perlretut (manpage)
Jump to navigation
Jump to search
computing: software: documentation: manpages: perlm: perlretut Perl regular expressions tutorial
Manpage
NAME
perlretut - Perl regular expressions tutorial
DESCRIPTION
Text has not been reformatted after transcription from man output |
This page provides a basic tutorial on understanding, creating and using
regular expressions in Perl. It serves as a complement to the reference
page on regular expressions perlre. Regular expressions are an integral
part of the "m//", "s///", "qr//" and "split" operators and so this tutorial
also overlaps with "Regexp Quote-Like Operators" in perlop and "split" in
perlfunc.
Perl is widely renowned for excellence in text processing, and regular
expressions are one of the big factors behind this fame. Perl regular
expressions display an efficiency and flexibility unknown in most other com-
puter languages. Mastering even the basics of regular expressions will
allow you to manipulate text with surprising ease.
What is a regular expression? A regular expression is simply a string that
describes a pattern. Patterns are in common use these days; examples are
the patterns typed into a search engine to find web pages and the patterns
used to list files in a directory, e.g., "ls *.txt" or "dir *.*". In Perl,
the patterns described by regular expressions are used to search strings,
extract desired parts of strings, and to do search and replace operations.
Regular expressions have the undeserved reputation of being abstract and
difficult to understand. Regular expressions are constructed using simple
concepts like conditionals and loops and are no more difficult to understand
than the corresponding "if" conditionals and "while" loops in the Perl lan-
guage itself. In fact, the main challenge in learning regular expressions
is just getting used to the terse notation used to express these concepts.
This tutorial flattens the learning curve by discussing regular expression
concepts, along with their notation, one at a time and with many examples.
The first part of the tutorial will progress from the simplest word searches
to the basic regular expression concepts. If you master the first part, you
will have all the tools needed to solve about 98% of your needs. The second
part of the tutorial is for those comfortable with the basics and hungry for
more power tools. It discusses the more advanced regular expression opera-
tors and introduces the latest cutting edge innovations in 5.6.0.
A note: to save time, âregular expressionâ is often abbreviated as regexp or
regex. Regexp is a more natural abbreviation than regex, but is harder to
pronounce. The Perl pod documentation is evenly split on regexp vs regex;
in Perl, there is more than one way to abbreviate it. Weâll use regexp in
this tutorial.
Part 1: The basics
Simple word matching
The simplest regexp is simply a word, or more generally, a string of charac-
ters. A regexp consisting of a word matches any string that contains that
word:
"Hello World" =~ /World/; # matches
What is this perl statement all about? "Hello World" is a simple double
quoted string. "World" is the regular expression and the "//" enclosing
"/World/" tells perl to search a string for a match. The operator "=~" as-
sociates the string with the regexp match and produces a true value if the
regexp matched, or false if the regexp did not match. In our case, "World"
matches the second word in "Hello World", so the expression is true.
Expressions like this are useful in conditionals:
if ("Hello World" =~ /World/) {
print "It matches\n";
}
else {
print "It doesnât match\n";
}
There are useful variations on this theme. The sense of the match can be
reversed by using "!~" operator:
if ("Hello World" !~ /World/) {
print "It doesnât match\n";
}
else {
print "It matches\n";
}
The literal string in the regexp can be replaced by a variable:
$greeting = "World";
if ("Hello World" =~ /$greeting/) {
print "It matches\n";
}
else {
print "It doesnât match\n";
}
If youâre matching against the special default variable $_, the "$_ =~" part
can be omitted:
$_ = "Hello World";
if (/World/) {
print "It matches\n";
}
else {
print "It doesnât match\n";
}
And finally, the "//" default delimiters for a match can be changed to arbi-
trary delimiters by putting an âmâ out front:
"Hello World" =~ m!World!; # matches, delimited by â!â
"Hello World" =~ m{World}; # matches, note the matching â{}â
"/usr/bin/perl" =~ m"/perl"; # matches after â/usr/binâ,
# â/â becomes an ordinary char
"/World/", "m!World!", and "m{World}" all represent the same thing. When,
e.g., "" is used as a delimiter, the forward slash â/â becomes an ordinary
character and can be used in a regexp without trouble.
Letâs consider how different regexps would match "Hello World":
"Hello World" =~ /world/; # doesnât match
"Hello World" =~ /o W/; # matches
"Hello World" =~ /oW/; # doesnât match
"Hello World" =~ /World /; # doesnât match
The first regexp "world" doesnât match because regexps are case-sensitive.
The second regexp matches because the substring âo Wâ occurs in the string
"Hello World" . The space character â â is treated like any other character
in a regexp and is needed to match in this case. The lack of a space char-
acter is the reason the third regexp âoWâ doesnât match. The fourth regexp
âWorld â doesnât match because there is a space at the end of the regexp,
but not at the end of the string. The lesson here is that regexps must
match a part of the string exactly in order for the statement to be true.
If a regexp matches in more than one place in the string, perl will always
match at the earliest possible point in the string:
"Hello World" =~ /o/; # matches âoâ in âHelloâ
"That hat is red" =~ /hat/; # matches âhatâ in âThatâ
With respect to character matching, there are a few more points you need to
know about. First of all, not all characters can be used âas isâ in a
match. Some characters, called metacharacters, are reserved for use in reg-
exp notation. The metacharacters are
{}[]()^$.â*+?\
The significance of each of these will be explained in the rest of the tuto-
rial, but for now, it is important only to know that a metacharacter can be
matched by putting a backslash before it:
"2+2=4" =~ /2+2/; # doesnât match, + is a metacharacter
"2+2=4" =~ /2\+2/; # matches, \+ is treated like an ordinary +
"The interval is [0,1)." =~ /[0,1)./ # is a syntax error!
