CD ROM Paradise Collection 4/CD ROM Paradise Collection 4 1995 Nov.iso/dskutil/msort112.zip/MSORT.DOC

Text File | 1995-01-21 | 16KB | 388 lines

                        MSORT -- Master Sort Program

                        SSORT -- Speedy Sort Program


                                Version 1.12

               This document and the associated software are
         Copyright 1994 by Martin Katz, Ph.D. all rights reserved.


                                  PURPOSE

MSORT and SSORT are superior text file sorting programs for DOS. MSORT and
SSORT are faster, easier to use and more powerful than the sort program
included with MS-DOS or PC-DOS. In addition, while DOS sort is limited to
small files, MSORT can sort huge files.

MSORT and SSORT allow you to sort files either alphabetically or numerically. You can also define the columns that define the sorting keys, sort in
reverse, ignore spaces while sorting, and treat upper and lower case as the
same, and much more. If you wish, MSORT will remove duplicate lines and/or
blank lines.

These are fully functional programs with no shareware reminders. MSORT and
SSORT are shareware and your support is necessary to continue development
of this product. If you find this software useful, you are obligated to
register it. Please See the LICENSE section below and the registration form
at the end of this file for more details. Please return the registration
form, even if you decide not to pay for this package.

Your comments and suggestions are welcome. Please email comments to me at
"[e-mail]" I can also be reached by sending a message to
"Martin Katz" on the RIME Shareware conference (private is preferred).


                                   LICENSE

Neither the author, nor the distributors of these programs assume
responsibility for any loss related directly or indirectly to the use of
this product. No warranty is implied or stated, including any warranty of
suitability for use. Use this product at your own risk.

If you have problems or questions, please send me a message via e-mail at
[e-mail] (I can also be reached on the RIME shareware
conference). Please include the program name, version number (shown at the
top of this file), a specific explanation of your problem, and your e-mail
address.

These programs may be freely distributed provided that they are not
changed, this file is included, and that no charge is made for this
software, except for media costs (limited to less than $10).

All distribution must include the following files:
   MSORT.DOC,
   MSORT.EXE, and
   SSORT.EXE.

The following additional files are included for convenience:
   FILE_ID.DIZ,
   HISTORY,
   README,
   REGISTER.FRM,
   SORTTAGS.BAT, and
   TESTFILE.COM.


                           USING SSORT and MSORT

You use the programs SSORT and MSORT in the same way. SSORT is faster, but
cannot sort large files (the data must all fit in memory). If SSORT cannot
sort your file, please use MSORT. This section will show SSORT in the
examples, but you can use MSORT the same way.

SSORT and MSORT assume that each line in your file ends with a line-feed
character (carriage-returns are ignored). Formatting and control characters
used in word processors should not be included in the input file (they are
just treated as characters to sort).

The simplest way to sort two files is
    SSORT (input file) (output file)

where (input file) and (output file) are file names. For example, to sort
the file PHONE.BOK and output to PHONE.SRT you would type
    SSORT PHONE.BOK PHONE.SRT

SSORT can also take its input from the DOS standard input and output to DOS
standard output (this is the manner required by the standard DOS sort
program). For example:
   SSORT < PHONE.BOK > PHONE.SRT

WARNING: If you use redirection, the input and output files must always
         be different. Otherwise, the input file will be destroyed. Use
         of the DOS SHARE program (or MS Windows) will prevent this
         destruction of data.


                                  OPTIONS

Options may begin with either a hyphen or a slash, and they may be either
upper or lower case. They are shown with hyphens below for ease of reading.


CONTROLLING THE SORTING ORDER

You can change the sorting order by using any of the following options
before the file names:

    -c   Case independent. Treat all letters as if they are upper case.
         Without this option, upper case comes before lower case.

    -#   Only letters will be used when comparing lines.

    -a   Only letters and digits will be used when comparing lines.

    -d   Only digits will be used when comparing lines.

    -w   Ignores spaces, tabs, etc. (Whitespace).

    -t   Ignore common short words (slow process).  This includes:
         a, an, am, are, do, if, is, I, me, my, that, the, this, we, you

    -n   Sort Numerically. This means that leading a leading plus sign,
         leading spaces, or leading zeroes will be ignored. If the number
         begins with a negative sign, it will be considered negative. If
         one of the remaining numbers is longer than the other, it is
         considered larger than the other.

