Contents of /code/trunk/README

README file for PCRE (Perl-compatible regular expressions)
----------------------------------------------------------

The distribution should contain the following files:

  ChangeLog         log of changes to the code
  Makefile          for building PCRE
  Performance       notes on performance
  README            this file
  Tech.Notes        notes on the encoding
  pcre.3            man page for the functions
  pcreposix.3       man page for the POSIX wrapper API
  maketables.c      auxiliary program for building chartables.c
  study.c           ) source of
  pcre.c            )   the functions
  pcreposix.c       )
  pcre.h            header for the external API
  pcreposix.h       header for the external POSIX wrapper API
  internal.h        header for internal use
  pcretest.c        test program
  pgrep.1           man page for pgrep
  pgrep.c           source of a grep utility that uses PCRE
  perltest          Perl test program
  testinput         test data, compatible with Perl
  testinput2        test data for error messages and non-Perl things
  testoutput        test results corresponding to testinput
  testoutput2       test results corresponding to testinput2

To build PCRE, edit Makefile for your system (it is a fairly simple make file)
and then run it. It builds a two libraries called libpcre.a and libpcreposix.a,
a test program called pcretest, and the pgrep command.

To test PCRE, run pcretest on the file testinput, and compare the output with
the contents of testoutput. There should be no differences. For example:

  pcretest testinput /tmp/anything
  diff /tmp/anything testoutput

Do the same with testinput2, comparing the output with testoutput2, but this
time using the -i flag for pcretest, i.e.

  pcretest -i testinput2 /tmp/anything
  diff /tmp/anything testoutput2

There are two sets of tests because the first set can also be fed directly into
the perltest program to check that Perl gives the same results. The second set
of tests check pcre_info(), pcre_study(), error detection and run-time flags
that are specific to PCRE, as well as the POSIX wrapper API.

To install PCRE, copy libpcre.a to any suitable library directory (e.g.
/usr/local/lib), pcre.h to any suitable include directory (e.g.
/usr/local/include), and pcre.3 to any suitable man directory (e.g.
/usr/local/man/man3).

To install the pgrep command, copy it to any suitable binary directory, (e.g.
/usr/local/bin) and pgrep.1 to any suitable man directory (e.g.
/usr/local/man/man1).

PCRE has its own native API, but a set of "wrapper" functions that are based on
the POSIX API are also supplied in the library libpcreposix.a. Note that this
just provides a POSIX calling interface to PCRE: the regular expressions
themselves still follow Perl syntax and semantics. The header file
for the POSIX-style functions is called pcreposix.h. The official POSIX name is
regex.h, but I didn't want to risk possible problems with existing files of
that name by distributing it that way. To use it with an existing program that
uses the POSIX API it will have to be renamed or pointed at by a link.


Character tables
----------------

PCRE uses four tables for manipulating and identifying characters. These are
compiled from a source file called chartables.c. This is not supplied in
the distribution, but is built by the program maketables (compiled from
maketables.c), which uses the ANSI C character handling functions such as
isalnum(), isalpha(), isupper(), islower(), etc. to build the table sources.
This means that the default C locale set in your system may affect the contents
of the tables. You can change the tables by editing chartables.c and then
re-building PCRE. If you do this, you should probably also edit Makefile to
ensure that the file doesn't ever get re-generated.

The first two tables pcre_lcc[] and pcre_fcc[] provide lower casing and a
case flipping functions, respectively. The pcre_cbits[] table consists of four
32-byte bit maps which identify digits, letters, "word" characters, and white
space, respectively. These are used when building 32-byte bit maps that
represent character classes.

The pcre_ctypes[] table has bits indicating various character types, as
follows:

    1   white space character
    2   letter
    4   decimal digit
    8   hexadecimal digit
   16   alphanumeric or '_'
  128   regular expression metacharacter or binary zero

You should not alter the set of characters that contain the 128 bit, as that
will cause PCRE to malfunction.


The pcretest program
--------------------

This program is intended for testing PCRE, but it can also be used for
experimenting with regular expressions.

If it is given two filename arguments, it reads from the first and writes to
the second. If it is given only one filename argument, it reads from that file
and writes to stdout. Otherwise, it reads from stdin and writes to stdout, and
prompts for each line of input.

The program handles any number of sets of input on a single input file. Each
set starts with a regular expression, and continues with any number of data
lines to be matched against the pattern. An empty line signals the end of the
set. The regular expressions are given enclosed in any non-alphameric
delimiters, for example

  /(a|bc)x+yz/

and may be followed by i, m, s, or x to set the PCRE_CASELESS, PCRE_MULTILINE,
PCRE_DOTALL, or PCRE_EXTENDED options, respectively. These options have the
same effect as they do in Perl.