"The interval is [0,1)." =~ /\[0,1\)\./ # matches
"/usr/bin/perl" =~ /\/usr\/bin\/perl/; # matches
In the last regexp, the forward slash â/â is also backslashed, because it is
used to delimit the regexp. This can lead to LTS (leaning toothpick syn-
drome), however, and it is often more readable to change delimiters.
"/usr/bin/perl" =~ m!/usr/bin/perl!; # easier to read
The backslash character â\â is a metacharacter itself and needs to be back-
slashed:
âC:\WIN32â =~ /C:\\WIN/; # matches
In addition to the metacharacters, there are some ASCII characters which
donât have printable character equivalents and are instead represented by
escape sequences. Common examples are "\t" for a tab, "\n" for a newline,
"\r" for a carriage return and "\a" for a bell. If your string is better
thought of as a sequence of arbitrary bytes, the octal escape sequence,
e.g., "\033", or hexadecimal escape sequence, e.g., "\x1B" may be a more
natural representation for your bytes. Here are some examples of escapes:
"1000\t2000" =~ m(0\t2) # matches
"1000\n2000" =~ /0\n20/ # matches
"1000\t2000" =~ /\000\t2/ # doesnât match, "0" ne "\000"
"cat" =~ /\143\x61\x74/ # matches, but a weird way to spell cat
If youâve been around Perl a while, all this talk of escape sequences may
seem familiar. Similar escape sequences are used in double-quoted strings
and in fact the regexps in Perl are mostly treated as double-quoted strings.
This means that variables can be used in regexps as well. Just like double-
quoted strings, the values of the variables in the regexp will be substi-
tuted in before the regexp is evaluated for matching purposes. So we have:
$foo = âhouseâ;
âhousecatâ =~ /$foo/; # matches
âcathouseâ =~ /cat$foo/; # matches
âhousecatâ =~ /${foo}cat/; # matches
So far, so good. With the knowledge above you can already perform searches
with just about any literal string regexp you can dream up. Here is a very
simple emulation of the Unix grep program:
% cat > simple_grep
#!/usr/bin/perl
$regexp = shift;
while (<>) {
print if /$regexp/;
}
^D
% chmod +x simple_grep
% simple_grep abba /usr/dict/words
Babbage
cabbage
cabbages
sabbath
Sabbathize
Sabbathizes
sabbatical
scabbard
scabbards
This program is easy to understand. "#!/usr/bin/perl" is the standard way
to invoke a perl program from the shell. "$regexp = shift;" saves the
first command line argument as the regexp to be used, leaving the rest of
the command line arguments to be treated as files. "while (<>)" loops over
all the lines in all the files. For each line, "print if /$regexp/;"
prints the line if the regexp matches the line. In this line, both "print"
and "/$regexp/" use the default variable $_ implicitly.
With all of the regexps above, if the regexp matched anywhere in the string,
it was considered a match. Sometimes, however, weâd like to specify where
in the string the regexp should try to match. To do this, we would use the
anchor metacharacters "^" and "$". The anchor "^" means match at the begin-
ning of the string and the anchor "$" means match at the end of the string,
or before a newline at the end of the string. Here is how they are used:
"housekeeper" =~ /keeper/; # matches
"housekeeper" =~ /^keeper/; # doesnât match
"housekeeper" =~ /keeper$/; # matches
"housekeeper\n" =~ /keeper$/; # matches
The second regexp doesnât match because "^" constrains "keeper" to match
only at the beginning of the string, but "housekeeper" has keeper starting
in the middle. The third regexp does match, since the "$" constrains
"keeper" to match only at the end of the string.
When both "^" and "$" are used at the same time, the regexp has to match
both the beginning and the end of the string, i.e., the regexp matches the
whole string. Consider
"keeper" =~ /^keep$/; # doesnât match
"keeper" =~ /^keeper$/; # matches
"" =~ /^$/; # ^$ matches an empty string
The first regexp doesnât match because the string has more to it than
"keep". Since the second regexp is exactly the string, it matches. Using
both "^" and "$" in a regexp forces the complete string to match, so it
gives you complete control over which strings match and which donât. Sup-
pose you are looking for a fellow named bert, off in a string by himself:
"dogbert" =~ /bert/; # matches, but not what you want
"dilbert" =~ /^bert/; # doesnât match, but ..
"bertram" =~ /^bert/; # matches, so still not good enough
"bertram" =~ /^bert$/; # doesnât match, good
"dilbert" =~ /^bert$/; # doesnât match, good
"bert" =~ /^bert$/; # matches, perfect
Of course, in the case of a literal string, one could just as easily use the
string equivalence "$string eq âbertâ" and it would be more efficient.
The "^...$" regexp really becomes useful when we add in the more powerful
regexp tools below.
Using character classes
Although one can already do quite a lot with the literal string regexps
above, weâve only scratched the surface of regular expression technology.
In this and subsequent sections we will introduce regexp concepts (and asso-
ciated metacharacter notations) that will allow a regexp to not just repre-
sent a single character sequence, but a whole class of them.
