KEYWORD SEARCH v1.44 - Search for keywords
within text & binary files. Grep, boolean,
prefix, suffix, whole word, case sensitive &
fuzzy searches, onscreen highlighting, line
numbers, totals, @filelists, subdirectory
scans, can read from stdin & hidden/system
files, accepts 4DOS 256-char command lines,
reads/converts Unix & Mac text files, more.

KWS V1.44 Copyright (c) 1991-1994 Copy Con Incorporated Page 1


DESCRIPTION
-----------
KWS (KeyWord Search) is a small, fast program that searches for key-
words within text and binary files. KWS has many features and options:

* Handles grep, boolean, prefix, suffix, whole word, case-sensitive,
and adjustable fuzzy searches.

* Supports wildcards and @filelists, and can span subdirectories.

* Can display matches in bold onscreen, and can also display line
numbers, file offsets, and totals.

* Can read data from standard input, from hidden/system files, and
from text files created on other platforms.

* Accepts input lines up to 2k in length, and can handle 4DOS/NDOS
256-character command lines.

* Much more...


REQUIREMENTS
------------
KWS will run on any PC compatible using DOS 2.0 or higher.


DISCLAIMER
----------
KWS is distributed on an "AS IS" basis without warranty, expressed or
implied. Considerable testing effort has been expended, but the user
is advised to check the program's suitability before relying on it.
The user assumes full risk as to the results of using this program. In
no event shall the author be liable for any incidental or consequent-
ial damages arising from the use, or inability to use this program.


REGISTRATION
------------
The unregistered version of KWS is freely distributable and fully
functional, and may be used for a 30-day trial period. If you plan to
continue using KWS, you are obligated to pay the modest registration
fee. Registered users will receive a serialized, nag-free version of
KWS, a free upgrade when available, and $5 upgrades thereafter.

All registration information is contained in the KWS.REG file. If that
file is missing, register by sending your name and address (please
print), along with $20 (cash, check, or money order in US funds) to:

Copy Con Incorporated
PO Box 23255
Baltimore MD 21203-5255

Please specify 3.5" or 5.25" diskette. MD residents add 5% sales tax.
Thank you for supporting Shareware.


KWS V1.44 Copyright (c) 1991-1994 Copy Con Incorporated Page 2


OPERATION
---------
KWS is a powerful, flexible program; it is also very easy to use. KWS
is activated from the command line, using the format

KWS filename search_string[...] [options]

The filename portion of the command line consists of one or more valid
filespecs to be searched. Wildcards are optional. (For a description
of wildcards and filespecs, see your DOS manual.) If more than one
filespec is used, filespecs must be separated by plus (+) characters,
as in "file1.txt+file2.txt".

KWS also has the ability to read filespecs from a file list. For
example, if the filename is in the format "@filelist.txt" then the
filespecs will be read from the file "filelist.txt". File lists
consist of one filespec per line, each optionally followed by a
semicolon and a comment. Each filespec in the file list is searched
relative to the current directory.

Search strings are comprised of keywords separated by semicolons. At
least one keyword from each search string must be present in a given
line for it to be displayed; for example,

KWS allfiles.lst fixed;hard disk;drive

will display the lines in file ALLFILES.LST that contain either words
FIXED or HARD (or both), and also contain either words DISK or DRIVE.

If a search string contains a keyword with embedded spaces or with a
redirection character (>,<, or |), the entire search string must be
enclosed in quotes; for example,

KWS temp.bat "del *;era *"

will search TEMP.BAT for phrases "del *" or "era *" (without quotes).

A logical 'NOT' search can be conducted by preceding a keyword with
an exclamation (!) character; for example,

KWS stuff.txt this !that

will display the lines in STUFF.TXT that contain the word THIS and not
the word THAT. If you wish to search for an exclamation as the first
(or only) character of a keyword, use two in a row; for example,

KWS stuff.txt this !!that

will display lines that contain both words THIS and !THAT. Exclamation
characters inside or following keywords (as in "HEY!" or "What!?") do
not have to be doubled.


