCEL 07-05-91 6:17p GMK 04-11-92 4:24p JUNK 02-17-92 8:31p TAP 03-21-92 11:50a VB 07-13-91 10:45p VISION 12-22-91 7:04p VISIONR 12-22-91 7:25p WP51 01-24-92 8:08p WPC 03-29-92 5:47p WPWIN 03-29-92 5:46p WPWINDOC 03-29-92 5:46p 18 file(s) 0 bytes 18792448 bytes free There are a couple of interesting items about these subdirectory listings. Instead of showing a file size, subdirectories are displayed with a "." This shows that these are directories, not files without extensions. You can see from the byte listing at the bottom of the display that subdirectory "files" have a zero-byte length. If you do a DIR command on a subdirectory, you'll see two special entries at the beginning of the listing: Volume in drive D is STACVOL_DSK Directory of D:\ARTICLES . 04-12-92 6:54p .. 04-12-92 6:54p MEMHELP BAK 19534 02-28-92 7:43p DEFINEKY BAT 1159 03-04-91 6:17p AUTOEXEC HLP 50345 08-20-91 8:44p CACHE HLP 6243 12-29-91 11:06a PROFILE KEJ 1361 01-26-92 7:48p XNAME LIB 4485 05-23-92 3:02p CACHE LOG 8687 12-22-91 7:54p DOSBOOK MST 5538 04-14-92 9:36p DELERASE OUT 13017 01-25-92 10:05a The "." stands for the current directory, here D:\ARTICLES. The double periods ".." stand for the parent directory, the directory that is immediately above this one. In this example the parent directory is the root directory (D:\). You can use this ".." shorthand to quickly move up one level in the directory tree, as we'll see below. Now let's look at the commands used to manage subdirectories and files. MD (or MKDIR) -- create a new subdirectory The MD Command (no one uses the long version, MKDIR) will create a new subdirectory. The syntax is: MD [d:][path]dirname where "dirname" is the subdirectory name. If you simply enter the name, the subdirectory will be created below the directory that you are currently in. Alternatively, you can show exactly where to create this directory by including drive and path information. For example, suppose you are in the C:\DOS subdirectory. The command MD UTILITY will create a C:\DOS\UTILITY subdirectory. With no path given, the new directory is created relative to the default directory. The command MD \UTILITY will create a C:\UTILITY directory directly off the root, no matter which subdirectory you are in on the C: drive when you issue the command. RD (or RMDIR) -- delete a directory To remove a subdirectory no longer needed, you can use the RD command (again, no one uses the longer RMDIR version). The syntax is: RD [d:][path]dirname where "dirname" is the directory to remove. You cannot remove the root directory, or directory you are currently in (the default directory). A directory must be empty, containing no files or subdirectories, before it can be removed. A hidden file may prevent you from removing a directory, since the file cannot be deleted with DEL or ERASE. If you have DOS 5.0, you can use the ATTRIB command to turn off the Hidden attribute, then delete the file normally. If you don't have DOS 5.0, you'll have to use a third party utility such as PCTools or Norton Utilities to unhide and delete such files. CD (or CHDIR) -- changes the default directory The CD Change Directory command is used to move around the directory tree. The syntax is: CD [d:][path][dirname] [..] [\] where "dirname" is the name of the directory that you want to move to. If you simply give the directory name, then that directory must be directly below the default directory. Otherwise you will receive an "Invalid directory" message. You can move to any directory on the drive by explicitly giving the path and directory name. Other uses of the CD command are: CD \ Will move you back to the root directory. CD .. Will move you up one level to the "parent" directory of the directory you are currently in. You can also use this as shorthand to move to another directory with the same parent, as shown below. CD Will show the name of the current directory. Let's look at some examples of the CD command. Let's say that our disk is organized like so and that you are in the ARTICLES subdirectory under the WP51 subdirectory: C:\ (ROOT) 旼컴컴컴컴쩡컴컴컨컴컴컴컫컴컴컴컴컴� 123R23 WP51 UTIL DOS � � � 旼컴컨컴커 BIN NORTON � � MEMOS ARTICLES <-- you are here � � 旼컴컴컴좆 旼컨컴컴커 � � � � 1991 1992 CCS LAWMUG CD .. will move you up to the parent, here the WP51 subdirectory (C:\WP51). CD \ will move you back to the root directory (C:\). CD CCS will move you down to the CCS subdirectory (C:\WP51\ARTICLES\CCS). Since there is no path specified, DOS looks for this subdirectory below the current directory. CD \UTIL\BIN will move you to the BIN subdirectory (C:\UTIL\BIN). Since the directory name starts with \, DOS begins looking for the UTIL subdirectory off the root. CD ..