One such concept is that of a character class. A character class allows a
set of possible characters, rather than just a single character, to match at
a particular point in a regexp. Character classes are denoted by brackets
"[...]", with the set of characters to be possibly matched inside. Here are
some examples:
/cat/; # matches âcatâ
/[bcr]at/; # matches âbat, âcatâ, or âratâ
/item[0123456789]/; # matches âitem0â or ... or âitem9â
"abc" =~ /[cab]/; # matches âaâ
In the last statement, even though âcâ is the first character in the class,
âaâ matches because the first character position in the string is the earli-
est point at which the regexp can match.
/[yY][eE][sS]/; # match âyesâ in a case-insensitive way
# âyesâ, âYesâ, âYESâ, etc.
This regexp displays a common task: perform a case-insensitive match. Perl
provides away of avoiding all those brackets by simply appending an âiâ to
the end of the match. Then "/[yY][eE][sS]/;" can be rewritten as "/yes/i;".
The âiâ stands for case-insensitive and is an example of a modifier of the
matching operation. We will meet other modifiers later in the tutorial.
We saw in the section above that there were ordinary characters, which rep-
resented themselves, and special characters, which needed a backslash "\" to
represent themselves. The same is true in a character class, but the sets
of ordinary and special characters inside a character class are different
than those outside a character class. The special characters for a charac-
ter class are "-]\^$". "]" is special because it denotes the end of a char-
acter class. "$" is special because it denotes a scalar variable. "\" is
special because it is used in escape sequences, just like above. Here is
how the special characters "]$\" are handled:
/[\]c]def/; # matches â]defâ or âcdefâ
$x = âbcrâ;
/[$x]at/; # matches âbatâ, âcatâ, or âratâ
/[\$x]at/; # matches â$atâ or âxatâ
/[\\$x]at/; # matches â\atâ, âbat, âcatâ, or âratâ
The last two are a little tricky. in "[\$x]", the backslash protects the
dollar sign, so the character class has two members "$" and "x". In
"[\\$x]", the backslash is protected, so $x is treated as a variable and
substituted in double quote fashion.
The special character â-â acts as a range operator within character classes,
so that a contiguous set of characters can be written as a range. With
ranges, the unwieldy "[0123456789]" and "[abc...xyz]" become the svelte
"[0-9]" and "[a-z]". Some examples are
/item[0-9]/; # matches âitem0â or ... or âitem9â
/[0-9bx-z]aa/; # matches â0aaâ, ..., â9aaâ,
# âbaaâ, âxaaâ, âyaaâ, or âzaaâ
/[0-9a-fA-F]/; # matches a hexadecimal digit
/[0-9a-zA-Z_]/; # matches a "word" character,
# like those in a perl variable name
If â-â is the first or last character in a character class, it is treated as
an ordinary character; "[-ab]", "[ab-]" and "[a\-b]" are all equivalent.
The special character "^" in the first position of a character class denotes
a negated character class, which matches any character but those in the
brackets. Both "[...]" and "[^...]" must match a character, or the match
fails. Then
/[^a]at/; # doesnât match âaatâ or âatâ, but matches
# all other âbatâ, âcat, â0atâ, â%atâ, etc.
/[^0-9]/; # matches a non-numeric character
/[a^]at/; # matches âaatâ or â^atâ; here â^â is ordinary
Now, even "[0-9]" can be a bother the write multiple times, so in the inter-
est of saving keystrokes and making regexps more readable, Perl has several
abbreviations for common character classes:
· \d is a digit and represents [0-9]
· \s is a whitespace character and represents [\ \t\r\n\f]
· \w is a word character (alphanumeric or _) and represents [0-9a-zA-Z_]
· \D is a negated \d; it represents any character but a digit [^0-9]
· \S is a negated \s; it represents any non-whitespace character [^\s]
· \W is a negated \w; it represents any non-word character [^\w]
· The period â.â matches any character but "\n"
The "\d\s\w\D\S\W" abbreviations can be used both inside and outside of
character classes. Here are some in use:
/\d\d:\d\d:\d\d/; # matches a hh:mm:ss time format
/[\d\s]/; # matches any digit or whitespace character
/\w\W\w/; # matches a word char, followed by a
# non-word char, followed by a word char
/..rt/; # matches any two chars, followed by ârtâ
/end\./; # matches âend.â
/end[.]/; # same thing, matches âend.â
Because a period is a metacharacter, it needs to be escaped to match as an
ordinary period. Because, for example, "\d" and "\w" are sets of characters,
it is incorrect to think of "[^\d\w]" as "[\D\W]"; in fact "[^\d\w]" is the
same as "[^\w]", which is the same as "[\W]". Think DeMorganâs laws.
An anchor useful in basic regexps is the word anchor "\b". This matches a
boundary between a word character and a non-word character "\w\W" or "\W\w":
$x = "Housecat catenates house and cat";
$x =~ /cat/; # matches cat in âhousecatâ
$x =~ /\bcat/; # matches cat in âcatenatesâ
$x =~ /cat\b/; # matches cat in âhousecatâ
$x =~ /\bcat\b/; # matches âcatâ at end of string
Note in the last example, the end of the string is considered a word bound-
ary.
