trunk/doc/pcrecompat.3

.TH PCRECOMPAT 3 "19 March 2013" "PCRE 8.33"
.SH NAME
PCRE - Perl-compatible regular expressions
.SH "DIFFERENCES BETWEEN PCRE AND PERL"
.rs
.sp
This document describes the differences in the ways that PCRE and Perl handle
regular expressions. The differences described here are with respect to Perl
versions 5.10 and above.
.P
1. PCRE has only a subset of Perl's Unicode support. Details of what it does
have are given in the
.\" HREF
\fBpcreunicode\fP
.\"
page.
.P
2. PCRE allows repeat quantifiers only on parenthesized assertions, but they do
not mean what you might think. For example, (?!a){3} does not assert that the
next three characters are not "a". It just asserts that the next character is
not "a" three times (in principle: PCRE optimizes this to run the assertion
just once). Perl allows repeat quantifiers on other assertions such as \eb, but
these do not seem to have any use.
.P
3. Capturing subpatterns that occur inside negative lookahead assertions are
counted, but their entries in the offsets vector are never set. Perl sometimes
(but not always) sets its numerical variables from inside negative assertions.
.P
4. Though binary zero characters are supported in the subject string, they are
not allowed in a pattern string because it is passed as a normal C string,
terminated by zero. The escape sequence \e0 can be used in the pattern to
represent a binary zero.
.P
5. The following Perl escape sequences are not supported: \el, \eu, \eL,
\eU, and \eN when followed by a character name or Unicode value. (\eN on its
own, matching a non-newline character, is supported.) In fact these are
implemented by Perl's general string-handling and are not part of its pattern
matching engine. If any of these are encountered by PCRE, an error is
generated by default. However, if the PCRE_JAVASCRIPT_COMPAT option is set,
\eU and \eu are interpreted as JavaScript interprets them.
.P
6. The Perl escape sequences \ep, \eP, and \eX are supported only if PCRE is
built with Unicode character property support. The properties that can be
tested with \ep and \eP are limited to the general category properties such as
Lu and Nd, script names such as Greek or Han, and the derived properties Any
and L&. PCRE does support the Cs (surrogate) property, which Perl does not; the
Perl documentation says "Because Perl hides the need for the user to understand
the internal representation of Unicode characters, there is no need to
implement the somewhat messy concept of surrogates."
.P
7. PCRE does support the \eQ...\eE escape for quoting substrings. Characters in
between are treated as literals. This is slightly different from Perl in that $
and @ are also handled as literals inside the quotes. In Perl, they cause
variable interpolation (but of course PCRE does not have variables). Note the
following examples:
.sp
    Pattern            PCRE matches      Perl matches
.sp
.\" JOIN
    \eQabc$xyz\eE        abc$xyz           abc followed by the
                                           contents of $xyz
    \eQabc\e$xyz\eE       abc\e$xyz          abc\e$xyz
    \eQabc\eE\e$\eQxyz\eE   abc$xyz           abc$xyz
.sp
The \eQ...\eE sequence is recognized both inside and outside character classes.
.P
8. Fairly obviously, PCRE does not support the (?{code}) and (??{code})
constructions. However, there is support for recursive patterns. This is not
available in Perl 5.8, but it is in Perl 5.10. Also, the PCRE "callout"
feature allows an external function to be called during pattern matching. See
the
.\" HREF
\fBpcrecallout\fP
.\"
documentation for details.
.P
9. Subpatterns that are called as subroutines (whether or not recursively) are
always treated as atomic groups in PCRE. This is like Python, but unlike Perl.
Captured values that are set outside a subroutine call can be reference from
inside in PCRE, but not in Perl. There is a discussion that explains these
differences in more detail in the
.\" HTML <a href="pcrepattern.html#recursiondifference">
.\" </a>
section on recursion differences from Perl
.\"
in the
.\" HREF
\fBpcrepattern\fP
.\"
page.