\MEMOS will move you to the MEMOS subdirectory (C:\WP51\MEMOS). You use the ".." as shorthand for the parent directory, here C:\WP51. It is a quick way to move to another subdirectory with the same parent. CD will simply display the current directory (C:\ARTICLES). PROMPT -- change to DOS prompt to show the current subdirectory In a complicated directory structure it is easy to get lost. You can use the PROMPT command can end subdirectory confusion by always displaying the name of the current directory. You've seen the default DOS prompt often, the infamous "C>". The PROMPT command can change "C>" to something a lot more informative. The syntax of the command is: PROMPT [prompt_text] where "prompt_text" is any string of text or special characters. These special characters will display system information as part of your prompt: $P the default directory $T the current time (HH:MM:SS.xx) $D the current date (MM-DD-YY) $V the DOS version $N the default drive $_ Carriage Return (moves the cursor to next line) $G the > character $L the < character $B the | character $Q the = character As you can see, the standard DOS prompt is $N$G, the default drive and the ">" sign. If you enter the PROMPT command with nothing after it, DOS resets the prompt to this default. The prompt is usually set in your AUTOEXEC.BAT file, though it can be reissued anytime from the DOS command line. Perhaps the most common prompt setting is the current drive and directory: PROMPT $P$G DOS will display the current drive and current directory, with the greater than sign. For example, if you are in the MEMOS subdirectory of the WP51 directory on your C: drive, your prompt will show: C:\WP51\MEMOS>_ There is one minor problem when having DOS display the current directory ($P). Each time you press Enter, DOS reads the disk to find the directory for the prompt. If your default drive is a floppy drive and you take the disk out (or you try to change to a drive without a disk), DOS will try to read the empty drive and you'll get an error message: Not ready reading drive A: Abort, Retry, Fail?_ At this point you can put a disk in and press R. Or you can simply press F, in which case your prompt will become: Current drive is no longer valid>_ Just change back to your hard drive and your prompt will reappear. By the way, there is an undocumented CONFIG.SYS setting in DOS 5.0 that will enter an "F" for you at any "Abort, Retry, Fail" messages. You do this by adding a "/F" switch to the SHELL=COMMAND.COM statement in CONFIG.SYS. Now anytime an "Abort, Retry, Fail" message occurs, DOS will automatically answer "Fail" and continue. Note that this switch is used with COMMAND.COM as your command processor, and most likely does not apply if you are using another processor (such as Norton's NDOS). PATH -- where to search for files The PATH command tells DOS the subdirectories to search when it can't find the specified program in the current directory. DOS will search this "path" of directories before it returns the "Bad Command or Filename" error message. Some application programs also use the path to find related files. The path is specified by listing the directory to search, separated by semicolons: PATH dirname1;dirname2;dirname3; .... A typical path might look like this: PATH C:\;C:\DOS;C:\UTIL;C:\WINDOWS;C:\PCTOOLS;C:\DOS\NORTON This path tells DOS to look in the following places (in the order listed) for the file specified, if it can't be found in the current directory: C:\ (the root directory of the C: drive) C:\DOS (the DOS subdirectory on the C: drive) C:\UTIL (the UTIL subdirectory on the C: drive) C:\WINDOWS (the WINDOWS subdirectory on the C: drive) C:\PCTOOLS (the PCTOOLS subdirectory on the C: drive) C:\DOS\NORTON (the NORTON subdirectory in the DOS subdirectory on the C: drive) The PATH is normally set in the AUTOEXEC.BAT file, and is limited to 127 characters in length (the limit of the DOS command line). If you have only a single hard drive, you can save a few characters by eliminating the "C:" drive identifier. The path above could be specified as: PATH \;\DOS;\UTIL;\WINDOWS;\PCTOOLS;\DOS\NORTON You'll want to put your important subdirectories on the path, with the most "important" first, but don't include every subdirectory. Long paths mean that DOS must search each of the listed subdirectories, including every time you make a typo or spelling mistake. Long paths can definitely slow you down, particularly with a sluggish hard drive. There are a couple of ways of viewing the current path. If you simply enter "PATH" at the DOS prompt, DOS will display the path. You can also enter the "SET" command to see the path, since DOS keeps the path as an environmental variable. The SET command will show all the environmental variables, including the PATH, PROMPT, and COMSPEC. (See the January 1992 Hard-Copy Beginner's Column for a discussion of the DOS Environment.) Since the path is an environmental variable, it can be accessed and manipulated in a batch file. This means you can dynamically change the path as needed. For example, I use Grammatik 5 as a grammar checker, but don't have it on my PATH statement. However when I want to use Grammatik, the Grammatik subdirectory (\GMK) must be on the path or the program can't run. So I've set up a batch file to put the \GMK subdirectory on the path, run Grammatik with WordPerfect, then remove \GMK from the path: @echo off set oldpath=%path% path=d:\gmk;%path% cd \wp51 gmkwp path=%oldpath% set oldpath= Let's take a closer look at this batch program. The second line creates a new environmental variable "oldpath" that contains the original path (to use an environment variable in a batch file, you surround it with percent signs). The next line creates a new path, with the "D:\GMK" subdirectory as the first directory listed; in essence, the original path is tacked on the end of \GMK. The 4th and 5th lines change to the WordPerfect subdirectory and run the Grammatik program. When Grammatik is done, the 6th line sets the path back to the original path, which was stored in the "oldpath" variable. The final line removes the "oldpath" variable from the environment, freeing up the memory it occupied.