You might wonder why â.â matches everything but "\n" - why not every charac-
ter? The reason is that often one is matching against lines and would like
to ignore the newline characters. For instance, while the string "\n" rep-
resents one line, we would like to think of as empty. Then
"" =~ /^$/; # matches
"\n" =~ /^$/; # matches, "\n" is ignored
"" =~ /./; # doesnât match; it needs a char
"" =~ /^.$/; # doesnât match; it needs a char
"\n" =~ /^.$/; # doesnât match; it needs a char other than "\n"
"a" =~ /^.$/; # matches
"a\n" =~ /^.$/; # matches, ignores the "\n"
This behavior is convenient, because we usually want to ignore newlines when
we count and match characters in a line. Sometimes, however, we want to
keep track of newlines. We might even want "^" and "$" to anchor at the
beginning and end of lines within the string, rather than just the beginning
and end of the string. Perl allows us to choose between ignoring and paying
attention to newlines by using the "//s" and "//m" modifiers. "//s" and
"//m" stand for single line and multi-line and they determine whether a
string is to be treated as one continuous string, or as a set of lines. The
two modifiers affect two aspects of how the regexp is interpreted: 1) how
the â.â character class is defined, and 2) where the anchors "^" and "$" are
able to match. Here are the four possible combinations:
· no modifiers (//): Default behavior. â.â matches any character except
"\n". "^" matches only at the beginning of the string and "$" matches
only at the end or before a newline at the end.
· s modifier (//s): Treat string as a single long line. â.â matches any
character, even "\n". "^" matches only at the beginning of the string
and "$" matches only at the end or before a newline at the end.
· m modifier (//m): Treat string as a set of multiple lines. â.â matches
any character except "\n". "^" and "$" are able to match at the start
or end of any line within the string.
· both s and m modifiers (//sm): Treat string as a single long line, but
detect multiple lines. â.â matches any character, even "\n". "^" and
"$", however, are able to match at the start or end of any line within
the string.
Here are examples of "//s" and "//m" in action:
$x = "There once was a girl\nWho programmed in Perl\n";
$x =~ /^Who/; # doesnât match, "Who" not at start of string
$x =~ /^Who/s; # doesnât match, "Who" not at start of string
$x =~ /^Who/m; # matches, "Who" at start of second line
$x =~ /^Who/sm; # matches, "Who" at start of second line
$x =~ /girl.Who/; # doesnât match, "." doesnât match "\n"
$x =~ /girl.Who/s; # matches, "." matches "\n"
$x =~ /girl.Who/m; # doesnât match, "." doesnât match "\n"
$x =~ /girl.Who/sm; # matches, "." matches "\n"
Most of the time, the default behavior is what is want, but "//s" and "//m"
are occasionally very useful. If "//m" is being used, the start of the
string can still be matched with "\A" and the end of string can still be
matched with the anchors "\Z" (matches both the end and the newline before,
like "$"), and "\z" (matches only the end):
$x =~ /^Who/m; # matches, "Who" at start of second line
$x =~ /\AWho/m; # doesnât match, "Who" is not at start of string
$x =~ /girl$/m; # matches, "girl" at end of first line
$x =~ /girl\Z/m; # doesnât match, "girl" is not at end of string
$x =~ /Perl\Z/m; # matches, "Perl" is at newline before end
$x =~ /Perl\z/m; # doesnât match, "Perl" is not at end of string
We now know how to create choices among classes of characters in a regexp.
What about choices among words or character strings? Such choices are
described in the next section.
Matching this or that
Sometimes we would like to our regexp to be able to match different possible
words or character strings. This is accomplished by using the alternation
metacharacter "â". To match "dog" or "cat", we form the regexp "dogâcat".
As before, perl will try to match the regexp at the earliest possible point
in the string. At each character position, perl will first try to match the
first alternative, "dog". If "dog" doesnât match, perl will then try the
next alternative, "cat". If "cat" doesnât match either, then the match
fails and perl moves to the next position in the string. Some examples:
"cats and dogs" =~ /catâdogâbird/; # matches "cat"
"cats and dogs" =~ /dogâcatâbird/; # matches "cat"
Even though "dog" is the first alternative in the second regexp, "cat" is
able to match earlier in the string.
"cats" =~ /câcaâcatâcats/; # matches "c"
"cats" =~ /catsâcatâcaâc/; # matches "cats"
Here, all the alternatives match at the first string position, so the first
alternative is the one that matches. If some of the alternatives are trun-
cations of the others, put the longest ones first to give them a chance to
match.
"cab" =~ /aâbâc/ # matches "c"
# /aâbâc/ == /[abc]/
The last example points out that character classes are like alternations of
characters. At a given character position, the first alternative that
allows the regexp match to succeed will be the one that matches.
Grouping things and hierarchical matching
Alternation allows a regexp to choose among alternatives, but by itself it
unsatisfying. The reason is that each alternative is a whole regexp, but
sometime we want alternatives for just part of a regexp. For instance, sup-
pose we want to search for housecats or housekeepers. The regexp "house-
catâhousekeeper" fits the bill, but is inefficient because we had to type
"house" twice. It would be nice to have parts of the regexp be constant,
like "house", and some parts have alternatives, like "catâkeeper".
The grouping metacharacters "()" solve this problem. Grouping allows parts
of a regexp to be treated as a single unit. Parts of a regexp are grouped
by enclosing them in parentheses. Thus we could solve the "housecatâhouse-
keeper" by forming the regexp as "house(catâkeeper)". The regexp
"house(catâkeeper)" means match "house" followed by either "cat" or
"keeper". Some more examples are
/(aâb)b/; # matches âabâ or âbbâ
/(acâb)b/; # matches âacbâ or âbbâ
/(^aâb)c/; # matches âacâ at start of string or âbcâ anywhere
/(aâ[bc])d/; # matches âadâ, âbdâ, or âcdâ
/house(catâ)/; # matches either âhousecatâ or âhouseâ
/house(cat(sâ)â)/; # matches either âhousecatsâ or âhousecatâ or
# âhouseâ. Note groups can be nested.
/(19â20â)\d\d/; # match years 19xx, 20xx, or the Y2K problem, xx
"20" =~ /(19â20â)\d\d/; # matches the null alternative â()\d\dâ,
# because â20\d\dâ canât match
Alternations behave the same way in groups as out of them: at a given string
position, the leftmost alternative that allows the regexp to match is taken.