.P
10. If any of the backtracking control verbs are used in a subpattern that is
called as a subroutine (whether or not recursively), their effect is confined
to that subpattern; it does not extend to the surrounding pattern. This is not
always the case in Perl. In particular, if (*THEN) is present in a group that
is called as a subroutine, its action is limited to that group, even if the
group does not contain any | characters. Note that such subpatterns are
processed as anchored at the point where they are tested.
.P
11. If a pattern contains more than one backtracking control verb, the first
one that is backtracked onto acts. For example, in the pattern
A(*COMMIT)B(*PRUNE)C a failure in B triggers (*COMMIT), but a failure in C
triggers (*PRUNE). Perl's behaviour is more complex; in many cases it is the
same as PCRE, but there are examples where it differs.
.P
12. Most backtracking verbs in assertions have their normal actions. They are 
not confined to the assertion.
.P
13. There are some differences that are concerned with the settings of captured
strings when part of a pattern is repeated. For example, matching "aba" against
the pattern /^(a(b)?)+$/ in Perl leaves $2 unset, but in PCRE it is set to "b".
.P
14. PCRE's handling of duplicate subpattern numbers and duplicate subpattern
names is not as general as Perl's. This is a consequence of the fact the PCRE
works internally just with numbers, using an external table to translate
between numbers and names. In particular, a pattern such as (?|(?<a>A)|(?<b)B),
where the two capturing parentheses have the same number but different names,
is not supported, and causes an error at compile time. If it were allowed, it
would not be possible to distinguish which parentheses matched, because both
names map to capturing subpattern number 1. To avoid this confusing situation,
an error is given at compile time.
.P
15. Perl recognizes comments in some places that PCRE does not, for example,
between the ( and ? at the start of a subpattern. If the /x modifier is set,
Perl allows white space between ( and ? but PCRE never does, even if the
PCRE_EXTENDED option is set.
.P
16. In PCRE, the upper/lower case character properties Lu and Ll are not 
affected when case-independent matching is specified. For example, \ep{Lu}
always matches an upper case letter. I think Perl has changed in this respect;
in the release at the time of writing (5.16), \ep{Lu} and \ep{Ll} match all
letters, regardless of case, when case independence is specified.
.P
17. PCRE provides some extensions to the Perl regular expression facilities.
Perl 5.10 includes new features that are not in earlier versions of Perl, some
of which (such as named parentheses) have been in PCRE for some time. This list
is with respect to Perl 5.10:
.sp
(a) Although lookbehind assertions in PCRE must match fixed length strings,
each alternative branch of a lookbehind assertion can match a different length
of string. Perl requires them all to have the same length.
.sp
(b) If PCRE_DOLLAR_ENDONLY is set and PCRE_MULTILINE is not set, the $
meta-character matches only at the very end of the string.
.sp
(c) If PCRE_EXTRA is set, a backslash followed by a letter with no special
meaning is faulted. Otherwise, like Perl, the backslash is quietly ignored.
(Perl can be made to issue a warning.)
.sp
(d) If PCRE_UNGREEDY is set, the greediness of the repetition quantifiers is
inverted, that is, by default they are not greedy, but if followed by a
question mark they are.
.sp
(e) PCRE_ANCHORED can be used at matching time to force a pattern to be tried
only at the first matching position in the subject string.
.sp
(f) The PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART, and
PCRE_NO_AUTO_CAPTURE options for \fBpcre_exec()\fP have no Perl equivalents.
.sp
(g) The \eR escape sequence can be restricted to match only CR, LF, or CRLF
by the PCRE_BSR_ANYCRLF option.
.sp
(h) The callout facility is PCRE-specific.
.sp
(i) The partial matching facility is PCRE-specific.
.sp
(j) Patterns compiled by PCRE can be saved and re-used at a later time, even on
different hosts that have the other endianness. However, this does not apply to
optimized data created by the just-in-time compiler.
