This file contains the PCRE man page that describes the regular expressions
-supported by PCRE version 6.2. Note that not all of the features are relevant
+supported by PCRE version 6.7. Note that not all of the features are relevant
in the context of Exim. In particular, the version of PCRE that is compiled
with Exim does not include UTF-8 support, there is no mechanism for changing
the options with which the PCRE functions are called, and features such as
If a pattern is compiled with the PCRE_EXTENDED option, whitespace in
the pattern (other than in a character class) and characters between a
- # outside a character class and the next newline character are ignored.
- An escaping backslash can be used to include a whitespace or # charac-
- ter as part of the pattern.
+ # outside a character class and the next newline are ignored. An escap-
+ ing backslash can be used to include a whitespace or # character as
+ part of the pattern.
If you want to remove the special meaning from a sequence of charac-
ters, you can do so by putting them between \Q and \E. This is differ-
\t tab (hex 09)
\ddd character with octal code ddd, or backreference
\xhh character with hex code hh
- \x{hhh..} character with hex code hhh... (UTF-8 mode only)
+ \x{hhh..} character with hex code hhh..
The precise effect of \cx is as follows: if x is a lower case letter,
it is converted to upper case. Then bit 6 of the character (hex 40) is
becomes hex 7B.
After \x, from zero to two hexadecimal digits are read (letters can be
- in upper or lower case). In UTF-8 mode, any number of hexadecimal dig-
- its may appear between \x{ and }, but the value of the character code
- must be less than 2**31 (that is, the maximum hexadecimal value is
- 7FFFFFFF). If characters other than hexadecimal digits appear between
- \x{ and }, or if there is no terminating }, this form of escape is not
- recognized. Instead, the initial \x will be interpreted as a basic
- hexadecimal escape, with no following digits, giving a character whose
- value is zero.
+ in upper or lower case). Any number of hexadecimal digits may appear
+ between \x{ and }, but the value of the character code must be less
+ than 256 in non-UTF-8 mode, and less than 2**31 in UTF-8 mode (that is,
+ the maximum hexadecimal value is 7FFFFFFF). If characters other than
+ hexadecimal digits appear between \x{ and }, or if there is no termi-
+ nating }, this form of escape is not recognized. Instead, the initial
+ \x will be interpreted as a basic hexadecimal escape, with no following
+ digits, giving a character whose value is zero.
Characters whose value is less than 256 can be defined by either of the
- two syntaxes for \x when PCRE is in UTF-8 mode. There is no difference
- in the way they are handled. For example, \xdc is exactly the same as
- \x{dc}.
+ two syntaxes for \x. There is no difference in the way they are han-
+ dled. For example, \xdc is exactly the same as \x{dc}.
- After \0 up to two further octal digits are read. In both cases, if
- there are fewer than two digits, just those that are present are used.
- Thus the sequence \0\x\07 specifies two binary zeros followed by a BEL
- character (code value 7). Make sure you supply two digits after the
- initial zero if the pattern character that follows is itself an octal
- digit.
+ After \0 up to two further octal digits are read. If there are fewer
+ than two digits, just those that are present are used. Thus the
+ sequence \0\x\07 specifies two binary zeros followed by a BEL character
+ (code value 7). Make sure you supply two digits after the initial zero
+ if the pattern character that follows is itself an octal digit.
The handling of a backslash followed by a digit other than 0 is compli-
cated. Outside a character class, PCRE reads it and any following dig-
Inside a character class, or if the decimal number is greater than 9
and there have not been that many capturing subpatterns, PCRE re-reads
- up to three octal digits following the backslash, and generates a sin-
- gle byte from the least significant 8 bits of the value. Any subsequent
- digits stand for themselves. For example:
+ up to three octal digits following the backslash, ane uses them to gen-
+ erate a data character. Any subsequent digits stand for themselves. In
+ non-UTF-8 mode, the value of a character specified in octal must be
+ less than \400. In UTF-8 mode, values up to \777 are permitted. For
+ example:
\040 is another way of writing a space
\40 is the same, provided there are fewer than 40
Note that octal values of 100 or greater must not be introduced by a
leading zero, because no more than three octal digits are ever read.
- All the sequences that define a single byte value or a single UTF-8
- character (in UTF-8 mode) can be used both inside and outside character
- classes. In addition, inside a character class, the sequence \b is
- interpreted as the backspace character (hex 08), and the sequence \X is
- interpreted as the character "X". Outside a character class, these
- sequences have different meanings (see below).
+ All the sequences that define a single character value can be used both
+ inside and outside character classes. In addition, inside a character
+ class, the sequence \b is interpreted as the backspace character (hex
+ 08), and the sequence \X is interpreted as the character "X". Outside a
+ character class, these sequences have different meanings (see below).
Generic character types
- The third use of backslash is for specifying generic character types.
+ The third use of backslash is for specifying generic character types.
The following are always recognized:
\d any decimal digit
\W any "non-word" character
Each pair of escape sequences partitions the complete set of characters
- into two disjoint sets. Any given character matches one, and only one,
+ into two disjoint sets. Any given character matches one, and only one,
of each pair.
These character type sequences can appear both inside and outside char-
- acter classes. They each match one character of the appropriate type.
- If the current matching point is at the end of the subject string, all
+ acter classes. They each match one character of the appropriate type.
+ If the current matching point is at the end of the subject string, all
of them fail, since there is no character to match.
- For compatibility with Perl, \s does not match the VT character (code
- 11). This makes it different from the the POSIX "space" class. The \s
- characters are HT (9), LF (10), FF (12), CR (13), and space (32).
+ For compatibility with Perl, \s does not match the VT character (code
+ 11). This makes it different from the the POSIX "space" class. The \s
+ characters are HT (9), LF (10), FF (12), CR (13), and space (32). (If
+ "use locale;" is included in a Perl script, \s may match the VT charac-
+ ter. In PCRE, it never does.)
A "word" character is an underscore or any character less than 256 that
- is a letter or digit. The definition of letters and digits is con-
- trolled by PCRE's low-valued character tables, and may vary if locale-
- specific matching is taking place (see "Locale support" in the pcreapi
- page). For example, in the "fr_FR" (French) locale, some character
- codes greater than 128 are used for accented letters, and these are
+ is a letter or digit. The definition of letters and digits is con-
+ trolled by PCRE's low-valued character tables, and may vary if locale-
+ specific matching is taking place (see "Locale support" in the pcreapi
+ page). For example, in the "fr_FR" (French) locale, some character
+ codes greater than 128 are used for accented letters, and these are
matched by \w.
- In UTF-8 mode, characters with values greater than 128 never match \d,
+ In UTF-8 mode, characters with values greater than 128 never match \d,
\s, or \w, and always match \D, \S, and \W. This is true even when Uni-
- code character property support is available.
+ code character property support is available. The use of locales with
+ Unicode is discouraged.
Unicode character properties
When PCRE is built with Unicode character property support, three addi-
- tional escape sequences to match generic character types are available
+ tional escape sequences to match character properties are available
when UTF-8 mode is selected. They are:
- \p{xx} a character with the xx property
- \P{xx} a character without the xx property
- \X an extended Unicode sequence
+ \p{xx} a character with the xx property
+ \P{xx} a character without the xx property
+ \X an extended Unicode sequence
The property names represented by xx above are limited to the Unicode
- general category properties. Each character has exactly one such prop-
- erty, specified by a two-letter abbreviation. For compatibility with
- Perl, negation can be specified by including a circumflex between the
- opening brace and the property name. For example, \p{^Lu} is the same
- as \P{Lu}.
-
- If only one letter is specified with \p or \P, it includes all the
- properties that start with that letter. In this case, in the absence of
- negation, the curly brackets in the escape sequence are optional; these
- two examples have the same effect:
+ script names, the general category properties, and "Any", which matches
+ any character (including newline). Other properties such as "InMusical-
+ Symbols" are not currently supported by PCRE. Note that \P{Any} does
+ not match any characters, so always causes a match failure.
+
+ Sets of Unicode characters are defined as belonging to certain scripts.
+ A character from one of these sets can be matched using a script name.
+ For example:
+
+ \p{Greek}
+ \P{Han}
+
+ Those that are not part of an identified script are lumped together as
+ "Common". The current list of scripts is:
+
+ Arabic, Armenian, Bengali, Bopomofo, Braille, Buginese, Buhid, Cana-
+ dian_Aboriginal, Cherokee, Common, Coptic, Cypriot, Cyrillic, Deseret,
+ Devanagari, Ethiopic, Georgian, Glagolitic, Gothic, Greek, Gujarati,
+ Gurmukhi, Han, Hangul, Hanunoo, Hebrew, Hiragana, Inherited, Kannada,
+ Katakana, Kharoshthi, Khmer, Lao, Latin, Limbu, Linear_B, Malayalam,
+ Mongolian, Myanmar, New_Tai_Lue, Ogham, Old_Italic, Old_Persian, Oriya,
+ Osmanya, Runic, Shavian, Sinhala, Syloti_Nagri, Syriac, Tagalog, Tag-
+ banwa, Tai_Le, Tamil, Telugu, Thaana, Thai, Tibetan, Tifinagh,
+ Ugaritic, Yi.
+
+ Each character has exactly one general category property, specified by
+ a two-letter abbreviation. For compatibility with Perl, negation can be
+ specified by including a circumflex between the opening brace and the
+ property name. For example, \p{^Lu} is the same as \P{Lu}.
+
+ If only one letter is specified with \p or \P, it includes all the gen-
+ eral category properties that start with that letter. In this case, in
+ the absence of negation, the curly brackets in the escape sequence are
+ optional; these two examples have the same effect:
\p{L}
\pL
- The following property codes are supported:
+ The following general category property codes are supported:
C Other
Cc Control
Zp Paragraph separator
Zs Space separator
- Extended properties such as "Greek" or "InMusicalSymbols" are not sup-
- ported by PCRE.
+ The special property L& is also supported: it matches a character that
+ has the Lu, Ll, or Lt property, in other words, a letter that is not
+ classified as a modifier or "other".
+
+ The long synonyms for these properties that Perl supports (such as
+ \p{Letter}) are not supported by PCRE, nor is it permitted to prefix
+ any of these properties with "Is".
+
+ No character that is in the Unicode table has the Cn (unassigned) prop-
+ erty. Instead, this property is assumed for any code point that is not
+ in the Unicode table.
Specifying caseless matching does not affect these escape sequences.
For example, \p{Lu} always matches only upper case letters.
However, if the startoffset argument of pcre_exec() is non-zero, indi-
cating that matching is to start at a point other than the beginning of
the subject, \A can never match. The difference between \Z and \z is
- that \Z matches before a newline that is the last character of the
- string as well as at the end of the string, whereas \z matches only at
- the end.
-
- The \G assertion is true only when the current matching position is at
- the start point of the match, as specified by the startoffset argument
- of pcre_exec(). It differs from \A when the value of startoffset is
- non-zero. By calling pcre_exec() multiple times with appropriate argu-
+ that \Z matches before a newline at the end of the string as well as at
+ the very end, whereas \z matches only at the end.
+
+ The \G assertion is true only when the current matching position is at
+ the start point of the match, as specified by the startoffset argument
+ of pcre_exec(). It differs from \A when the value of startoffset is
+ non-zero. By calling pcre_exec() multiple times with appropriate argu-
ments, you can mimic Perl's /g option, and it is in this kind of imple-
mentation where \G can be useful.
- Note, however, that PCRE's interpretation of \G, as the start of the
+ Note, however, that PCRE's interpretation of \G, as the start of the
current match, is subtly different from Perl's, which defines it as the
- end of the previous match. In Perl, these can be different when the
- previously matched string was empty. Because PCRE does just one match
+ end of the previous match. In Perl, these can be different when the
+ previously matched string was empty. Because PCRE does just one match
at a time, it cannot reproduce this behaviour.
- If all the alternatives of a pattern begin with \G, the expression is
+ If all the alternatives of a pattern begin with \G, the expression is
anchored to the starting match position, and the "anchored" flag is set
in the compiled regular expression.
CIRCUMFLEX AND DOLLAR
Outside a character class, in the default matching mode, the circumflex
- character is an assertion that is true only if the current matching
- point is at the start of the subject string. If the startoffset argu-
- ment of pcre_exec() is non-zero, circumflex can never match if the
- PCRE_MULTILINE option is unset. Inside a character class, circumflex
+ character is an assertion that is true only if the current matching
+ point is at the start of the subject string. If the startoffset argu-
+ ment of pcre_exec() is non-zero, circumflex can never match if the
+ PCRE_MULTILINE option is unset. Inside a character class, circumflex
has an entirely different meaning (see below).
- Circumflex need not be the first character of the pattern if a number
- of alternatives are involved, but it should be the first thing in each
- alternative in which it appears if the pattern is ever to match that
- branch. If all possible alternatives start with a circumflex, that is,
- if the pattern is constrained to match only at the start of the sub-
- ject, it is said to be an "anchored" pattern. (There are also other
+ Circumflex need not be the first character of the pattern if a number
+ of alternatives are involved, but it should be the first thing in each
+ alternative in which it appears if the pattern is ever to match that
+ branch. If all possible alternatives start with a circumflex, that is,
+ if the pattern is constrained to match only at the start of the sub-
+ ject, it is said to be an "anchored" pattern. (There are also other
constructs that can cause a pattern to be anchored.)
- A dollar character is an assertion that is true only if the current
- matching point is at the end of the subject string, or immediately
- before a newline character that is the last character in the string (by
- default). Dollar need not be the last character of the pattern if a
- number of alternatives are involved, but it should be the last item in
- any branch in which it appears. Dollar has no special meaning in a
- character class.
+ A dollar character is an assertion that is true only if the current
+ matching point is at the end of the subject string, or immediately
+ before a newline at the end of the string (by default). Dollar need not
+ be the last character of the pattern if a number of alternatives are
+ involved, but it should be the last item in any branch in which it
+ appears. Dollar has no special meaning in a character class.
The meaning of dollar can be changed so that it matches only at the
very end of the string, by setting the PCRE_DOLLAR_ENDONLY option at
compile time. This does not affect the \Z assertion.
The meanings of the circumflex and dollar characters are changed if the
- PCRE_MULTILINE option is set. When this is the case, they match immedi-
- ately after and immediately before an internal newline character,
- respectively, in addition to matching at the start and end of the sub-
- ject string. For example, the pattern /^abc$/ matches the subject
- string "def\nabc" (where \n represents a newline character) in multi-
- line mode, but not otherwise. Consequently, patterns that are anchored
- in single line mode because all branches start with ^ are not anchored
- in multiline mode, and a match for circumflex is possible when the
- startoffset argument of pcre_exec() is non-zero. The PCRE_DOL-
- LAR_ENDONLY option is ignored if PCRE_MULTILINE is set.
-
- Note that the sequences \A, \Z, and \z can be used to match the start
- and end of the subject in both modes, and if all branches of a pattern
- start with \A it is always anchored, whether PCRE_MULTILINE is set or
- not.
+ PCRE_MULTILINE option is set. When this is the case, a circumflex
+ matches immediately after internal newlines as well as at the start of
+ the subject string. It does not match after a newline that ends the
+ string. A dollar matches before any newlines in the string, as well as
+ at the very end, when PCRE_MULTILINE is set. When newline is specified
+ as the two-character sequence CRLF, isolated CR and LF characters do
+ not indicate newlines.
+
+ For example, the pattern /^abc$/ matches the subject string "def\nabc"
+ (where \n represents a newline) in multiline mode, but not otherwise.
+ Consequently, patterns that are anchored in single line mode because
+ all branches start with ^ are not anchored in multiline mode, and a
+ match for circumflex is possible when the startoffset argument of
+ pcre_exec() is non-zero. The PCRE_DOLLAR_ENDONLY option is ignored if
+ PCRE_MULTILINE is set.
+
+ Note that the sequences \A, \Z, and \z can be used to match the start
+ and end of the subject in both modes, and if all branches of a pattern
+ start with \A it is always anchored, whether or not PCRE_MULTILINE is
+ set.
FULL STOP (PERIOD, DOT)
Outside a character class, a dot in the pattern matches any one charac-
- ter in the subject, including a non-printing character, but not (by
- default) newline. In UTF-8 mode, a dot matches any UTF-8 character,
- which might be more than one byte long, except (by default) newline. If
- the PCRE_DOTALL option is set, dots match newlines as well. The han-
- dling of dot is entirely independent of the handling of circumflex and
- dollar, the only relationship being that they both involve newline
- characters. Dot has no special meaning in a character class.
+ ter in the subject string except (by default) a character that signi-
+ fies the end of a line. In UTF-8 mode, the matched character may be
+ more than one byte long. When a line ending is defined as a single
+ character (CR or LF), dot never matches that character; when the two-
+ character sequence CRLF is used, dot does not match CR if it is immedi-
+ ately followed by LF, but otherwise it matches all characters (includ-
+ ing isolated CRs and LFs).
+
+ The behaviour of dot with regard to newlines can be changed. If the
+ PCRE_DOTALL option is set, a dot matches any one character, without
+ exception. If newline is defined as the two-character sequence CRLF, it
+ takes two dots to match it.
+
+ The handling of dot is entirely independent of the handling of circum-
+ flex and dollar, the only relationship being that they both involve
+ newlines. Dot has no special meaning in a character class.
MATCHING A SINGLE BYTE
Outside a character class, the escape sequence \C matches any one byte,
- both in and out of UTF-8 mode. Unlike a dot, it can match a newline.
- The feature is provided in Perl in order to match individual bytes in
- UTF-8 mode. Because it breaks up UTF-8 characters into individual
+ both in and out of UTF-8 mode. Unlike a dot, it always matches CR and
+ LF. The feature is provided in Perl in order to match individual bytes
+ in UTF-8 mode. Because it breaks up UTF-8 characters into individual
bytes, what remains in the string may be a malformed UTF-8 string. For
this reason, the \C escape sequence is best avoided.
PCRE is compiled with Unicode property support as well as with UTF-8
support.
- The newline character is never treated in any special way in character
- classes, whatever the setting of the PCRE_DOTALL or PCRE_MULTILINE
- options is. A class such as [^a] will always match a newline.
+ Characters that might indicate line breaks (CR and LF) are never
+ treated in any special way when matching character classes, whatever
+ line-ending sequence is in use, and whatever setting of the PCRE_DOTALL
+ and PCRE_MULTILINE options is used. A class such as [^a] always matches
+ one of these characters.
The minus (hyphen) character can be used to specify a range of charac-
ters in a character class. For example, [d-m] matches any letter
matches either "gilbert" or "sullivan". Any number of alternatives may
appear, and an empty alternative is permitted (matching the empty
- string). The matching process tries each alternative in turn, from
- left to right, and the first one that succeeds is used. If the alterna-
- tives are within a subpattern (defined below), "succeeds" means match-
- ing the rest of the main pattern as well as the alternative in the sub-
- pattern.
+ string). The matching process tries each alternative in turn, from left
+ to right, and the first one that succeeds is used. If the alternatives
+ are within a subpattern (defined below), "succeeds" means matching the
+ rest of the main pattern as well as the alternative in the subpattern.
INTERNAL OPTION SETTING
the effects of option settings happen at compile time. There would be
some very weird behaviour otherwise.
- The PCRE-specific options PCRE_UNGREEDY and PCRE_EXTRA can be changed
- in the same way as the Perl-compatible options by using the characters
- U and X respectively. The (?X) flag setting is special in that it must
- always occur earlier in the pattern than any of the additional features
- it turns on, even when it is at top level. It is best to put it at the
- start.
+ The PCRE-specific options PCRE_DUPNAMES, PCRE_UNGREEDY, and PCRE_EXTRA
+ can be changed in the same way as the Perl-compatible options by using
+ the characters J, U and X respectively.
SUBPATTERNS
cat(aract|erpillar|)
- matches one of the words "cat", "cataract", or "caterpillar". Without
- the parentheses, it would match "cataract", "erpillar" or the empty
+ matches one of the words "cat", "cataract", or "caterpillar". Without
+ the parentheses, it would match "cataract", "erpillar" or the empty
string.
- 2. It sets up the subpattern as a capturing subpattern. This means
- that, when the whole pattern matches, that portion of the subject
+ 2. It sets up the subpattern as a capturing subpattern. This means
+ that, when the whole pattern matches, that portion of the subject
string that matched the subpattern is passed back to the caller via the
- ovector argument of pcre_exec(). Opening parentheses are counted from
- left to right (starting from 1) to obtain numbers for the capturing
+ ovector argument of pcre_exec(). Opening parentheses are counted from
+ left to right (starting from 1) to obtain numbers for the capturing
subpatterns.
- For example, if the string "the red king" is matched against the pat-
+ For example, if the string "the red king" is matched against the pat-
tern
the ((red|white) (king|queen))
the captured substrings are "red king", "red", and "king", and are num-
bered 1, 2, and 3, respectively.
- The fact that plain parentheses fulfil two functions is not always
- helpful. There are often times when a grouping subpattern is required
- without a capturing requirement. If an opening parenthesis is followed
- by a question mark and a colon, the subpattern does not do any captur-
- ing, and is not counted when computing the number of any subsequent
- capturing subpatterns. For example, if the string "the white queen" is
+ The fact that plain parentheses fulfil two functions is not always
+ helpful. There are often times when a grouping subpattern is required
+ without a capturing requirement. If an opening parenthesis is followed
+ by a question mark and a colon, the subpattern does not do any captur-
+ ing, and is not counted when computing the number of any subsequent
+ capturing subpatterns. For example, if the string "the white queen" is
matched against the pattern
the ((?:red|white) (king|queen))
the captured substrings are "white queen" and "queen", and are numbered
- 1 and 2. The maximum number of capturing subpatterns is 65535, and the
- maximum depth of nesting of all subpatterns, both capturing and non-
+ 1 and 2. The maximum number of capturing subpatterns is 65535, and the
+ maximum depth of nesting of all subpatterns, both capturing and non-
capturing, is 200.
- As a convenient shorthand, if any option settings are required at the
- start of a non-capturing subpattern, the option letters may appear
+ As a convenient shorthand, if any option settings are required at the
+ start of a non-capturing subpattern, the option letters may appear
between the "?" and the ":". Thus the two patterns
(?i:saturday|sunday)
(?:(?i)saturday|sunday)
match exactly the same set of strings. Because alternative branches are
- tried from left to right, and options are not reset until the end of
- the subpattern is reached, an option setting in one branch does affect
- subsequent branches, so the above patterns match "SUNDAY" as well as
+ tried from left to right, and options are not reset until the end of
+ the subpattern is reached, an option setting in one branch does affect
+ subsequent branches, so the above patterns match "SUNDAY" as well as
"Saturday".
NAMED SUBPATTERNS
- Identifying capturing parentheses by number is simple, but it can be
- very hard to keep track of the numbers in complicated regular expres-
- sions. Furthermore, if an expression is modified, the numbers may
- change. To help with this difficulty, PCRE supports the naming of sub-
- patterns, something that Perl does not provide. The Python syntax
- (?P<name>...) is used. Names consist of alphanumeric characters and
- underscores, and must be unique within a pattern.
-
- Named capturing parentheses are still allocated numbers as well as
+ Identifying capturing parentheses by number is simple, but it can be
+ very hard to keep track of the numbers in complicated regular expres-
+ sions. Furthermore, if an expression is modified, the numbers may
+ change. To help with this difficulty, PCRE supports the naming of sub-
+ patterns, something that Perl does not provide. The Python syntax
+ (?P<name>...) is used. References to capturing parentheses from other
+ parts of the pattern, such as backreferences, recursion, and condi-
+ tions, can be made by name as well as by number.
+
+ Names consist of up to 32 alphanumeric characters and underscores.
+ Named capturing parentheses are still allocated numbers as well as
names. The PCRE API provides function calls for extracting the name-to-
- number translation table from a compiled pattern. There is also a con-
- venience function for extracting a captured substring by name. For fur-
- ther details see the pcreapi documentation.
+ number translation table from a compiled pattern. There is also a con-
+ venience function for extracting a captured substring by name.
+
+ By default, a name must be unique within a pattern, but it is possible
+ to relax this constraint by setting the PCRE_DUPNAMES option at compile
+ time. This can be useful for patterns where only one instance of the
+ named parentheses can match. Suppose you want to match the name of a
+ weekday, either as a 3-letter abbreviation or as the full name, and in
+ both cases you want to extract the abbreviation. This pattern (ignoring
+ the line breaks) does the job:
+
+ (?P<DN>Mon|Fri|Sun)(?:day)?|
+ (?P<DN>Tue)(?:sday)?|
+ (?P<DN>Wed)(?:nesday)?|
+ (?P<DN>Thu)(?:rsday)?|
+ (?P<DN>Sat)(?:urday)?
+
+ There are five capturing substrings, but only one is ever set after a
+ match. The convenience function for extracting the data by name
+ returns the substring for the first, and in this example, the only,
+ subpattern of that name that matched. This saves searching to find
+ which numbered subpattern it was. If you make a reference to a non-
+ unique named subpattern from elsewhere in the pattern, the one that
+ corresponds to the lowest number is used. For further details of the
+ interfaces for handling named subpatterns, see the pcreapi documenta-
+ tion.
REPETITION
meaning or processing of a possessive quantifier and the equivalent
atomic group.
- The possessive quantifier syntax is an extension to the Perl syntax. It
- originates in Sun's Java package.
+ The possessive quantifier syntax is an extension to the Perl syntax.
+ Jeffrey Friedl originated the idea (and the name) in the first edition
+ of his book. Mike McCloskey liked it, so implemented it when he built
+ Sun's Java package, and PCRE copied it from there.
When a pattern contains an unlimited repeat inside a subpattern that
can itself be repeated an unlimited number of times, the use of an
it is always taken as a back reference, and causes an error only if
there are not that many capturing left parentheses in the entire pat-
tern. In other words, the parentheses that are referenced need not be
- to the left of the reference for numbers less than 10. See the subsec-
- tion entitled "Non-printing characters" above for further details of
- the handling of digits following a backslash.
+ to the left of the reference for numbers less than 10. A "forward back
+ reference" of this type can make sense when a repetition is involved
+ and the subpattern to the right has participated in an earlier itera-
+ tion.
- A back reference matches whatever actually matched the capturing sub-
- pattern in the current subject string, rather than anything matching
+ It is not possible to have a numerical "forward back reference" to sub-
+ pattern whose number is 10 or more. However, a back reference to any
+ subpattern is possible using named parentheses (see below). See also
+ the subsection entitled "Non-printing characters" above for further
+ details of the handling of digits following a backslash.
+
+ A back reference matches whatever actually matched the capturing sub-
+ pattern in the current subject string, rather than anything matching
the subpattern itself (see "Subpatterns as subroutines" below for a way
of doing that). So the pattern
(sens|respons)e and \1ibility
- matches "sense and sensibility" and "response and responsibility", but
- not "sense and responsibility". If caseful matching is in force at the
- time of the back reference, the case of letters is relevant. For exam-
+ matches "sense and sensibility" and "response and responsibility", but
+ not "sense and responsibility". If caseful matching is in force at the
+ time of the back reference, the case of letters is relevant. For exam-
ple,
((?i)rah)\s+\1
- matches "rah rah" and "RAH RAH", but not "RAH rah", even though the
+ matches "rah rah" and "RAH RAH", but not "RAH rah", even though the
original capturing subpattern is matched caselessly.
- Back references to named subpatterns use the Python syntax (?P=name).
+ Back references to named subpatterns use the Python syntax (?P=name).
We could rewrite the above example as follows:
- (?<p1>(?i)rah)\s+(?P=p1)
+ (?P<p1>(?i)rah)\s+(?P=p1)
+
+ A subpattern that is referenced by name may appear in the pattern
+ before or after the reference.
There may be more than one back reference to the same subpattern. If a
subpattern has not actually been used in a particular match, any back
does find an occurrence of "bar" that is not preceded by "foo". The
contents of a lookbehind assertion are restricted such that all the
strings it matches must have a fixed length. However, if there are sev-
- eral alternatives, they do not all have to have the same fixed length.
- Thus
+ eral top-level alternatives, they do not all have to have the same
+ fixed length. Thus
(?<=bullock|donkey)
tives in the subpattern, a compile-time error occurs.
There are three kinds of condition. If the text between the parentheses
- consists of a sequence of digits, the condition is satisfied if the
- capturing subpattern of that number has previously matched. The number
- must be greater than zero. Consider the following pattern, which con-
- tains non-significant white space to make it more readable (assume the
- PCRE_EXTENDED option) and to divide it into three parts for ease of
- discussion:
+ consists of a sequence of digits, or a sequence of alphanumeric charac-
+ ters and underscores, the condition is satisfied if the capturing sub-
+ pattern of that number or name has previously matched. There is a pos-
+ sible ambiguity here, because subpattern names may consist entirely of
+ digits. PCRE looks first for a named subpattern; if it cannot find one
+ and the text consists entirely of digits, it looks for a subpattern of
+ that number, which must be greater than zero. Using subpattern names
+ that consist entirely of digits is not recommended.