KWS V1.44 Copyright (c) 1991-1994 Copy Con Incorporated Page 3


OPERATION (continued)
---------------------
A fuzzy search can be conducted by preceding a keyword with a tilde
(~) character; lines with words that are similar to the keyword will
be displayed. This option is useful if you aren't sure of a keyword's
correct spelling, or if a keyword may be spelled incorrectly in the
files to be searched; for example,

KWS gnu.txt ~copyrite

will display lines with words similar to COPYRITE. (The similarity
percentage can be adjusted with the /F option.) If you wish to combine
a fuzzy search with a logical 'NOT' search, use the 'NOT' character
before the tilde (as in "!~THE"). Fuzzy searches cannot be combined
with prefix/suffix/whole word searches (described below).

Whole word searches can be conducted by putting underscore (_)
characters before and after the keyword; for example,

KWS files.lst _c_ source

will display lines in FILES.LST that contain a lone 'C' and the word
SOURCE. Similarly, prefixes can be searched for by putting an
underscore before a keyword, and suffixes can be searched for by
putting an underscore after a keyword. If you wish to search for an
underscore as the first or last (or only) character of a keyword, use
two underscores in a row, as with the logical 'NOT' exclamation
character above. If you wish to combine a prefix/whole word search
with a logical 'NOT' search, use the 'NOT' character before the
underscore (as in "!_the_"). Prefix/suffix/whole word searches cannot
be combined with fuzzy searches (described above).

Up to 10 search strings are allowed, with up to 10 keywords per
string.


OPTIONS
-------
Options can appear anywhere on the command line after the filespec.
They can be in any order, and are not case-sensitive. Options can be
separated by spaces (/L3 /P /T) or strung together (/L3/P/T or /L3PT).

Option Description

/A When this option is used, all binary files will automatically
be shown with hexadecimal file offsets.

/B,B0 These options make all matches appear in bold type onscreen.
/B uses direct screen writes; /B0 uses BIOS screen writes,
which are slower, but compatible with more video systems.
When using either option, output cannot be routed using DOS
redirection; use the /O option to redirect output.


KWS V1.44 Copyright (c) 1991-1994 Copy Con Incorporated Page 4


OPTIONS (continued)
-------------------
/C Search will be case-sensitive. Case-sensitive searches are
approximately 20% faster than case-insensitive searches.

/D Control characters (ASCII value less than 32) in text files
will be replaced with spaces before being viewed or routed.

/E The command line will be echoed as the first output line.
Useful mainly with redirected files to avoid repetition, and
for testing batch files.

/Fn This option controls the similarity percentage for fuzzy
searches. Similarity defaults to 65 percent, and can be
adjusted from 0 (display almost every line) to 100 (display
exact matches).

/G This option puts KWS into regular expression mode, where
certain characters within keywords have special meaning:

* An asterisk matches zero or more characters; for example,
E*T matches ET, EAT, or EDIT.

? A question mark matches any one character; for example,
E?T matches EAT, EDT, or EST.

^ A circumflex at the start of a keyword matches the start
of a line.

$ A dollar sign at the end of a keyword matches the end of a
line.

[] A string enclosed in brackets matches any character in
that string, but no others. If the first character in the
string is a circumflex (^), the expression matches any
character except the characters in the string. For
example, [XYZ] matches X, Y, or Z, while [^XYZ] matches
any character except X, Y, or Z. A range of characters can
be specified with two characters separated by a hyphen
(-); these can be combined to form expressions like
[A-BD-Z?], which matches the ? character and any letter
except C.

\ A backslash tells KWS to search for the literal character
that follows it; for example, \* searches for an asterisk
instead of searching for zero or more characters.

/H Hidden and/or system files will also be searched.

/I Ignore the config file, if present (see CONFIGURATION FILE).
If stringing options together, this option must be the first
in the string; for example, use "/ip" instead of "/pi".


KWS V1.44 Copyright (c) 1991-1994 Copy Con Incorporated Page 5


OPTIONS (continued)
-------------------
/J[n] Use the viewer defined in the config file to display matches
(see CONFIGURATION FILE). The number, if specified, is the
number of lines before the match to start viewing from.

/Kc Change the keyword separator character from a semicolon to
the specified character. This option may be needed to pass
multiple keyword parameters to a batch file (as some versions
of DOS remove semicolons from batch parameters), or if a
semicolon is part of a keyword to be searched.

/L[n] Line numbers or file offset position will be displayed with
output. If the specified width begins with a zero, then the
file offset will be displayed; otherwise, line numbers will
be displayed. If no width is specified, the numbers will be
left-justified, with a colon between the number and the text;
otherwise, the numbers will be right-justified and zero-
padded to the left, followed by the colon and text. If the
specified width is too small for a number to fit, the number
will NOT be truncated.