So in the last example at the first string position, "20" matches the second
alternative, but there is nothing left over to match the next two digits
"\d\d". So perl moves on to the next alternative, which is the null alter-
native and that works, since "20" is two digits.
The process of trying one alternative, seeing if it matches, and moving on
to the next alternative if it doesnât, is called backtracking. The term
âbacktrackingâ comes from the idea that matching a regexp is like a walk in
the woods. Successfully matching a regexp is like arriving at a destina-
tion. There are many possible trailheads, one for each string position, and
each one is tried in order, left to right. From each trailhead there may be
many paths, some of which get you there, and some which are dead ends. When
you walk along a trail and hit a dead end, you have to backtrack along the
trail to an earlier point to try another trail. If you hit your destina-
tion, you stop immediately and forget about trying all the other trails.
You are persistent, and only if you have tried all the trails from all the
trailheads and not arrived at your destination, do you declare failure. To
be concrete, here is a step-by-step analysis of what perl does when it tries
to match the regexp
"abcde" =~ /(abdâabc)(dfâdâde)/;
0 Start with the first letter in the string âaâ.
1 Try the first alternative in the first group âabdâ.
2 Match âaâ followed by âbâ. So far so good.
3 âdâ in the regexp doesnât match âcâ in the string - a dead end. So
backtrack two characters and pick the second alternative in the first
group âabcâ.
4 Match âaâ followed by âbâ followed by âcâ. We are on a roll and have
satisfied the first group. Set $1 to âabcâ.
5 Move on to the second group and pick the first alternative âdfâ.
6 Match the âdâ.
7 âfâ in the regexp doesnât match âeâ in the string, so a dead end. Back-
track one character and pick the second alternative in the second group
âdâ.
8 âdâ matches. The second grouping is satisfied, so set $2 to âdâ.
9 We are at the end of the regexp, so we are done! We have matched âabcdâ
out of the string "abcde".
There are a couple of things to note about this analysis. First, the third
alternative in the second group âdeâ also allows a match, but we stopped
before we got to it - at a given character position, leftmost wins. Second,
we were able to get a match at the first character position of the string
âaâ. If there were no matches at the first position, perl would move to the
second character position âbâ and attempt the match all over again. Only
when all possible paths at all possible character positions have been
exhausted does perl give up and declare "$string =~ /(abdâabc)(dfâdâde)/;"
to be false.
Even with all this work, regexp matching happens remarkably fast. To speed
things up, during compilation stage, perl compiles the regexp into a compact
sequence of opcodes that can often fit inside a processor cache. When the
code is executed, these opcodes can then run at full throttle and search
very quickly.
Extracting matches
The grouping metacharacters "()" also serve another completely different
function: they allow the extraction of the parts of a string that matched.
This is very useful to find out what matched and for text processing in gen-
eral. For each grouping, the part that matched inside goes into the special
variables $1, $2, etc. They can be used just as ordinary variables:
# extract hours, minutes, seconds
if ($time =~ /(\d\d):(\d\d):(\d\d)/) { # match hh:mm:ss format
$hours = $1;
$minutes = $2;
$seconds = $3;
}
Now, we know that in scalar context, "$time =~ /(\d\d):(\d\d):(\d\d)/"
returns a true or false value. In list context, however, it returns the
list of matched values "($1,$2,$3)". So we could write the code more com-
pactly as
# extract hours, minutes, seconds
($hours, $minutes, $second) = ($time =~ /(\d\d):(\d\d):(\d\d)/);
If the groupings in a regexp are nested, $1 gets the group with the leftmost
opening parenthesis, $2 the next opening parenthesis, etc. For example,
here is a complex regexp and the matching variables indicated below it:
/(ab(cdâef)((gi)âj))/;
1 2 34
so that if the regexp matched, e.g., $2 would contain âcdâ or âefâ. For con-
venience, perl sets $+ to the string held by the highest numbered $1, $2,
... that got assigned (and, somewhat related, $^N to the value of the $1,
$2, ... most-recently assigned; i.e. the $1, $2, ... associated with the
rightmost closing parenthesis used in the match).
Closely associated with the matching variables $1, $2, ... are the backref-
erences "\1", "\2", ... . Backreferences are simply matching variables that
can be used inside a regexp. This is a really nice feature - what matches
later in a regexp can depend on what matched earlier in the regexp. Suppose
we wanted to look for doubled words in text, like âthe theâ. The following
regexp finds all 3-letter doubles with a space in between:
/(\w\w\w)\s\1/;
The grouping assigns a value to \1, so that the same 3 letter sequence is
used for both parts. Here are some words with repeated parts:
% simple_grep â^(\w\w\w\wâ\w\w\wâ\w\wâ\w)\1$â /usr/dict/words
beriberi
booboo
coco
mama
murmur
papa
The regexp has a single grouping which considers 4-letter combinations, then
3-letter combinations, etc. and uses "\1" to look for a repeat. Although
$1 and "\1" represent the same thing, care should be taken to use matched
variables $1, $2, ... only outside a regexp and backreferences "\1", "\2",
... only inside a regexp; not doing so may lead to surprising and/or unde-
fined results.