+
+ Consider the following pattern, which contains non-significant white
+ space to make it more readable (assume the PCRE_EXTENDED option) and to
+ divide it into three parts for ease of discussion:
( \( )? [^()]+ (?(1) \) )
tern is executed and a closing parenthesis is required. Otherwise,
since no-pattern is not present, the subpattern matches nothing. In
other words, this pattern matches a sequence of non-parentheses,
- optionally enclosed in parentheses.
+ optionally enclosed in parentheses. Rewriting it to use a named subpat-
+ tern gives this:
+
+ (?P<OPEN> \( )? [^()]+ (?(OPEN) \) )
- If the condition is the string (R), it is satisfied if a recursive call
- to the pattern or subpattern has been made. At "top level", the condi-
- tion is false. This is a PCRE extension. Recursive patterns are
- described in the next section.
+ If the condition is the string (R), and there is no subpattern with the
+ name R, the condition is satisfied if a recursive call to the pattern
+ or subpattern has been made. At "top level", the condition is false.
+ This is a PCRE extension. Recursive patterns are described in the next
+ section.
If the condition is not a sequence of digits or (R), it must be an
assertion. This may be a positive or negative lookahead or lookbehind
at all.
If the PCRE_EXTENDED option is set, an unescaped # character outside a
- character class introduces a comment that continues up to the next new-
- line character in the pattern.
+ character class introduces a comment that continues to immediately
+ after the next newline in the pattern.
RECURSIVE PATTERNS
tion.) The special item (?R) is a recursive call of the entire regular
expression.
- For example, this PCRE pattern solves the nested parentheses problem
- (assume the PCRE_EXTENDED option is set so that white space is
- ignored):
+ A recursive subpattern call is always treated as an atomic group. That
+ is, once it has matched some of the subject string, it is never re-
+ entered, even if it contains untried alternatives and there is a subse-
+ quent matching failure.
+
+ This PCRE pattern solves the nested parentheses problem (assume the
+ PCRE_EXTENDED option is set so that white space is ignored):
\( ( (?>[^()]+) | (?R) )* \)
First it matches an opening parenthesis. Then it matches any number of
substrings which can either be a sequence of non-parentheses, or a
- recursive match of the pattern itself (that is a correctly parenthe-
+ recursive match of the pattern itself (that is, a correctly parenthe-
sized substring). Finally there is a closing parenthesis.
If this were part of a larger pattern, you would not want to recurse
(sens|respons)e and (?1)ibility
is used, it does match "sense and responsibility" as well as the other
- two strings. Such references must, however, follow the subpattern to
- which they refer.
+ two strings. Such references, if given numerically, must follow the
+ subpattern to which they refer. However, named references can refer to
+ later subpatterns.
+
+ Like recursive subpatterns, a "subroutine" call is always treated as an
+ atomic group. That is, once it has matched some of the subject string,
+ it is never re-entered, even if it contains untried alternatives and
+ there is a subsequent matching failure.
CALLOUTS
gether. A complete description of the interface to the callout function
is given in the pcrecallout documentation.
-Last updated: 28 February 2005
-Copyright (c) 1997-2005 University of Cambridge.
+Last updated: 06 June 2006
+Copyright (c) 1997-2006 University of Cambridge.
+++ /dev/null
-/* $Cambridge: exim/src/src/pcre/internal.h,v 1.2 2005/06/15 08:57:10 ph10 Exp $ */
-
-/*************************************************
-* Perl-Compatible Regular Expressions *
-*************************************************/
-
-
-/* This is a library of functions to support regular expressions whose syntax
-and semantics are as close as possible to those of the Perl 5 language. See
-the file doc/Tech.Notes for some information on the internals.
-
-Written by: Philip Hazel <ph10@cam.ac.uk>
-
- Copyright (c) 1997-2004 University of Cambridge
-
------------------------------------------------------------------------------
-Redistribution and use in source and binary forms, with or without
-modification, are permitted provided that the following conditions are met:
-
- * Redistributions of source code must retain the above copyright notice,
- this list of conditions and the following disclaimer.
-
- * Redistributions in binary form must reproduce the above copyright
- notice, this list of conditions and the following disclaimer in the
- documentation and/or other materials provided with the distribution.
-
- * Neither the name of the University of Cambridge nor the names of its
- contributors may be used to endorse or promote products derived from
- this software without specific prior written permission.
-
-THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
-AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
-IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
-ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
-LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
-CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
-SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
-INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
-CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
-ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
-POSSIBILITY OF SUCH DAMAGE.
------------------------------------------------------------------------------
-*/
-
-/* This header contains definitions that are shared between the different
-modules, but which are not relevant to the outside. */
-
-/* Get the definitions provided by running "configure" */
-
-#include "config.h"
-
-/* Standard C headers plus the external interface definition. The only time
-setjmp and stdarg are used is when NO_RECURSE is set. */
-
-#include <ctype.h>
-#include <limits.h>
-#include <setjmp.h>
-#include <stdarg.h>
-#include <stddef.h>
-#include <stdio.h>
-#include <stdlib.h>
-#include <string.h>
-
-#ifndef PCRE_SPY
-#define PCRE_DEFINITION /* Win32 __declspec(export) trigger for .dll */
-#endif
-
-/* We need to have types that specify unsigned 16-bit and 32-bit integers. We
-cannot determine these outside the compilation (e.g. by running a program as
-part of "configure") because PCRE is often cross-compiled for use on other
-systems. Instead we make use of the maximum sizes that are available at
-preprocessor time in standard C environments. */
-
-#if USHRT_MAX == 65535
- typedef unsigned short pcre_uint16;
-#elif UINT_MAX == 65535
- typedef unsigned int pcre_uint16;
-#else
- #error Cannot determine a type for 16-bit unsigned integers
-#endif
-
-#if UINT_MAX == 4294967295
- typedef unsigned int pcre_uint32;
-#elif ULONG_MAX == 4294967295
- typedef unsigned long int pcre_uint32;
-#else
- #error Cannot determine a type for 32-bit unsigned integers
-#endif
-
-/* All character handling must be done as unsigned characters. Otherwise there
-are problems with top-bit-set characters and functions such as isspace().
-However, we leave the interface to the outside world as char *, because that
-should make things easier for callers. We define a short type for unsigned char
-to save lots of typing. I tried "uchar", but it causes problems on Digital
-Unix, where it is defined in sys/types, so use "uschar" instead. */
-
-typedef unsigned char uschar;
-
-/* Include the public PCRE header */
-
-#include "pcre.h"
-
-/* When compiling for use with the Virtual Pascal compiler, these functions
-need to have their names changed. PCRE must be compiled with the -DVPCOMPAT
-option on the command line. */
-
-#ifdef VPCOMPAT
-#define strncmp(s1,s2,m) _strncmp(s1,s2,m)
-#define memcpy(d,s,n) _memcpy(d,s,n)
-#define memmove(d,s,n) _memmove(d,s,n)
-#define memset(s,c,n) _memset(s,c,n)
-#else /* VPCOMPAT */
-
-/* To cope with SunOS4 and other systems that lack memmove() but have bcopy(),
-define a macro for memmove() if HAVE_MEMMOVE is false, provided that HAVE_BCOPY
-is set. Otherwise, include an emulating function for those systems that have
-neither (there some non-Unix environments where this is the case). This assumes
-that all calls to memmove are moving strings upwards in store, which is the
-case in PCRE. */
-
-#if ! HAVE_MEMMOVE
-#undef memmove /* some systems may have a macro */
-#if HAVE_BCOPY
-#define memmove(a, b, c) bcopy(b, a, c)
-#else /* HAVE_BCOPY */
-void *
-pcre_memmove(unsigned char *dest, const unsigned char *src, size_t n)
-{
-int i;
-dest += n;
-src += n;
-for (i = 0; i < n; ++i) *(--dest) = *(--src);
-}
-#define memmove(a, b, c) pcre_memmove(a, b, c)
-#endif /* not HAVE_BCOPY */
-#endif /* not HAVE_MEMMOVE */
-#endif /* not VPCOMPAT */
-
-
-/* PCRE keeps offsets in its compiled code as 2-byte quantities (always stored
-in big-endian order) by default. These are used, for example, to link from the
-start of a subpattern to its alternatives and its end. The use of 2 bytes per
-offset limits the size of the compiled regex to around 64K, which is big enough
-for almost everybody. However, I received a request for an even bigger limit.
-For this reason, and also to make the code easier to maintain, the storing and
-loading of offsets from the byte string is now handled by the macros that are
-defined here.
-
-The macros are controlled by the value of LINK_SIZE. This defaults to 2 in
-the config.h file, but can be overridden by using -D on the command line. This
-is automated on Unix systems via the "configure" command. */
-
-#if LINK_SIZE == 2
-
-#define PUT(a,n,d) \
- (a[n] = (d) >> 8), \
- (a[(n)+1] = (d) & 255)
-
-#define GET(a,n) \
- (((a)[n] << 8) | (a)[(n)+1])
-
-#define MAX_PATTERN_SIZE (1 << 16)
-
-
-#elif LINK_SIZE == 3
-
-#define PUT(a,n,d) \
- (a[n] = (d) >> 16), \
- (a[(n)+1] = (d) >> 8), \
- (a[(n)+2] = (d) & 255)
-
-#define GET(a,n) \
- (((a)[n] << 16) | ((a)[(n)+1] << 8) | (a)[(n)+2])
-
-#define MAX_PATTERN_SIZE (1 << 24)
-
-
-#elif LINK_SIZE == 4
-
-#define PUT(a,n,d) \
- (a[n] = (d) >> 24), \
- (a[(n)+1] = (d) >> 16), \
- (a[(n)+2] = (d) >> 8), \
- (a[(n)+3] = (d) & 255)
-
-#define GET(a,n) \
- (((a)[n] << 24) | ((a)[(n)+1] << 16) | ((a)[(n)+2] << 8) | (a)[(n)+3])
-
-#define MAX_PATTERN_SIZE (1 << 30) /* Keep it positive */
-
-
-#else
-#error LINK_SIZE must be either 2, 3, or 4
-#endif
-
-
-/* Convenience macro defined in terms of the others */
-
-#define PUTINC(a,n,d) PUT(a,n,d), a += LINK_SIZE
-
-
-/* PCRE uses some other 2-byte quantities that do not change when the size of
-offsets changes. There are used for repeat counts and for other things such as
-capturing parenthesis numbers in back references. */
-
-#define PUT2(a,n,d) \
- a[n] = (d) >> 8; \
- a[(n)+1] = (d) & 255
-
-#define GET2(a,n) \
- (((a)[n] << 8) | (a)[(n)+1])
-
-#define PUT2INC(a,n,d) PUT2(a,n,d), a += 2
-
-
-/* In case there is no definition of offsetof() provided - though any proper
-Standard C system should have one. */
-
-#ifndef offsetof
-#define offsetof(p_type,field) ((size_t)&(((p_type *)0)->field))
-#endif
-
-
-/* These are the public options that can change during matching. */
-
-#define PCRE_IMS (PCRE_CASELESS|PCRE_MULTILINE|PCRE_DOTALL)
-
-/* Private options flags start at the most significant end of the four bytes,
-but skip the top bit so we can use ints for convenience without getting tangled
-with negative values. The public options defined in pcre.h start at the least
-significant end. Make sure they don't overlap, though now that we have expanded
-to four bytes, there is plenty of space. */
-
-#define PCRE_FIRSTSET 0x40000000 /* first_byte is set */
-#define PCRE_REQCHSET 0x20000000 /* req_byte is set */
-#define PCRE_STARTLINE 0x10000000 /* start after \n for multiline */
-#define PCRE_ICHANGED 0x08000000 /* i option changes within regex */
-#define PCRE_NOPARTIAL 0x04000000 /* can't use partial with this regex */
-
-/* Options for the "extra" block produced by pcre_study(). */
-
-#define PCRE_STUDY_MAPPED 0x01 /* a map of starting chars exists */
-
-/* Masks for identifying the public options which are permitted at compile
-time, run time or study time, respectively. */
-
-#define PUBLIC_OPTIONS \
- (PCRE_CASELESS|PCRE_EXTENDED|PCRE_ANCHORED|PCRE_MULTILINE| \
- PCRE_DOTALL|PCRE_DOLLAR_ENDONLY|PCRE_EXTRA|PCRE_UNGREEDY|PCRE_UTF8| \
- PCRE_NO_AUTO_CAPTURE|PCRE_NO_UTF8_CHECK|PCRE_AUTO_CALLOUT)
-
-#define PUBLIC_EXEC_OPTIONS \
- (PCRE_ANCHORED|PCRE_NOTBOL|PCRE_NOTEOL|PCRE_NOTEMPTY|PCRE_NO_UTF8_CHECK| \
- PCRE_PARTIAL)
-
-#define PUBLIC_STUDY_OPTIONS 0 /* None defined */
-
-/* Magic number to provide a small check against being handed junk. */
-
-#define MAGIC_NUMBER 0x50435245UL /* 'PCRE' */
-
-/* Negative values for the firstchar and reqchar variables */
-
-#define REQ_UNSET (-2)
-#define REQ_NONE (-1)
-
-/* Flags added to firstbyte or reqbyte; a "non-literal" item is either a
-variable-length repeat, or a anything other than literal characters. */
-
-#define REQ_CASELESS 0x0100 /* indicates caselessness */
-#define REQ_VARY 0x0200 /* reqbyte followed non-literal item */
-
-/* Miscellaneous definitions */
-
-typedef int BOOL;
-
-#define FALSE 0
-#define TRUE 1
-
-/* Escape items that are just an encoding of a particular data value. Note that
-ESC_n is defined as yet another macro, which is set in config.h to either \n
-(the default) or \r (which some people want). */
-
-#ifndef ESC_e
-#define ESC_e 27
-#endif
-
-#ifndef ESC_f
-#define ESC_f '\f'
-#endif
-
-#ifndef ESC_n
-#define ESC_n NEWLINE
-#endif
-
-#ifndef ESC_r
-#define ESC_r '\r'
-#endif
-
-/* We can't officially use ESC_t because it is a POSIX reserved identifier
-(presumably because of all the others like size_t). */
-
-#ifndef ESC_tee
-#define ESC_tee '\t'
-#endif
-
-/* These are escaped items that aren't just an encoding of a particular data
-value such as \n. They must have non-zero values, as check_escape() returns
-their negation. Also, they must appear in the same order as in the opcode
-definitions below, up to ESC_z. There's a dummy for OP_ANY because it
-corresponds to "." rather than an escape sequence. The final one must be
-ESC_REF as subsequent values are used for \1, \2, \3, etc. There is are two
-tests in the code for an escape greater than ESC_b and less than ESC_Z to
-detect the types that may be repeated. These are the types that consume
-characters. If any new escapes are put in between that don't consume a
-character, that code will have to change. */
-
-enum { ESC_A = 1, ESC_G, ESC_B, ESC_b, ESC_D, ESC_d, ESC_S, ESC_s, ESC_W,
- ESC_w, ESC_dum1, ESC_C, ESC_P, ESC_p, ESC_X, ESC_Z, ESC_z, ESC_E,
- ESC_Q, ESC_REF };
-
-/* Flag bits and data types for the extended class (OP_XCLASS) for classes that
-contain UTF-8 characters with values greater than 255. */
-
-#define XCL_NOT 0x01 /* Flag: this is a negative class */
-#define XCL_MAP 0x02 /* Flag: a 32-byte map is present */
-
-#define XCL_END 0 /* Marks end of individual items */
-#define XCL_SINGLE 1 /* Single item (one multibyte char) follows */
-#define XCL_RANGE 2 /* A range (two multibyte chars) follows */
-#define XCL_PROP 3 /* Unicode property (one property code) follows */
-#define XCL_NOTPROP 4 /* Unicode inverted property (ditto) */
-
-
-/* Opcode table: OP_BRA must be last, as all values >= it are used for brackets
-that extract substrings. Starting from 1 (i.e. after OP_END), the values up to
-OP_EOD must correspond in order to the list of escapes immediately above.
-Note that whenever this list is updated, the two macro definitions that follow
-must also be updated to match. */
-
-enum {
- OP_END, /* 0 End of pattern */
-
- /* Values corresponding to backslashed metacharacters */
-
- OP_SOD, /* 1 Start of data: \A */
- OP_SOM, /* 2 Start of match (subject + offset): \G */
- OP_NOT_WORD_BOUNDARY, /* 3 \B */
- OP_WORD_BOUNDARY, /* 4 \b */
- OP_NOT_DIGIT, /* 5 \D */
- OP_DIGIT, /* 6 \d */
- OP_NOT_WHITESPACE, /* 7 \S */
- OP_WHITESPACE, /* 8 \s */
- OP_NOT_WORDCHAR, /* 9 \W */
- OP_WORDCHAR, /* 10 \w */
- OP_ANY, /* 11 Match any character */
- OP_ANYBYTE, /* 12 Match any byte (\C); different to OP_ANY for UTF-8 */
- OP_NOTPROP, /* 13 \P (not Unicode property) */
- OP_PROP, /* 14 \p (Unicode property) */
- OP_EXTUNI, /* 15 \X (extended Unicode sequence */
- OP_EODN, /* 16 End of data or \n at end of data: \Z. */
- OP_EOD, /* 17 End of data: \z */
-
- OP_OPT, /* 18 Set runtime options */
- OP_CIRC, /* 19 Start of line - varies with multiline switch */
- OP_DOLL, /* 20 End of line - varies with multiline switch */
- OP_CHAR, /* 21 Match one character, casefully */
- OP_CHARNC, /* 22 Match one character, caselessly */
- OP_NOT, /* 23 Match anything but the following char */
-
- OP_STAR, /* 24 The maximizing and minimizing versions of */
- OP_MINSTAR, /* 25 all these opcodes must come in pairs, with */
- OP_PLUS, /* 26 the minimizing one second. */
- OP_MINPLUS, /* 27 This first set applies to single characters */
- OP_QUERY, /* 28 */
- OP_MINQUERY, /* 29 */
- OP_UPTO, /* 30 From 0 to n matches */
- OP_MINUPTO, /* 31 */
- OP_EXACT, /* 32 Exactly n matches */
-
- OP_NOTSTAR, /* 33 The maximizing and minimizing versions of */
- OP_NOTMINSTAR, /* 34 all these opcodes must come in pairs, with */
- OP_NOTPLUS, /* 35 the minimizing one second. */
- OP_NOTMINPLUS, /* 36 This set applies to "not" single characters */
- OP_NOTQUERY, /* 37 */
- OP_NOTMINQUERY, /* 38 */
- OP_NOTUPTO, /* 39 From 0 to n matches */
- OP_NOTMINUPTO, /* 40 */
- OP_NOTEXACT, /* 41 Exactly n matches */
-
- OP_TYPESTAR, /* 42 The maximizing and minimizing versions of */
- OP_TYPEMINSTAR, /* 43 all these opcodes must come in pairs, with */
- OP_TYPEPLUS, /* 44 the minimizing one second. These codes must */
- OP_TYPEMINPLUS, /* 45 be in exactly the same order as those above. */
- OP_TYPEQUERY, /* 46 This set applies to character types such as \d */
- OP_TYPEMINQUERY, /* 47 */
- OP_TYPEUPTO, /* 48 From 0 to n matches */
- OP_TYPEMINUPTO, /* 49 */
- OP_TYPEEXACT, /* 50 Exactly n matches */
-
- OP_CRSTAR, /* 51 The maximizing and minimizing versions of */
- OP_CRMINSTAR, /* 52 all these opcodes must come in pairs, with */
- OP_CRPLUS, /* 53 the minimizing one second. These codes must */
- OP_CRMINPLUS, /* 54 be in exactly the same order as those above. */
- OP_CRQUERY, /* 55 These are for character classes and back refs */
- OP_CRMINQUERY, /* 56 */
- OP_CRRANGE, /* 57 These are different to the three sets above. */
- OP_CRMINRANGE, /* 58 */
-
- OP_CLASS, /* 59 Match a character class, chars < 256 only */
- OP_NCLASS, /* 60 Same, but the bitmap was created from a negative
- class - the difference is relevant only when a UTF-8
- character > 255 is encountered. */
-
- OP_XCLASS, /* 61 Extended class for handling UTF-8 chars within the
- class. This does both positive and negative. */
-
- OP_REF, /* 62 Match a back reference */
- OP_RECURSE, /* 63 Match a numbered subpattern (possibly recursive) */
- OP_CALLOUT, /* 64 Call out to external function if provided */
-
- OP_ALT, /* 65 Start of alternation */
- OP_KET, /* 66 End of group that doesn't have an unbounded repeat */
- OP_KETRMAX, /* 67 These two must remain together and in this */
- OP_KETRMIN, /* 68 order. They are for groups the repeat for ever. */
-
- /* The assertions must come before ONCE and COND */
-
- OP_ASSERT, /* 69 Positive lookahead */
- OP_ASSERT_NOT, /* 70 Negative lookahead */
- OP_ASSERTBACK, /* 71 Positive lookbehind */
- OP_ASSERTBACK_NOT, /* 72 Negative lookbehind */
- OP_REVERSE, /* 73 Move pointer back - used in lookbehind assertions */
-
- /* ONCE and COND must come after the assertions, with ONCE first, as there's
- a test for >= ONCE for a subpattern that isn't an assertion. */
-
- OP_ONCE, /* 74 Once matched, don't back up into the subpattern */
- OP_COND, /* 75 Conditional group */
- OP_CREF, /* 76 Used to hold an extraction string number (cond ref) */
-
- OP_BRAZERO, /* 77 These two must remain together and in this */
- OP_BRAMINZERO, /* 78 order. */
-
- OP_BRANUMBER, /* 79 Used for extracting brackets whose number is greater
- than can fit into an opcode. */
-
- OP_BRA /* 80 This and greater values are used for brackets that
- extract substrings up to EXTRACT_BASIC_MAX. After
- that, use is made of OP_BRANUMBER. */
-};
-
-/* WARNING WARNING WARNING: There is an implicit assumption in pcre.c and
-study.c that all opcodes are less than 128 in value. This makes handling UTF-8
-character sequences easier. */
-
-/* The highest extraction number before we have to start using additional
-bytes. (Originally PCRE didn't have support for extraction counts highter than
-this number.) The value is limited by the number of opcodes left after OP_BRA,
-i.e. 255 - OP_BRA. We actually set it a bit lower to leave room for additional
-opcodes. */
-
-#define EXTRACT_BASIC_MAX 100
-
-
-/* This macro defines textual names for all the opcodes. There are used only
-for debugging, in pcre.c when DEBUG is defined, and also in pcretest.c. The
-macro is referenced only in printint.c. */
-
-#define OP_NAME_LIST \
- "End", "\\A", "\\G", "\\B", "\\b", "\\D", "\\d", \
- "\\S", "\\s", "\\W", "\\w", "Any", "Anybyte", \
- "notprop", "prop", "extuni", \
- "\\Z", "\\z", \
- "Opt", "^", "$", "char", "charnc", "not", \
- "*", "*?", "+", "+?", "?", "??", "{", "{", "{", \
- "*", "*?", "+", "+?", "?", "??", "{", "{", "{", \
- "*", "*?", "+", "+?", "?", "??", "{", "{", "{", \
- "*", "*?", "+", "+?", "?", "??", "{", "{", \
- "class", "nclass", "xclass", "Ref", "Recurse", "Callout", \
- "Alt", "Ket", "KetRmax", "KetRmin", "Assert", "Assert not", \
- "AssertB", "AssertB not", "Reverse", "Once", "Cond", "Cond ref",\
- "Brazero", "Braminzero", "Branumber", "Bra"
-
-
-/* This macro defines the length of fixed length operations in the compiled
-regex. The lengths are used when searching for specific things, and also in the
-debugging printing of a compiled regex. We use a macro so that it can be
-incorporated both into pcre.c and pcretest.c without being publicly exposed.
-
-As things have been extended, some of these are no longer fixed lenths, but are
-minima instead. For example, the length of a single-character repeat may vary
-in UTF-8 mode. The code that uses this table must know about such things. */
-
-#define OP_LENGTHS \
- 1, /* End */ \
- 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, /* \A, \G, \B, \B, \D, \d, \S, \s, \W, \w */ \
- 1, 1, /* Any, Anybyte */ \
- 2, 2, 1, /* NOTPROP, PROP, EXTUNI */ \
- 1, 1, 2, 1, 1, /* \Z, \z, Opt, ^, $ */ \
- 2, /* Char - the minimum length */ \
- 2, /* Charnc - the minimum length */ \
- 2, /* not */ \
- /* Positive single-char repeats ** These are */ \
- 2, 2, 2, 2, 2, 2, /* *, *?, +, +?, ?, ?? ** minima in */ \
- 4, 4, 4, /* upto, minupto, exact ** UTF-8 mode */ \
- /* Negative single-char repeats - only for chars < 256 */ \
- 2, 2, 2, 2, 2, 2, /* NOT *, *?, +, +?, ?, ?? */ \
- 4, 4, 4, /* NOT upto, minupto, exact */ \
- /* Positive type repeats */ \
- 2, 2, 2, 2, 2, 2, /* Type *, *?, +, +?, ?, ?? */ \
- 4, 4, 4, /* Type upto, minupto, exact */ \
- /* Character class & ref repeats */ \
- 1, 1, 1, 1, 1, 1, /* *, *?, +, +?, ?, ?? */ \
- 5, 5, /* CRRANGE, CRMINRANGE */ \
- 33, /* CLASS */ \
- 33, /* NCLASS */ \
- 0, /* XCLASS - variable length */ \
- 3, /* REF */ \
- 1+LINK_SIZE, /* RECURSE */ \
- 2+2*LINK_SIZE, /* CALLOUT */ \
- 1+LINK_SIZE, /* Alt */ \
- 1+LINK_SIZE, /* Ket */ \
- 1+LINK_SIZE, /* KetRmax */ \
- 1+LINK_SIZE, /* KetRmin */ \
- 1+LINK_SIZE, /* Assert */ \
- 1+LINK_SIZE, /* Assert not */ \
- 1+LINK_SIZE, /* Assert behind */ \
- 1+LINK_SIZE, /* Assert behind not */ \
- 1+LINK_SIZE, /* Reverse */ \
- 1+LINK_SIZE, /* Once */ \
- 1+LINK_SIZE, /* COND */ \
- 3, /* CREF */ \
- 1, 1, /* BRAZERO, BRAMINZERO */ \
- 3, /* BRANUMBER */ \
- 1+LINK_SIZE /* BRA */ \
-
-
-/* A magic value for OP_CREF to indicate the "in recursion" condition. */
-
-#define CREF_RECURSE 0xffff
-
-/* The texts of compile-time error messages are defined as macros here so that
-they can be accessed by the POSIX wrapper and converted into error codes. Yes,
-I could have used error codes in the first place, but didn't feel like changing
-just to accommodate the POSIX wrapper. */
-
-#define ERR1 "\\ at end of pattern"
-#define ERR2 "\\c at end of pattern"
-#define ERR3 "unrecognized character follows \\"
-#define ERR4 "numbers out of order in {} quantifier"
-#define ERR5 "number too big in {} quantifier"
-#define ERR6 "missing terminating ] for character class"
-#define ERR7 "invalid escape sequence in character class"
-#define ERR8 "range out of order in character class"
-#define ERR9 "nothing to repeat"
-#define ERR10 "operand of unlimited repeat could match the empty string"
-#define ERR11 "internal error: unexpected repeat"
-#define ERR12 "unrecognized character after (?"