There are also some upper case options that do not match Perl options: /A, /E,
and /X set PCRE_ANCHORED, PCRE_DOLLAR_ENDONLY, and PCRE_EXTRA respectively.
The /D option is a PCRE debugging feature. It causes the internal form of
compiled regular expressions to be output after compilation. The /S option
causes pcre_study() to be called after the expression has been compiled, and
the results used when the expression is matched. If /I is present as well as
/S, then pcre_study() is called with the PCRE_CASELESS option.

Finally, the /P option causes pcretest to call PCRE via the POSIX wrapper API
rather than its native API. When this is done, all other options except /i and
/m are ignored. REG_ICASE is set if /i is present, and REG_NEWLINE is set if /m
is present. The wrapper functions force PCRE_DOLLAR_ENDONLY always, and
PCRE_DOTALL unless REG_NEWLINE is set.

A regular expression can extend over several lines of input; the newlines are
included in it. See the testinput file for many examples.

Before each data line is passed to pcre_exec(), leading and trailing whitespace
is removed, and it is then scanned for \ escapes. The following are recognized:

  \a     alarm (= BEL)
  \b     backspace
  \e     escape
  \f     formfeed
  \n     newline
  \r     carriage return
  \t     tab
  \v     vertical tab
  \nnn   octal character (up to 3 octal digits)
  \xhh   hexadecimal character (up to 2 hex digits)

  \A     pass the PCRE_ANCHORED option to pcre_exec()
  \B     pass the PCRE_NOTBOL option to pcre_exec()
  \E     pass the PCRE_DOLLAR_ENDONLY option to pcre_exec()
  \I     pass the PCRE_CASELESS option to pcre_exec()
  \M     pass the PCRE_MULTILINE option to pcre_exec()
  \S     pass the PCRE_DOTALL option to pcre_exec()
  \Odd   set the size of the output vector passed to pcre_exec() to dd
           (any number of decimal digits)
  \Z     pass the PCRE_NOTEOL option to pcre_exec()

A backslash followed by anything else just escapes the anything else. If the
very last character is a backslash, it is ignored. This gives a way of passing
an empty line as data, since a real empty line terminates the data input.

If /P was present on the regex, causing the POSIX wrapper API to be used, only
\B, and \Z have any effect, causing REG_NOTBOL and REG_NOTEOL to be passed to
regexec() respectively.

When a match succeeds, pcretest outputs the list of identified substrings that
pcre_exec() returns, starting with number 0 for the string that matched the
whole pattern. Here is an example of an interactive pcretest run.

  $ pcretest
  Testing Perl-Compatible Regular Expressions
  PCRE version 0.90 08-Sep-1997

      re> /^abc(\d+)/
    data> abc123
   0: abc123
   1: 123
    data> xyz
  No match

Note that while patterns can be continued over several lines (a plain ">"
prompt is used for continuations), data lines may not. However newlines can be
included in data by means of the \n escape.

If the -p option is given to pcretest, it is equivalent to adding /P to each
regular expression: the POSIX wrapper API is used to call PCRE. None of the
following flags has any effect in this case.

If the option -d is given to pcretest, it is equivalent to adding /D to each
regular expression: the internal form is output after compilation.

If the option -i (for "information") is given to pcretest, it calls pcre_info()
after compiling an expression, and outputs the information it gets back. If the
pattern is studied, the results of that are also output.

If the option -s is given to pcretest, it outputs the size of each compiled
pattern after it has been compiled.

If the -t option is given, each compile, study, and match is run 2000 times
while being timed, and the resulting time per compile or match is output in
milliseconds. Do not set -t with -s, because you will then get the size output
2000 times and the timing will be distorted.


The perltest program
--------------------

The perltest program tests Perl's regular expressions; it has the same
specification as pcretest, and so can be given identical input, except that
input patterns can be followed only by Perl's lower case options.

The data lines are processed as Perl strings, so if they contain $ or @
characters, these have to be escaped. For this reason, all such characters in
the testinput file are escaped so that it can be used for perltest as well as
for pcretest, and the special upper case options such as /A that pcretest
recognizes are not used in this file. The output should be identical, apart
from the initial identifying banner.

The testinput2 file is not suitable for feeding to Perltest, since it does
make use of the special upper case options and escapes that pcretest uses to
test additional features of PCRE.