In addition to what was matched, Perl 5.6.0 also provides the positions of
what was matched with the "@-" and "@+" arrays. "$-[0]" is the position of
the start of the entire match and $+[0] is the position of the end. Simi-
larly, "$-[n]" is the position of the start of the $n match and $+[n] is the
position of the end. If $n is undefined, so are "$-[n]" and $+[n]. Then this
code
$x = "Mmm...donut, thought Homer";
$x =~ /^(MmmâYech)\.\.\.(donutâpeas)/; # matches
foreach $expr (1..$#-) {
print "Match $expr: â${$expr}â at position ($-[$expr],$+[$expr])\n";
}
prints
Match 1: âMmmâ at position (0,3)
Match 2: âdonutâ at position (6,11)
Even if there are no groupings in a regexp, it is still possible to find out
what exactly matched in a string. If you use them, perl will set $â to the
part of the string before the match, will set $& to the part of the string
that matched, and will set $â to the part of the string after the match. An
example:
$x = "the cat caught the mouse";
$x =~ /cat/; # $â = âthe â, $& = âcatâ, $â = â caught the mouseâ
$x =~ /the/; # $â = ââ, $& = âtheâ, $â = â cat caught the mouseâ
In the second match, "$â = ââ" because the regexp matched at the first
character position in the string and stopped, it never saw the second âtheâ.
It is important to note that using $â and $â slows down regexp matching
quite a bit, and $& slows it down to a lesser extent, because if they are
used in one regexp in a program, they are generated for <all> regexps in the
program. So if raw performance is a goal of your application, they should
be avoided. If you need them, use "@-" and "@+" instead:
$â is the same as substr( $x, 0, $-[0] )
$& is the same as substr( $x, $-[0], $+[0]-$-[0] )
$â is the same as substr( $x, $+[0] )
Matching repetitions
The examples in the previous section display an annoying weakness. We were
only matching 3-letter words, or syllables of 4 letters or less. Weâd like
to be able to match words or syllables of any length, without writing out
tedious alternatives like "\w\w\w\wâ\w\w\wâ\w\wâ\w".
This is exactly the problem the quantifier metacharacters "?", "*", "+", and
"{}" were created for. They allow us to determine the number of repeats of
a portion of a regexp we consider to be a match. Quantifiers are put imme-
diately after the character, character class, or grouping that we want to
specify. They have the following meanings:
· "a?" = match âaâ 1 or 0 times
· "a*" = match âaâ 0 or more times, i.e., any number of times
· "a+" = match âaâ 1 or more times, i.e., at least once
· "a{n,m}" = match at least "n" times, but not more than "m" times.
· "a{n,}" = match at least "n" or more times
· "a{n}" = match exactly "n" times
Here are some examples:
/[a-z]+\s+\d*/; # match a lowercase word, at least some space, and
# any number of digits
/(\w+)\s+\1/; # match doubled words of arbitrary length
/y(es)?/i; # matches âyâ, âYâ, or a case-insensitive âyesâ
$year =~ /\d{2,4}/; # make sure year is at least 2 but not more
# than 4 digits
$year =~ /\d{4}â\d{2}/; # better match; throw out 3 digit dates
$year =~ /\d{2}(\d{2})?/; # same thing written differently. However,
# this produces $1 and the other does not.
% simple_grep â^(\w+)\1$â /usr/dict/words # isnât this easier?
beriberi
booboo
coco
mama
murmur
papa
For all of these quantifiers, perl will try to match as much of the string
as possible, while still allowing the regexp to succeed. Thus with
"/a?.../", perl will first try to match the regexp with the "a" present; if
that fails, perl will try to match the regexp without the "a" present. For
the quantifier "*", we get the following:
$x = "the cat in the hat";
$x =~ /^(.*)(cat)(.*)$/; # matches,
# $1 = âthe â
# $2 = âcatâ
# $3 = â in the hatâ
Which is what we might expect, the match finds the only "cat" in the string
and locks onto it. Consider, however, this regexp:
$x =~ /^(.*)(at)(.*)$/; # matches,
# $1 = âthe cat in the hâ
# $2 = âatâ
# $3 = ââ (0 matches)
One might initially guess that perl would find the "at" in "cat" and stop
there, but that wouldnât give the longest possible string to the first quan-
tifier ".*". Instead, the first quantifier ".*" grabs as much of the string
as possible while still having the regexp match. In this example, that
means having the "at" sequence with the final "at" in the string. The other
important principle illustrated here is that when there are two or more ele-
ments in a regexp, the leftmost quantifier, if there is one, gets to grab as
much the string as possible, leaving the rest of the regexp to fight over
scraps. Thus in our example, the first quantifier ".*" grabs most of the
string, while the second quantifier ".*" gets the empty string. Quanti-
fiers that grab as much of the string as possible are called maximal match
or greedy quantifiers.
When a regexp can match a string in several different ways, we can use the
principles above to predict which way the regexp will match:
· Principle 0: Taken as a whole, any regexp will be matched at the earli-
est possible position in the string.
· Principle 1: In an alternation "aâbâc...", the leftmost alternative that
allows a match for the whole regexp will be the one used.
· Principle 2: The maximal matching quantifiers "?", "*", "+" and "{n,m}"
will in general match as much of the string as possible while still
allowing the whole regexp to match.