-#define ERR13 "POSIX named classes are supported only within a class"
-#define ERR14 "missing )"
-#define ERR15 "reference to non-existent subpattern"
-#define ERR16 "erroffset passed as NULL"
-#define ERR17 "unknown option bit(s) set"
-#define ERR18 "missing ) after comment"
-#define ERR19 "parentheses nested too deeply"
-#define ERR20 "regular expression too large"
-#define ERR21 "failed to get memory"
-#define ERR22 "unmatched parentheses"
-#define ERR23 "internal error: code overflow"
-#define ERR24 "unrecognized character after (?<"
-#define ERR25 "lookbehind assertion is not fixed length"
-#define ERR26 "malformed number after (?("
-#define ERR27 "conditional group contains more than two branches"
-#define ERR28 "assertion expected after (?("
-#define ERR29 "(?R or (?digits must be followed by )"
-#define ERR30 "unknown POSIX class name"
-#define ERR31 "POSIX collating elements are not supported"
-#define ERR32 "this version of PCRE is not compiled with PCRE_UTF8 support"
-#define ERR33 "spare error"
-#define ERR34 "character value in \\x{...} sequence is too large"
-#define ERR35 "invalid condition (?(0)"
-#define ERR36 "\\C not allowed in lookbehind assertion"
-#define ERR37 "PCRE does not support \\L, \\l, \\N, \\U, or \\u"
-#define ERR38 "number after (?C is > 255"
-#define ERR39 "closing ) for (?C expected"
-#define ERR40 "recursive call could loop indefinitely"
-#define ERR41 "unrecognized character after (?P"
-#define ERR42 "syntax error after (?P"
-#define ERR43 "two named groups have the same name"
-#define ERR44 "invalid UTF-8 string"
-#define ERR45 "support for \\P, \\p, and \\X has not been compiled"
-#define ERR46 "malformed \\P or \\p sequence"
-#define ERR47 "unknown property name after \\P or \\p"
-
-/* The real format of the start of the pcre block; the index of names and the
-code vector run on as long as necessary after the end. We store an explicit
-offset to the name table so that if a regex is compiled on one host, saved, and
-then run on another where the size of pointers is different, all might still
-be well. For the case of compiled-on-4 and run-on-8, we include an extra
-pointer that is always NULL. For future-proofing, we also include a few dummy
-fields - even though you can never get this planning right!
-
-NOTE NOTE NOTE:
-Because people can now save and re-use compiled patterns, any additions to this
-structure should be made at the end, and something earlier (e.g. a new
-flag in the options or one of the dummy fields) should indicate that the new
-fields are present. Currently PCRE always sets the dummy fields to zero.
-NOTE NOTE NOTE:
-*/
-
-typedef struct real_pcre {
- pcre_uint32 magic_number;
- pcre_uint32 size; /* Total that was malloced */
- pcre_uint32 options;
- pcre_uint32 dummy1; /* For future use, maybe */
-
- pcre_uint16 top_bracket;
- pcre_uint16 top_backref;
- pcre_uint16 first_byte;
- pcre_uint16 req_byte;
- pcre_uint16 name_table_offset; /* Offset to name table that follows */
- pcre_uint16 name_entry_size; /* Size of any name items */
- pcre_uint16 name_count; /* Number of name items */
- pcre_uint16 dummy2; /* For future use, maybe */
-
- const unsigned char *tables; /* Pointer to tables or NULL for std */
- const unsigned char *nullpad; /* NULL padding */
-} real_pcre;
-
-/* The format of the block used to store data from pcre_study(). The same
-remark (see NOTE above) about extending this structure applies. */
-
-typedef struct pcre_study_data {
- pcre_uint32 size; /* Total that was malloced */
- pcre_uint32 options;
- uschar start_bits[32];
-} pcre_study_data;
-
-/* Structure for passing "static" information around between the functions
-doing the compiling, so that they are thread-safe. */
-
-typedef struct compile_data {
- const uschar *lcc; /* Points to lower casing table */
- const uschar *fcc; /* Points to case-flipping table */
- const uschar *cbits; /* Points to character type table */
- const uschar *ctypes; /* Points to table of type maps */
- const uschar *start_code; /* The start of the compiled code */
- const uschar *start_pattern; /* The start of the pattern */
- uschar *name_table; /* The name/number table */
- int names_found; /* Number of entries so far */
- int name_entry_size; /* Size of each entry */
- int top_backref; /* Maximum back reference */
- unsigned int backref_map; /* Bitmap of low back refs */
- int req_varyopt; /* "After variable item" flag for reqbyte */
- BOOL nopartial; /* Set TRUE if partial won't work */
-} compile_data;
-
-/* Structure for maintaining a chain of pointers to the currently incomplete
-branches, for testing for left recursion. */
-
-typedef struct branch_chain {
- struct branch_chain *outer;
- uschar *current;
-} branch_chain;
-
-/* Structure for items in a linked list that represents an explicit recursive
-call within the pattern. */
-
-typedef struct recursion_info {
- struct recursion_info *prevrec; /* Previous recursion record (or NULL) */
- int group_num; /* Number of group that was called */
- const uschar *after_call; /* "Return value": points after the call in the expr */
- const uschar *save_start; /* Old value of md->start_match */
- int *offset_save; /* Pointer to start of saved offsets */
- int saved_max; /* Number of saved offsets */
-} recursion_info;
-
-/* When compiling in a mode that doesn't use recursive calls to match(),
-a structure is used to remember local variables on the heap. It is defined in
-pcre.c, close to the match() function, so that it is easy to keep it in step
-with any changes of local variable. However, the pointer to the current frame
-must be saved in some "static" place over a longjmp(). We declare the
-structure here so that we can put a pointer in the match_data structure.
-NOTE: This isn't used for a "normal" compilation of pcre. */
-
-struct heapframe;
-
-/* Structure for passing "static" information around between the functions
-doing the matching, so that they are thread-safe. */
-
-typedef struct match_data {
- unsigned long int match_call_count; /* As it says */
- unsigned long int match_limit;/* As it says */
- int *offset_vector; /* Offset vector */
- int offset_end; /* One past the end */
- int offset_max; /* The maximum usable for return data */
- const uschar *lcc; /* Points to lower casing table */
- const uschar *ctypes; /* Points to table of type maps */
- BOOL offset_overflow; /* Set if too many extractions */
- BOOL notbol; /* NOTBOL flag */
- BOOL noteol; /* NOTEOL flag */
- BOOL utf8; /* UTF8 flag */
- BOOL endonly; /* Dollar not before final \n */
- BOOL notempty; /* Empty string match not wanted */
- BOOL partial; /* PARTIAL flag */
- BOOL hitend; /* Hit the end of the subject at some point */
- const uschar *start_code; /* For use when recursing */
- const uschar *start_subject; /* Start of the subject string */
- const uschar *end_subject; /* End of the subject string */
- const uschar *start_match; /* Start of this match attempt */
- const uschar *end_match_ptr; /* Subject position at end match */
- int end_offset_top; /* Highwater mark at end of match */
- int capture_last; /* Most recent capture number */
- int start_offset; /* The start offset value */
- recursion_info *recursive; /* Linked list of recursion data */
- void *callout_data; /* To pass back to callouts */
- struct heapframe *thisframe; /* Used only when compiling for no recursion */
-} match_data;
-
-/* Bit definitions for entries in the pcre_ctypes table. */
-
-#define ctype_space 0x01
-#define ctype_letter 0x02
-#define ctype_digit 0x04
-#define ctype_xdigit 0x08
-#define ctype_word 0x10 /* alphameric or '_' */
-#define ctype_meta 0x80 /* regexp meta char or zero (end pattern) */
-
-/* Offsets for the bitmap tables in pcre_cbits. Each table contains a set
-of bits for a class map. Some classes are built by combining these tables. */
-
-#define cbit_space 0 /* [:space:] or \s */
-#define cbit_xdigit 32 /* [:xdigit:] */
-#define cbit_digit 64 /* [:digit:] or \d */
-#define cbit_upper 96 /* [:upper:] */
-#define cbit_lower 128 /* [:lower:] */
-#define cbit_word 160 /* [:word:] or \w */
-#define cbit_graph 192 /* [:graph:] */
-#define cbit_print 224 /* [:print:] */
-#define cbit_punct 256 /* [:punct:] */
-#define cbit_cntrl 288 /* [:cntrl:] */
-#define cbit_length 320 /* Length of the cbits table */
-
-/* Offsets of the various tables from the base tables pointer, and
-total length. */
-
-#define lcc_offset 0
-#define fcc_offset 256
-#define cbits_offset 512
-#define ctypes_offset (cbits_offset + cbit_length)
-#define tables_length (ctypes_offset + 256)
-
-/* End of internal.h */
+++ /dev/null
-/* $Cambridge: exim/src/src/pcre/pcre.c,v 1.2 2005/06/15 08:57:10 ph10 Exp $ */
-
-/*************************************************
-* Perl-Compatible Regular Expressions *
-*************************************************/
-
-/*
-This is a library of functions to support regular expressions whose syntax
-and semantics are as close as possible to those of the Perl 5 language. See
-the file Tech.Notes for some information on the internals.
-
-Written by: Philip Hazel <ph10@cam.ac.uk>
-
- Copyright (c) 1997-2004 University of Cambridge
-
------------------------------------------------------------------------------
-Redistribution and use in source and binary forms, with or without
-modification, are permitted provided that the following conditions are met:
-
- * Redistributions of source code must retain the above copyright notice,
- this list of conditions and the following disclaimer.
-
- * Redistributions in binary form must reproduce the above copyright
- notice, this list of conditions and the following disclaimer in the
- documentation and/or other materials provided with the distribution.
-
- * Neither the name of the University of Cambridge nor the names of its
- contributors may be used to endorse or promote products derived from
- this software without specific prior written permission.
-
-THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
-AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
-IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
-ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
-LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
-CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
-SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
-INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
-CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
-ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
-POSSIBILITY OF SUCH DAMAGE.
------------------------------------------------------------------------------
-*/
-
-
-/* Define DEBUG to get debugging output on stdout. */
-/* #define DEBUG */
-
-/* Use a macro for debugging printing, 'cause that eliminates the use of #ifdef
-inline, and there are *still* stupid compilers about that don't like indented
-pre-processor statements. I suppose it's only been 10 years... */
-
-#ifdef DEBUG
-#define DPRINTF(p) printf p
-#else
-#define DPRINTF(p) /*nothing*/
-#endif
-
-/* Include the internals header, which itself includes "config.h", the Standard
-C headers, and the external pcre header. */
-
-#include "internal.h"
-
-/* If Unicode Property support is wanted, include a private copy of the
-function that does it, and the table that translates names to numbers. */
-
-#ifdef SUPPORT_UCP
-#include "ucp.c"
-#include "ucptypetable.c"
-#endif
-
-/* Maximum number of items on the nested bracket stacks at compile time. This
-applies to the nesting of all kinds of parentheses. It does not limit
-un-nested, non-capturing parentheses. This number can be made bigger if
-necessary - it is used to dimension one int and one unsigned char vector at
-compile time. */
-
-#define BRASTACK_SIZE 200
-
-
-/* Maximum number of ints of offset to save on the stack for recursive calls.
-If the offset vector is bigger, malloc is used. This should be a multiple of 3,
-because the offset vector is always a multiple of 3 long. */
-
-#define REC_STACK_SAVE_MAX 30
-
-
-/* The maximum remaining length of subject we are prepared to search for a
-req_byte match. */
-
-#define REQ_BYTE_MAX 1000
-
-
-/* Table of sizes for the fixed-length opcodes. It's defined in a macro so that
-the definition is next to the definition of the opcodes in internal.h. */
-
-static const uschar OP_lengths[] = { OP_LENGTHS };
-
-/* Min and max values for the common repeats; for the maxima, 0 => infinity */
-
-static const char rep_min[] = { 0, 0, 1, 1, 0, 0 };
-static const char rep_max[] = { 0, 0, 0, 0, 1, 1 };
-
-/* Table for handling escaped characters in the range '0'-'z'. Positive returns
-are simple data values; negative values are for special things like \d and so
-on. Zero means further processing is needed (for things like \x), or the escape
-is invalid. */
-
-#if !EBCDIC /* This is the "normal" table for ASCII systems */
-static const short int escapes[] = {
- 0, 0, 0, 0, 0, 0, 0, 0, /* 0 - 7 */
- 0, 0, ':', ';', '<', '=', '>', '?', /* 8 - ? */
- '@', -ESC_A, -ESC_B, -ESC_C, -ESC_D, -ESC_E, 0, -ESC_G, /* @ - G */
- 0, 0, 0, 0, 0, 0, 0, 0, /* H - O */
--ESC_P, -ESC_Q, 0, -ESC_S, 0, 0, 0, -ESC_W, /* P - W */
--ESC_X, 0, -ESC_Z, '[', '\\', ']', '^', '_', /* X - _ */
- '`', 7, -ESC_b, 0, -ESC_d, ESC_e, ESC_f, 0, /* ` - g */
- 0, 0, 0, 0, 0, 0, ESC_n, 0, /* h - o */
--ESC_p, 0, ESC_r, -ESC_s, ESC_tee, 0, 0, -ESC_w, /* p - w */
- 0, 0, -ESC_z /* x - z */
-};
-
-#else /* This is the "abnormal" table for EBCDIC systems */
-static const short int escapes[] = {
-/* 48 */ 0, 0, 0, '.', '<', '(', '+', '|',
-/* 50 */ '&', 0, 0, 0, 0, 0, 0, 0,
-/* 58 */ 0, 0, '!', '$', '*', ')', ';', '~',
-/* 60 */ '-', '/', 0, 0, 0, 0, 0, 0,
-/* 68 */ 0, 0, '|', ',', '%', '_', '>', '?',
-/* 70 */ 0, 0, 0, 0, 0, 0, 0, 0,
-/* 78 */ 0, '`', ':', '#', '@', '\'', '=', '"',
-/* 80 */ 0, 7, -ESC_b, 0, -ESC_d, ESC_e, ESC_f, 0,
-/* 88 */ 0, 0, 0, '{', 0, 0, 0, 0,
-/* 90 */ 0, 0, 0, 'l', 0, ESC_n, 0, -ESC_p,
-/* 98 */ 0, ESC_r, 0, '}', 0, 0, 0, 0,
-/* A0 */ 0, '~', -ESC_s, ESC_tee, 0, 0, -ESC_w, 0,
-/* A8 */ 0,-ESC_z, 0, 0, 0, '[', 0, 0,
-/* B0 */ 0, 0, 0, 0, 0, 0, 0, 0,
-/* B8 */ 0, 0, 0, 0, 0, ']', '=', '-',
-/* C0 */ '{',-ESC_A, -ESC_B, -ESC_C, -ESC_D,-ESC_E, 0, -ESC_G,
-/* C8 */ 0, 0, 0, 0, 0, 0, 0, 0,
-/* D0 */ '}', 0, 0, 0, 0, 0, 0, -ESC_P,
-/* D8 */-ESC_Q, 0, 0, 0, 0, 0, 0, 0,
-/* E0 */ '\\', 0, -ESC_S, 0, 0, 0, -ESC_W, -ESC_X,
-/* E8 */ 0,-ESC_Z, 0, 0, 0, 0, 0, 0,
-/* F0 */ 0, 0, 0, 0, 0, 0, 0, 0,
-/* F8 */ 0, 0, 0, 0, 0, 0, 0, 0
-};
-#endif
-
-
-/* Tables of names of POSIX character classes and their lengths. The list is
-terminated by a zero length entry. The first three must be alpha, upper, lower,
-as this is assumed for handling case independence. */
-
-static const char *const posix_names[] = {
- "alpha", "lower", "upper",
- "alnum", "ascii", "blank", "cntrl", "digit", "graph",
- "print", "punct", "space", "word", "xdigit" };
-
-static const uschar posix_name_lengths[] = {
- 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 4, 6, 0 };
-
-/* Table of class bit maps for each POSIX class; up to three may be combined
-to form the class. The table for [:blank:] is dynamically modified to remove
-the vertical space characters. */
-
-static const int posix_class_maps[] = {
- cbit_lower, cbit_upper, -1, /* alpha */
- cbit_lower, -1, -1, /* lower */
- cbit_upper, -1, -1, /* upper */
- cbit_digit, cbit_lower, cbit_upper, /* alnum */
- cbit_print, cbit_cntrl, -1, /* ascii */
- cbit_space, -1, -1, /* blank - a GNU extension */
- cbit_cntrl, -1, -1, /* cntrl */
- cbit_digit, -1, -1, /* digit */
- cbit_graph, -1, -1, /* graph */
- cbit_print, -1, -1, /* print */
- cbit_punct, -1, -1, /* punct */
- cbit_space, -1, -1, /* space */
- cbit_word, -1, -1, /* word - a Perl extension */
- cbit_xdigit,-1, -1 /* xdigit */
-};
-
-/* Table to identify digits and hex digits. This is used when compiling
-patterns. Note that the tables in chartables are dependent on the locale, and
-may mark arbitrary characters as digits - but the PCRE compiling code expects
-to handle only 0-9, a-z, and A-Z as digits when compiling. That is why we have
-a private table here. It costs 256 bytes, but it is a lot faster than doing
-character value tests (at least in some simple cases I timed), and in some
-applications one wants PCRE to compile efficiently as well as match
-efficiently.
-
-For convenience, we use the same bit definitions as in chartables:
-
- 0x04 decimal digit
- 0x08 hexadecimal digit
-
-Then we can use ctype_digit and ctype_xdigit in the code. */
-
-#if !EBCDIC /* This is the "normal" case, for ASCII systems */
-static const unsigned char digitab[] =
- {
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 0- 7 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 8- 15 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 16- 23 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 24- 31 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* - ' */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* ( - / */
- 0x0c,0x0c,0x0c,0x0c,0x0c,0x0c,0x0c,0x0c, /* 0 - 7 */
- 0x0c,0x0c,0x00,0x00,0x00,0x00,0x00,0x00, /* 8 - ? */
- 0x00,0x08,0x08,0x08,0x08,0x08,0x08,0x00, /* @ - G */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* H - O */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* P - W */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* X - _ */
- 0x00,0x08,0x08,0x08,0x08,0x08,0x08,0x00, /* ` - g */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* h - o */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* p - w */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* x -127 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 128-135 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 136-143 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 144-151 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 152-159 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 160-167 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 168-175 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 176-183 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 184-191 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 192-199 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 200-207 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 208-215 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 216-223 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 224-231 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 232-239 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 240-247 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00};/* 248-255 */
-
-#else /* This is the "abnormal" case, for EBCDIC systems */
-static const unsigned char digitab[] =
- {
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 0- 7 0 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 8- 15 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 16- 23 10 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 24- 31 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 32- 39 20 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 40- 47 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 48- 55 30 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 56- 63 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* - 71 40 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 72- | */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* & - 87 50 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 88- ¬ */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* - -103 60 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 104- ? */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 112-119 70 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 120- " */
- 0x00,0x08,0x08,0x08,0x08,0x08,0x08,0x00, /* 128- g 80 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* h -143 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 144- p 90 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* q -159 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 160- x A0 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* y -175 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* ^ -183 B0 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 184-191 */
- 0x00,0x08,0x08,0x08,0x08,0x08,0x08,0x00, /* { - G C0 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* H -207 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* } - P D0 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* Q -223 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* \ - X E0 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* Y -239 */
- 0x0c,0x0c,0x0c,0x0c,0x0c,0x0c,0x0c,0x0c, /* 0 - 7 F0 */
- 0x0c,0x0c,0x00,0x00,0x00,0x00,0x00,0x00};/* 8 -255 */
-
-static const unsigned char ebcdic_chartab[] = { /* chartable partial dup */
- 0x80,0x00,0x00,0x00,0x00,0x01,0x00,0x00, /* 0- 7 */
- 0x00,0x00,0x00,0x00,0x01,0x01,0x00,0x00, /* 8- 15 */
- 0x00,0x00,0x00,0x00,0x00,0x01,0x00,0x00, /* 16- 23 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 24- 31 */
- 0x00,0x00,0x00,0x00,0x00,0x01,0x00,0x00, /* 32- 39 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 40- 47 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 48- 55 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 56- 63 */
- 0x01,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* - 71 */
- 0x00,0x00,0x00,0x80,0x00,0x80,0x80,0x80, /* 72- | */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* & - 87 */
- 0x00,0x00,0x00,0x80,0x80,0x80,0x00,0x00, /* 88- ¬ */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* - -103 */
- 0x00,0x00,0x00,0x00,0x00,0x10,0x00,0x80, /* 104- ? */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 112-119 */
- 0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* 120- " */
- 0x00,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x12, /* 128- g */
- 0x12,0x12,0x00,0x00,0x00,0x00,0x00,0x00, /* h -143 */
- 0x00,0x12,0x12,0x12,0x12,0x12,0x12,0x12, /* 144- p */
- 0x12,0x12,0x00,0x00,0x00,0x00,0x00,0x00, /* q -159 */
- 0x00,0x00,0x12,0x12,0x12,0x12,0x12,0x12, /* 160- x */
- 0x12,0x12,0x00,0x00,0x00,0x00,0x00,0x00, /* y -175 */
- 0x80,0x00,0x00,0x00,0x00,0x00,0x00,0x00, /* ^ -183 */
- 0x00,0x00,0x80,0x00,0x00,0x00,0x00,0x00, /* 184-191 */
- 0x80,0x1a,0x1a,0x1a,0x1a,0x1a,0x1a,0x12, /* { - G */
- 0x12,0x12,0x00,0x00,0x00,0x00,0x00,0x00, /* H -207 */
- 0x00,0x12,0x12,0x12,0x12,0x12,0x12,0x12, /* } - P */
- 0x12,0x12,0x00,0x00,0x00,0x00,0x00,0x00, /* Q -223 */
- 0x00,0x00,0x12,0x12,0x12,0x12,0x12,0x12, /* \ - X */
- 0x12,0x12,0x00,0x00,0x00,0x00,0x00,0x00, /* Y -239 */
- 0x1c,0x1c,0x1c,0x1c,0x1c,0x1c,0x1c,0x1c, /* 0 - 7 */
- 0x1c,0x1c,0x00,0x00,0x00,0x00,0x00,0x00};/* 8 -255 */
-#endif
-
-
-/* Definition to allow mutual recursion */
-
-static BOOL
- compile_regex(int, int, int *, uschar **, const uschar **, const char **,
- BOOL, int, int *, int *, branch_chain *, compile_data *);
-
-/* Structure for building a chain of data that actually lives on the
-stack, for holding the values of the subject pointer at the start of each
-subpattern, so as to detect when an empty string has been matched by a
-subpattern - to break infinite loops. When NO_RECURSE is set, these blocks
-are on the heap, not on the stack. */
-
-typedef struct eptrblock {
- struct eptrblock *epb_prev;
- const uschar *epb_saved_eptr;
-} eptrblock;
-
-/* Flag bits for the match() function */
-
-#define match_condassert 0x01 /* Called to check a condition assertion */
-#define match_isgroup 0x02 /* Set if start of bracketed group */
-
-/* Non-error returns from the match() function. Error returns are externally
-defined PCRE_ERROR_xxx codes, which are all negative. */
-
-#define MATCH_MATCH 1
-#define MATCH_NOMATCH 0
-
-
-
-/*************************************************
-* Global variables *
-*************************************************/
-
-/* PCRE is thread-clean and doesn't use any global variables in the normal
-sense. However, it calls memory allocation and free functions via the four
-indirections below, and it can optionally do callouts. These values can be
-changed by the caller, but are shared between all threads. However, when
-compiling for Virtual Pascal, things are done differently (see pcre.in). */
-
-#ifndef VPCOMPAT
-#ifdef __cplusplus
-extern "C" void *(*pcre_malloc)(size_t) = malloc;
-extern "C" void (*pcre_free)(void *) = free;
-extern "C" void *(*pcre_stack_malloc)(size_t) = malloc;
-extern "C" void (*pcre_stack_free)(void *) = free;
-extern "C" int (*pcre_callout)(pcre_callout_block *) = NULL;
-#else
-void *(*pcre_malloc)(size_t) = malloc;
-void (*pcre_free)(void *) = free;
-void *(*pcre_stack_malloc)(size_t) = malloc;
-void (*pcre_stack_free)(void *) = free;
-int (*pcre_callout)(pcre_callout_block *) = NULL;
-#endif
-#endif
-
-
-/*************************************************
-* Macros and tables for character handling *
-*************************************************/
-
-/* When UTF-8 encoding is being used, a character is no longer just a single
-byte. The macros for character handling generate simple sequences when used in
-byte-mode, and more complicated ones for UTF-8 characters. */
-
-#ifndef SUPPORT_UTF8
-#define GETCHAR(c, eptr) c = *eptr;
-#define GETCHARINC(c, eptr) c = *eptr++;
-#define GETCHARINCTEST(c, eptr) c = *eptr++;
-#define GETCHARLEN(c, eptr, len) c = *eptr;
-#define BACKCHAR(eptr)
-
-#else /* SUPPORT_UTF8 */
-
-/* Get the next UTF-8 character, not advancing the pointer. This is called when
-we know we are in UTF-8 mode. */
-
-#define GETCHAR(c, eptr) \
- c = *eptr; \
- if ((c & 0xc0) == 0xc0) \
- { \
- int gcii; \
- int gcaa = utf8_table4[c & 0x3f]; /* Number of additional bytes */ \
- int gcss = 6*gcaa; \
- c = (c & utf8_table3[gcaa]) << gcss; \
- for (gcii = 1; gcii <= gcaa; gcii++) \
- { \
- gcss -= 6; \
- c |= (eptr[gcii] & 0x3f) << gcss; \
- } \
- }
-
-/* Get the next UTF-8 character, advancing the pointer. This is called when we
-know we are in UTF-8 mode. */
-
-#define GETCHARINC(c, eptr) \
- c = *eptr++; \
- if ((c & 0xc0) == 0xc0) \
- { \
- int gcaa = utf8_table4[c & 0x3f]; /* Number of additional bytes */ \
- int gcss = 6*gcaa; \
- c = (c & utf8_table3[gcaa]) << gcss; \
- while (gcaa-- > 0) \
- { \
- gcss -= 6; \
- c |= (*eptr++ & 0x3f) << gcss; \
- } \
- }
-
-/* Get the next character, testing for UTF-8 mode, and advancing the pointer */
-
-#define GETCHARINCTEST(c, eptr) \
- c = *eptr++; \
- if (md->utf8 && (c & 0xc0) == 0xc0) \
- { \
- int gcaa = utf8_table4[c & 0x3f]; /* Number of additional bytes */ \
- int gcss = 6*gcaa; \
- c = (c & utf8_table3[gcaa]) << gcss; \
- while (gcaa-- > 0) \
- { \
- gcss -= 6; \
- c |= (*eptr++ & 0x3f) << gcss; \
- } \
- }
-
-/* Get the next UTF-8 character, not advancing the pointer, incrementing length
-if there are extra bytes. This is called when we know we are in UTF-8 mode. */
-
-#define GETCHARLEN(c, eptr, len) \
- c = *eptr; \
- if ((c & 0xc0) == 0xc0) \
- { \
- int gcii; \
- int gcaa = utf8_table4[c & 0x3f]; /* Number of additional bytes */ \
- int gcss = 6*gcaa; \
- c = (c & utf8_table3[gcaa]) << gcss; \
- for (gcii = 1; gcii <= gcaa; gcii++) \
- { \
- gcss -= 6; \
- c |= (eptr[gcii] & 0x3f) << gcss; \
- } \
- len += gcaa; \
- }
-
-/* If the pointer is not at the start of a character, move it back until
-it is. Called only in UTF-8 mode. */
-
-#define BACKCHAR(eptr) while((*eptr & 0xc0) == 0x80) eptr--;
-
-#endif
-
-
-
-/*************************************************
-* Default character tables *
-*************************************************/
-
-/* A default set of character tables is included in the PCRE binary. Its source
-is built by the maketables auxiliary program, which uses the default C ctypes
-functions, and put in the file chartables.c. These tables are used by PCRE
-whenever the caller of pcre_compile() does not provide an alternate set of
-tables. */
-
-#include "chartables.c"
-
-
-
-#ifdef SUPPORT_UTF8
-/*************************************************
-* Tables for UTF-8 support *
-*************************************************/
-
-/* These are the breakpoints for different numbers of bytes in a UTF-8
-character. */
-
-static const int utf8_table1[] =
- { 0x7f, 0x7ff, 0xffff, 0x1fffff, 0x3ffffff, 0x7fffffff};
-
-/* These are the indicator bits and the mask for the data bits to set in the
-first byte of a character, indexed by the number of additional bytes. */
-
-static const int utf8_table2[] = { 0, 0xc0, 0xe0, 0xf0, 0xf8, 0xfc};
-static const int utf8_table3[] = { 0xff, 0x1f, 0x0f, 0x07, 0x03, 0x01};
-
-/* Table of the number of extra characters, indexed by the first character
-masked with 0x3f. The highest number for a valid UTF-8 character is in fact
-0x3d. */
-
-static const uschar utf8_table4[] = {
- 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
- 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,
- 2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,
- 3,3,3,3,3,3,3,3,4,4,4,4,5,5,5,5 };
-
-
-/*************************************************
-* Convert character value to UTF-8 *
-*************************************************/
-
-/* This function takes an integer value in the range 0 - 0x7fffffff
-and encodes it as a UTF-8 character in 0 to 6 bytes.