/M[n] Stop searching after n matches per file. /M alone will cause
KWS to print only the filenames of successful searches, and
these can be redirected to a file to be used as a file list.

/N Skip over binary files when searching multiple files.

/Ofn Route output to the specified device/filename. Output will be
in addition to, not instead of, output displayed onscreen. To
route output without viewing, use DOS redirection instead of
this option (see your DOS manual for details). Do not append
other options on to the end of this option, as they will be
mistaken as part of the device/filename.

/P[n] Pause between each screen of text. If no height is specified,
KWS will autodetect your screen size, and pause accordingly;
otherwise, KWS will pause after every n lines. While paused,
pressing the 'C' key will display the current and any
remaining files continuously. Pressing the Escape key while
paused will skip the current file and go on to the next.

/R Instead of reading from text files, KWS will read from
standard input. When in this mode, the /V option will be
inactive. Do not specify a filename when using this option.

/S Subdirectories will also be searched.

/T,T0 Display total lines found per file. /T will display lines and
totals; /T0 will display totals only. For wildcard and file
list searches, a grand total will also be displayed.

/U Unix output format: All lines of output are preceded by the
name of the file that contained the matching line.


KWS V1.44 Copyright (c) 1991-1994 Copy Con Incorporated Page 6


OPTIONS (continued)
-------------------
/V If KWS cannot determine if a file is binary or text, text is
assumed; using this option will tell KWS to treat all files
as binary. This option is inactive when using the /R option.

/Wn Normally, all keywords must be on the same line for the line
to be displayed; use this option to look at more than one
line at a time. If all keywords are found within the
specified window size (maximum of 32767 lines), then the
lines will be displayed, starting with the first line to
contain a keyword, and ending with the last line to contain a
keyword. This option is inactive when searching binary files.

/X Line numbers and offsets will be displayed in hexadecimal.

/Z Verbose mode: KWS prints the filename of every file searched,
each matching line is preceded by its line number, and a
count of matching lines in each file is given, even if the
count is zero.


CONFIGURATION FILE
------------------
Under DOS 3.0 or higher, a default configuration for KWS can be stored
in a text file. If a file named KWS.CFG is in the same directory as
KWS.EXE, KWS will read default configuration values from the CFG file.
The config file can be ignored by using the /I option.

Each line of the file consists of a configuration item, an equal sign,
and the item's value.

Item Description

OPTIONS For specifying default command line option values. As with
command line options, case is not important, and the options
may be strung together. Any options enabled in the config
file can be disabled on the command line by specifying the
option with a minus sign appended. For example, with a config
file that reads

options=/b /p

bold keywords and screen pausing will be in effect whenever
KWS is run. To disable bold keywords, "/b-" can be specified
on the command line.

VIEWER Determines what program is run when the /J option is used.
The value specified is the command line to execute. If "!F"
or "!f" is encountered on the line, it will be replaced with
the current file name before executing, and "!L" or "!l" will
be replaced with the current line number. For example, if a
match is found on line 30 of FILES.LST, and

viewer=c:\utils\list.com !f /#:!l /l-


KWS V1.44 Copyright (c) 1991-1994 Copy Con Incorporated Page 7


CONFIGURATION FILE (continued)
------------------------------
is in your config file, KWS will execute:

C:\UTILS\LIST.COM FILES.LST /#:30 /l-


EXAMPLES
--------
KWS catalog.lst dvorak keyboard /p /u
will take the lines in file CATALOG.LST that contain the words
DVORAK and KEYBOARD, and display them one screen at a time (P), in
UNIX output format (U).

KWS catalog.lst arc;lzh shell /l5 /oprn
will take the lines in file CATALOG.LST that contain either ARC or
LZH, and also contain SHELL, and display them onscreen with line
numbers (L5), while sending them to the printer (OPRN).

KWS allfiles.lst /c !.ZIP >nozip.lst
will make a file called NOZIP.LST, which will contain all of the
lines in ALLFILES.LST that do NOT contain the word .ZIP in caps (C).
Output will not be seen onscreen.

KWS \dl\*.lst MIDI /ptc
will search all .LST files in the DL directory for the word MIDI in
capital letters (C), and displays the filenames and lines onscreen,
one screen at a time (P), with totals (T) for each file.