Philip Hazel <ph10@cam.ac.uk>
October 1997
1	nigel	3	README file for PCRE (Perl-compatible regular expressions)
2			----------------------------------------------------------
3
4			The distribution should contain the following files:
5
6			ChangeLog log of changes to the code
7			Makefile for building PCRE
8			Performance notes on performance
9			README this file
10			Tech.Notes notes on the encoding
11			pcre.3 man page for the functions
12			pcreposix.3 man page for the POSIX wrapper API
13			maketables.c auxiliary program for building chartables.c
14			study.c ) source of
15			pcre.c ) the functions
16			pcreposix.c )
17			pcre.h header for the external API
18			pcreposix.h header for the external POSIX wrapper API
19			internal.h header for internal use
20			pcretest.c test program
21			pgrep.1 man page for pgrep
22			pgrep.c source of a grep utility that uses PCRE
23			perltest Perl test program
24			testinput test data, compatible with Perl
25			testinput2 test data for error messages and non-Perl things
26			testoutput test results corresponding to testinput
27			testoutput2 test results corresponding to testinput2
28
29			To build PCRE, edit Makefile for your system (it is a fairly simple make file)
30			and then run it. It builds a two libraries called libpcre.a and libpcreposix.a,
31			a test program called pcretest, and the pgrep command.
32
33			To test PCRE, run pcretest on the file testinput, and compare the output with
34			the contents of testoutput. There should be no differences. For example:
35
36			pcretest testinput /tmp/anything
37			diff /tmp/anything testoutput
38
39			Do the same with testinput2, comparing the output with testoutput2, but this
40			time using the -i flag for pcretest, i.e.
41
42			pcretest -i testinput2 /tmp/anything
43			diff /tmp/anything testoutput2
44
45			There are two sets of tests because the first set can also be fed directly into
46			the perltest program to check that Perl gives the same results. The second set
47			of tests check pcre_info(), pcre_study(), error detection and run-time flags
48			that are specific to PCRE, as well as the POSIX wrapper API.
49
50			To install PCRE, copy libpcre.a to any suitable library directory (e.g.
51			/usr/local/lib), pcre.h to any suitable include directory (e.g.
52			/usr/local/include), and pcre.3 to any suitable man directory (e.g.
53			/usr/local/man/man3).
54
55			To install the pgrep command, copy it to any suitable binary directory, (e.g.
56			/usr/local/bin) and pgrep.1 to any suitable man directory (e.g.
57			/usr/local/man/man1).
58
59			PCRE has its own native API, but a set of "wrapper" functions that are based on
60			the POSIX API are also supplied in the library libpcreposix.a. Note that this
61			just provides a POSIX calling interface to PCRE: the regular expressions
62			themselves still follow Perl syntax and semantics. The header file
63			for the POSIX-style functions is called pcreposix.h. The official POSIX name is
64			regex.h, but I didn't want to risk possible problems with existing files of
65			that name by distributing it that way. To use it with an existing program that
66			uses the POSIX API it will have to be renamed or pointed at by a link.
67
68
69			Character tables
70			----------------
71
72			PCRE uses four tables for manipulating and identifying characters. These are
73			compiled from a source file called chartables.c. This is not supplied in
74			the distribution, but is built by the program maketables (compiled from
75			maketables.c), which uses the ANSI C character handling functions such as
76			isalnum(), isalpha(), isupper(), islower(), etc. to build the table sources.
77			This means that the default C locale set in your system may affect the contents
78			of the tables. You can change the tables by editing chartables.c and then
79			re-building PCRE. If you do this, you should probably also edit Makefile to
80			ensure that the file doesn't ever get re-generated.
81
82			The first two tables pcre_lcc[] and pcre_fcc[] provide lower casing and a
83			case flipping functions, respectively. The pcre_cbits[] table consists of four
84			32-byte bit maps which identify digits, letters, "word" characters, and white
85			space, respectively. These are used when building 32-byte bit maps that
86			represent character classes.
87
88			The pcre_ctypes[] table has bits indicating various character types, as
89			follows:
90
91			1 white space character
92			2 letter
93			4 decimal digit
94			8 hexadecimal digit
95			16 alphanumeric or '_'
96			128 regular expression metacharacter or binary zero
97
98			You should not alter the set of characters that contain the 128 bit, as that
99			will cause PCRE to malfunction.
100
101
102			The pcretest program
103			--------------------
104
105			This program is intended for testing PCRE, but it can also be used for
106			experimenting with regular expressions.
107
108			If it is given two filename arguments, it reads from the first and writes to
109			the second. If it is given only one filename argument, it reads from that file
110			and writes to stdout. Otherwise, it reads from stdin and writes to stdout, and
111			prompts for each line of input.
112
113			The program handles any number of sets of input on a single input file. Each
114			set starts with a regular expression, and continues with any number of data
115			lines to be matched against the pattern. An empty line signals the end of the
116			set. The regular expressions are given enclosed in any non-alphameric