-
-Arguments:
- cvalue the character value
- buffer pointer to buffer for result - at least 6 bytes long
-
-Returns: number of characters placed in the buffer
-*/
-
-static int
-ord2utf8(int cvalue, uschar *buffer)
-{
-register int i, j;
-for (i = 0; i < sizeof(utf8_table1)/sizeof(int); i++)
- if (cvalue <= utf8_table1[i]) break;
-buffer += i;
-for (j = i; j > 0; j--)
- {
- *buffer-- = 0x80 | (cvalue & 0x3f);
- cvalue >>= 6;
- }
-*buffer = utf8_table2[i] | cvalue;
-return i + 1;
-}
-#endif
-
-
-
-/*************************************************
-* Print compiled regex *
-*************************************************/
-
-/* The code for doing this is held in a separate file that is also included in
-pcretest.c. It defines a function called print_internals(). */
-
-#ifdef DEBUG
-#include "printint.c"
-#endif
-
-
-
-/*************************************************
-* Return version string *
-*************************************************/
-
-#define STRING(a) # a
-#define XSTRING(s) STRING(s)
-
-EXPORT const char *
-pcre_version(void)
-{
-return XSTRING(PCRE_MAJOR) "." XSTRING(PCRE_MINOR) " " XSTRING(PCRE_DATE);
-}
-
-
-
-
-/*************************************************
-* Flip bytes in an integer *
-*************************************************/
-
-/* This function is called when the magic number in a regex doesn't match in
-order to flip its bytes to see if we are dealing with a pattern that was
-compiled on a host of different endianness. If so, this function is used to
-flip other byte values.
-
-Arguments:
- value the number to flip
- n the number of bytes to flip (assumed to be 2 or 4)
-
-Returns: the flipped value
-*/
-
-static long int
-byteflip(long int value, int n)
-{
-if (n == 2) return ((value & 0x00ff) << 8) | ((value & 0xff00) >> 8);
-return ((value & 0x000000ff) << 24) |
- ((value & 0x0000ff00) << 8) |
- ((value & 0x00ff0000) >> 8) |
- ((value & 0xff000000) >> 24);
-}
-
-
-
-/*************************************************
-* Test for a byte-flipped compiled regex *
-*************************************************/
-
-/* This function is called from pce_exec() and also from pcre_fullinfo(). Its
-job is to test whether the regex is byte-flipped - that is, it was compiled on
-a system of opposite endianness. The function is called only when the native
-MAGIC_NUMBER test fails. If the regex is indeed flipped, we flip all the
-relevant values into a different data block, and return it.
-
-Arguments:
- re points to the regex
- study points to study data, or NULL
- internal_re points to a new regex block
- internal_study points to a new study block
-
-Returns: the new block if is is indeed a byte-flipped regex
- NULL if it is not
-*/
-
-static real_pcre *
-try_flipped(const real_pcre *re, real_pcre *internal_re,
- const pcre_study_data *study, pcre_study_data *internal_study)
-{
-if (byteflip(re->magic_number, sizeof(re->magic_number)) != MAGIC_NUMBER)
- return NULL;
-
-*internal_re = *re; /* To copy other fields */
-internal_re->size = byteflip(re->size, sizeof(re->size));
-internal_re->options = byteflip(re->options, sizeof(re->options));
-internal_re->top_bracket = byteflip(re->top_bracket, sizeof(re->top_bracket));
-internal_re->top_backref = byteflip(re->top_backref, sizeof(re->top_backref));
-internal_re->first_byte = byteflip(re->first_byte, sizeof(re->first_byte));
-internal_re->req_byte = byteflip(re->req_byte, sizeof(re->req_byte));
-internal_re->name_table_offset = byteflip(re->name_table_offset,
- sizeof(re->name_table_offset));
-internal_re->name_entry_size = byteflip(re->name_entry_size,
- sizeof(re->name_entry_size));
-internal_re->name_count = byteflip(re->name_count, sizeof(re->name_count));
-
-if (study != NULL)
- {
- *internal_study = *study; /* To copy other fields */
- internal_study->size = byteflip(study->size, sizeof(study->size));
- internal_study->options = byteflip(study->options, sizeof(study->options));
- }
-
-return internal_re;
-}
-
-
-
-/*************************************************
-* (Obsolete) Return info about compiled pattern *
-*************************************************/
-
-/* This is the original "info" function. It picks potentially useful data out
-of the private structure, but its interface was too rigid. It remains for
-backwards compatibility. The public options are passed back in an int - though
-the re->options field has been expanded to a long int, all the public options
-at the low end of it, and so even on 16-bit systems this will still be OK.
-Therefore, I haven't changed the API for pcre_info().
-
-Arguments:
- argument_re points to compiled code
- optptr where to pass back the options
- first_byte where to pass back the first character,
- or -1 if multiline and all branches start ^,
- or -2 otherwise
-
-Returns: number of capturing subpatterns
- or negative values on error
-*/
-
-EXPORT int
-pcre_info(const pcre *argument_re, int *optptr, int *first_byte)
-{
-real_pcre internal_re;
-const real_pcre *re = (const real_pcre *)argument_re;
-if (re == NULL) return PCRE_ERROR_NULL;
-if (re->magic_number != MAGIC_NUMBER)
- {
- re = try_flipped(re, &internal_re, NULL, NULL);
- if (re == NULL) return PCRE_ERROR_BADMAGIC;
- }
-if (optptr != NULL) *optptr = (int)(re->options & PUBLIC_OPTIONS);
-if (first_byte != NULL)
- *first_byte = ((re->options & PCRE_FIRSTSET) != 0)? re->first_byte :
- ((re->options & PCRE_STARTLINE) != 0)? -1 : -2;
-return re->top_bracket;
-}
-
-
-
-/*************************************************
-* Return info about compiled pattern *
-*************************************************/
-
-/* This is a newer "info" function which has an extensible interface so
-that additional items can be added compatibly.
-
-Arguments:
- argument_re points to compiled code
- extra_data points extra data, or NULL
- what what information is required
- where where to put the information
-
-Returns: 0 if data returned, negative on error
-*/
-
-EXPORT int
-pcre_fullinfo(const pcre *argument_re, const pcre_extra *extra_data, int what,
- void *where)
-{
-real_pcre internal_re;
-pcre_study_data internal_study;
-const real_pcre *re = (const real_pcre *)argument_re;
-const pcre_study_data *study = NULL;
-
-if (re == NULL || where == NULL) return PCRE_ERROR_NULL;
-
-if (extra_data != NULL && (extra_data->flags & PCRE_EXTRA_STUDY_DATA) != 0)
- study = (const pcre_study_data *)extra_data->study_data;
-
-if (re->magic_number != MAGIC_NUMBER)
- {
- re = try_flipped(re, &internal_re, study, &internal_study);
- if (re == NULL) return PCRE_ERROR_BADMAGIC;
- if (study != NULL) study = &internal_study;
- }
-
-switch (what)
- {
- case PCRE_INFO_OPTIONS:
- *((unsigned long int *)where) = re->options & PUBLIC_OPTIONS;
- break;
-
- case PCRE_INFO_SIZE:
- *((size_t *)where) = re->size;
- break;
-
- case PCRE_INFO_STUDYSIZE:
- *((size_t *)where) = (study == NULL)? 0 : study->size;
- break;
-
- case PCRE_INFO_CAPTURECOUNT:
- *((int *)where) = re->top_bracket;
- break;
-
- case PCRE_INFO_BACKREFMAX:
- *((int *)where) = re->top_backref;
- break;
-
- case PCRE_INFO_FIRSTBYTE:
- *((int *)where) =
- ((re->options & PCRE_FIRSTSET) != 0)? re->first_byte :
- ((re->options & PCRE_STARTLINE) != 0)? -1 : -2;
- break;
-
- /* Make sure we pass back the pointer to the bit vector in the external
- block, not the internal copy (with flipped integer fields). */
-
- case PCRE_INFO_FIRSTTABLE:
- *((const uschar **)where) =
- (study != NULL && (study->options & PCRE_STUDY_MAPPED) != 0)?
- ((const pcre_study_data *)extra_data->study_data)->start_bits : NULL;
- break;
-
- case PCRE_INFO_LASTLITERAL:
- *((int *)where) =
- ((re->options & PCRE_REQCHSET) != 0)? re->req_byte : -1;
- break;
-
- case PCRE_INFO_NAMEENTRYSIZE:
- *((int *)where) = re->name_entry_size;
- break;
-
- case PCRE_INFO_NAMECOUNT:
- *((int *)where) = re->name_count;
- break;
-
- case PCRE_INFO_NAMETABLE:
- *((const uschar **)where) = (const uschar *)re + re->name_table_offset;
- break;
-
- case PCRE_INFO_DEFAULT_TABLES:
- *((const uschar **)where) = (const uschar *)pcre_default_tables;
- break;
-
- default: return PCRE_ERROR_BADOPTION;
- }
-
-return 0;
-}
-
-
-
-/*************************************************
-* Return info about what features are configured *
-*************************************************/
-
-/* This is function which has an extensible interface so that additional items
-can be added compatibly.
-
-Arguments:
- what what information is required
- where where to put the information
-
-Returns: 0 if data returned, negative on error
-*/
-
-EXPORT int
-pcre_config(int what, void *where)
-{
-switch (what)
- {
- case PCRE_CONFIG_UTF8:
-#ifdef SUPPORT_UTF8
- *((int *)where) = 1;
-#else
- *((int *)where) = 0;
-#endif
- break;
-
- case PCRE_CONFIG_UNICODE_PROPERTIES:
-#ifdef SUPPORT_UCP
- *((int *)where) = 1;
-#else
- *((int *)where) = 0;
-#endif
- break;
-
- case PCRE_CONFIG_NEWLINE:
- *((int *)where) = NEWLINE;
- break;
-
- case PCRE_CONFIG_LINK_SIZE:
- *((int *)where) = LINK_SIZE;
- break;
-
- case PCRE_CONFIG_POSIX_MALLOC_THRESHOLD:
- *((int *)where) = POSIX_MALLOC_THRESHOLD;
- break;
-
- case PCRE_CONFIG_MATCH_LIMIT:
- *((unsigned int *)where) = MATCH_LIMIT;
- break;
-
- case PCRE_CONFIG_STACKRECURSE:
-#ifdef NO_RECURSE
- *((int *)where) = 0;
-#else
- *((int *)where) = 1;
-#endif
- break;
-
- default: return PCRE_ERROR_BADOPTION;
- }
-
-return 0;
-}
-
-
-
-#ifdef DEBUG
-/*************************************************
-* Debugging function to print chars *
-*************************************************/
-
-/* Print a sequence of chars in printable format, stopping at the end of the
-subject if the requested.
-
-Arguments:
- p points to characters
- length number to print
- is_subject TRUE if printing from within md->start_subject
- md pointer to matching data block, if is_subject is TRUE
-
-Returns: nothing
-*/
-
-static void
-pchars(const uschar *p, int length, BOOL is_subject, match_data *md)
-{
-int c;
-if (is_subject && length > md->end_subject - p) length = md->end_subject - p;
-while (length-- > 0)
- if (isprint(c = *(p++))) printf("%c", c); else printf("\\x%02x", c);
-}
-#endif
-
-
-
-
-/*************************************************
-* Handle escapes *
-*************************************************/
-
-/* This function is called when a \ has been encountered. It either returns a
-positive value for a simple escape such as \n, or a negative value which
-encodes one of the more complicated things such as \d. When UTF-8 is enabled,
-a positive value greater than 255 may be returned. On entry, ptr is pointing at
-the \. On exit, it is on the final character of the escape sequence.
-
-Arguments:
- ptrptr points to the pattern position pointer
- errorptr points to the pointer to the error message
- bracount number of previous extracting brackets
- options the options bits
- isclass TRUE if inside a character class
-
-Returns: zero or positive => a data character
- negative => a special escape sequence
- on error, errorptr is set
-*/
-
-static int
-check_escape(const uschar **ptrptr, const char **errorptr, int bracount,
- int options, BOOL isclass)
-{
-const uschar *ptr = *ptrptr;
-int c, i;
-
-/* If backslash is at the end of the pattern, it's an error. */
-
-c = *(++ptr);
-if (c == 0) *errorptr = ERR1;
-
-/* Non-alphamerics are literals. For digits or letters, do an initial lookup in
-a table. A non-zero result is something that can be returned immediately.
-Otherwise further processing may be required. */
-
-#if !EBCDIC /* ASCII coding */
-else if (c < '0' || c > 'z') {} /* Not alphameric */
-else if ((i = escapes[c - '0']) != 0) c = i;
-
-#else /* EBCDIC coding */
-else if (c < 'a' || (ebcdic_chartab[c] & 0x0E) == 0) {} /* Not alphameric */
-else if ((i = escapes[c - 0x48]) != 0) c = i;
-#endif
-
-/* Escapes that need further processing, or are illegal. */
-
-else
- {
- const uschar *oldptr;
- switch (c)
- {
- /* A number of Perl escapes are not handled by PCRE. We give an explicit
- error. */
-
- case 'l':
- case 'L':
- case 'N':
- case 'u':
- case 'U':
- *errorptr = ERR37;
- break;
-
- /* The handling of escape sequences consisting of a string of digits
- starting with one that is not zero is not straightforward. By experiment,
- the way Perl works seems to be as follows:
-
- Outside a character class, the digits are read as a decimal number. If the
- number is less than 10, or if there are that many previous extracting
- left brackets, then it is a back reference. Otherwise, up to three octal
- digits are read to form an escaped byte. Thus \123 is likely to be octal
- 123 (cf \0123, which is octal 012 followed by the literal 3). If the octal
- value is greater than 377, the least significant 8 bits are taken. Inside a
- character class, \ followed by a digit is always an octal number. */
-
- case '1': case '2': case '3': case '4': case '5':
- case '6': case '7': case '8': case '9':
-
- if (!isclass)
- {
- oldptr = ptr;
- c -= '0';
- while ((digitab[ptr[1]] & ctype_digit) != 0)
- c = c * 10 + *(++ptr) - '0';
- if (c < 10 || c <= bracount)
- {
- c = -(ESC_REF + c);
- break;
- }
- ptr = oldptr; /* Put the pointer back and fall through */
- }
-
- /* Handle an octal number following \. If the first digit is 8 or 9, Perl
- generates a binary zero byte and treats the digit as a following literal.
- Thus we have to pull back the pointer by one. */
-
- if ((c = *ptr) >= '8')
- {
- ptr--;
- c = 0;
- break;
- }
-
- /* \0 always starts an octal number, but we may drop through to here with a
- larger first octal digit. */
-
- case '0':
- c -= '0';
- while(i++ < 2 && ptr[1] >= '0' && ptr[1] <= '7')
- c = c * 8 + *(++ptr) - '0';
- c &= 255; /* Take least significant 8 bits */
- break;
-
- /* \x is complicated when UTF-8 is enabled. \x{ddd} is a character number
- which can be greater than 0xff, but only if the ddd are hex digits. */
-
- case 'x':
-#ifdef SUPPORT_UTF8
- if (ptr[1] == '{' && (options & PCRE_UTF8) != 0)
- {
- const uschar *pt = ptr + 2;
- register int count = 0;
- c = 0;
- while ((digitab[*pt] & ctype_xdigit) != 0)
- {
- int cc = *pt++;
- count++;
-#if !EBCDIC /* ASCII coding */
- if (cc >= 'a') cc -= 32; /* Convert to upper case */
- c = c * 16 + cc - ((cc < 'A')? '0' : ('A' - 10));
-#else /* EBCDIC coding */
- if (cc >= 'a' && cc <= 'z') cc += 64; /* Convert to upper case */
- c = c * 16 + cc - ((cc >= '0')? '0' : ('A' - 10));
-#endif
- }
- if (*pt == '}')
- {
- if (c < 0 || count > 8) *errorptr = ERR34;
- ptr = pt;
- break;
- }
- /* If the sequence of hex digits does not end with '}', then we don't
- recognize this construct; fall through to the normal \x handling. */
- }
-#endif
-
- /* Read just a single hex char */
-
- c = 0;
- while (i++ < 2 && (digitab[ptr[1]] & ctype_xdigit) != 0)
- {
- int cc; /* Some compilers don't like ++ */
- cc = *(++ptr); /* in initializers */
-#if !EBCDIC /* ASCII coding */
- if (cc >= 'a') cc -= 32; /* Convert to upper case */
- c = c * 16 + cc - ((cc < 'A')? '0' : ('A' - 10));
-#else /* EBCDIC coding */
- if (cc <= 'z') cc += 64; /* Convert to upper case */
- c = c * 16 + cc - ((cc >= '0')? '0' : ('A' - 10));
-#endif
- }
- break;
-
- /* Other special escapes not starting with a digit are straightforward */
-
- case 'c':
- c = *(++ptr);
- if (c == 0)
- {
- *errorptr = ERR2;
- return 0;
- }
-
- /* A letter is upper-cased; then the 0x40 bit is flipped. This coding
- is ASCII-specific, but then the whole concept of \cx is ASCII-specific.
- (However, an EBCDIC equivalent has now been added.) */
-
-#if !EBCDIC /* ASCII coding */
- if (c >= 'a' && c <= 'z') c -= 32;
- c ^= 0x40;
-#else /* EBCDIC coding */
- if (c >= 'a' && c <= 'z') c += 64;
- c ^= 0xC0;
-#endif
- break;
-
- /* PCRE_EXTRA enables extensions to Perl in the matter of escapes. Any
- other alphameric following \ is an error if PCRE_EXTRA was set; otherwise,
- for Perl compatibility, it is a literal. This code looks a bit odd, but
- there used to be some cases other than the default, and there may be again
- in future, so I haven't "optimized" it. */
-
- default:
- if ((options & PCRE_EXTRA) != 0) switch(c)
- {
- default:
- *errorptr = ERR3;
- break;
- }
- break;
- }
- }
-
-*ptrptr = ptr;
-return c;
-}
-
-
-
-#ifdef SUPPORT_UCP
-/*************************************************
-* Handle \P and \p *
-*************************************************/
-
-/* This function is called after \P or \p has been encountered, provided that
-PCRE is compiled with support for Unicode properties. On entry, ptrptr is
-pointing at the P or p. On exit, it is pointing at the final character of the
-escape sequence.
-
-Argument:
- ptrptr points to the pattern position pointer
- negptr points to a boolean that is set TRUE for negation else FALSE
- errorptr points to the pointer to the error message
-
-Returns: value from ucp_type_table, or -1 for an invalid type
-*/
-
-static int
-get_ucp(const uschar **ptrptr, BOOL *negptr, const char **errorptr)
-{
-int c, i, bot, top;
-const uschar *ptr = *ptrptr;
-char name[4];
-
-c = *(++ptr);
-if (c == 0) goto ERROR_RETURN;
-
-*negptr = FALSE;
-
-/* \P or \p can be followed by a one- or two-character name in {}, optionally
-preceded by ^ for negation. */
-
-if (c == '{')
- {
- if (ptr[1] == '^')
- {
- *negptr = TRUE;
- ptr++;
- }
- for (i = 0; i <= 2; i++)
- {
- c = *(++ptr);
- if (c == 0) goto ERROR_RETURN;
- if (c == '}') break;
- name[i] = c;
- }
- if (c !='}') /* Try to distinguish error cases */
- {
- while (*(++ptr) != 0 && *ptr != '}');
- if (*ptr == '}') goto UNKNOWN_RETURN; else goto ERROR_RETURN;
- }
- name[i] = 0;
- }
-
-/* Otherwise there is just one following character */
-
-else
- {
- name[0] = c;
- name[1] = 0;
- }
-
-*ptrptr = ptr;
-
-/* Search for a recognized property name using binary chop */
-
-bot = 0;
-top = sizeof(utt)/sizeof(ucp_type_table);
-
-while (bot < top)
- {
- i = (bot + top)/2;
- c = strcmp(name, utt[i].name);
- if (c == 0) return utt[i].value;
- if (c > 0) bot = i + 1; else top = i;
- }
-
-UNKNOWN_RETURN:
-*errorptr = ERR47;
-*ptrptr = ptr;
-return -1;
-
-ERROR_RETURN:
-*errorptr = ERR46;
-*ptrptr = ptr;
-return -1;
-}
-#endif
-
-
-
-
-/*************************************************
-* Check for counted repeat *
-*************************************************/
-
-/* This function is called when a '{' is encountered in a place where it might
-start a quantifier. It looks ahead to see if it really is a quantifier or not.
-It is only a quantifier if it is one of the forms {ddd} {ddd,} or {ddd,ddd}
-where the ddds are digits.
-
-Arguments:
- p pointer to the first char after '{'
-
-Returns: TRUE or FALSE
-*/
-
-static BOOL
-is_counted_repeat(const uschar *p)
-{
-if ((digitab[*p++] & ctype_digit) == 0) return FALSE;
-while ((digitab[*p] & ctype_digit) != 0) p++;
-if (*p == '}') return TRUE;
-
-if (*p++ != ',') return FALSE;
-if (*p == '}') return TRUE;
-
-if ((digitab[*p++] & ctype_digit) == 0) return FALSE;
-while ((digitab[*p] & ctype_digit) != 0) p++;
-
-return (*p == '}');
-}
-
-
-
-/*************************************************
-* Read repeat counts *
-*************************************************/
-
-/* Read an item of the form {n,m} and return the values. This is called only
-after is_counted_repeat() has confirmed that a repeat-count quantifier exists,
-so the syntax is guaranteed to be correct, but we need to check the values.
-
-Arguments:
- p pointer to first char after '{'
- minp pointer to int for min
- maxp pointer to int for max
- returned as -1 if no max
- errorptr points to pointer to error message
-
-Returns: pointer to '}' on success;
- current ptr on error, with errorptr set
-*/
-
-static const uschar *
-read_repeat_counts(const uschar *p, int *minp, int *maxp, const char **errorptr)
-{
-int min = 0;
-int max = -1;
-
-while ((digitab[*p] & ctype_digit) != 0) min = min * 10 + *p++ - '0';
-
-if (*p == '}') max = min; else
- {
- if (*(++p) != '}')
- {
- max = 0;
- while((digitab[*p] & ctype_digit) != 0) max = max * 10 + *p++ - '0';
- if (max < min)
- {
- *errorptr = ERR4;
- return p;
- }
- }
- }
-
-/* Do paranoid checks, then fill in the required variables, and pass back the
-pointer to the terminating '}'. */
-
-if (min > 65535 || max > 65535)
- *errorptr = ERR5;
-else
- {
- *minp = min;
- *maxp = max;
- }
-return p;
-}
-
-
-
-/*************************************************
-* Find first significant op code *
-*************************************************/
-
-/* This is called by several functions that scan a compiled expression looking
-for a fixed first character, or an anchoring op code etc. It skips over things
-that do not influence this. For some calls, a change of option is important.
-For some calls, it makes sense to skip negative forward and all backward
-assertions, and also the \b assertion; for others it does not.
-
-Arguments:
- code pointer to the start of the group
- options pointer to external options
- optbit the option bit whose changing is significant, or
- zero if none are
- skipassert TRUE if certain assertions are to be skipped
-
-Returns: pointer to the first significant opcode
-*/
-
-static const uschar*
-first_significant_code(const uschar *code, int *options, int optbit,
- BOOL skipassert)
-{
-for (;;)
- {
- switch ((int)*code)
- {
- case OP_OPT:
- if (optbit > 0 && ((int)code[1] & optbit) != (*options & optbit))
- *options = (int)code[1];
- code += 2;
- break;
-
- case OP_ASSERT_NOT:
- case OP_ASSERTBACK:
- case OP_ASSERTBACK_NOT:
- if (!skipassert) return code;
- do code += GET(code, 1); while (*code == OP_ALT);
- code += OP_lengths[*code];
- break;
-
- case OP_WORD_BOUNDARY:
- case OP_NOT_WORD_BOUNDARY:
- if (!skipassert) return code;
- /* Fall through */
-
- case OP_CALLOUT:
- case OP_CREF:
- case OP_BRANUMBER:
- code += OP_lengths[*code];
- break;
-
- default:
- return code;
- }
- }
-/* Control never reaches here */
-}
-
-
-
-
-/*************************************************
-* Find the fixed length of a pattern *
-*************************************************/
-
-/* Scan a pattern and compute the fixed length of subject that will match it,
-if the length is fixed. This is needed for dealing with backward assertions.
-In UTF8 mode, the result is in characters rather than bytes.
-
-Arguments:
- code points to the start of the pattern (the bracket)
- options the compiling options
-
-Returns: the fixed length, or -1 if there is no fixed length,
- or -2 if \C was encountered
-*/
-
-static int
-find_fixedlength(uschar *code, int options)
-{
-int length = -1;
-
-register int branchlength = 0;
-register uschar *cc = code + 1 + LINK_SIZE;
-
-/* Scan along the opcodes for this branch. If we get to the end of the
-branch, check the length against that of the other branches. */
-
-for (;;)
- {
- int d;
- register int op = *cc;
- if (op >= OP_BRA) op = OP_BRA;
-
- switch (op)
- {
- case OP_BRA:
- case OP_ONCE:
- case OP_COND:
- d = find_fixedlength(cc, options);
- if (d < 0) return d;
- branchlength += d;
- do cc += GET(cc, 1); while (*cc == OP_ALT);
- cc += 1 + LINK_SIZE;
- break;
-
- /* Reached end of a branch; if it's a ket it is the end of a nested
- call. If it's ALT it is an alternation in a nested call. If it is
- END it's the end of the outer call. All can be handled by the same code. */
-
- case OP_ALT:
- case OP_KET:
- case OP_KETRMAX:
- case OP_KETRMIN:
- case OP_END:
- if (length < 0) length = branchlength;
- else if (length != branchlength) return -1;
- if (*cc != OP_ALT) return length;
- cc += 1 + LINK_SIZE;
- branchlength = 0;
- break;
-
- /* Skip over assertive subpatterns */
-
- case OP_ASSERT:
- case OP_ASSERT_NOT:
- case OP_ASSERTBACK:
- case OP_ASSERTBACK_NOT:
- do cc += GET(cc, 1); while (*cc == OP_ALT);
- /* Fall through */
-
- /* Skip over things that don't match chars */
-
- case OP_REVERSE:
- case OP_BRANUMBER:
- case OP_CREF:
- case OP_OPT:
- case OP_CALLOUT:
- case OP_SOD:
- case OP_SOM:
- case OP_EOD:
- case OP_EODN:
- case OP_CIRC:
- case OP_DOLL:
- case OP_NOT_WORD_BOUNDARY:
- case OP_WORD_BOUNDARY:
- cc += OP_lengths[*cc];
- break;
-
- /* Handle literal characters */
-
- case OP_CHAR:
- case OP_CHARNC:
- branchlength++;
- cc += 2;
-#ifdef SUPPORT_UTF8
- if ((options & PCRE_UTF8) != 0)
- {
- while ((*cc & 0xc0) == 0x80) cc++;
- }
-#endif
- break;
-
- /* Handle exact repetitions. The count is already in characters, but we
- need to skip over a multibyte character in UTF8 mode. */
-
- case OP_EXACT:
- branchlength += GET2(cc,1);
- cc += 4;
-#ifdef SUPPORT_UTF8
- if ((options & PCRE_UTF8) != 0)
- {
- while((*cc & 0x80) == 0x80) cc++;
- }
-#endif
- break;
-
- case OP_TYPEEXACT:
- branchlength += GET2(cc,1);
- cc += 4;
- break;
-
- /* Handle single-char matchers */
-
- case OP_PROP:
- case OP_NOTPROP:
- cc++;
- /* Fall through */
-
- case OP_NOT_DIGIT:
- case OP_DIGIT:
- case OP_NOT_WHITESPACE:
- case OP_WHITESPACE:
- case OP_NOT_WORDCHAR:
- case OP_WORDCHAR:
- case OP_ANY:
- branchlength++;
- cc++;
- break;
-
- /* The single-byte matcher isn't allowed */
-
- case OP_ANYBYTE:
- return -2;
-
- /* Check a class for variable quantification */
-
-#ifdef SUPPORT_UTF8
- case OP_XCLASS:
- cc += GET(cc, 1) - 33;
- /* Fall through */
-#endif
-
- case OP_CLASS:
- case OP_NCLASS:
- cc += 33;
-
- switch (*cc)
- {
- case OP_CRSTAR:
- case OP_CRMINSTAR:
- case OP_CRQUERY:
- case OP_CRMINQUERY:
- return -1;
-
- case OP_CRRANGE:
- case OP_CRMINRANGE:
- if (GET2(cc,1) != GET2(cc,3)) return -1;
- branchlength += GET2(cc,1);
- cc += 5;
- break;
-
- default:
- branchlength++;
- }
- break;
-
- /* Anything else is variable length */
-
- default:
- return -1;
- }
- }
-/* Control never gets here */
-}
-
-
-
-
-/*************************************************
-* Scan compiled regex for numbered bracket *
-*************************************************/
-
-/* This little function scans through a compiled pattern until it finds a
-capturing bracket with the given number.