KWS *.bak _the_
will search the backup files for lines that contain the word THE,
but will not display them if THE is part of another word (as in
THESIS or OTHER).

KWS temp.bak _the_; /b
will display every line in TEMP.BAK, highlighting (B) any occurrence
of the word THE.

KWS c:\*.bat _del;_era /s/h
will search all (S) of the batch files on the C: drive, including
those with their hidden and/or system attributes set (H), to see if
any of them delete/erase files.

dir | KWS -92 /r
pipes the output from the DIR command into KWS (R), which will
display the directory entries from 1992.

KWS ; /r /l3 two.txt
creates file TWO.TXT which is a duplicate of file ONE.TXT, except
that TWO.TXT will have 3-digit line numbers (L3). (A search for the
keyword separator will find every line.) The file ONE.TXT is read
from the standard input device (R).


KWS V1.44 Copyright (c) 1991-1994 Copy Con Incorporated Page 8


EXAMPLES (continued)
--------------------
KWS unix.man ; >unix.doc

converts a Unix format text file (LF delimited) into a DOS format
text file (CR/LF delimited). KWS can also read CR delimited files.

KWS *.* ~copyrite /nf70b
will search the text files (N) in the current directory for words
similar to COPYRITE, highlighting (B) any close calls (F70).

KWS *.* ~copyrite !1993 /nf70b
will display the same lines as the above example, except for lines
containing the word 1993.

KWS allfiles.lst _c_;assem;asm src;source /pb
will display all lines in ALLFILES.LST that have to do with C or
assembler source code. Keywords will be in bold (B), and output will
be paused after each screen (P).

KWS allfiles.lst _c_;assem;asm src;source !c++ /pb
will display the same as above, except C++ files will not be shown.

KWS \*.com+\*.exe ; /hal6p /od:com.txt
examines all (H) .COM and .EXE files in the root directory, using
hex (A) file offsets 6 characters wide (L6), pausing at each screen
(P), while also sending output to file COM.TXT on the D: drive
(OD:COM.TXT).

KWS telix.use download file /w2
searches a Telix usage log, looking at two lines at a time (W2), to
see what files were downloaded.

KWS kws.doc ; /pdolpt2
displays this file, one screen at a time (P), removing embedded
formfeed characters (D), while also sending output to printer LPT2
(OLPT2).

KWS c:\*.bat ; /sm >filelist.txt
will create a file named FILELIST.TXT containing the names of all
(S) of the batch files on the C: drive. This file is used in the
next example.

KWS @filelist.txt label "if %;goto"
will search the filenames in FILELIST.TXT, and display lines in
those files that contain the word LABEL and either the word GOTO or
the phrase "IF %" (without the quotes). Note that this example has
only two search strings, not three; if any words in a search string
contain spaces, then the entire search string must be in quotes.

KWS @filelist.txt label "if %;goto" /j3
will conduct the same search as above, but the program specified in
the configuration file under VIEWER will be used to show each match.
Viewing will start 3 lines before the actual match.


KWS V1.44 Copyright (c) 1991-1994 Copy Con Incorporated Page 9


RETURN CODES
------------
The DOS errorlevel will be set if a search is unsuccessful:

Level Description

1 Invalid command line options

2 None of the files contain the search string(s)

3 Ctrl-Break interrupt

4 Invalid directory specified

5 No files match filespec

6 Unable to open output file


SUPPORT
-------
In the event of a major upgrade, notice will be sent to registered
users, detailing changes since their last upgrade. Minor evaluation
upgrades will be released through BBSes, distributors, user groups, et
cetera. Registered users will receive a free upgrade upon request, and
subsequent upgrades can be obtained for $5 shipping.

If you have questions, comments, suggestions, bug reports, et cetera,
feel free to contact me (RD Thomas) by any of the following methods:

Internet: [email protected]
RIME: Shareware conference (->837 or ->1520)
FIDO: Shareware conference
BBSes: GreyHawk (301-596-7064 or 410-290-3260 or 410-720-5083)
The Programmer's Corner (301-596-7692 or 410-995-6873)
US Mail: PO Box 23255, Baltimore MD 21203-5255

Looking forward to hearing from you.


To print this document, type "COPY KWS.DOC PRN" at the DOS prompt.