.sp
(k) The alternative matching functions (\fBpcre_dfa_exec()\fP,
\fBpcre16_dfa_exec()\fP and \fBpcre32_dfa_exec()\fP,) match in a different way
and are not Perl-compatible.
.sp
(l) PCRE recognizes some special sequences such as (*CR) at the start of
a pattern that set overall options that cannot be changed within the pattern.
.
.
.SH AUTHOR
.rs
.sp
.nf
Philip Hazel
University Computing Service
Cambridge CB2 3QH, England.
.fi
.
.
.SH REVISION
.rs
.sp
.nf
Last updated: 19 March 2013
Copyright (c) 1997-2013 University of Cambridge.
.fi
1	ph10	1297	.TH PCRECOMPAT 3 "19 March 2013" "PCRE 8.33"
2	nigel	63	.SH NAME
3			PCRE - Perl-compatible regular expressions
4	nigel	75	.SH "DIFFERENCES BETWEEN PCRE AND PERL"
5	nigel	63	.rs
6			.sp
7			This document describes the differences in the ways that PCRE and Perl handle
8	ph10	456	regular expressions. The differences described here are with respect to Perl
9	ph10	562	versions 5.10 and above.
10	nigel	75	.P
11	ph10	858	1. PCRE has only a subset of Perl's Unicode support. Details of what it does
12			have are given in the
13	nigel	73	.\" HREF
14	ph10	678	\fBpcreunicode\fP
15	nigel	73	.\"
16			page.
17	nigel	75	.P
18	ph10	637	2. PCRE allows repeat quantifiers only on parenthesized assertions, but they do
19			not mean what you might think. For example, (?!a){3} does not assert that the
20			next three characters are not "a". It just asserts that the next character is
21	ph10	654	not "a" three times (in principle: PCRE optimizes this to run the assertion
22			just once). Perl allows repeat quantifiers on other assertions such as \eb, but
23	ph10	637	these do not seem to have any use.
24	nigel	75	.P
25	nigel	73	3. Capturing subpatterns that occur inside negative lookahead assertions are
26	ph10	1292	counted, but their entries in the offsets vector are never set. Perl sometimes
27			(but not always) sets its numerical variables from inside negative assertions.
28	nigel	75	.P
29	nigel	73	4. Though binary zero characters are supported in the subject string, they are
30	nigel	63	not allowed in a pattern string because it is passed as a normal C string,
31	nigel	75	terminated by zero. The escape sequence \e0 can be used in the pattern to
32	nigel	63	represent a binary zero.
33	nigel	75	.P
34			5. The following Perl escape sequences are not supported: \el, \eu, \eL,
35	ph10	654	\eU, and \eN when followed by a character name or Unicode value. (\eN on its
36	ph10	634	own, matching a non-newline character, is supported.) In fact these are
37			implemented by Perl's general string-handling and are not part of its pattern
38			matching engine. If any of these are encountered by PCRE, an error is
39	ph10	836	generated by default. However, if the PCRE_JAVASCRIPT_COMPAT option is set,
40	ph10	745	\eU and \eu are interpreted as JavaScript interprets them.
41	nigel	75	.P
42			6. The Perl escape sequences \ep, \eP, and \eX are supported only if PCRE is
43			built with Unicode character property support. The properties that can be
44			tested with \ep and \eP are limited to the general category properties such as
45	nigel	87	Lu and Nd, script names such as Greek or Han, and the derived properties Any
46	ph10	450	and L&. PCRE does support the Cs (surrogate) property, which Perl does not; the
47			Perl documentation says "Because Perl hides the need for the user to understand
48			the internal representation of Unicode characters, there is no need to
49			implement the somewhat messy concept of surrogates."