-
-Arguments:
- code points to start of expression
- utf8 TRUE in UTF-8 mode
- number the required bracket number
-
-Returns: pointer to the opcode for the bracket, or NULL if not found
-*/
-
-static const uschar *
-find_bracket(const uschar *code, BOOL utf8, int number)
-{
-#ifndef SUPPORT_UTF8
-utf8 = utf8; /* Stop pedantic compilers complaining */
-#endif
-
-for (;;)
- {
- register int c = *code;
- if (c == OP_END) return NULL;
- else if (c > OP_BRA)
- {
- int n = c - OP_BRA;
- if (n > EXTRACT_BASIC_MAX) n = GET2(code, 2+LINK_SIZE);
- if (n == number) return (uschar *)code;
- code += OP_lengths[OP_BRA];
- }
- else
- {
- code += OP_lengths[c];
-
-#ifdef SUPPORT_UTF8
-
- /* In UTF-8 mode, opcodes that are followed by a character may be followed
- by a multi-byte character. The length in the table is a minimum, so we have
- to scan along to skip the extra bytes. All opcodes are less than 128, so we
- can use relatively efficient code. */
-
- if (utf8) switch(c)
- {
- case OP_CHAR:
- case OP_CHARNC:
- case OP_EXACT:
- case OP_UPTO:
- case OP_MINUPTO:
- case OP_STAR:
- case OP_MINSTAR:
- case OP_PLUS:
- case OP_MINPLUS:
- case OP_QUERY:
- case OP_MINQUERY:
- while ((*code & 0xc0) == 0x80) code++;
- break;
-
- /* XCLASS is used for classes that cannot be represented just by a bit
- map. This includes negated single high-valued characters. The length in
- the table is zero; the actual length is stored in the compiled code. */
-
- case OP_XCLASS:
- code += GET(code, 1) + 1;
- break;
- }
-#endif
- }
- }
-}
-
-
-
-/*************************************************
-* Scan compiled regex for recursion reference *
-*************************************************/
-
-/* This little function scans through a compiled pattern until it finds an
-instance of OP_RECURSE.
-
-Arguments:
- code points to start of expression
- utf8 TRUE in UTF-8 mode
-
-Returns: pointer to the opcode for OP_RECURSE, or NULL if not found
-*/
-
-static const uschar *
-find_recurse(const uschar *code, BOOL utf8)
-{
-#ifndef SUPPORT_UTF8
-utf8 = utf8; /* Stop pedantic compilers complaining */
-#endif
-
-for (;;)
- {
- register int c = *code;
- if (c == OP_END) return NULL;
- else if (c == OP_RECURSE) return code;
- else if (c > OP_BRA)
- {
- code += OP_lengths[OP_BRA];
- }
- else
- {
- code += OP_lengths[c];
-
-#ifdef SUPPORT_UTF8
-
- /* In UTF-8 mode, opcodes that are followed by a character may be followed
- by a multi-byte character. The length in the table is a minimum, so we have
- to scan along to skip the extra bytes. All opcodes are less than 128, so we
- can use relatively efficient code. */
-
- if (utf8) switch(c)
- {
- case OP_CHAR:
- case OP_CHARNC:
- case OP_EXACT:
- case OP_UPTO:
- case OP_MINUPTO:
- case OP_STAR:
- case OP_MINSTAR:
- case OP_PLUS:
- case OP_MINPLUS:
- case OP_QUERY:
- case OP_MINQUERY:
- while ((*code & 0xc0) == 0x80) code++;
- break;
-
- /* XCLASS is used for classes that cannot be represented just by a bit
- map. This includes negated single high-valued characters. The length in
- the table is zero; the actual length is stored in the compiled code. */
-
- case OP_XCLASS:
- code += GET(code, 1) + 1;
- break;
- }
-#endif
- }
- }
-}
-
-
-
-/*************************************************
-* Scan compiled branch for non-emptiness *
-*************************************************/
-
-/* This function scans through a branch of a compiled pattern to see whether it
-can match the empty string or not. It is called only from could_be_empty()
-below. Note that first_significant_code() skips over assertions. If we hit an
-unclosed bracket, we return "empty" - this means we've struck an inner bracket
-whose current branch will already have been scanned.
-
-Arguments:
- code points to start of search
- endcode points to where to stop
- utf8 TRUE if in UTF8 mode
-
-Returns: TRUE if what is matched could be empty
-*/
-
-static BOOL
-could_be_empty_branch(const uschar *code, const uschar *endcode, BOOL utf8)
-{
-register int c;
-for (code = first_significant_code(code + 1 + LINK_SIZE, NULL, 0, TRUE);
- code < endcode;
- code = first_significant_code(code + OP_lengths[c], NULL, 0, TRUE))
- {
- const uschar *ccode;
-
- c = *code;
-
- if (c >= OP_BRA)
- {
- BOOL empty_branch;
- if (GET(code, 1) == 0) return TRUE; /* Hit unclosed bracket */
-
- /* Scan a closed bracket */
-
- empty_branch = FALSE;
- do
- {
- if (!empty_branch && could_be_empty_branch(code, endcode, utf8))
- empty_branch = TRUE;
- code += GET(code, 1);
- }
- while (*code == OP_ALT);
- if (!empty_branch) return FALSE; /* All branches are non-empty */
- code += 1 + LINK_SIZE;
- c = *code;
- }
-
- else switch (c)
- {
- /* Check for quantifiers after a class */
-
-#ifdef SUPPORT_UTF8
- case OP_XCLASS:
- ccode = code + GET(code, 1);
- goto CHECK_CLASS_REPEAT;
-#endif
-
- case OP_CLASS:
- case OP_NCLASS:
- ccode = code + 33;
-
-#ifdef SUPPORT_UTF8
- CHECK_CLASS_REPEAT:
-#endif
-
- switch (*ccode)
- {
- case OP_CRSTAR: /* These could be empty; continue */
- case OP_CRMINSTAR:
- case OP_CRQUERY:
- case OP_CRMINQUERY:
- break;
-
- default: /* Non-repeat => class must match */
- case OP_CRPLUS: /* These repeats aren't empty */
- case OP_CRMINPLUS:
- return FALSE;
-
- case OP_CRRANGE:
- case OP_CRMINRANGE:
- if (GET2(ccode, 1) > 0) return FALSE; /* Minimum > 0 */
- break;
- }
- break;
-
- /* Opcodes that must match a character */
-
- case OP_PROP:
- case OP_NOTPROP:
- case OP_EXTUNI:
- case OP_NOT_DIGIT:
- case OP_DIGIT:
- case OP_NOT_WHITESPACE:
- case OP_WHITESPACE:
- case OP_NOT_WORDCHAR:
- case OP_WORDCHAR:
- case OP_ANY:
- case OP_ANYBYTE:
- case OP_CHAR:
- case OP_CHARNC:
- case OP_NOT:
- case OP_PLUS:
- case OP_MINPLUS:
- case OP_EXACT:
- case OP_NOTPLUS:
- case OP_NOTMINPLUS:
- case OP_NOTEXACT:
- case OP_TYPEPLUS:
- case OP_TYPEMINPLUS:
- case OP_TYPEEXACT:
- return FALSE;
-
- /* End of branch */
-
- case OP_KET:
- case OP_KETRMAX:
- case OP_KETRMIN:
- case OP_ALT:
- return TRUE;
-
- /* In UTF-8 mode, STAR, MINSTAR, QUERY, MINQUERY, UPTO, and MINUPTO may be
- followed by a multibyte character */
-
-#ifdef SUPPORT_UTF8
- case OP_STAR:
- case OP_MINSTAR:
- case OP_QUERY:
- case OP_MINQUERY:
- case OP_UPTO:
- case OP_MINUPTO:
- if (utf8) while ((code[2] & 0xc0) == 0x80) code++;
- break;
-#endif
- }
- }
-
-return TRUE;
-}
-
-
-
-/*************************************************
-* Scan compiled regex for non-emptiness *
-*************************************************/
-
-/* This function is called to check for left recursive calls. We want to check
-the current branch of the current pattern to see if it could match the empty
-string. If it could, we must look outwards for branches at other levels,
-stopping when we pass beyond the bracket which is the subject of the recursion.
-
-Arguments:
- code points to start of the recursion
- endcode points to where to stop (current RECURSE item)
- bcptr points to the chain of current (unclosed) branch starts
- utf8 TRUE if in UTF-8 mode
-
-Returns: TRUE if what is matched could be empty
-*/
-
-static BOOL
-could_be_empty(const uschar *code, const uschar *endcode, branch_chain *bcptr,
- BOOL utf8)
-{
-while (bcptr != NULL && bcptr->current >= code)
- {
- if (!could_be_empty_branch(bcptr->current, endcode, utf8)) return FALSE;
- bcptr = bcptr->outer;
- }
-return TRUE;
-}
-
-
-
-/*************************************************
-* Check for POSIX class syntax *
-*************************************************/
-
-/* This function is called when the sequence "[:" or "[." or "[=" is
-encountered in a character class. It checks whether this is followed by an
-optional ^ and then a sequence of letters, terminated by a matching ":]" or
-".]" or "=]".
-
-Argument:
- ptr pointer to the initial [
- endptr where to return the end pointer
- cd pointer to compile data
-
-Returns: TRUE or FALSE
-*/
-
-static BOOL
-check_posix_syntax(const uschar *ptr, const uschar **endptr, compile_data *cd)
-{
-int terminator; /* Don't combine these lines; the Solaris cc */
-terminator = *(++ptr); /* compiler warns about "non-constant" initializer. */
-if (*(++ptr) == '^') ptr++;
-while ((cd->ctypes[*ptr] & ctype_letter) != 0) ptr++;
-if (*ptr == terminator && ptr[1] == ']')
- {
- *endptr = ptr;
- return TRUE;
- }
-return FALSE;
-}
-
-
-
-
-/*************************************************
-* Check POSIX class name *
-*************************************************/
-
-/* This function is called to check the name given in a POSIX-style class entry
-such as [:alnum:].
-
-Arguments:
- ptr points to the first letter
- len the length of the name
-
-Returns: a value representing the name, or -1 if unknown
-*/
-
-static int
-check_posix_name(const uschar *ptr, int len)
-{
-register int yield = 0;
-while (posix_name_lengths[yield] != 0)
- {
- if (len == posix_name_lengths[yield] &&
- strncmp((const char *)ptr, posix_names[yield], len) == 0) return yield;
- yield++;
- }
-return -1;
-}
-
-
-/*************************************************
-* Adjust OP_RECURSE items in repeated group *
-*************************************************/
-
-/* OP_RECURSE items contain an offset from the start of the regex to the group
-that is referenced. This means that groups can be replicated for fixed
-repetition simply by copying (because the recursion is allowed to refer to
-earlier groups that are outside the current group). However, when a group is
-optional (i.e. the minimum quantifier is zero), OP_BRAZERO is inserted before
-it, after it has been compiled. This means that any OP_RECURSE items within it
-that refer to the group itself or any contained groups have to have their
-offsets adjusted. That is the job of this function. Before it is called, the
-partially compiled regex must be temporarily terminated with OP_END.
-
-Arguments:
- group points to the start of the group
- adjust the amount by which the group is to be moved
- utf8 TRUE in UTF-8 mode
- cd contains pointers to tables etc.
-
-Returns: nothing
-*/
-
-static void
-adjust_recurse(uschar *group, int adjust, BOOL utf8, compile_data *cd)
-{
-uschar *ptr = group;
-while ((ptr = (uschar *)find_recurse(ptr, utf8)) != NULL)
- {
- int offset = GET(ptr, 1);
- if (cd->start_code + offset >= group) PUT(ptr, 1, offset + adjust);
- ptr += 1 + LINK_SIZE;
- }
-}
-
-
-
-/*************************************************
-* Insert an automatic callout point *
-*************************************************/
-
-/* This function is called when the PCRE_AUTO_CALLOUT option is set, to insert
-callout points before each pattern item.
-
-Arguments:
- code current code pointer
- ptr current pattern pointer
- cd pointers to tables etc
-
-Returns: new code pointer
-*/
-
-static uschar *
-auto_callout(uschar *code, const uschar *ptr, compile_data *cd)
-{
-*code++ = OP_CALLOUT;
-*code++ = 255;
-PUT(code, 0, ptr - cd->start_pattern); /* Pattern offset */
-PUT(code, LINK_SIZE, 0); /* Default length */
-return code + 2*LINK_SIZE;
-}
-
-
-
-/*************************************************
-* Complete a callout item *
-*************************************************/
-
-/* A callout item contains the length of the next item in the pattern, which
-we can't fill in till after we have reached the relevant point. This is used
-for both automatic and manual callouts.
-
-Arguments:
- previous_callout points to previous callout item
- ptr current pattern pointer
- cd pointers to tables etc
-
-Returns: nothing
-*/
-
-static void
-complete_callout(uschar *previous_callout, const uschar *ptr, compile_data *cd)
-{
-int length = ptr - cd->start_pattern - GET(previous_callout, 2);
-PUT(previous_callout, 2 + LINK_SIZE, length);
-}
-
-
-
-#ifdef SUPPORT_UCP
-/*************************************************
-* Get othercase range *
-*************************************************/
-
-/* This function is passed the start and end of a class range, in UTF-8 mode
-with UCP support. It searches up the characters, looking for internal ranges of
-characters in the "other" case. Each call returns the next one, updating the
-start address.
-
-Arguments:
- cptr points to starting character value; updated
- d end value
- ocptr where to put start of othercase range
- odptr where to put end of othercase range
-
-Yield: TRUE when range returned; FALSE when no more
-*/
-
-static BOOL
-get_othercase_range(int *cptr, int d, int *ocptr, int *odptr)
-{
-int c, chartype, othercase, next;
-
-for (c = *cptr; c <= d; c++)
- {
- if (ucp_findchar(c, &chartype, &othercase) == ucp_L && othercase != 0) break;
- }
-
-if (c > d) return FALSE;
-
-*ocptr = othercase;
-next = othercase + 1;
-
-for (++c; c <= d; c++)
- {
- if (ucp_findchar(c, &chartype, &othercase) != ucp_L || othercase != next)
- break;
- next++;
- }
-
-*odptr = next - 1;
-*cptr = c;
-
-return TRUE;
-}
-#endif /* SUPPORT_UCP */
-
-
-/*************************************************
-* Compile one branch *
-*************************************************/
-
-/* Scan the pattern, compiling it into the code vector. If the options are
-changed during the branch, the pointer is used to change the external options
-bits.
-
-Arguments:
- optionsptr pointer to the option bits
- brackets points to number of extracting brackets used
- codeptr points to the pointer to the current code point
- ptrptr points to the current pattern pointer
- errorptr points to pointer to error message
- firstbyteptr set to initial literal character, or < 0 (REQ_UNSET, REQ_NONE)
- reqbyteptr set to the last literal character required, else < 0
- bcptr points to current branch chain
- cd contains pointers to tables etc.
-
-Returns: TRUE on success
- FALSE, with *errorptr set on error
-*/
-
-static BOOL
-compile_branch(int *optionsptr, int *brackets, uschar **codeptr,
- const uschar **ptrptr, const char **errorptr, int *firstbyteptr,
- int *reqbyteptr, branch_chain *bcptr, compile_data *cd)
-{
-int repeat_type, op_type;
-int repeat_min = 0, repeat_max = 0; /* To please picky compilers */
-int bravalue = 0;
-int greedy_default, greedy_non_default;
-int firstbyte, reqbyte;
-int zeroreqbyte, zerofirstbyte;
-int req_caseopt, reqvary, tempreqvary;
-int condcount = 0;
-int options = *optionsptr;
-int after_manual_callout = 0;
-register int c;
-register uschar *code = *codeptr;
-uschar *tempcode;
-BOOL inescq = FALSE;
-BOOL groupsetfirstbyte = FALSE;
-const uschar *ptr = *ptrptr;
-const uschar *tempptr;
-uschar *previous = NULL;
-uschar *previous_callout = NULL;
-uschar classbits[32];
-
-#ifdef SUPPORT_UTF8
-BOOL class_utf8;
-BOOL utf8 = (options & PCRE_UTF8) != 0;
-uschar *class_utf8data;
-uschar utf8_char[6];
-#else
-BOOL utf8 = FALSE;
-#endif
-
-/* Set up the default and non-default settings for greediness */
-
-greedy_default = ((options & PCRE_UNGREEDY) != 0);
-greedy_non_default = greedy_default ^ 1;
-
-/* Initialize no first byte, no required byte. REQ_UNSET means "no char
-matching encountered yet". It gets changed to REQ_NONE if we hit something that
-matches a non-fixed char first char; reqbyte just remains unset if we never
-find one.
-
-When we hit a repeat whose minimum is zero, we may have to adjust these values
-to take the zero repeat into account. This is implemented by setting them to
-zerofirstbyte and zeroreqbyte when such a repeat is encountered. The individual
-item types that can be repeated set these backoff variables appropriately. */
-
-firstbyte = reqbyte = zerofirstbyte = zeroreqbyte = REQ_UNSET;
-
-/* The variable req_caseopt contains either the REQ_CASELESS value or zero,
-according to the current setting of the caseless flag. REQ_CASELESS is a bit
-value > 255. It is added into the firstbyte or reqbyte variables to record the
-case status of the value. This is used only for ASCII characters. */
-
-req_caseopt = ((options & PCRE_CASELESS) != 0)? REQ_CASELESS : 0;
-
-/* Switch on next character until the end of the branch */
-
-for (;; ptr++)
- {
- BOOL negate_class;
- BOOL possessive_quantifier;
- BOOL is_quantifier;
- int class_charcount;
- int class_lastchar;
- int newoptions;
- int recno;
- int skipbytes;
- int subreqbyte;
- int subfirstbyte;
- int mclength;
- uschar mcbuffer[8];
-
- /* Next byte in the pattern */
-
- c = *ptr;
-
- /* If in \Q...\E, check for the end; if not, we have a literal */
-
- if (inescq && c != 0)
- {
- if (c == '\\' && ptr[1] == 'E')
- {
- inescq = FALSE;
- ptr++;
- continue;
- }
- else
- {
- if (previous_callout != NULL)
- {
- complete_callout(previous_callout, ptr, cd);
- previous_callout = NULL;
- }
- if ((options & PCRE_AUTO_CALLOUT) != 0)
- {
- previous_callout = code;
- code = auto_callout(code, ptr, cd);
- }
- goto NORMAL_CHAR;
- }
- }
-
- /* Fill in length of a previous callout, except when the next thing is
- a quantifier. */
-
- is_quantifier = c == '*' || c == '+' || c == '?' ||
- (c == '{' && is_counted_repeat(ptr+1));
-
- if (!is_quantifier && previous_callout != NULL &&
- after_manual_callout-- <= 0)
- {
- complete_callout(previous_callout, ptr, cd);
- previous_callout = NULL;
- }
-
- /* In extended mode, skip white space and comments */
-
- if ((options & PCRE_EXTENDED) != 0)
- {
- if ((cd->ctypes[c] & ctype_space) != 0) continue;
- if (c == '#')
- {
- /* The space before the ; is to avoid a warning on a silly compiler
- on the Macintosh. */
- while ((c = *(++ptr)) != 0 && c != NEWLINE) ;
- if (c != 0) continue; /* Else fall through to handle end of string */
- }
- }
-
- /* No auto callout for quantifiers. */
-
- if ((options & PCRE_AUTO_CALLOUT) != 0 && !is_quantifier)
- {
- previous_callout = code;
- code = auto_callout(code, ptr, cd);
- }
-
- switch(c)
- {
- /* The branch terminates at end of string, |, or ). */
-
- case 0:
- case '|':
- case ')':
- *firstbyteptr = firstbyte;
- *reqbyteptr = reqbyte;
- *codeptr = code;
- *ptrptr = ptr;
- return TRUE;
-
- /* Handle single-character metacharacters. In multiline mode, ^ disables
- the setting of any following char as a first character. */
-
- case '^':
- if ((options & PCRE_MULTILINE) != 0)
- {
- if (firstbyte == REQ_UNSET) firstbyte = REQ_NONE;
- }
- previous = NULL;
- *code++ = OP_CIRC;
- break;
-
- case '$':
- previous = NULL;
- *code++ = OP_DOLL;
- break;
-
- /* There can never be a first char if '.' is first, whatever happens about
- repeats. The value of reqbyte doesn't change either. */
-
- case '.':
- if (firstbyte == REQ_UNSET) firstbyte = REQ_NONE;
- zerofirstbyte = firstbyte;
- zeroreqbyte = reqbyte;
- previous = code;
- *code++ = OP_ANY;
- break;
-
- /* Character classes. If the included characters are all < 255 in value, we
- build a 32-byte bitmap of the permitted characters, except in the special
- case where there is only one such character. For negated classes, we build
- the map as usual, then invert it at the end. However, we use a different
- opcode so that data characters > 255 can be handled correctly.
-
- If the class contains characters outside the 0-255 range, a different
- opcode is compiled. It may optionally have a bit map for characters < 256,
- but those above are are explicitly listed afterwards. A flag byte tells
- whether the bitmap is present, and whether this is a negated class or not.
- */
-
- case '[':
- previous = code;
-
- /* PCRE supports POSIX class stuff inside a class. Perl gives an error if
- they are encountered at the top level, so we'll do that too. */
-
- if ((ptr[1] == ':' || ptr[1] == '.' || ptr[1] == '=') &&
- check_posix_syntax(ptr, &tempptr, cd))
- {
- *errorptr = (ptr[1] == ':')? ERR13 : ERR31;
- goto FAILED;
- }
-
- /* If the first character is '^', set the negation flag and skip it. */
-
- if ((c = *(++ptr)) == '^')
- {
- negate_class = TRUE;
- c = *(++ptr);
- }
- else
- {
- negate_class = FALSE;
- }
-
- /* Keep a count of chars with values < 256 so that we can optimize the case
- of just a single character (as long as it's < 256). For higher valued UTF-8
- characters, we don't yet do any optimization. */
-
- class_charcount = 0;
- class_lastchar = -1;
-
-#ifdef SUPPORT_UTF8
- class_utf8 = FALSE; /* No chars >= 256 */
- class_utf8data = code + LINK_SIZE + 34; /* For UTF-8 items */
-#endif
-
- /* Initialize the 32-char bit map to all zeros. We have to build the
- map in a temporary bit of store, in case the class contains only 1
- character (< 256), because in that case the compiled code doesn't use the
- bit map. */
-
- memset(classbits, 0, 32 * sizeof(uschar));
-
- /* Process characters until ] is reached. By writing this as a "do" it
- means that an initial ] is taken as a data character. The first pass
- through the regex checked the overall syntax, so we don't need to be very
- strict here. At the start of the loop, c contains the first byte of the
- character. */
-
- do
- {
-#ifdef SUPPORT_UTF8
- if (utf8 && c > 127)
- { /* Braces are required because the */
- GETCHARLEN(c, ptr, ptr); /* macro generates multiple statements */
- }
-#endif
-
- /* Inside \Q...\E everything is literal except \E */
-
- if (inescq)
- {
- if (c == '\\' && ptr[1] == 'E')
- {
- inescq = FALSE;
- ptr++;
- continue;
- }
- else goto LONE_SINGLE_CHARACTER;
- }
-
- /* Handle POSIX class names. Perl allows a negation extension of the
- form [:^name:]. A square bracket that doesn't match the syntax is
- treated as a literal. We also recognize the POSIX constructions
- [.ch.] and [=ch=] ("collating elements") and fault them, as Perl
- 5.6 and 5.8 do. */
-
- if (c == '[' &&
- (ptr[1] == ':' || ptr[1] == '.' || ptr[1] == '=') &&
- check_posix_syntax(ptr, &tempptr, cd))
- {
- BOOL local_negate = FALSE;
- int posix_class, i;
- register const uschar *cbits = cd->cbits;
-
- if (ptr[1] != ':')
- {
- *errorptr = ERR31;
- goto FAILED;
- }
-
- ptr += 2;
- if (*ptr == '^')
- {
- local_negate = TRUE;
- ptr++;
- }
-
- posix_class = check_posix_name(ptr, tempptr - ptr);
- if (posix_class < 0)
- {
- *errorptr = ERR30;
- goto FAILED;
- }
-
- /* If matching is caseless, upper and lower are converted to
- alpha. This relies on the fact that the class table starts with
- alpha, lower, upper as the first 3 entries. */
-
- if ((options & PCRE_CASELESS) != 0 && posix_class <= 2)
- posix_class = 0;
-
- /* Or into the map we are building up to 3 of the static class
- tables, or their negations. The [:blank:] class sets up the same
- chars as the [:space:] class (all white space). We remove the vertical
- white space chars afterwards. */
-
- posix_class *= 3;
- for (i = 0; i < 3; i++)
- {
- BOOL blankclass = strncmp((char *)ptr, "blank", 5) == 0;
- int taboffset = posix_class_maps[posix_class + i];
- if (taboffset < 0) break;
- if (local_negate)
- {
- if (i == 0)
- for (c = 0; c < 32; c++) classbits[c] |= ~cbits[c+taboffset];
- else
- for (c = 0; c < 32; c++) classbits[c] &= ~cbits[c+taboffset];
- if (blankclass) classbits[1] |= 0x3c;
- }
- else
- {
- for (c = 0; c < 32; c++) classbits[c] |= cbits[c+taboffset];
- if (blankclass) classbits[1] &= ~0x3c;
- }
- }
-
- ptr = tempptr + 1;
- class_charcount = 10; /* Set > 1; assumes more than 1 per class */
- continue; /* End of POSIX syntax handling */
- }
-
- /* Backslash may introduce a single character, or it may introduce one
- of the specials, which just set a flag. Escaped items are checked for
- validity in the pre-compiling pass. The sequence \b is a special case.
- Inside a class (and only there) it is treated as backspace. Elsewhere
- it marks a word boundary. Other escapes have preset maps ready to
- or into the one we are building. We assume they have more than one
- character in them, so set class_charcount bigger than one. */
-
- if (c == '\\')
- {
- c = check_escape(&ptr, errorptr, *brackets, options, TRUE);
-
- if (-c == ESC_b) c = '\b'; /* \b is backslash in a class */
- else if (-c == ESC_X) c = 'X'; /* \X is literal X in a class */
- else if (-c == ESC_Q) /* Handle start of quoted string */
- {
- if (ptr[1] == '\\' && ptr[2] == 'E')
- {
- ptr += 2; /* avoid empty string */
- }
- else inescq = TRUE;
- continue;
- }
-
- if (c < 0)
- {
- register const uschar *cbits = cd->cbits;
- class_charcount += 2; /* Greater than 1 is what matters */
- switch (-c)
- {
- case ESC_d:
- for (c = 0; c < 32; c++) classbits[c] |= cbits[c+cbit_digit];
- continue;
-
- case ESC_D:
- for (c = 0; c < 32; c++) classbits[c] |= ~cbits[c+cbit_digit];
- continue;
-
- case ESC_w:
- for (c = 0; c < 32; c++) classbits[c] |= cbits[c+cbit_word];
- continue;
-
- case ESC_W:
- for (c = 0; c < 32; c++) classbits[c] |= ~cbits[c+cbit_word];
- continue;
-
- case ESC_s:
- for (c = 0; c < 32; c++) classbits[c] |= cbits[c+cbit_space];
- classbits[1] &= ~0x08; /* Perl 5.004 onwards omits VT from \s */
- continue;
-
- case ESC_S:
- for (c = 0; c < 32; c++) classbits[c] |= ~cbits[c+cbit_space];
- classbits[1] |= 0x08; /* Perl 5.004 onwards omits VT from \s */
- continue;
-
-#ifdef SUPPORT_UCP
- case ESC_p:
- case ESC_P:
- {
- BOOL negated;
- int property = get_ucp(&ptr, &negated, errorptr);
- if (property < 0) goto FAILED;
- class_utf8 = TRUE;
- *class_utf8data++ = ((-c == ESC_p) != negated)?