· Principle 3: If there are two or more elements in a regexp, the leftmost
greedy quantifier, if any, will match as much of the string as possible
while still allowing the whole regexp to match. The next leftmost
greedy quantifier, if any, will try to match as much of the string
remaining available to it as possible, while still allowing the whole
regexp to match. And so on, until all the regexp elements are satis-
fied.
As we have seen above, Principle 0 overrides the others - the regexp will be
matched as early as possible, with the other principles determining how the
regexp matches at that earliest character position.
Here is an example of these principles in action:
$x = "The programming republic of Perl";
$x =~ /^(.+)(eâr)(.*)$/; # matches,
# $1 = âThe programming republic of Peâ
# $2 = ârâ
# $3 = âlâ
This regexp matches at the earliest string position, âTâ. One might think
that "e", being leftmost in the alternation, would be matched, but "r" pro-
duces the longest string in the first quantifier.
$x =~ /(m{1,2})(.*)$/; # matches,
# $1 = âmmâ
# $2 = âing republic of Perlâ
Here, The earliest possible match is at the first âmâ in "programming".
"m{1,2}" is the first quantifier, so it gets to match a maximal "mm".
$x =~ /.*(m{1,2})(.*)$/; # matches,
# $1 = âmâ
# $2 = âing republic of Perlâ
Here, the regexp matches at the start of the string. The first quantifier
".*" grabs as much as possible, leaving just a single âmâ for the second
quantifier "m{1,2}".
$x =~ /(.?)(m{1,2})(.*)$/; # matches,
# $1 = âaâ
# $2 = âmmâ
# $3 = âing republic of Perlâ
Here, ".?" eats its maximal one character at the earliest possible position
in the string, âaâ in "programming", leaving "m{1,2}" the opportunity to
match both "m"âs. Finally,
"aXXXb" =~ /(X*)/; # matches with $1 = ââ
because it can match zero copies of âXâ at the beginning of the string. If
you definitely want to match at least one âXâ, use "X+", not "X*".
Sometimes greed is not good. At times, we would like quantifiers to match a
minimal piece of string, rather than a maximal piece. For this purpose,
Larry Wall created the minimal match or non-greedy quantifiers "??","*?",
"+?", and "{}?". These are the usual quantifiers with a "?" appended to
them. They have the following meanings:
· "a??" = match âaâ 0 or 1 times. Try 0 first, then 1.
· "a*?" = match âaâ 0 or more times, i.e., any number of times, but as few
times as possible
· "a+?" = match âaâ 1 or more times, i.e., at least once, but as few times
as possible
· "a{n,m}?" = match at least "n" times, not more than "m" times, as few
times as possible
· "a{n,}?" = match at least "n" times, but as few times as possible
· "a{n}?" = match exactly "n" times. Because we match exactly "n" times,
"a{n}?" is equivalent to "a{n}" and is just there for notational consis-
tency.
Letâs look at the example above, but with minimal quantifiers:
$x = "The programming republic of Perl";
$x =~ /^(.+?)(eâr)(.*)$/; # matches,
# $1 = âThâ
# $2 = âeâ
# $3 = â programming republic of Perlâ
The minimal string that will allow both the start of the string "^" and the
alternation to match is "Th", with the alternation "eâr" matching "e". The
second quantifier ".*" is free to gobble up the rest of the string.
$x =~ /(m{1,2}?)(.*?)$/; # matches,
# $1 = âmâ
# $2 = âming republic of Perlâ
The first string position that this regexp can match is at the first âmâ in
"programming". At this position, the minimal "m{1,2}?" matches just one
âmâ. Although the second quantifier ".*?" would prefer to match no charac-
ters, it is constrained by the end-of-string anchor "$" to match the rest of
the string.
$x =~ /(.*?)(m{1,2}?)(.*)$/; # matches,
# $1 = âThe prograâ
# $2 = âmâ
# $3 = âming republic of Perlâ
In this regexp, you might expect the first minimal quantifier ".*?" to
match the empty string, because it is not constrained by a "^" anchor to
match the beginning of the word. Principle 0 applies here, however.
Because it is possible for the whole regexp to match at the start of the
string, it will match at the start of the string. Thus the first quantifier
has to match everything up to the first "m". The second minimal quantifier
matches just one "m" and the third quantifier matches the rest of the
string.
$x =~ /(.??)(m{1,2})(.*)$/; # matches,
# $1 = âaâ
# $2 = âmmâ
# $3 = âing republic of Perlâ
Just as in the previous regexp, the first quantifier ".??" can match earli-
est at position âaâ, so it does. The second quantifier is greedy, so it
matches "mm", and the third matches the rest of the string.
We can modify principle 3 above to take into account non-greedy quantifiers:
· Principle 3: If there are two or more elements in a regexp, the leftmost
greedy (non-greedy) quantifier, if any, will match as much (little) of
the string as possible while still allowing the whole regexp to match.
The next leftmost greedy (non-greedy) quantifier, if any, will try to
match as much (little) of the string remaining available to it as possi-
ble, while still allowing the whole regexp to match. And so on, until
all the regexp elements are satisfied.
Just like alternation, quantifiers are also susceptible to backtracking.
Here is a step-by-step analysis of the example
$x = "the cat in the hat";
$x =~ /^(.*)(at)(.*)$/; # matches,
# $1 = âthe cat in the hâ
# $2 = âatâ
# $3 = ââ (0 matches)
0 Start with the first letter in the string âtâ.
1 The first quantifier â.*â starts out by matching the whole string âthe
cat in the hatâ.
2 âaâ in the regexp element âatâ doesnât match the end of the string.
Backtrack one character.
