The Use of UNIX Regular Expressions

Regular Expressions are a standard function of the UNIX operating system.
They are used to search for or replace text located within files or from the
keyboard. The UNIX commands GREP (Global Regular Expression Print) and
SUBS (SUBstitue Strings) are used to manipulate text via regular expressions.

There are various characters that the reuglar expression processor sees as
having special meaning. These special characters are:

.       period      matches any alpha-numeric character

^       caret       matches the beginning of the line
                          grep ^#include
                    will find all line beginning with #include

$       dollar      matches the end of the line
                          grep }$
                    will find all lines ending with a close brace

*       asterisk    Repeats any number of the last pattern. Includes
                    zero repetitions.
                          grep  test*
                    finds lines with at least one occurence of test.

[]      brakets     specifies a set of given characters.
                          grep ^[abcd0123]
                    will find lines begining with any one of the
                    specified characters.

-       hyphen      denotes a range of characters.
                          grep [A-Z0-9]$
                    finds lines ending in a capital letter or a number.

\       backslash   removes the special characteristics of a character.
                          grep \$
                    finds dollar signs in a line of text.

NOTE:   When the caret (^) is used as the first character after a left
        square braket, it reverses the meaning of the search.
                grep [^A-Z]&
        will find all lines not ending with a capital letter.
                grep ^[^A-Z]
        will find all lines not begining with a capital letter. Note that
        in this case the caret is used in both contexts.

Pattern Segments

One of the powerful functions of the SUBS command is the ability to
rearrange the internal structure of text lines. This is done by the use of
pattern segments. While searching through the lines of text, certain patterns
may be given identification numbers. The command can then selectively delete
or rearrange the numbered segments.

\(      denotes the beginning of a segment.
\)      denotes the end of a segment.
\#\     denotes the segement in the second part of the SUBS command.

The best way to understand this usage is by example.

Say that you need to prepare a list of all 'C' source code but you want
to replace the '.C' extent with '.OBJ' the following SUBS command will
do just that.

        SUBS "\(.*\)\.c" "\1\.OBJ"

In this command the use of the '.*' within the first segment is a powerful
construction. It is the same as the DOS '*' wild card. The period followed by
the asterisk means repeat any character any number of times. The name of the
file is the first part of the search. The \1\ calls for the replacement of
only the file name followed by the .OBJ .