50	nigel	75	.P
51	ph10	1011	7. PCRE does support the \eQ...\eE escape for quoting substrings. Characters in
52	nigel	63	between are treated as literals. This is slightly different from Perl in that $
53			and @ are also handled as literals inside the quotes. In Perl, they cause
54			variable interpolation (but of course PCRE does not have variables). Note the
55			following examples:
56	nigel	75	.sp
57	nigel	63	Pattern PCRE matches Perl matches
58	nigel	75	.sp
59			.\" JOIN
60			\eQabc$xyz\eE abc$xyz abc followed by the
61	nigel	63	contents of $xyz
62	nigel	75	\eQabc\e$xyz\eE abc\e$xyz abc\e$xyz
63			\eQabc\eE\e$\eQxyz\eE abc$xyz abc$xyz
64			.sp
65			The \eQ...\eE sequence is recognized both inside and outside character classes.
66			.P
67	ph10	1011	8. Fairly obviously, PCRE does not support the (?{code}) and (??{code})
68	nigel	93	constructions. However, there is support for recursive patterns. This is not
69	ph10	453	available in Perl 5.8, but it is in Perl 5.10. Also, the PCRE "callout"
70	nigel	93	feature allows an external function to be called during pattern matching. See
71			the
72	nigel	75	.\" HREF
73			\fBpcrecallout\fP
74			.\"
75			documentation for details.
76			.P
77	ph10	1011	9. Subpatterns that are called as subroutines (whether or not recursively) are
78	ph10	716	always treated as atomic groups in PCRE. This is like Python, but unlike Perl.
79	ph10	724	Captured values that are set outside a subroutine call can be reference from
80			inside in PCRE, but not in Perl. There is a discussion that explains these
81			differences in more detail in the
82	ph10	453	.\" HTML <a href="pcrepattern.html#recursiondifference">
83			.\" </a>
84			section on recursion differences from Perl
85			.\"
86			in the
87			.\" HREF
88	ph10	456	\fBpcrepattern\fP
89	ph10	453	.\"
90			page.
91	nigel	93	.P
92	ph10	1297	10. If any of the backtracking control verbs are used in a subpattern that is
93			called as a subroutine (whether or not recursively), their effect is confined
94			to that subpattern; it does not extend to the surrounding pattern. This is not
95			always the case in Perl. In particular, if (*THEN) is present in a group that
96			is called as a subroutine, its action is limited to that group, even if the
97			group does not contain any \| characters. Note that such subpatterns are
98			processed as anchored at the point where they are tested.
99	ph10	716	.P
100	ph10	1293	11. If a pattern contains more than one backtracking control verb, the first
101			one that is backtracked onto acts. For example, in the pattern
102			A(COMMIT)B(PRUNE)C a failure in B triggers (*COMMIT), but a failure in C
103			triggers (*PRUNE). Perl's behaviour is more complex; in many cases it is the
104			same as PCRE, but there are examples where it differs.
105			.P
106	ph10	1297	12. Most backtracking verbs in assertions have their normal actions. They are
107			not confined to the assertion.
108			.P
109			13. There are some differences that are concerned with the settings of captured
110	nigel	63	strings when part of a pattern is repeated. For example, matching "aba" against
111			the pattern /^(a(b)?)+$/ in Perl leaves $2 unset, but in PCRE it is set to "b".
112	nigel	75	.P
113	ph10	1297	14. PCRE's handling of duplicate subpattern numbers and duplicate subpattern
114	ph10	459	names is not as general as Perl's. This is a consequence of the fact the PCRE
115			works internally just with numbers, using an external table to translate
116			between numbers and names. In particular, a pattern such as (?\|(?<a>A)\|(?<b)B),
117			where the two capturing parentheses have the same number but different names,
118			is not supported, and causes an error at compile time. If it were allowed, it
119			would not be possible to distinguish which parentheses matched, because both
120			names map to capturing subpattern number 1. To avoid this confusing situation,
121			an error is given at compile time.