    -r   Sort in Reverse order

    -s   Sort by line length (size).

    +N   (N is a decimal number) Ignore characters before text column N
         when comparing lines to sort. The first text column is column 1.
         -+N and /+N are the same as +N.

  +N.M   (N and M are decimal numbers) Defines a field to sort beginning at
         column N, and M characters wide. You may define up to 10 fields
         when sorting. Later fields are only checked if the lines are equal
         in earlier fields. -+N.M and /+N.M are the same as +N.M

         The sort order options (-# -a, -c, -d, -n, -r, -s,  or -w) apply
         to the field specified immediately after the options. That is, you
         first state the sort order options and then you define the field.
         If you have more than one field, you should state the sort order
         options for each.


SIMPLE EXAMPLE

You may combine the options, but they must be separated by spaces. For
example:
    SSORT /C /W PHONE.BOK PHONE.SRT
will sort the phone book, ignoring all spaces and tabs, and treating upper
and lower case as equivalent.


MORE COMPLETE PHONE BOOK EXAMPLE

As a more complicated example, assume that each line in the phone book
begins with the name (up to 20 characters), the phone number begins at
column 25, and the address begins at column 40 on the line. You might type
    SSORT -c -w +1.20 -w -c +40 -d -n +25.10 PHONE.BOK PHONE.SRT

The text will be sorted by the name (columns 1-20), ignoring spaces and
tabs, and treating upper and lower case as the same. If two lines contain
the same name, then the addresses are compared, again ignoring spaces and
treating upper and lower case as equivalent. Notice that the address field
begins at column 40, because no length is specified, comparison will extend
to the end of the line. If two lines are still equal, then the phone number
is checked (columns 25-34). The phone number is compared numerically, and
any punctuation is ignored.


REMOVING DUPLICATE OR BLANK LINES

    -b   Remove all Blank lines from the output.

    -u   Only output Unique lines. That is, if a line appears more than
         once, only output the first copy. The sort order options affect how
         this option works. For example, if the -c option is given, then two
         lines will be considered the same, even if one of them is all lower
         case and the other is all upper case.


ADVANCED OPTIONS

These options are provided for completeness, but should not normally be
needed:

 -oNAME  Specifies the Output file name (no space is allowed between the -o
         and the name).

      -  An option that is just a hyphen or slash means to read the input
         from the standard input. If an output file is specified without
         -o, then the output file name must come after this hyphen or slash.

    -lN  Specifies an end of line character other than line-feed. N is the
         decimal ASCII code for the character to use for the line end.
         There must not be a space between -l and the number. When this
         option is specified, the file is assumed to be binary, not text.


                             USING MSORT

The options for MSORT are exactly the same as the options for SSORT.
SSORT tends to be somewhat faster than MSORT, but can only sort files
that can be read into conventional memory.

MSORT will create two temporary files. If you set a TMP (or TEMP)
environment variable the directory specified (if it exists) will be
used for the temporary files. Otherwise, the current directory will be
used. MSORT will fail if it cannot create these temporary files.


                             WHY IS IT SO FAST?

MSORT is designed to use memory wisely to speed up the sorting. It uses a
unique combination of methods (algorithms) that allow it to take advantage
of the ordering of files that are "almost sorted," while also working well
when the data is more or less random.


                                 LIMITATIONS

SSORT is limited to files up to 16000 lines long. In addition, the program
and entire file must fit in available conventional memory.

MSORT is limited by available disk space. MSORT requires disk space for the
input file, the output file, and two temporary files. Thus, if the input
file is N bytes long, you must have at least 3N bytes of disk space
available. MSORT cannot sort files longer than two billion lines, but it is
unlikely that this limit will ever be reached.

Neither MSORT nor SSORT makes use of extended or expanded memory. However,
a disk cache will speed up MSORT on large files.

The length of a line should not exceed 30,000 characters.

In a network or multiprocessing environment, these programs will only work
if they have the privilege of reading the input file, creating and writing
to the output file, and (in the case of MSORT) creating, reading, and
writing to the temporary files.

WARNING:  The input and output files must always be different.
          Otherwise, the input file can be destroyed. SSORT cannot always
          detect this, so be careful. Use of the SHARE program before
          running SSORT or MSORT will prevent destruction of data (SHARE is
          not necessary when running MS-Windows 3.x or a multi-tasking
          operating system).