- XCL_PROP : XCL_NOTPROP;
- *class_utf8data++ = property;
- class_charcount -= 2; /* Not a < 256 character */
- }
- continue;
-#endif
-
- /* Unrecognized escapes are faulted if PCRE is running in its
- strict mode. By default, for compatibility with Perl, they are
- treated as literals. */
-
- default:
- if ((options & PCRE_EXTRA) != 0)
- {
- *errorptr = ERR7;
- goto FAILED;
- }
- c = *ptr; /* The final character */
- class_charcount -= 2; /* Undo the default count from above */
- }
- }
-
- /* Fall through if we have a single character (c >= 0). This may be
- > 256 in UTF-8 mode. */
-
- } /* End of backslash handling */
-
- /* A single character may be followed by '-' to form a range. However,
- Perl does not permit ']' to be the end of the range. A '-' character
- here is treated as a literal. */
-
- if (ptr[1] == '-' && ptr[2] != ']')
- {
- int d;
- ptr += 2;
-
-#ifdef SUPPORT_UTF8
- if (utf8)
- { /* Braces are required because the */
- GETCHARLEN(d, ptr, ptr); /* macro generates multiple statements */
- }
- else
-#endif
- d = *ptr; /* Not UTF-8 mode */
-
- /* The second part of a range can be a single-character escape, but
- not any of the other escapes. Perl 5.6 treats a hyphen as a literal
- in such circumstances. */
-
- if (d == '\\')
- {
- const uschar *oldptr = ptr;
- d = check_escape(&ptr, errorptr, *brackets, options, TRUE);
-
- /* \b is backslash; \X is literal X; any other special means the '-'
- was literal */
-
- if (d < 0)
- {
- if (d == -ESC_b) d = '\b';
- else if (d == -ESC_X) d = 'X'; else
- {
- ptr = oldptr - 2;
- goto LONE_SINGLE_CHARACTER; /* A few lines below */
- }
- }
- }
-
- /* The check that the two values are in the correct order happens in
- the pre-pass. Optimize one-character ranges */
-
- if (d == c) goto LONE_SINGLE_CHARACTER; /* A few lines below */
-
- /* In UTF-8 mode, if the upper limit is > 255, or > 127 for caseless
- matching, we have to use an XCLASS with extra data items. Caseless
- matching for characters > 127 is available only if UCP support is
- available. */
-
-#ifdef SUPPORT_UTF8
- if (utf8 && (d > 255 || ((options & PCRE_CASELESS) != 0 && d > 127)))
- {
- class_utf8 = TRUE;
-
- /* With UCP support, we can find the other case equivalents of
- the relevant characters. There may be several ranges. Optimize how
- they fit with the basic range. */
-
-#ifdef SUPPORT_UCP
- if ((options & PCRE_CASELESS) != 0)
- {
- int occ, ocd;
- int cc = c;
- int origd = d;
- while (get_othercase_range(&cc, origd, &occ, &ocd))
- {
- if (occ >= c && ocd <= d) continue; /* Skip embedded ranges */
-
- if (occ < c && ocd >= c - 1) /* Extend the basic range */
- { /* if there is overlap, */
- c = occ; /* noting that if occ < c */
- continue; /* we can't have ocd > d */
- } /* because a subrange is */
- if (ocd > d && occ <= d + 1) /* always shorter than */
- { /* the basic range. */
- d = ocd;
- continue;
- }
-
- if (occ == ocd)
- {
- *class_utf8data++ = XCL_SINGLE;
- }
- else
- {
- *class_utf8data++ = XCL_RANGE;
- class_utf8data += ord2utf8(occ, class_utf8data);
- }
- class_utf8data += ord2utf8(ocd, class_utf8data);
- }
- }
-#endif /* SUPPORT_UCP */
-
- /* Now record the original range, possibly modified for UCP caseless
- overlapping ranges. */
-
- *class_utf8data++ = XCL_RANGE;
- class_utf8data += ord2utf8(c, class_utf8data);
- class_utf8data += ord2utf8(d, class_utf8data);
-
- /* With UCP support, we are done. Without UCP support, there is no
- caseless matching for UTF-8 characters > 127; we can use the bit map
- for the smaller ones. */
-
-#ifdef SUPPORT_UCP
- continue; /* With next character in the class */
-#else
- if ((options & PCRE_CASELESS) == 0 || c > 127) continue;
-
- /* Adjust upper limit and fall through to set up the map */
-
- d = 127;
-
-#endif /* SUPPORT_UCP */
- }
-#endif /* SUPPORT_UTF8 */
-
- /* We use the bit map for all cases when not in UTF-8 mode; else
- ranges that lie entirely within 0-127 when there is UCP support; else
- for partial ranges without UCP support. */
-
- for (; c <= d; c++)
- {
- classbits[c/8] |= (1 << (c&7));
- if ((options & PCRE_CASELESS) != 0)
- {
- int uc = cd->fcc[c]; /* flip case */
- classbits[uc/8] |= (1 << (uc&7));
- }
- class_charcount++; /* in case a one-char range */
- class_lastchar = c;
- }
-
- continue; /* Go get the next char in the class */
- }
-
- /* Handle a lone single character - we can get here for a normal
- non-escape char, or after \ that introduces a single character or for an
- apparent range that isn't. */
-
- LONE_SINGLE_CHARACTER:
-
- /* Handle a character that cannot go in the bit map */
-
-#ifdef SUPPORT_UTF8
- if (utf8 && (c > 255 || ((options & PCRE_CASELESS) != 0 && c > 127)))
- {
- class_utf8 = TRUE;
- *class_utf8data++ = XCL_SINGLE;
- class_utf8data += ord2utf8(c, class_utf8data);
-
-#ifdef SUPPORT_UCP
- if ((options & PCRE_CASELESS) != 0)
- {
- int chartype;
- int othercase;
- if (ucp_findchar(c, &chartype, &othercase) >= 0 && othercase > 0)
- {
- *class_utf8data++ = XCL_SINGLE;
- class_utf8data += ord2utf8(othercase, class_utf8data);
- }
- }
-#endif /* SUPPORT_UCP */
-
- }
- else
-#endif /* SUPPORT_UTF8 */
-
- /* Handle a single-byte character */
- {
- classbits[c/8] |= (1 << (c&7));
- if ((options & PCRE_CASELESS) != 0)
- {
- c = cd->fcc[c]; /* flip case */
- classbits[c/8] |= (1 << (c&7));
- }
- class_charcount++;
- class_lastchar = c;
- }
- }
-
- /* Loop until ']' reached; the check for end of string happens inside the
- loop. This "while" is the end of the "do" above. */
-
- while ((c = *(++ptr)) != ']' || inescq);
-
- /* If class_charcount is 1, we saw precisely one character whose value is
- less than 256. In non-UTF-8 mode we can always optimize. In UTF-8 mode, we
- can optimize the negative case only if there were no characters >= 128
- because OP_NOT and the related opcodes like OP_NOTSTAR operate on
- single-bytes only. This is an historical hangover. Maybe one day we can
- tidy these opcodes to handle multi-byte characters.
-
- The optimization throws away the bit map. We turn the item into a
- 1-character OP_CHAR[NC] if it's positive, or OP_NOT if it's negative. Note
- that OP_NOT does not support multibyte characters. In the positive case, it
- can cause firstbyte to be set. Otherwise, there can be no first char if
- this item is first, whatever repeat count may follow. In the case of
- reqbyte, save the previous value for reinstating. */
-
-#ifdef SUPPORT_UTF8
- if (class_charcount == 1 &&
- (!utf8 ||
- (!class_utf8 && (!negate_class || class_lastchar < 128))))
-
-#else
- if (class_charcount == 1)
-#endif
- {
- zeroreqbyte = reqbyte;
-
- /* The OP_NOT opcode works on one-byte characters only. */
-
- if (negate_class)
- {
- if (firstbyte == REQ_UNSET) firstbyte = REQ_NONE;
- zerofirstbyte = firstbyte;
- *code++ = OP_NOT;
- *code++ = class_lastchar;
- break;
- }
-
- /* For a single, positive character, get the value into mcbuffer, and
- then we can handle this with the normal one-character code. */
-
-#ifdef SUPPORT_UTF8
- if (utf8 && class_lastchar > 127)
- mclength = ord2utf8(class_lastchar, mcbuffer);
- else
-#endif
- {
- mcbuffer[0] = class_lastchar;
- mclength = 1;
- }
- goto ONE_CHAR;
- } /* End of 1-char optimization */
-
- /* The general case - not the one-char optimization. If this is the first
- thing in the branch, there can be no first char setting, whatever the
- repeat count. Any reqbyte setting must remain unchanged after any kind of
- repeat. */
-
- if (firstbyte == REQ_UNSET) firstbyte = REQ_NONE;
- zerofirstbyte = firstbyte;
- zeroreqbyte = reqbyte;
-
- /* If there are characters with values > 255, we have to compile an
- extended class, with its own opcode. If there are no characters < 256,
- we can omit the bitmap. */
-
-#ifdef SUPPORT_UTF8
- if (class_utf8)
- {
- *class_utf8data++ = XCL_END; /* Marks the end of extra data */
- *code++ = OP_XCLASS;
- code += LINK_SIZE;
- *code = negate_class? XCL_NOT : 0;
-
- /* If the map is required, install it, and move on to the end of
- the extra data */
-
- if (class_charcount > 0)
- {
- *code++ |= XCL_MAP;
- memcpy(code, classbits, 32);
- code = class_utf8data;
- }
-
- /* If the map is not required, slide down the extra data. */
-
- else
- {
- int len = class_utf8data - (code + 33);
- memmove(code + 1, code + 33, len);
- code += len + 1;
- }
-
- /* Now fill in the complete length of the item */
-
- PUT(previous, 1, code - previous);
- break; /* End of class handling */
- }
-#endif
-
- /* If there are no characters > 255, negate the 32-byte map if necessary,
- and copy it into the code vector. If this is the first thing in the branch,
- there can be no first char setting, whatever the repeat count. Any reqbyte
- setting must remain unchanged after any kind of repeat. */
-
- if (negate_class)
- {
- *code++ = OP_NCLASS;
- for (c = 0; c < 32; c++) code[c] = ~classbits[c];
- }
- else
- {
- *code++ = OP_CLASS;
- memcpy(code, classbits, 32);
- }
- code += 32;
- break;
-
- /* Various kinds of repeat; '{' is not necessarily a quantifier, but this
- has been tested above. */
-
- case '{':
- if (!is_quantifier) goto NORMAL_CHAR;
- ptr = read_repeat_counts(ptr+1, &repeat_min, &repeat_max, errorptr);
- if (*errorptr != NULL) goto FAILED;
- goto REPEAT;
-
- case '*':
- repeat_min = 0;
- repeat_max = -1;
- goto REPEAT;
-
- case '+':
- repeat_min = 1;
- repeat_max = -1;
- goto REPEAT;
-
- case '?':
- repeat_min = 0;
- repeat_max = 1;
-
- REPEAT:
- if (previous == NULL)
- {
- *errorptr = ERR9;
- goto FAILED;
- }
-
- if (repeat_min == 0)
- {
- firstbyte = zerofirstbyte; /* Adjust for zero repeat */
- reqbyte = zeroreqbyte; /* Ditto */
- }
-
- /* Remember whether this is a variable length repeat */
-
- reqvary = (repeat_min == repeat_max)? 0 : REQ_VARY;
-
- op_type = 0; /* Default single-char op codes */
- possessive_quantifier = FALSE; /* Default not possessive quantifier */
-
- /* Save start of previous item, in case we have to move it up to make space
- for an inserted OP_ONCE for the additional '+' extension. */
-
- tempcode = previous;
-
- /* If the next character is '+', we have a possessive quantifier. This
- implies greediness, whatever the setting of the PCRE_UNGREEDY option.
- If the next character is '?' this is a minimizing repeat, by default,
- but if PCRE_UNGREEDY is set, it works the other way round. We change the
- repeat type to the non-default. */
-
- if (ptr[1] == '+')
- {
- repeat_type = 0; /* Force greedy */
- possessive_quantifier = TRUE;
- ptr++;
- }
- else if (ptr[1] == '?')
- {
- repeat_type = greedy_non_default;
- ptr++;
- }
- else repeat_type = greedy_default;
-
- /* If previous was a recursion, we need to wrap it inside brackets so that
- it can be replicated if necessary. */
-
- if (*previous == OP_RECURSE)
- {
- memmove(previous + 1 + LINK_SIZE, previous, 1 + LINK_SIZE);
- code += 1 + LINK_SIZE;
- *previous = OP_BRA;
- PUT(previous, 1, code - previous);
- *code = OP_KET;
- PUT(code, 1, code - previous);
- code += 1 + LINK_SIZE;
- }
-
- /* If previous was a character match, abolish the item and generate a
- repeat item instead. If a char item has a minumum of more than one, ensure
- that it is set in reqbyte - it might not be if a sequence such as x{3} is
- the first thing in a branch because the x will have gone into firstbyte
- instead. */
-
- if (*previous == OP_CHAR || *previous == OP_CHARNC)
- {
- /* Deal with UTF-8 characters that take up more than one byte. It's
- easier to write this out separately than try to macrify it. Use c to
- hold the length of the character in bytes, plus 0x80 to flag that it's a
- length rather than a small character. */
-
-#ifdef SUPPORT_UTF8
- if (utf8 && (code[-1] & 0x80) != 0)
- {
- uschar *lastchar = code - 1;
- while((*lastchar & 0xc0) == 0x80) lastchar--;
- c = code - lastchar; /* Length of UTF-8 character */
- memcpy(utf8_char, lastchar, c); /* Save the char */
- c |= 0x80; /* Flag c as a length */
- }
- else
-#endif
-
- /* Handle the case of a single byte - either with no UTF8 support, or
- with UTF-8 disabled, or for a UTF-8 character < 128. */
-
- {
- c = code[-1];
- if (repeat_min > 1) reqbyte = c | req_caseopt | cd->req_varyopt;
- }
-
- goto OUTPUT_SINGLE_REPEAT; /* Code shared with single character types */
- }
-
- /* If previous was a single negated character ([^a] or similar), we use
- one of the special opcodes, replacing it. The code is shared with single-
- character repeats by setting opt_type to add a suitable offset into
- repeat_type. OP_NOT is currently used only for single-byte chars. */
-
- else if (*previous == OP_NOT)
- {
- op_type = OP_NOTSTAR - OP_STAR; /* Use "not" opcodes */
- c = previous[1];
- goto OUTPUT_SINGLE_REPEAT;
- }
-
- /* If previous was a character type match (\d or similar), abolish it and
- create a suitable repeat item. The code is shared with single-character
- repeats by setting op_type to add a suitable offset into repeat_type. Note
- the the Unicode property types will be present only when SUPPORT_UCP is
- defined, but we don't wrap the little bits of code here because it just
- makes it horribly messy. */
-
- else if (*previous < OP_EODN)
- {
- uschar *oldcode;
- int prop_type;
- op_type = OP_TYPESTAR - OP_STAR; /* Use type opcodes */
- c = *previous;
-
- OUTPUT_SINGLE_REPEAT:
- prop_type = (*previous == OP_PROP || *previous == OP_NOTPROP)?
- previous[1] : -1;
-
- oldcode = code;
- code = previous; /* Usually overwrite previous item */
-
- /* If the maximum is zero then the minimum must also be zero; Perl allows
- this case, so we do too - by simply omitting the item altogether. */
-
- if (repeat_max == 0) goto END_REPEAT;
-
- /* All real repeats make it impossible to handle partial matching (maybe
- one day we will be able to remove this restriction). */
-
- if (repeat_max != 1) cd->nopartial = TRUE;
-
- /* Combine the op_type with the repeat_type */
-
- repeat_type += op_type;
-
- /* A minimum of zero is handled either as the special case * or ?, or as
- an UPTO, with the maximum given. */
-
- if (repeat_min == 0)
- {
- if (repeat_max == -1) *code++ = OP_STAR + repeat_type;
- else if (repeat_max == 1) *code++ = OP_QUERY + repeat_type;
- else
- {
- *code++ = OP_UPTO + repeat_type;
- PUT2INC(code, 0, repeat_max);
- }
- }
-
- /* A repeat minimum of 1 is optimized into some special cases. If the
- maximum is unlimited, we use OP_PLUS. Otherwise, the original item it
- left in place and, if the maximum is greater than 1, we use OP_UPTO with
- one less than the maximum. */
-
- else if (repeat_min == 1)
- {
- if (repeat_max == -1)
- *code++ = OP_PLUS + repeat_type;
- else
- {
- code = oldcode; /* leave previous item in place */
- if (repeat_max == 1) goto END_REPEAT;
- *code++ = OP_UPTO + repeat_type;
- PUT2INC(code, 0, repeat_max - 1);
- }
- }
-
- /* The case {n,n} is just an EXACT, while the general case {n,m} is
- handled as an EXACT followed by an UPTO. */
-
- else
- {
- *code++ = OP_EXACT + op_type; /* NB EXACT doesn't have repeat_type */
- PUT2INC(code, 0, repeat_min);
-
- /* If the maximum is unlimited, insert an OP_STAR. Before doing so,
- we have to insert the character for the previous code. For a repeated
- Unicode property match, there is an extra byte that defines the
- required property. In UTF-8 mode, long characters have their length in
- c, with the 0x80 bit as a flag. */
-
- if (repeat_max < 0)
- {
-#ifdef SUPPORT_UTF8
- if (utf8 && c >= 128)
- {
- memcpy(code, utf8_char, c & 7);
- code += c & 7;
- }
- else
-#endif
- {
- *code++ = c;
- if (prop_type >= 0) *code++ = prop_type;
- }
- *code++ = OP_STAR + repeat_type;
- }
-
- /* Else insert an UPTO if the max is greater than the min, again
- preceded by the character, for the previously inserted code. */
-
- else if (repeat_max != repeat_min)
- {
-#ifdef SUPPORT_UTF8
- if (utf8 && c >= 128)
- {
- memcpy(code, utf8_char, c & 7);
- code += c & 7;
- }
- else
-#endif
- *code++ = c;
- if (prop_type >= 0) *code++ = prop_type;
- repeat_max -= repeat_min;
- *code++ = OP_UPTO + repeat_type;
- PUT2INC(code, 0, repeat_max);
- }
- }
-
- /* The character or character type itself comes last in all cases. */
-
-#ifdef SUPPORT_UTF8
- if (utf8 && c >= 128)
- {
- memcpy(code, utf8_char, c & 7);
- code += c & 7;
- }
- else
-#endif
- *code++ = c;
-
- /* For a repeated Unicode property match, there is an extra byte that
- defines the required property. */
-
-#ifdef SUPPORT_UCP
- if (prop_type >= 0) *code++ = prop_type;
-#endif
- }
-
- /* If previous was a character class or a back reference, we put the repeat
- stuff after it, but just skip the item if the repeat was {0,0}. */
-
- else if (*previous == OP_CLASS ||
- *previous == OP_NCLASS ||
-#ifdef SUPPORT_UTF8
- *previous == OP_XCLASS ||
-#endif
- *previous == OP_REF)
- {
- if (repeat_max == 0)
- {
- code = previous;
- goto END_REPEAT;
- }
-
- /* All real repeats make it impossible to handle partial matching (maybe
- one day we will be able to remove this restriction). */
-
- if (repeat_max != 1) cd->nopartial = TRUE;
-
- if (repeat_min == 0 && repeat_max == -1)
- *code++ = OP_CRSTAR + repeat_type;
- else if (repeat_min == 1 && repeat_max == -1)
- *code++ = OP_CRPLUS + repeat_type;
- else if (repeat_min == 0 && repeat_max == 1)
- *code++ = OP_CRQUERY + repeat_type;
- else
- {
- *code++ = OP_CRRANGE + repeat_type;
- PUT2INC(code, 0, repeat_min);
- if (repeat_max == -1) repeat_max = 0; /* 2-byte encoding for max */
- PUT2INC(code, 0, repeat_max);
- }
- }
-
- /* If previous was a bracket group, we may have to replicate it in certain
- cases. */
-
- else if (*previous >= OP_BRA || *previous == OP_ONCE ||
- *previous == OP_COND)
- {
- register int i;
- int ketoffset = 0;
- int len = code - previous;
- uschar *bralink = NULL;
-
- /* If the maximum repeat count is unlimited, find the end of the bracket
- by scanning through from the start, and compute the offset back to it
- from the current code pointer. There may be an OP_OPT setting following
- the final KET, so we can't find the end just by going back from the code
- pointer. */
-
- if (repeat_max == -1)
- {
- register uschar *ket = previous;
- do ket += GET(ket, 1); while (*ket != OP_KET);
- ketoffset = code - ket;
- }
-
- /* The case of a zero minimum is special because of the need to stick
- OP_BRAZERO in front of it, and because the group appears once in the
- data, whereas in other cases it appears the minimum number of times. For
- this reason, it is simplest to treat this case separately, as otherwise
- the code gets far too messy. There are several special subcases when the
- minimum is zero. */
-
- if (repeat_min == 0)
- {
- /* If the maximum is also zero, we just omit the group from the output
- altogether. */
-
- if (repeat_max == 0)
- {
- code = previous;
- goto END_REPEAT;
- }
-
- /* If the maximum is 1 or unlimited, we just have to stick in the
- BRAZERO and do no more at this point. However, we do need to adjust
- any OP_RECURSE calls inside the group that refer to the group itself or
- any internal group, because the offset is from the start of the whole
- regex. Temporarily terminate the pattern while doing this. */
-
- if (repeat_max <= 1)
- {
- *code = OP_END;
- adjust_recurse(previous, 1, utf8, cd);
- memmove(previous+1, previous, len);
- code++;
- *previous++ = OP_BRAZERO + repeat_type;
- }
-
- /* If the maximum is greater than 1 and limited, we have to replicate
- in a nested fashion, sticking OP_BRAZERO before each set of brackets.
- The first one has to be handled carefully because it's the original
- copy, which has to be moved up. The remainder can be handled by code
- that is common with the non-zero minimum case below. We have to
- adjust the value or repeat_max, since one less copy is required. Once
- again, we may have to adjust any OP_RECURSE calls inside the group. */
-
- else
- {
- int offset;
- *code = OP_END;
- adjust_recurse(previous, 2 + LINK_SIZE, utf8, cd);
- memmove(previous + 2 + LINK_SIZE, previous, len);
- code += 2 + LINK_SIZE;
- *previous++ = OP_BRAZERO + repeat_type;
- *previous++ = OP_BRA;
-
- /* We chain together the bracket offset fields that have to be
- filled in later when the ends of the brackets are reached. */
-
- offset = (bralink == NULL)? 0 : previous - bralink;
- bralink = previous;
- PUTINC(previous, 0, offset);
- }
-
- repeat_max--;
- }
-
- /* If the minimum is greater than zero, replicate the group as many
- times as necessary, and adjust the maximum to the number of subsequent
- copies that we need. If we set a first char from the group, and didn't
- set a required char, copy the latter from the former. */
-
- else
- {
- if (repeat_min > 1)
- {
- if (groupsetfirstbyte && reqbyte < 0) reqbyte = firstbyte;
- for (i = 1; i < repeat_min; i++)
- {
- memcpy(code, previous, len);
- code += len;
- }
- }
- if (repeat_max > 0) repeat_max -= repeat_min;
- }
-
- /* This code is common to both the zero and non-zero minimum cases. If
- the maximum is limited, it replicates the group in a nested fashion,
- remembering the bracket starts on a stack. In the case of a zero minimum,
- the first one was set up above. In all cases the repeat_max now specifies
- the number of additional copies needed. */
-
- if (repeat_max >= 0)
- {
- for (i = repeat_max - 1; i >= 0; i--)
- {
- *code++ = OP_BRAZERO + repeat_type;
-
- /* All but the final copy start a new nesting, maintaining the
- chain of brackets outstanding. */
-
- if (i != 0)
- {
- int offset;
- *code++ = OP_BRA;
- offset = (bralink == NULL)? 0 : code - bralink;
- bralink = code;
- PUTINC(code, 0, offset);
- }
-
- memcpy(code, previous, len);
- code += len;
- }
-
- /* Now chain through the pending brackets, and fill in their length
- fields (which are holding the chain links pro tem). */
-
- while (bralink != NULL)
- {
- int oldlinkoffset;
- int offset = code - bralink + 1;
- uschar *bra = code - offset;
- oldlinkoffset = GET(bra, 1);
- bralink = (oldlinkoffset == 0)? NULL : bralink - oldlinkoffset;
- *code++ = OP_KET;
- PUTINC(code, 0, offset);
- PUT(bra, 1, offset);
- }
- }
-
- /* If the maximum is unlimited, set a repeater in the final copy. We
- can't just offset backwards from the current code point, because we
- don't know if there's been an options resetting after the ket. The
- correct offset was computed above. */
-
- else code[-ketoffset] = OP_KETRMAX + repeat_type;
- }
-
- /* Else there's some kind of shambles */
-
- else
- {
- *errorptr = ERR11;
- goto FAILED;
- }
-
- /* If the character following a repeat is '+', we wrap the entire repeated
- item inside OP_ONCE brackets. This is just syntactic sugar, taken from
- Sun's Java package. The repeated item starts at tempcode, not at previous,
- which might be the first part of a string whose (former) last char we
- repeated. However, we don't support '+' after a greediness '?'. */
-
- if (possessive_quantifier)
- {
- int len = code - tempcode;
- memmove(tempcode + 1+LINK_SIZE, tempcode, len);
- code += 1 + LINK_SIZE;
- len += 1 + LINK_SIZE;
- tempcode[0] = OP_ONCE;
- *code++ = OP_KET;
- PUTINC(code, 0, len);
- PUT(tempcode, 1, len);
- }
-
- /* In all case we no longer have a previous item. We also set the
- "follows varying string" flag for subsequently encountered reqbytes if
- it isn't already set and we have just passed a varying length item. */
-
- END_REPEAT:
- previous = NULL;
- cd->req_varyopt |= reqvary;
- break;
-
-
- /* Start of nested bracket sub-expression, or comment or lookahead or
- lookbehind or option setting or condition. First deal with special things
- that can come after a bracket; all are introduced by ?, and the appearance
- of any of them means that this is not a referencing group. They were
- checked for validity in the first pass over the string, so we don't have to
- check for syntax errors here. */
-
- case '(':
- newoptions = options;
- skipbytes = 0;
-
- if (*(++ptr) == '?')