122	ph10	456	.P
123	ph10	1297	15. Perl recognizes comments in some places that PCRE does not, for example,
124	ph10	654	between the ( and ? at the start of a subpattern. If the /x modifier is set,
125	ph10	968	Perl allows white space between ( and ? but PCRE never does, even if the
126	ph10	635	PCRE_EXTENDED option is set.
127	ph10	562	.P
128	ph10	1297	16. In PCRE, the upper/lower case character properties Lu and Ll are not
129	ph10	1258	affected when case-independent matching is specified. For example, \ep{Lu}
130			always matches an upper case letter. I think Perl has changed in this respect;
131			in the release at the time of writing (5.16), \ep{Lu} and \ep{Ll} match all
132			letters, regardless of case, when case independence is specified.
133			.P
134	ph10	1297	17. PCRE provides some extensions to the Perl regular expression facilities.
135	ph10	456	Perl 5.10 includes new features that are not in earlier versions of Perl, some
136			of which (such as named parentheses) have been in PCRE for some time. This list
137			is with respect to Perl 5.10:
138			.sp
139			(a) Although lookbehind assertions in PCRE must match fixed length strings,
140			each alternative branch of a lookbehind assertion can match a different length
141			of string. Perl requires them all to have the same length.
142			.sp
143	nigel	63	(b) If PCRE_DOLLAR_ENDONLY is set and PCRE_MULTILINE is not set, the $
144			meta-character matches only at the very end of the string.
145	nigel	75	.sp
146	nigel	63	(c) If PCRE_EXTRA is set, a backslash followed by a letter with no special
147	ph10	181	meaning is faulted. Otherwise, like Perl, the backslash is quietly ignored.
148			(Perl can be made to issue a warning.)
149	nigel	75	.sp
150	nigel	63	(d) If PCRE_UNGREEDY is set, the greediness of the repetition quantifiers is
151			inverted, that is, by default they are not greedy, but if followed by a
152			question mark they are.
153	nigel	75	.sp
154			(e) PCRE_ANCHORED can be used at matching time to force a pattern to be tried
155			only at the first matching position in the subject string.
156			.sp
157	ph10	442	(f) The PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART, and
158			PCRE_NO_AUTO_CAPTURE options for \fBpcre_exec()\fP have no Perl equivalents.
159	nigel	75	.sp
160	ph10	231	(g) The \eR escape sequence can be restricted to match only CR, LF, or CRLF
161			by the PCRE_BSR_ANYCRLF option.
162	nigel	75	.sp
163	ph10	231	(h) The callout facility is PCRE-specific.
164	nigel	75	.sp
165	ph10	231	(i) The partial matching facility is PCRE-specific.
166			.sp
167			(j) Patterns compiled by PCRE can be saved and re-used at a later time, even on
168	ph10	733	different hosts that have the other endianness. However, this does not apply to
169	ph10	716	optimized data created by the just-in-time compiler.
170	nigel	77	.sp
171	chpe	1055	(k) The alternative matching functions (\fBpcre_dfa_exec()\fP,
172			\fBpcre16_dfa_exec()\fP and \fBpcre32_dfa_exec()\fP,) match in a different way
173			and are not Perl-compatible.
174	ph10	231	.sp
175			(l) PCRE recognizes some special sequences such as (*CR) at the start of
176			a pattern that set overall options that cannot be changed within the pattern.
177	ph10	99	.
178			.
179			.SH AUTHOR
180			.rs
181			.sp
182			.nf
183			Philip Hazel
184			University Computing Service
185			Cambridge CB2 3QH, England.
186			.fi
187			.
188			.
189			.SH REVISION
190			.rs
191			.sp
192			.nf
193	ph10	1297	Last updated: 19 March 2013
194	ph10	1258	Copyright (c) 1997-2013 University of Cambridge.
195	ph10	99	.fi
Name	Value
svn:eol-style	native
svn:keywords	"Author Date Id Revision Url"