                              SORTING TAGLINES

SORTTAGS.BAT is a sample batch file that can be used to sort a file of
taglines. You can use it as follows:
    SORTTAGS TagsFile SortedTags
That, is you give it the file name of the existing file and a filename for
the sorted tags. If you do not give a name for an output file, the old file
is renamed as TAGS.OLD, and the sorted output will have the same name as the
original file.

SORTTAGS.BAT uses msort with the following options:
    msort -# -c -t -u -b

This means:
    Digits, spaces, and punctuation will be ignored (-#).

    Upper and lower case letters are considered the same (-c).

    Common small words (like "the") will be ignored (-t) -- this is a slow
    process.

    If two lines are equivalent then only one of the lines will be kept in
    the output (-u). Because of the other options, lines are considered
    equivalent even if they differ in digits, spaces, punctuation, common
    small words, or case (upper or lower).

    Finally, blank lines will be removed (-b).


LONG TAGLINES

Some programs limit the length of taglines to 57 characters. If you sort a
tagline file by length, then all of the long lines will be at the bottom.
You can then use an editor to remove the long lines. You might want to keep
them in a separate file. To sort a file named MYTAGS by length, you could
type:
    MSORT -s MYTAGS BYLENGTH.TAG

You could then edit BYLENGTH.TAG to remove the long lines. As a last step,
you could use SORTTAGS to sort the remaining taglines (please make a backup
before sorting):
    COPY MYTAGS MYTAGS.BAK
    SORTTAGS BYLENGTH.TAG MYTAGS
    DEL BYLENGTH.TAG


                              TESTFILE PROGRAM

The TESTFILE program is a small utility for creating files of numbered
lines. These files are useful for testing sorting and editing programs.
The program has no command line arguments and outputs to standard output.

Here is a sample execution of the program:
  C:\MSORT> TESTFILE > TEST

  Enter first number to print (0...99999999): 1
  Enter last  number to print (0...99999999): 10
  Enter minimum width of numbers to print (0...10): 3
  Enter minimum line width in characters (0...1024): 40
  Enter maximum line width in characters (0...1024): 40
  Enter the ASCII value of character to fill out the line with: 46

This creates a file called TEST that contains ten lines (1 - 10). Each line
will be exactly 40 characters long. Since the lines are long enough, each
line will begin and end with the line number. The line numbers will be
padded out on the left with zeros. The space between the line numbers will
be filled by periods (ASCII 46). The TEST file will look like this:

001..................................001
002..................................002
003..................................003
004..................................004
005..................................005
006..................................006
007..................................007
008..................................008
009..................................009
010..................................010


                              ABOUT THE AUTHOR

Martin Katz is a Professor of Computer Science at California State
University, Fullerton. His specialties include advanced user interfaces,
software engineering, data structures and algorithms, and programming
languages. He teaches a wide variety of courses for students with varying
backgrounds. As a consultant to industry, Dr. Katz has been involved in
development of a variety of software products.

E-Mail address (internet): [e-mail]

_____________________________________________________________________


                    MSORT/SSORT for DOS  (version 1.12)
                             REGISTRATION FORM

This program is not free, it is shareware. Shareware programs are
distributed on the honor system, assuming that honorable users will pay for
the product if it is found to be useful.  If you like the program you are
obligated to register it within 30 days. Registration will support
development of future shareware products.

Please type or print clearly:

MAIL TO:                           FROM:
Martin Katz, Ph.D.                 Name: __________________________________
6452 Adobe Circle Road, South
Irvine, CA  92715               Address: __________________________________

                                   City: __________________________________

                        State, ZIP Code: __________________________________

                     Country if not USA: __________________________________

                Company (if applicable): __________________________________

                        Daytime Phone #: __________________________________

                      Registration date: __________________________________


                             REQUESTED DONATION

Personal, educational, or nonprofit use          $10.00 (one computer)
Nonprofit site license (any number of users)    $100.00

Commercial use                                   $50.00 (per computer)
Commercial site-license (any number of users)   $500.00

TOTAL ENCLOSED US FUNDS                   US $______.____

Please send this form with a check or money order (drawn on a USA bank),
made out to "Martin Katz, Ph.D." Sorry, no credit cards or purchase orders
accepted.  Please do not send cash through the mail.