- {
- int set, unset;
- int *optset;
-
- switch (*(++ptr))
- {
- case '#': /* Comment; skip to ket */
- ptr++;
- while (*ptr != ')') ptr++;
- continue;
-
- case ':': /* Non-extracting bracket */
- bravalue = OP_BRA;
- ptr++;
- break;
-
- case '(':
- bravalue = OP_COND; /* Conditional group */
-
- /* Condition to test for recursion */
-
- if (ptr[1] == 'R')
- {
- code[1+LINK_SIZE] = OP_CREF;
- PUT2(code, 2+LINK_SIZE, CREF_RECURSE);
- skipbytes = 3;
- ptr += 3;
- }
-
- /* Condition to test for a numbered subpattern match. We know that
- if a digit follows ( then there will just be digits until ) because
- the syntax was checked in the first pass. */
-
- else if ((digitab[ptr[1]] && ctype_digit) != 0)
- {
- int condref; /* Don't amalgamate; some compilers */
- condref = *(++ptr) - '0'; /* grumble at autoincrement in declaration */
- while (*(++ptr) != ')') condref = condref*10 + *ptr - '0';
- if (condref == 0)
- {
- *errorptr = ERR35;
- goto FAILED;
- }
- ptr++;
- code[1+LINK_SIZE] = OP_CREF;
- PUT2(code, 2+LINK_SIZE, condref);
- skipbytes = 3;
- }
- /* For conditions that are assertions, we just fall through, having
- set bravalue above. */
- break;
-
- case '=': /* Positive lookahead */
- bravalue = OP_ASSERT;
- ptr++;
- break;
-
- case '!': /* Negative lookahead */
- bravalue = OP_ASSERT_NOT;
- ptr++;
- break;
-
- case '<': /* Lookbehinds */
- switch (*(++ptr))
- {
- case '=': /* Positive lookbehind */
- bravalue = OP_ASSERTBACK;
- ptr++;
- break;
-
- case '!': /* Negative lookbehind */
- bravalue = OP_ASSERTBACK_NOT;
- ptr++;
- break;
- }
- break;
-
- case '>': /* One-time brackets */
- bravalue = OP_ONCE;
- ptr++;
- break;
-
- case 'C': /* Callout - may be followed by digits; */
- previous_callout = code; /* Save for later completion */
- after_manual_callout = 1; /* Skip one item before completing */
- *code++ = OP_CALLOUT; /* Already checked that the terminating */
- { /* closing parenthesis is present. */
- int n = 0;
- while ((digitab[*(++ptr)] & ctype_digit) != 0)
- n = n * 10 + *ptr - '0';
- if (n > 255)
- {
- *errorptr = ERR38;
- goto FAILED;
- }
- *code++ = n;
- PUT(code, 0, ptr - cd->start_pattern + 1); /* Pattern offset */
- PUT(code, LINK_SIZE, 0); /* Default length */
- code += 2 * LINK_SIZE;
- }
- previous = NULL;
- continue;
-
- case 'P': /* Named subpattern handling */
- if (*(++ptr) == '<') /* Definition */
- {
- int i, namelen;
- uschar *slot = cd->name_table;
- const uschar *name; /* Don't amalgamate; some compilers */
- name = ++ptr; /* grumble at autoincrement in declaration */
-
- while (*ptr++ != '>');
- namelen = ptr - name - 1;
-
- for (i = 0; i < cd->names_found; i++)
- {
- int crc = memcmp(name, slot+2, namelen);
- if (crc == 0)
- {
- if (slot[2+namelen] == 0)
- {
- *errorptr = ERR43;
- goto FAILED;
- }
- crc = -1; /* Current name is substring */
- }
- if (crc < 0)
- {
- memmove(slot + cd->name_entry_size, slot,
- (cd->names_found - i) * cd->name_entry_size);
- break;
- }
- slot += cd->name_entry_size;
- }
-
- PUT2(slot, 0, *brackets + 1);
- memcpy(slot + 2, name, namelen);
- slot[2+namelen] = 0;
- cd->names_found++;
- goto NUMBERED_GROUP;
- }
-
- if (*ptr == '=' || *ptr == '>') /* Reference or recursion */
- {
- int i, namelen;
- int type = *ptr++;
- const uschar *name = ptr;
- uschar *slot = cd->name_table;
-
- while (*ptr != ')') ptr++;
- namelen = ptr - name;
-
- for (i = 0; i < cd->names_found; i++)
- {
- if (strncmp((char *)name, (char *)slot+2, namelen) == 0) break;
- slot += cd->name_entry_size;
- }
- if (i >= cd->names_found)
- {
- *errorptr = ERR15;
- goto FAILED;
- }
-
- recno = GET2(slot, 0);
-
- if (type == '>') goto HANDLE_RECURSION; /* A few lines below */
-
- /* Back reference */
-
- previous = code;
- *code++ = OP_REF;
- PUT2INC(code, 0, recno);
- cd->backref_map |= (recno < 32)? (1 << recno) : 1;
- if (recno > cd->top_backref) cd->top_backref = recno;
- continue;
- }
-
- /* Should never happen */
- break;
-
- case 'R': /* Pattern recursion */
- ptr++; /* Same as (?0) */
- /* Fall through */
-
- /* Recursion or "subroutine" call */
-
- case '0': case '1': case '2': case '3': case '4':
- case '5': case '6': case '7': case '8': case '9':
- {
- const uschar *called;
- recno = 0;
- while((digitab[*ptr] & ctype_digit) != 0)
- recno = recno * 10 + *ptr++ - '0';
-
- /* Come here from code above that handles a named recursion */
-
- HANDLE_RECURSION:
-
- previous = code;
-
- /* Find the bracket that is being referenced. Temporarily end the
- regex in case it doesn't exist. */
-
- *code = OP_END;
- called = (recno == 0)?
- cd->start_code : find_bracket(cd->start_code, utf8, recno);
-
- if (called == NULL)
- {
- *errorptr = ERR15;
- goto FAILED;
- }
-
- /* If the subpattern is still open, this is a recursive call. We
- check to see if this is a left recursion that could loop for ever,
- and diagnose that case. */
-
- if (GET(called, 1) == 0 && could_be_empty(called, code, bcptr, utf8))
- {
- *errorptr = ERR40;
- goto FAILED;
- }
-
- /* Insert the recursion/subroutine item */
-
- *code = OP_RECURSE;
- PUT(code, 1, called - cd->start_code);
- code += 1 + LINK_SIZE;
- }
- continue;
-
- /* Character after (? not specially recognized */
-
- default: /* Option setting */
- set = unset = 0;
- optset = &set;
-
- while (*ptr != ')' && *ptr != ':')
- {
- switch (*ptr++)
- {
- case '-': optset = &unset; break;
-
- case 'i': *optset |= PCRE_CASELESS; break;
- case 'm': *optset |= PCRE_MULTILINE; break;
- case 's': *optset |= PCRE_DOTALL; break;
- case 'x': *optset |= PCRE_EXTENDED; break;
- case 'U': *optset |= PCRE_UNGREEDY; break;
- case 'X': *optset |= PCRE_EXTRA; break;
- }
- }
-
- /* Set up the changed option bits, but don't change anything yet. */
-
- newoptions = (options | set) & (~unset);
-
- /* If the options ended with ')' this is not the start of a nested
- group with option changes, so the options change at this level. Compile
- code to change the ims options if this setting actually changes any of
- them. We also pass the new setting back so that it can be put at the
- start of any following branches, and when this group ends (if we are in
- a group), a resetting item can be compiled.
-
- Note that if this item is right at the start of the pattern, the
- options will have been abstracted and made global, so there will be no
- change to compile. */
-
- if (*ptr == ')')
- {
- if ((options & PCRE_IMS) != (newoptions & PCRE_IMS))
- {
- *code++ = OP_OPT;
- *code++ = newoptions & PCRE_IMS;
- }
-
- /* Change options at this level, and pass them back for use
- in subsequent branches. Reset the greedy defaults and the case
- value for firstbyte and reqbyte. */
-
- *optionsptr = options = newoptions;
- greedy_default = ((newoptions & PCRE_UNGREEDY) != 0);
- greedy_non_default = greedy_default ^ 1;
- req_caseopt = ((options & PCRE_CASELESS) != 0)? REQ_CASELESS : 0;
-
- previous = NULL; /* This item can't be repeated */
- continue; /* It is complete */
- }
-
- /* If the options ended with ':' we are heading into a nested group
- with possible change of options. Such groups are non-capturing and are
- not assertions of any kind. All we need to do is skip over the ':';
- the newoptions value is handled below. */
-
- bravalue = OP_BRA;
- ptr++;
- }
- }
-
- /* If PCRE_NO_AUTO_CAPTURE is set, all unadorned brackets become
- non-capturing and behave like (?:...) brackets */
-
- else if ((options & PCRE_NO_AUTO_CAPTURE) != 0)
- {
- bravalue = OP_BRA;
- }
-
- /* Else we have a referencing group; adjust the opcode. If the bracket
- number is greater than EXTRACT_BASIC_MAX, we set the opcode one higher, and
- arrange for the true number to follow later, in an OP_BRANUMBER item. */
-
- else
- {
- NUMBERED_GROUP:
- if (++(*brackets) > EXTRACT_BASIC_MAX)
- {
- bravalue = OP_BRA + EXTRACT_BASIC_MAX + 1;
- code[1+LINK_SIZE] = OP_BRANUMBER;
- PUT2(code, 2+LINK_SIZE, *brackets);
- skipbytes = 3;
- }
- else bravalue = OP_BRA + *brackets;
- }
-
- /* Process nested bracketed re. Assertions may not be repeated, but other
- kinds can be. We copy code into a non-register variable in order to be able
- to pass its address because some compilers complain otherwise. Pass in a
- new setting for the ims options if they have changed. */
-
- previous = (bravalue >= OP_ONCE)? code : NULL;
- *code = bravalue;
- tempcode = code;
- tempreqvary = cd->req_varyopt; /* Save value before bracket */
-
- if (!compile_regex(
- newoptions, /* The complete new option state */
- options & PCRE_IMS, /* The previous ims option state */
- brackets, /* Extracting bracket count */
- &tempcode, /* Where to put code (updated) */
- &ptr, /* Input pointer (updated) */
- errorptr, /* Where to put an error message */
- (bravalue == OP_ASSERTBACK ||
- bravalue == OP_ASSERTBACK_NOT), /* TRUE if back assert */
- skipbytes, /* Skip over OP_COND/OP_BRANUMBER */
- &subfirstbyte, /* For possible first char */
- &subreqbyte, /* For possible last char */
- bcptr, /* Current branch chain */
- cd)) /* Tables block */
- goto FAILED;
-
- /* At the end of compiling, code is still pointing to the start of the
- group, while tempcode has been updated to point past the end of the group
- and any option resetting that may follow it. The pattern pointer (ptr)
- is on the bracket. */
-
- /* If this is a conditional bracket, check that there are no more than
- two branches in the group. */
-
- else if (bravalue == OP_COND)
- {
- uschar *tc = code;
- condcount = 0;
-
- do {
- condcount++;
- tc += GET(tc,1);
- }
- while (*tc != OP_KET);
-
- if (condcount > 2)
- {
- *errorptr = ERR27;
- goto FAILED;
- }
-
- /* If there is just one branch, we must not make use of its firstbyte or
- reqbyte, because this is equivalent to an empty second branch. */
-
- if (condcount == 1) subfirstbyte = subreqbyte = REQ_NONE;
- }
-
- /* Handle updating of the required and first characters. Update for normal
- brackets of all kinds, and conditions with two branches (see code above).
- If the bracket is followed by a quantifier with zero repeat, we have to
- back off. Hence the definition of zeroreqbyte and zerofirstbyte outside the
- main loop so that they can be accessed for the back off. */
-
- zeroreqbyte = reqbyte;
- zerofirstbyte = firstbyte;
- groupsetfirstbyte = FALSE;
-
- if (bravalue >= OP_BRA || bravalue == OP_ONCE || bravalue == OP_COND)
- {
- /* If we have not yet set a firstbyte in this branch, take it from the
- subpattern, remembering that it was set here so that a repeat of more
- than one can replicate it as reqbyte if necessary. If the subpattern has
- no firstbyte, set "none" for the whole branch. In both cases, a zero
- repeat forces firstbyte to "none". */
-
- if (firstbyte == REQ_UNSET)
- {
- if (subfirstbyte >= 0)
- {
- firstbyte = subfirstbyte;
- groupsetfirstbyte = TRUE;
- }
- else firstbyte = REQ_NONE;
- zerofirstbyte = REQ_NONE;
- }
-
- /* If firstbyte was previously set, convert the subpattern's firstbyte
- into reqbyte if there wasn't one, using the vary flag that was in
- existence beforehand. */
-
- else if (subfirstbyte >= 0 && subreqbyte < 0)
- subreqbyte = subfirstbyte | tempreqvary;
-
- /* If the subpattern set a required byte (or set a first byte that isn't
- really the first byte - see above), set it. */
-
- if (subreqbyte >= 0) reqbyte = subreqbyte;
- }
-
- /* For a forward assertion, we take the reqbyte, if set. This can be
- helpful if the pattern that follows the assertion doesn't set a different
- char. For example, it's useful for /(?=abcde).+/. We can't set firstbyte
- for an assertion, however because it leads to incorrect effect for patterns
- such as /(?=a)a.+/ when the "real" "a" would then become a reqbyte instead
- of a firstbyte. This is overcome by a scan at the end if there's no
- firstbyte, looking for an asserted first char. */
-
- else if (bravalue == OP_ASSERT && subreqbyte >= 0) reqbyte = subreqbyte;
-
- /* Now update the main code pointer to the end of the group. */
-
- code = tempcode;
-
- /* Error if hit end of pattern */
-
- if (*ptr != ')')
- {
- *errorptr = ERR14;
- goto FAILED;
- }
- break;
-
- /* Check \ for being a real metacharacter; if not, fall through and handle
- it as a data character at the start of a string. Escape items are checked
- for validity in the pre-compiling pass. */
-
- case '\\':
- tempptr = ptr;
- c = check_escape(&ptr, errorptr, *brackets, options, FALSE);
-
- /* Handle metacharacters introduced by \. For ones like \d, the ESC_ values
- are arranged to be the negation of the corresponding OP_values. For the
- back references, the values are ESC_REF plus the reference number. Only
- back references and those types that consume a character may be repeated.
- We can test for values between ESC_b and ESC_Z for the latter; this may
- have to change if any new ones are ever created. */
-
- if (c < 0)
- {
- if (-c == ESC_Q) /* Handle start of quoted string */
- {
- if (ptr[1] == '\\' && ptr[2] == 'E') ptr += 2; /* avoid empty string */
- else inescq = TRUE;
- continue;
- }
-
- /* For metasequences that actually match a character, we disable the
- setting of a first character if it hasn't already been set. */
-
- if (firstbyte == REQ_UNSET && -c > ESC_b && -c < ESC_Z)
- firstbyte = REQ_NONE;
-
- /* Set values to reset to if this is followed by a zero repeat. */
-
- zerofirstbyte = firstbyte;
- zeroreqbyte = reqbyte;
-
- /* Back references are handled specially */
-
- if (-c >= ESC_REF)
- {
- int number = -c - ESC_REF;
- previous = code;
- *code++ = OP_REF;
- PUT2INC(code, 0, number);
- }
-
- /* So are Unicode property matches, if supported. We know that get_ucp
- won't fail because it was tested in the pre-pass. */
-
-#ifdef SUPPORT_UCP
- else if (-c == ESC_P || -c == ESC_p)
- {
- BOOL negated;
- int value = get_ucp(&ptr, &negated, errorptr);
- previous = code;
- *code++ = ((-c == ESC_p) != negated)? OP_PROP : OP_NOTPROP;
- *code++ = value;
- }
-#endif
-
- /* For the rest, we can obtain the OP value by negating the escape
- value */
-
- else
- {
- previous = (-c > ESC_b && -c < ESC_Z)? code : NULL;
- *code++ = -c;
- }
- continue;
- }
-
- /* We have a data character whose value is in c. In UTF-8 mode it may have
- a value > 127. We set its representation in the length/buffer, and then
- handle it as a data character. */
-
-#ifdef SUPPORT_UTF8
- if (utf8 && c > 127)
- mclength = ord2utf8(c, mcbuffer);
- else
-#endif
-
- {
- mcbuffer[0] = c;
- mclength = 1;
- }
-
- goto ONE_CHAR;
-
- /* Handle a literal character. It is guaranteed not to be whitespace or #
- when the extended flag is set. If we are in UTF-8 mode, it may be a
- multi-byte literal character. */
-
- default:
- NORMAL_CHAR:
- mclength = 1;
- mcbuffer[0] = c;
-
-#ifdef SUPPORT_UTF8
- if (utf8 && (c & 0xc0) == 0xc0)
- {
- while ((ptr[1] & 0xc0) == 0x80)
- mcbuffer[mclength++] = *(++ptr);
- }
-#endif
-
- /* At this point we have the character's bytes in mcbuffer, and the length
- in mclength. When not in UTF-8 mode, the length is always 1. */
-
- ONE_CHAR:
- previous = code;
- *code++ = ((options & PCRE_CASELESS) != 0)? OP_CHARNC : OP_CHAR;
- for (c = 0; c < mclength; c++) *code++ = mcbuffer[c];
-
- /* Set the first and required bytes appropriately. If no previous first
- byte, set it from this character, but revert to none on a zero repeat.
- Otherwise, leave the firstbyte value alone, and don't change it on a zero
- repeat. */
-
- if (firstbyte == REQ_UNSET)
- {
- zerofirstbyte = REQ_NONE;
- zeroreqbyte = reqbyte;
-
- /* If the character is more than one byte long, we can set firstbyte
- only if it is not to be matched caselessly. */
-
- if (mclength == 1 || req_caseopt == 0)
- {
- firstbyte = mcbuffer[0] | req_caseopt;
- if (mclength != 1) reqbyte = code[-1] | cd->req_varyopt;
- }
- else firstbyte = reqbyte = REQ_NONE;
- }
-
- /* firstbyte was previously set; we can set reqbyte only the length is
- 1 or the matching is caseful. */
-
- else
- {
- zerofirstbyte = firstbyte;
- zeroreqbyte = reqbyte;
- if (mclength == 1 || req_caseopt == 0)
- reqbyte = code[-1] | req_caseopt | cd->req_varyopt;
- }
-
- break; /* End of literal character handling */
- }
- } /* end of big loop */
-
-/* Control never reaches here by falling through, only by a goto for all the
-error states. Pass back the position in the pattern so that it can be displayed
-to the user for diagnosing the error. */
-
-FAILED:
-*ptrptr = ptr;
-return FALSE;
-}
-
-
-
-
-/*************************************************
-* Compile sequence of alternatives *
-*************************************************/
-
-/* On entry, ptr is pointing past the bracket character, but on return
-it points to the closing bracket, or vertical bar, or end of string.
-The code variable is pointing at the byte into which the BRA operator has been
-stored. If the ims options are changed at the start (for a (?ims: group) or
-during any branch, we need to insert an OP_OPT item at the start of every
-following branch to ensure they get set correctly at run time, and also pass
-the new options into every subsequent branch compile.
-
-Argument:
- options option bits, including any changes for this subpattern
- oldims previous settings of ims option bits
- brackets -> int containing the number of extracting brackets used
- codeptr -> the address of the current code pointer
- ptrptr -> the address of the current pattern pointer
- errorptr -> pointer to error message
- lookbehind TRUE if this is a lookbehind assertion
- skipbytes skip this many bytes at start (for OP_COND, OP_BRANUMBER)
- firstbyteptr place to put the first required character, or a negative number
- reqbyteptr place to put the last required character, or a negative number
- bcptr pointer to the chain of currently open branches
- cd points to the data block with tables pointers etc.
-
-Returns: TRUE on success
-*/
-
-static BOOL
-compile_regex(int options, int oldims, int *brackets, uschar **codeptr,
- const uschar **ptrptr, const char **errorptr, BOOL lookbehind, int skipbytes,
- int *firstbyteptr, int *reqbyteptr, branch_chain *bcptr, compile_data *cd)
-{
-const uschar *ptr = *ptrptr;
-uschar *code = *codeptr;
-uschar *last_branch = code;
-uschar *start_bracket = code;
-uschar *reverse_count = NULL;
-int firstbyte, reqbyte;
-int branchfirstbyte, branchreqbyte;
-branch_chain bc;
-
-bc.outer = bcptr;
-bc.current = code;
-
-firstbyte = reqbyte = REQ_UNSET;
-
-/* Offset is set zero to mark that this bracket is still open */
-
-PUT(code, 1, 0);
-code += 1 + LINK_SIZE + skipbytes;
-
-/* Loop for each alternative branch */
-
-for (;;)
- {
- /* Handle a change of ims options at the start of the branch */
-
- if ((options & PCRE_IMS) != oldims)
- {
- *code++ = OP_OPT;
- *code++ = options & PCRE_IMS;
- }
-
- /* Set up dummy OP_REVERSE if lookbehind assertion */
-
- if (lookbehind)
- {
- *code++ = OP_REVERSE;
- reverse_count = code;
- PUTINC(code, 0, 0);
- }
-
- /* Now compile the branch */
-
- if (!compile_branch(&options, brackets, &code, &ptr, errorptr,
- &branchfirstbyte, &branchreqbyte, &bc, cd))
- {
- *ptrptr = ptr;
- return FALSE;
- }
-
- /* If this is the first branch, the firstbyte and reqbyte values for the
- branch become the values for the regex. */
-
- if (*last_branch != OP_ALT)
- {
- firstbyte = branchfirstbyte;
- reqbyte = branchreqbyte;
- }
-
- /* If this is not the first branch, the first char and reqbyte have to
- match the values from all the previous branches, except that if the previous
- value for reqbyte didn't have REQ_VARY set, it can still match, and we set
- REQ_VARY for the regex. */
-
- else
- {
- /* If we previously had a firstbyte, but it doesn't match the new branch,
- we have to abandon the firstbyte for the regex, but if there was previously
- no reqbyte, it takes on the value of the old firstbyte. */
-
- if (firstbyte >= 0 && firstbyte != branchfirstbyte)
- {
- if (reqbyte < 0) reqbyte = firstbyte;
- firstbyte = REQ_NONE;
- }
-
- /* If we (now or from before) have no firstbyte, a firstbyte from the
- branch becomes a reqbyte if there isn't a branch reqbyte. */
-
- if (firstbyte < 0 && branchfirstbyte >= 0 && branchreqbyte < 0)
- branchreqbyte = branchfirstbyte;
-
- /* Now ensure that the reqbytes match */
-
- if ((reqbyte & ~REQ_VARY) != (branchreqbyte & ~REQ_VARY))
- reqbyte = REQ_NONE;
- else reqbyte |= branchreqbyte; /* To "or" REQ_VARY */
- }
-
- /* If lookbehind, check that this branch matches a fixed-length string,
- and put the length into the OP_REVERSE item. Temporarily mark the end of
- the branch with OP_END. */
-
- if (lookbehind)
- {
- int length;
- *code = OP_END;
- length = find_fixedlength(last_branch, options);
- DPRINTF(("fixed length = %d\n", length));
- if (length < 0)
- {
- *errorptr = (length == -2)? ERR36 : ERR25;
- *ptrptr = ptr;
- return FALSE;
- }
- PUT(reverse_count, 0, length);
- }
-
- /* Reached end of expression, either ')' or end of pattern. Go back through
- the alternative branches and reverse the chain of offsets, with the field in
- the BRA item now becoming an offset to the first alternative. If there are
- no alternatives, it points to the end of the group. The length in the
- terminating ket is always the length of the whole bracketed item. If any of
- the ims options were changed inside the group, compile a resetting op-code
- following, except at the very end of the pattern. Return leaving the pointer
- at the terminating char. */
-
- if (*ptr != '|')
- {
- int length = code - last_branch;
- do
- {
- int prev_length = GET(last_branch, 1);
- PUT(last_branch, 1, length);
- length = prev_length;
- last_branch -= length;
- }
- while (length > 0);
-
- /* Fill in the ket */
-
- *code = OP_KET;
- PUT(code, 1, code - start_bracket);
- code += 1 + LINK_SIZE;
-
- /* Resetting option if needed */
-
- if ((options & PCRE_IMS) != oldims && *ptr == ')')
- {
- *code++ = OP_OPT;
- *code++ = oldims;
- }
-
- /* Set values to pass back */
-
- *codeptr = code;
- *ptrptr = ptr;
- *firstbyteptr = firstbyte;
- *reqbyteptr = reqbyte;
- return TRUE;
- }
-
- /* Another branch follows; insert an "or" node. Its length field points back
- to the previous branch while the bracket remains open. At the end the chain
- is reversed. It's done like this so that the start of the bracket has a
- zero offset until it is closed, making it possible to detect recursion. */
-
- *code = OP_ALT;
- PUT(code, 1, code - last_branch);
- bc.current = last_branch = code;
- code += 1 + LINK_SIZE;
- ptr++;
- }
-/* Control never reaches here */
-}
-
-
-
-
-/*************************************************
-* Check for anchored expression *
-*************************************************/
-
-/* Try to find out if this is an anchored regular expression. Consider each
-alternative branch. If they all start with OP_SOD or OP_CIRC, or with a bracket
-all of whose alternatives start with OP_SOD or OP_CIRC (recurse ad lib), then
-it's anchored. However, if this is a multiline pattern, then only OP_SOD
-counts, since OP_CIRC can match in the middle.
-
-We can also consider a regex to be anchored if OP_SOM starts all its branches.
-This is the code for \G, which means "match at start of match position, taking
-into account the match offset".
-
-A branch is also implicitly anchored if it starts with .* and DOTALL is set,
-because that will try the rest of the pattern at all possible matching points,
-so there is no point trying again.... er ....
-
-.... except when the .* appears inside capturing parentheses, and there is a
-subsequent back reference to those parentheses. We haven't enough information
-to catch that case precisely.
-
-At first, the best we could do was to detect when .* was in capturing brackets
-and the highest back reference was greater than or equal to that level.
-However, by keeping a bitmap of the first 31 back references, we can catch some
-of the more common cases more precisely.
-
-Arguments:
- code points to start of expression (the bracket)
- options points to the options setting
- bracket_map a bitmap of which brackets we are inside while testing; this
- handles up to substring 31; after that we just have to take
- the less precise approach
- backref_map the back reference bitmap
-
-Returns: TRUE or FALSE
-*/
-
-static BOOL
-is_anchored(register const uschar *code, int *options, unsigned int bracket_map,
- unsigned int backref_map)
-{
-do {
- const uschar *scode =
- first_significant_code(code + 1+LINK_SIZE, options, PCRE_MULTILINE, FALSE);
- register int op = *scode;
-
- /* Capturing brackets */
-
- if (op > OP_BRA)
- {
- int new_map;
- op -= OP_BRA;
- if (op > EXTRACT_BASIC_MAX) op = GET2(scode, 2+LINK_SIZE);
- new_map = bracket_map | ((op < 32)? (1 << op) : 1);
- if (!is_anchored(scode, options, new_map, backref_map)) return FALSE;
- }
-
- /* Other brackets */
-
- else if (op == OP_BRA || op == OP_ASSERT || op == OP_ONCE || op == OP_COND)
- {
- if (!is_anchored(scode, options, bracket_map, backref_map)) return FALSE;
- }
-
- /* .* is not anchored unless DOTALL is set and it isn't in brackets that
- are or may be referenced. */
-
- else if ((op == OP_TYPESTAR || op == OP_TYPEMINSTAR) &&
- (*options & PCRE_DOTALL) != 0)
- {
- if (scode[1] != OP_ANY || (bracket_map & backref_map) != 0) return FALSE;
- }
-
- /* Check for explicit anchoring */
-
- else if (op != OP_SOD && op != OP_SOM &&
- ((*options & PCRE_MULTILINE) != 0 || op != OP_CIRC))
- return FALSE;
- code += GET(code, 1);
- }
-while (*code == OP_ALT); /* Loop for each alternative */
-return TRUE;
-}
-
-
-
-/*************************************************
-* Check for starting with ^ or .* *
-*************************************************/
-
-/* This is called to find out if every branch starts with ^ or .* so that
-"first char" processing can be done to speed things up in multiline
-matching and for non-DOTALL patterns that start with .* (which must start at
-the beginning or after \n). As in the case of is_anchored() (see above), we
-have to take account of back references to capturing brackets that contain .*
-because in that case we can't make the assumption.
-
-Arguments:
- code points to start of expression (the bracket)
- bracket_map a bitmap of which brackets we are inside while testing; this
- handles up to substring 31; after that we just have to take
- the less precise approach
- backref_map the back reference bitmap
-
-Returns: TRUE or FALSE
-*/
-
-static BOOL
-is_startline(const uschar *code, unsigned int bracket_map,
- unsigned int backref_map)
-{
-do {
- const uschar *scode = first_significant_code(code + 1+LINK_SIZE, NULL, 0,
- FALSE);
- register int op = *scode;
-
- /* Capturing brackets */
-
- if (op > OP_BRA)
- {
- int new_map;
- op -= OP_BRA;
- if (op > EXTRACT_BASIC_MAX) op = GET2(scode, 2+LINK_SIZE);
- new_map = bracket_map | ((op < 32)? (1 << op) : 1);
- if (!is_startline(scode, new_map, backref_map)) return FALSE;
- }
-
- /* Other brackets */
-
- else if (op == OP_BRA || op == OP_ASSERT || op == OP_ONCE || op == OP_COND)
- { if (!is_startline(scode, bracket_map, backref_map)) return FALSE; }
-
- /* .* means "start at start or after \n" if it isn't in brackets that
- may be referenced. */
-
- else if (op == OP_TYPESTAR || op == OP_TYPEMINSTAR)
- {
- if (scode[1] != OP_ANY || (bracket_map & backref_map) != 0) return FALSE;
- }
-
- /* Check for explicit circumflex */
-
- else if (op != OP_CIRC) return FALSE;
-
- /* Move on to the next alternative */
-
- code += GET(code, 1);
- }
-while (*code == OP_ALT); /* Loop for each alternative */
-return TRUE;
-}
-
-
-
-/*************************************************
-* Check for asserted fixed first char *
-*************************************************/
-
-/* During compilation, the "first char" settings from forward assertions are
-discarded, because they can cause conflicts with actual literals that follow.
-However, if we end up without a first char setting for an unanchored pattern,
-it is worth scanning the regex to see if there is an initial asserted first
-char. If all branches start with the same asserted char, or with a bracket all
-of whose alternatives start with the same asserted char (recurse ad lib), then
-we return that char, otherwise -1.