117			delimiters, for example
118
119			/(a\|bc)x+yz/
120
121			and may be followed by i, m, s, or x to set the PCRE_CASELESS, PCRE_MULTILINE,
122			PCRE_DOTALL, or PCRE_EXTENDED options, respectively. These options have the
123			same effect as they do in Perl.
124
125			There are also some upper case options that do not match Perl options: /A, /E,
126			and /X set PCRE_ANCHORED, PCRE_DOLLAR_ENDONLY, and PCRE_EXTRA respectively.
127			The /D option is a PCRE debugging feature. It causes the internal form of
128			compiled regular expressions to be output after compilation. The /S option
129			causes pcre_study() to be called after the expression has been compiled, and
130			the results used when the expression is matched. If /I is present as well as
131			/S, then pcre_study() is called with the PCRE_CASELESS option.
132
133			Finally, the /P option causes pcretest to call PCRE via the POSIX wrapper API
134			rather than its native API. When this is done, all other options except /i and
135			/m are ignored. REG_ICASE is set if /i is present, and REG_NEWLINE is set if /m
136			is present. The wrapper functions force PCRE_DOLLAR_ENDONLY always, and
137			PCRE_DOTALL unless REG_NEWLINE is set.
138
139			A regular expression can extend over several lines of input; the newlines are
140			included in it. See the testinput file for many examples.
141
142			Before each data line is passed to pcre_exec(), leading and trailing whitespace
143			is removed, and it is then scanned for \ escapes. The following are recognized:
144
145			\a alarm (= BEL)
146			\b backspace
147			\e escape
148			\f formfeed
149			\n newline
150			\r carriage return
151			\t tab
152			\v vertical tab
153			\nnn octal character (up to 3 octal digits)
154			\xhh hexadecimal character (up to 2 hex digits)
155
156			\A pass the PCRE_ANCHORED option to pcre_exec()
157			\B pass the PCRE_NOTBOL option to pcre_exec()
158			\E pass the PCRE_DOLLAR_ENDONLY option to pcre_exec()
159			\I pass the PCRE_CASELESS option to pcre_exec()
160			\M pass the PCRE_MULTILINE option to pcre_exec()
161			\S pass the PCRE_DOTALL option to pcre_exec()
162			\Odd set the size of the output vector passed to pcre_exec() to dd
163			(any number of decimal digits)
164			\Z pass the PCRE_NOTEOL option to pcre_exec()
165
166			A backslash followed by anything else just escapes the anything else. If the
167			very last character is a backslash, it is ignored. This gives a way of passing
168			an empty line as data, since a real empty line terminates the data input.
169
170			If /P was present on the regex, causing the POSIX wrapper API to be used, only
171			\B, and \Z have any effect, causing REG_NOTBOL and REG_NOTEOL to be passed to
172			regexec() respectively.
173
174			When a match succeeds, pcretest outputs the list of identified substrings that
175			pcre_exec() returns, starting with number 0 for the string that matched the
176			whole pattern. Here is an example of an interactive pcretest run.
177
178			$ pcretest
179			Testing Perl-Compatible Regular Expressions
180			PCRE version 0.90 08-Sep-1997
181
182			re> /^abc(\d+)/
183			data> abc123
184			0: abc123
185			1: 123
186			data> xyz
187			No match
188
189			Note that while patterns can be continued over several lines (a plain ">"
190			prompt is used for continuations), data lines may not. However newlines can be
191			included in data by means of the \n escape.
192
193			If the -p option is given to pcretest, it is equivalent to adding /P to each
194			regular expression: the POSIX wrapper API is used to call PCRE. None of the
195			following flags has any effect in this case.
196
197			If the option -d is given to pcretest, it is equivalent to adding /D to each
198			regular expression: the internal form is output after compilation.
199
200			If the option -i (for "information") is given to pcretest, it calls pcre_info()
201			after compiling an expression, and outputs the information it gets back. If the
202			pattern is studied, the results of that are also output.
203
204			If the option -s is given to pcretest, it outputs the size of each compiled
205			pattern after it has been compiled.
206
207			If the -t option is given, each compile, study, and match is run 2000 times
208			while being timed, and the resulting time per compile or match is output in
209			milliseconds. Do not set -t with -s, because you will then get the size output
210			2000 times and the timing will be distorted.
211
212
213
214			The perltest program
215			--------------------
216
217			The perltest program tests Perl's regular expressions; it has the same
218			specification as pcretest, and so can be given identical input, except that
219			input patterns can be followed only by Perl's lower case options.
220
221			The data lines are processed as Perl strings, so if they contain $ or @
222			characters, these have to be escaped. For this reason, all such characters in
223			the testinput file are escaped so that it can be used for perltest as well as
224			for pcretest, and the special upper case options such as /A that pcretest
225			recognizes are not used in this file. The output should be identical, apart
226			from the initial identifying banner.
227
228			The testinput2 file is not suitable for feeding to Perltest, since it does
229			make use of the special upper case options and escapes that pcretest uses to
230			test additional features of PCRE.
231
232			Philip Hazel <ph10@cam.ac.uk>
233			October 1997