-
-Arguments:
- code points to start of expression (the bracket)
- options pointer to the options (used to check casing changes)
- inassert TRUE if in an assertion
-
-Returns: -1 or the fixed first char
-*/
-
-static int
-find_firstassertedchar(const uschar *code, int *options, BOOL inassert)
-{
-register int c = -1;
-do {
- int d;
- const uschar *scode =
- first_significant_code(code + 1+LINK_SIZE, options, PCRE_CASELESS, TRUE);
- register int op = *scode;
-
- if (op >= OP_BRA) op = OP_BRA;
-
- switch(op)
- {
- default:
- return -1;
-
- case OP_BRA:
- case OP_ASSERT:
- case OP_ONCE:
- case OP_COND:
- if ((d = find_firstassertedchar(scode, options, op == OP_ASSERT)) < 0)
- return -1;
- if (c < 0) c = d; else if (c != d) return -1;
- break;
-
- case OP_EXACT: /* Fall through */
- scode += 2;
-
- case OP_CHAR:
- case OP_CHARNC:
- case OP_PLUS:
- case OP_MINPLUS:
- if (!inassert) return -1;
- if (c < 0)
- {
- c = scode[1];
- if ((*options & PCRE_CASELESS) != 0) c |= REQ_CASELESS;
- }
- else if (c != scode[1]) return -1;
- break;
- }
-
- code += GET(code, 1);
- }
-while (*code == OP_ALT);
-return c;
-}
-
-
-
-
-#ifdef SUPPORT_UTF8
-/*************************************************
-* Validate a UTF-8 string *
-*************************************************/
-
-/* This function is called (optionally) at the start of compile or match, to
-validate that a supposed UTF-8 string is actually valid. The early check means
-that subsequent code can assume it is dealing with a valid string. The check
-can be turned off for maximum performance, but then consequences of supplying
-an invalid string are then undefined.
-
-Arguments:
- string points to the string
- length length of string, or -1 if the string is zero-terminated
-
-Returns: < 0 if the string is a valid UTF-8 string
- >= 0 otherwise; the value is the offset of the bad byte
-*/
-
-static int
-valid_utf8(const uschar *string, int length)
-{
-register const uschar *p;
-
-if (length < 0)
- {
- for (p = string; *p != 0; p++);
- length = p - string;
- }
-
-for (p = string; length-- > 0; p++)
- {
- register int ab;
- register int c = *p;
- if (c < 128) continue;
- if ((c & 0xc0) != 0xc0) return p - string;
- ab = utf8_table4[c & 0x3f]; /* Number of additional bytes */
- if (length < ab) return p - string;
- length -= ab;
-
- /* Check top bits in the second byte */
- if ((*(++p) & 0xc0) != 0x80) return p - string;
-
- /* Check for overlong sequences for each different length */
- switch (ab)
- {
- /* Check for xx00 000x */
- case 1:
- if ((c & 0x3e) == 0) return p - string;
- continue; /* We know there aren't any more bytes to check */
-
- /* Check for 1110 0000, xx0x xxxx */
- case 2:
- if (c == 0xe0 && (*p & 0x20) == 0) return p - string;
- break;
-
- /* Check for 1111 0000, xx00 xxxx */
- case 3:
- if (c == 0xf0 && (*p & 0x30) == 0) return p - string;
- break;
-
- /* Check for 1111 1000, xx00 0xxx */
- case 4:
- if (c == 0xf8 && (*p & 0x38) == 0) return p - string;
- break;
-
- /* Check for leading 0xfe or 0xff, and then for 1111 1100, xx00 00xx */
- case 5:
- if (c == 0xfe || c == 0xff ||
- (c == 0xfc && (*p & 0x3c) == 0)) return p - string;
- break;
- }
-
- /* Check for valid bytes after the 2nd, if any; all must start 10 */
- while (--ab > 0)
- {
- if ((*(++p) & 0xc0) != 0x80) return p - string;
- }
- }
-
-return -1;
-}
-#endif
-
-
-
-/*************************************************
-* Compile a Regular Expression *
-*************************************************/
-
-/* This function takes a string and returns a pointer to a block of store
-holding a compiled version of the expression.
-
-Arguments:
- pattern the regular expression
- options various option bits
- errorptr pointer to pointer to error text
- erroroffset ptr offset in pattern where error was detected
- tables pointer to character tables or NULL
-
-Returns: pointer to compiled data block, or NULL on error,
- with errorptr and erroroffset set
-*/
-
-EXPORT pcre *
-pcre_compile(const char *pattern, int options, const char **errorptr,
- int *erroroffset, const unsigned char *tables)
-{
-real_pcre *re;
-int length = 1 + LINK_SIZE; /* For initial BRA plus length */
-int c, firstbyte, reqbyte;
-int bracount = 0;
-int branch_extra = 0;
-int branch_newextra;
-int item_count = -1;
-int name_count = 0;
-int max_name_size = 0;
-int lastitemlength = 0;
-#ifdef SUPPORT_UTF8
-BOOL utf8;
-BOOL class_utf8;
-#endif
-BOOL inescq = FALSE;
-unsigned int brastackptr = 0;
-size_t size;
-uschar *code;
-const uschar *codestart;
-const uschar *ptr;
-compile_data compile_block;
-int brastack[BRASTACK_SIZE];
-uschar bralenstack[BRASTACK_SIZE];
-
-/* We can't pass back an error message if errorptr is NULL; I guess the best we
-can do is just return NULL. */
-
-if (errorptr == NULL) return NULL;
-*errorptr = NULL;
-
-/* However, we can give a message for this error */
-
-if (erroroffset == NULL)
- {
- *errorptr = ERR16;
- return NULL;
- }
-*erroroffset = 0;
-
-/* Can't support UTF8 unless PCRE has been compiled to include the code. */
-
-#ifdef SUPPORT_UTF8
-utf8 = (options & PCRE_UTF8) != 0;
-if (utf8 && (options & PCRE_NO_UTF8_CHECK) == 0 &&
- (*erroroffset = valid_utf8((uschar *)pattern, -1)) >= 0)
- {
- *errorptr = ERR44;
- return NULL;
- }
-#else
-if ((options & PCRE_UTF8) != 0)
- {
- *errorptr = ERR32;
- return NULL;
- }
-#endif
-
-if ((options & ~PUBLIC_OPTIONS) != 0)
- {
- *errorptr = ERR17;
- return NULL;
- }
-
-/* Set up pointers to the individual character tables */
-
-if (tables == NULL) tables = pcre_default_tables;
-compile_block.lcc = tables + lcc_offset;
-compile_block.fcc = tables + fcc_offset;
-compile_block.cbits = tables + cbits_offset;
-compile_block.ctypes = tables + ctypes_offset;
-
-/* Maximum back reference and backref bitmap. This is updated for numeric
-references during the first pass, but for named references during the actual
-compile pass. The bitmap records up to 31 back references to help in deciding
-whether (.*) can be treated as anchored or not. */
-
-compile_block.top_backref = 0;
-compile_block.backref_map = 0;
-
-/* Reflect pattern for debugging output */
-
-DPRINTF(("------------------------------------------------------------------\n"));
-DPRINTF(("%s\n", pattern));
-
-/* The first thing to do is to make a pass over the pattern to compute the
-amount of store required to hold the compiled code. This does not have to be
-perfect as long as errors are overestimates. At the same time we can detect any
-flag settings right at the start, and extract them. Make an attempt to correct
-for any counted white space if an "extended" flag setting appears late in the
-pattern. We can't be so clever for #-comments. */
-
-ptr = (const uschar *)(pattern - 1);
-while ((c = *(++ptr)) != 0)
- {
- int min, max;
- int class_optcount;
- int bracket_length;
- int duplength;
-
- /* If we are inside a \Q...\E sequence, all chars are literal */
-
- if (inescq)
- {
- if ((options & PCRE_AUTO_CALLOUT) != 0) length += 2 + 2*LINK_SIZE;
- goto NORMAL_CHAR;
- }
-
- /* Otherwise, first check for ignored whitespace and comments */
-
- if ((options & PCRE_EXTENDED) != 0)
- {
- if ((compile_block.ctypes[c] & ctype_space) != 0) continue;
- if (c == '#')
- {
- /* The space before the ; is to avoid a warning on a silly compiler
- on the Macintosh. */
- while ((c = *(++ptr)) != 0 && c != NEWLINE) ;
- if (c == 0) break;
- continue;
- }
- }
-
- item_count++; /* Is zero for the first non-comment item */
-
- /* Allow space for auto callout before every item except quantifiers. */
-
- if ((options & PCRE_AUTO_CALLOUT) != 0 &&
- c != '*' && c != '+' && c != '?' &&
- (c != '{' || !is_counted_repeat(ptr + 1)))
- length += 2 + 2*LINK_SIZE;
-
- switch(c)
- {
- /* A backslashed item may be an escaped data character or it may be a
- character type. */
-
- case '\\':
- c = check_escape(&ptr, errorptr, bracount, options, FALSE);
- if (*errorptr != NULL) goto PCRE_ERROR_RETURN;
-
- lastitemlength = 1; /* Default length of last item for repeats */
-
- if (c >= 0) /* Data character */
- {
- length += 2; /* For a one-byte character */
-
-#ifdef SUPPORT_UTF8
- if (utf8 && c > 127)
- {
- int i;
- for (i = 0; i < sizeof(utf8_table1)/sizeof(int); i++)
- if (c <= utf8_table1[i]) break;
- length += i;
- lastitemlength += i;
- }
-#endif
-
- continue;
- }
-
- /* If \Q, enter "literal" mode */
-
- if (-c == ESC_Q)
- {
- inescq = TRUE;
- continue;
- }
-
- /* \X is supported only if Unicode property support is compiled */
-
-#ifndef SUPPORT_UCP
- if (-c == ESC_X)
- {
- *errorptr = ERR45;
- goto PCRE_ERROR_RETURN;
- }
-#endif
-
- /* \P and \p are for Unicode properties, but only when the support has
- been compiled. Each item needs 2 bytes. */
-
- else if (-c == ESC_P || -c == ESC_p)
- {
-#ifdef SUPPORT_UCP
- BOOL negated;
- length += 2;
- lastitemlength = 2;
- if (get_ucp(&ptr, &negated, errorptr) < 0) goto PCRE_ERROR_RETURN;
- continue;
-#else
- *errorptr = ERR45;
- goto PCRE_ERROR_RETURN;
-#endif
- }
-
- /* Other escapes need one byte */
-
- length++;
-
- /* A back reference needs an additional 2 bytes, plus either one or 5
- bytes for a repeat. We also need to keep the value of the highest
- back reference. */
-
- if (c <= -ESC_REF)
- {
- int refnum = -c - ESC_REF;
- compile_block.backref_map |= (refnum < 32)? (1 << refnum) : 1;
- if (refnum > compile_block.top_backref)
- compile_block.top_backref = refnum;
- length += 2; /* For single back reference */
- if (ptr[1] == '{' && is_counted_repeat(ptr+2))
- {
- ptr = read_repeat_counts(ptr+2, &min, &max, errorptr);
- if (*errorptr != NULL) goto PCRE_ERROR_RETURN;
- if ((min == 0 && (max == 1 || max == -1)) ||
- (min == 1 && max == -1))
- length++;
- else length += 5;
- if (ptr[1] == '?') ptr++;
- }
- }
- continue;
-
- case '^': /* Single-byte metacharacters */
- case '.':
- case '$':
- length++;
- lastitemlength = 1;
- continue;
-
- case '*': /* These repeats won't be after brackets; */
- case '+': /* those are handled separately */
- case '?':
- length++;
- goto POSESSIVE; /* A few lines below */
-
- /* This covers the cases of braced repeats after a single char, metachar,
- class, or back reference. */
-
- case '{':
- if (!is_counted_repeat(ptr+1)) goto NORMAL_CHAR;
- ptr = read_repeat_counts(ptr+1, &min, &max, errorptr);
- if (*errorptr != NULL) goto PCRE_ERROR_RETURN;
-
- /* These special cases just insert one extra opcode */
-
- if ((min == 0 && (max == 1 || max == -1)) ||
- (min == 1 && max == -1))
- length++;
-
- /* These cases might insert additional copies of a preceding character. */
-
- else
- {
- if (min != 1)
- {
- length -= lastitemlength; /* Uncount the original char or metachar */
- if (min > 0) length += 3 + lastitemlength;
- }
- length += lastitemlength + ((max > 0)? 3 : 1);
- }
-
- if (ptr[1] == '?') ptr++; /* Needs no extra length */
-
- POSESSIVE: /* Test for possessive quantifier */
- if (ptr[1] == '+')
- {
- ptr++;
- length += 2 + 2*LINK_SIZE; /* Allow for atomic brackets */
- }
- continue;
-
- /* An alternation contains an offset to the next branch or ket. If any ims
- options changed in the previous branch(es), and/or if we are in a
- lookbehind assertion, extra space will be needed at the start of the
- branch. This is handled by branch_extra. */
-
- case '|':
- length += 1 + LINK_SIZE + branch_extra;
- continue;
-
- /* A character class uses 33 characters provided that all the character
- values are less than 256. Otherwise, it uses a bit map for low valued
- characters, and individual items for others. Don't worry about character
- types that aren't allowed in classes - they'll get picked up during the
- compile. A character class that contains only one single-byte character
- uses 2 or 3 bytes, depending on whether it is negated or not. Notice this
- where we can. (In UTF-8 mode we can do this only for chars < 128.) */
-
- case '[':
- if (*(++ptr) == '^')
- {
- class_optcount = 10; /* Greater than one */
- ptr++;
- }
- else class_optcount = 0;
-
-#ifdef SUPPORT_UTF8
- class_utf8 = FALSE;
-#endif
-
- /* Written as a "do" so that an initial ']' is taken as data */
-
- if (*ptr != 0) do
- {
- /* Inside \Q...\E everything is literal except \E */
-
- if (inescq)
- {
- if (*ptr != '\\' || ptr[1] != 'E') goto GET_ONE_CHARACTER;
- inescq = FALSE;
- ptr += 1;
- continue;
- }
-
- /* Outside \Q...\E, check for escapes */
-
- if (*ptr == '\\')
- {
- c = check_escape(&ptr, errorptr, bracount, options, TRUE);
- if (*errorptr != NULL) goto PCRE_ERROR_RETURN;
-
- /* \b is backspace inside a class; \X is literal */
-
- if (-c == ESC_b) c = '\b';
- else if (-c == ESC_X) c = 'X';
-
- /* \Q enters quoting mode */
-
- else if (-c == ESC_Q)
- {
- inescq = TRUE;
- continue;
- }
-
- /* Handle escapes that turn into characters */
-
- if (c >= 0) goto NON_SPECIAL_CHARACTER;
-
- /* Escapes that are meta-things. The normal ones just affect the
- bit map, but Unicode properties require an XCLASS extended item. */
-
- else
- {
- class_optcount = 10; /* \d, \s etc; make sure > 1 */
-#ifdef SUPPORT_UTF8
- if (-c == ESC_p || -c == ESC_P)
- {
- if (!class_utf8)
- {
- class_utf8 = TRUE;
- length += LINK_SIZE + 2;
- }
- length += 2;
- }
-#endif
- }
- }
-
- /* Check the syntax for POSIX stuff. The bits we actually handle are
- checked during the real compile phase. */
-
- else if (*ptr == '[' && check_posix_syntax(ptr, &ptr, &compile_block))
- {
- ptr++;
- class_optcount = 10; /* Make sure > 1 */
- }
-
- /* Anything else increments the possible optimization count. We have to
- detect ranges here so that we can compute the number of extra ranges for
- caseless wide characters when UCP support is available. If there are wide
- characters, we are going to have to use an XCLASS, even for single
- characters. */
-
- else
- {
- int d;
-
- GET_ONE_CHARACTER:
-
-#ifdef SUPPORT_UTF8
- if (utf8)
- {
- int extra = 0;
- GETCHARLEN(c, ptr, extra);
- ptr += extra;
- }
- else c = *ptr;
-#else
- c = *ptr;
-#endif
-
- /* Come here from handling \ above when it escapes to a char value */
-
- NON_SPECIAL_CHARACTER:
- class_optcount++;
-
- d = -1;
- if (ptr[1] == '-')
- {
- uschar const *hyptr = ptr++;
- if (ptr[1] == '\\')
- {
- ptr++;
- d = check_escape(&ptr, errorptr, bracount, options, TRUE);
- if (*errorptr != NULL) goto PCRE_ERROR_RETURN;
- if (-d == ESC_b) d = '\b'; /* backspace */
- else if (-d == ESC_X) d = 'X'; /* literal X in a class */
- }
- else if (ptr[1] != 0 && ptr[1] != ']')
- {
- ptr++;
-#ifdef SUPPORT_UTF8
- if (utf8)
- {
- int extra = 0;
- GETCHARLEN(d, ptr, extra);
- ptr += extra;
- }
- else
-#endif
- d = *ptr;
- }
- if (d < 0) ptr = hyptr; /* go back to hyphen as data */
- }
-
- /* If d >= 0 we have a range. In UTF-8 mode, if the end is > 255, or >
- 127 for caseless matching, we will need to use an XCLASS. */
-
- if (d >= 0)
- {
- class_optcount = 10; /* Ensure > 1 */
- if (d < c)
- {
- *errorptr = ERR8;
- goto PCRE_ERROR_RETURN;
- }
-
-#ifdef SUPPORT_UTF8
- if (utf8 && (d > 255 || ((options & PCRE_CASELESS) != 0 && d > 127)))
- {
- uschar buffer[6];
- if (!class_utf8) /* Allow for XCLASS overhead */
- {
- class_utf8 = TRUE;
- length += LINK_SIZE + 2;
- }
-
-#ifdef SUPPORT_UCP
- /* If we have UCP support, find out how many extra ranges are
- needed to map the other case of characters within this range. We
- have to mimic the range optimization here, because extending the
- range upwards might push d over a boundary that makes is use
- another byte in the UTF-8 representation. */
-
- if ((options & PCRE_CASELESS) != 0)
- {
- int occ, ocd;
- int cc = c;
- int origd = d;
- while (get_othercase_range(&cc, origd, &occ, &ocd))
- {
- if (occ >= c && ocd <= d) continue; /* Skip embedded */
-
- if (occ < c && ocd >= c - 1) /* Extend the basic range */
- { /* if there is overlap, */
- c = occ; /* noting that if occ < c */
- continue; /* we can't have ocd > d */
- } /* because a subrange is */
- if (ocd > d && occ <= d + 1) /* always shorter than */
- { /* the basic range. */
- d = ocd;
- continue;
- }
-
- /* An extra item is needed */
-
- length += 1 + ord2utf8(occ, buffer) +
- ((occ == ocd)? 0 : ord2utf8(ocd, buffer));
- }
- }
-#endif /* SUPPORT_UCP */
-
- /* The length of the (possibly extended) range */
-
- length += 1 + ord2utf8(c, buffer) + ord2utf8(d, buffer);
- }
-#endif /* SUPPORT_UTF8 */
-
- }
-
- /* We have a single character. There is nothing to be done unless we
- are in UTF-8 mode. If the char is > 255, or 127 when caseless, we must
- allow for an XCL_SINGLE item, doubled for caselessness if there is UCP
- support. */
-
- else
- {
-#ifdef SUPPORT_UTF8
- if (utf8 && (c > 255 || ((options & PCRE_CASELESS) != 0 && c > 127)))
- {
- uschar buffer[6];
- class_optcount = 10; /* Ensure > 1 */
- if (!class_utf8) /* Allow for XCLASS overhead */
- {
- class_utf8 = TRUE;
- length += LINK_SIZE + 2;
- }
-#ifdef SUPPORT_UCP
- length += (((options & PCRE_CASELESS) != 0)? 2 : 1) *
- (1 + ord2utf8(c, buffer));
-#else /* SUPPORT_UCP */
- length += 1 + ord2utf8(c, buffer);
-#endif /* SUPPORT_UCP */
- }
-#endif /* SUPPORT_UTF8 */
- }
- }
- }
- while (*(++ptr) != 0 && (inescq || *ptr != ']')); /* Concludes "do" above */
-
- if (*ptr == 0) /* Missing terminating ']' */
- {
- *errorptr = ERR6;
- goto PCRE_ERROR_RETURN;
- }
-
- /* We can optimize when there was only one optimizable character. Repeats
- for positive and negated single one-byte chars are handled by the general
- code. Here, we handle repeats for the class opcodes. */
-
- if (class_optcount == 1) length += 3; else
- {
- length += 33;
-
- /* A repeat needs either 1 or 5 bytes. If it is a possessive quantifier,
- we also need extra for wrapping the whole thing in a sub-pattern. */
-
- if (*ptr != 0 && ptr[1] == '{' && is_counted_repeat(ptr+2))
- {
- ptr = read_repeat_counts(ptr+2, &min, &max, errorptr);
- if (*errorptr != NULL) goto PCRE_ERROR_RETURN;
- if ((min == 0 && (max == 1 || max == -1)) ||
- (min == 1 && max == -1))
- length++;
- else length += 5;
- if (ptr[1] == '+')
- {
- ptr++;
- length += 2 + 2*LINK_SIZE;
- }
- else if (ptr[1] == '?') ptr++;
- }
- }
- continue;
-
- /* Brackets may be genuine groups or special things */
-
- case '(':
- branch_newextra = 0;
- bracket_length = 1 + LINK_SIZE;
-
- /* Handle special forms of bracket, which all start (? */
-
- if (ptr[1] == '?')
- {
- int set, unset;
- int *optset;
-
- switch (c = ptr[2])
- {
- /* Skip over comments entirely */
- case '#':
- ptr += 3;
- while (*ptr != 0 && *ptr != ')') ptr++;
- if (*ptr == 0)
- {
- *errorptr = ERR18;
- goto PCRE_ERROR_RETURN;
- }
- continue;
-
- /* Non-referencing groups and lookaheads just move the pointer on, and
- then behave like a non-special bracket, except that they don't increment
- the count of extracting brackets. Ditto for the "once only" bracket,
- which is in Perl from version 5.005. */
-
- case ':':
- case '=':
- case '!':
- case '>':
- ptr += 2;
- break;
-
- /* (?R) specifies a recursive call to the regex, which is an extension
- to provide the facility which can be obtained by (?p{perl-code}) in
- Perl 5.6. In Perl 5.8 this has become (??{perl-code}).
-
- From PCRE 4.00, items such as (?3) specify subroutine-like "calls" to
- the appropriate numbered brackets. This includes both recursive and
- non-recursive calls. (?R) is now synonymous with (?0). */
-
- case 'R':
- ptr++;
-
- case '0': case '1': case '2': case '3': case '4':
- case '5': case '6': case '7': case '8': case '9':
- ptr += 2;
- if (c != 'R')
- while ((digitab[*(++ptr)] & ctype_digit) != 0);
- if (*ptr != ')')
- {
- *errorptr = ERR29;
- goto PCRE_ERROR_RETURN;
- }
- length += 1 + LINK_SIZE;
-
- /* If this item is quantified, it will get wrapped inside brackets so
- as to use the code for quantified brackets. We jump down and use the
- code that handles this for real brackets. */
-
- if (ptr[1] == '+' || ptr[1] == '*' || ptr[1] == '?' || ptr[1] == '{')
- {
- length += 2 + 2 * LINK_SIZE; /* to make bracketed */
- duplength = 5 + 3 * LINK_SIZE;
- goto HANDLE_QUANTIFIED_BRACKETS;
- }
- continue;
-
- /* (?C) is an extension which provides "callout" - to provide a bit of
- the functionality of the Perl (?{...}) feature. An optional number may
- follow (default is zero). */
-
- case 'C':
- ptr += 2;
- while ((digitab[*(++ptr)] & ctype_digit) != 0);
- if (*ptr != ')')
- {
- *errorptr = ERR39;
- goto PCRE_ERROR_RETURN;
- }
- length += 2 + 2*LINK_SIZE;
- continue;
-
- /* Named subpatterns are an extension copied from Python */
-
- case 'P':
- ptr += 3;
- if (*ptr == '<')
- {
- const uschar *p; /* Don't amalgamate; some compilers */
- p = ++ptr; /* grumble at autoincrement in declaration */
- while ((compile_block.ctypes[*ptr] & ctype_word) != 0) ptr++;
- if (*ptr != '>')
- {
- *errorptr = ERR42;
- goto PCRE_ERROR_RETURN;
- }
- name_count++;
- if (ptr - p > max_name_size) max_name_size = (ptr - p);
- break;
- }
-
- if (*ptr == '=' || *ptr == '>')
- {
- while ((compile_block.ctypes[*(++ptr)] & ctype_word) != 0);
- if (*ptr != ')')
- {
- *errorptr = ERR42;
- goto PCRE_ERROR_RETURN;
- }
- break;
- }
-
- /* Unknown character after (?P */
-
- *errorptr = ERR41;
- goto PCRE_ERROR_RETURN;
-
- /* Lookbehinds are in Perl from version 5.005 */
-
- case '<':
- ptr += 3;
- if (*ptr == '=' || *ptr == '!')
- {
- branch_newextra = 1 + LINK_SIZE;
- length += 1 + LINK_SIZE; /* For the first branch */
- break;
- }
- *errorptr = ERR24;
- goto PCRE_ERROR_RETURN;
-
- /* Conditionals are in Perl from version 5.005. The bracket must either
- be followed by a number (for bracket reference) or by an assertion
- group, or (a PCRE extension) by 'R' for a recursion test. */
-
- case '(':
- if (ptr[3] == 'R' && ptr[4] == ')')
- {
- ptr += 4;
- length += 3;
- }
- else if ((digitab[ptr[3]] & ctype_digit) != 0)
- {
- ptr += 4;
- length += 3;
- while ((digitab[*ptr] & ctype_digit) != 0) ptr++;
- if (*ptr != ')')
- {
- *errorptr = ERR26;
- goto PCRE_ERROR_RETURN;
- }
- }
- else /* An assertion must follow */
- {
- ptr++; /* Can treat like ':' as far as spacing is concerned */
- if (ptr[2] != '?' ||
- (ptr[3] != '=' && ptr[3] != '!' && ptr[3] != '<') )
- {
- ptr += 2; /* To get right offset in message */
- *errorptr = ERR28;
- goto PCRE_ERROR_RETURN;
- }
- }
- break;
-
- /* Else loop checking valid options until ) is met. Anything else is an
- error. If we are without any brackets, i.e. at top level, the settings
- act as if specified in the options, so massage the options immediately.
- This is for backward compatibility with Perl 5.004. */
-
- default:
- set = unset = 0;
- optset = &set;
- ptr += 2;
-
- for (;; ptr++)
- {
- c = *ptr;
- switch (c)
- {
- case 'i':
- *optset |= PCRE_CASELESS;
- continue;
-
- case 'm':
- *optset |= PCRE_MULTILINE;
- continue;
-
- case 's':
- *optset |= PCRE_DOTALL;
- continue;
-
- case 'x':
- *optset |= PCRE_EXTENDED;
- continue;
-
- case 'X':
- *optset |= PCRE_EXTRA;
- continue;
-
- case 'U':
- *optset |= PCRE_UNGREEDY;
- continue;
-
- case '-':
- optset = &unset;
- continue;
-
- /* A termination by ')' indicates an options-setting-only item; if
- this is at the very start of the pattern (indicated by item_count
- being zero), we use it to set the global options. This is helpful
- when analyzing the pattern for first characters, etc. Otherwise
- nothing is done here and it is handled during the compiling
- process.
-
- [Historical note: Up to Perl 5.8, options settings at top level
- were always global settings, wherever they appeared in the pattern.
- That is, they were equivalent to an external setting. From 5.8
- onwards, they apply only to what follows (which is what you might
- expect).] */
-
- case ')':
- if (item_count == 0)
- {
- options = (options | set) & (~unset);
- set = unset = 0; /* To save length */
- item_count--; /* To allow for several */
- }
-
- /* Fall through */
-
- /* A termination by ':' indicates the start of a nested group with
- the given options set. This is again handled at compile time, but
- we must allow for compiled space if any of the ims options are
- set. We also have to allow for resetting space at the end of
- the group, which is why 4 is added to the length and not just 2.
- If there are several changes of options within the same group, this
- will lead to an over-estimate on the length, but this shouldn't
- matter very much. We also have to allow for resetting options at
- the start of any alternations, which we do by setting
- branch_newextra to 2. Finally, we record whether the case-dependent
- flag ever changes within the regex. This is used by the "required
- character" code. */
-
- case ':':
- if (((set|unset) & PCRE_IMS) != 0)
- {
- length += 4;
- branch_newextra = 2;
- if (((set|unset) & PCRE_CASELESS) != 0) options |= PCRE_ICHANGED;
- }
- goto END_OPTIONS;
-
- /* Unrecognized option character */
-
- default:
- *errorptr = ERR12;
- goto PCRE_ERROR_RETURN;
- }
- }
-
- /* If we hit a closing bracket, that's it - this is a freestanding
- option-setting. We need to ensure that branch_extra is updated if
- necessary. The only values branch_newextra can have here are 0 or 2.
- If the value is 2, then branch_extra must either be 2 or 5, depending
- on whether this is a lookbehind group or not. */
-
- END_OPTIONS:
- if (c == ')')
- {
- if (branch_newextra == 2 &&
- (branch_extra == 0 || branch_extra == 1+LINK_SIZE))
- branch_extra += branch_newextra;
- continue;
-
projects / exim.git / commitdiff