3 âaâ in the regexp element âatâ still doesnât match the last letter of
the string âtâ, so backtrack one more character.
4 Now we can match the âaâ and the âtâ.
5 Move on to the third element â.*â. Since we are at the end of the
string and â.*â can match 0 times, assign it the empty string.
6 We are done!
Most of the time, all this moving forward and backtracking happens quickly
and searching is fast. There are some pathological regexps, however, whose
execution time exponentially grows with the size of the string. A typical
structure that blows up in your face is of the form
/(aâb+)*/;
The problem is the nested indeterminate quantifiers. There are many differ-
ent ways of partitioning a string of length n between the "+" and "*": one
repetition with "b+" of length n, two repetitions with the first "b+" length
k and the second with length n-k, m repetitions whose bits add up to length
n, etc. In fact there are an exponential number of ways to partition a
string as a function of length. A regexp may get lucky and match early in
the process, but if there is no match, perl will try every possibility
before giving up. So be careful with nested "*"âs, "{n,m}"âs, and "+"âs.
The book Mastering regular expressions by Jeffrey Friedl gives a wonderful
discussion of this and other efficiency issues.
Building a regexp
At this point, we have all the basic regexp concepts covered, so letâs give
a more involved example of a regular expression. We will build a regexp
that matches numbers.
The first task in building a regexp is to decide what we want to match and
what we want to exclude. In our case, we want to match both integers and
floating point numbers and we want to reject any string that isnât a number.
The next task is to break the problem down into smaller problems that are
easily converted into a regexp.
The simplest case is integers. These consist of a sequence of digits, with
an optional sign in front. The digits we can represent with "\d+" and the
sign can be matched with "[+-]". Thus the integer regexp is
/[+-]?\d+/; # matches integers
A floating point number potentially has a sign, an integral part, a decimal
point, a fractional part, and an exponent. One or more of these parts is
optional, so we need to check out the different possibilities. Floating
point numbers which are in proper form include 123., 0.345, .34, -1e6, and
25.4E-72. As with integers, the sign out front is completely optional and
can be matched by "[+-]?". We can see that if there is no exponent, float-
ing point numbers must have a decimal point, otherwise they are integers.
We might be tempted to model these with "\d*\.\d*", but this would also
match just a single decimal point, which is not a number. So the three
cases of floating point number sans exponent are
/[+-]?\d+\./; # 1., 321., etc.
/[+-]?\.\d+/; # .1, .234, etc.
/[+-]?\d+\.\d+/; # 1.0, 30.56, etc.
These can be combined into a single regexp with a three-way alternation:
/[+-]?(\d+\.\d+â\d+\.â\.\d+)/; # floating point, no exponent
In this alternation, it is important to put â\d+\.\d+â before â\d+\.â. If
â\d+\.â were first, the regexp would happily match that and ignore the frac-
tional part of the number.
Now consider floating point numbers with exponents. The key observation
here is that both integers and numbers with decimal points are allowed in
front of an exponent. Then exponents, like the overall sign, are indepen-
dent of whether we are matching numbers with or without decimal points, and
can be âdecoupledâ from the mantissa. The overall form of the regexp now
becomes clear:
/^(optional sign)(integer â f.p. mantissa)(optional exponent)$/;
The exponent is an "e" or "E", followed by an integer. So the exponent reg-
exp is
/[eE][+-]?\d+/; # exponent
Putting all the parts together, we get a regexp that matches numbers:
/^[+-]?(\d+\.\d+â\d+\.â\.\d+â\d+)([eE][+-]?\d+)?$/; # Ta da!
Long regexps like this may impress your friends, but can be hard to deci-
pher. In complex situations like this, the "//x" modifier for a match is
invaluable. It allows one to put nearly arbitrary whitespace and comments
into a regexp without affecting their meaning. Using it, we can rewrite our
âextendedâ regexp in the more pleasing form
/^
[+-]? # first, match an optional sign
( # then match integers or f.p. mantissas:
\d+\.\d+ # mantissa of the form a.b
â\d+\. # mantissa of the form a.
â\.\d+ # mantissa of the form .b
â\d+ # integer of the form a
)
([eE][+-]?\d+)? # finally, optionally match an exponent
$/x;
If whitespace is mostly irrelevant, how does one include space characters in
an extended regexp? The answer is to backslash it â\ â or put it in a char-
acter class "[ ]" . The same thing goes for pound signs, use "\#" or "[#]".
For instance, Perl allows a space between the sign and the mantissa/integer,
and we could add this to our regexp as follows:
/^
[+-]?\ * # first, match an optional sign *and space*
( # then match integers or f.p. mantissas:
\d+\.\d+ # mantissa of the form a.b
â\d+\. # mantissa of the form a.
â\.\d+ # mantissa of the form .b
â\d+ # integer of the form a
)
([eE][+-]?\d+)? # finally, optionally match an exponent
$/x;
In this form, it is easier to see a way to simplify the alternation. Alter-
natives 1, 2, and 4 all start with "\d+", so it could be factored out:
/^
[+-]?\ * # first, match an optional sign
( # then match integers or f.p. mantissas:
\d+ # start out with a ...
(
\.\d* # mantissa of the form a.b or a.
)? # ? takes care of integers of the form a
â\.\d+ # mantissa of the form .b
)
([eE][+-]?\d+)? # finally, optionally match an exponent
$/x;
or written in the compact form,
/^[+-]?\ *(\d+(\.\d*)?â\.\d+)([eE][+-]?\d+)?$/;
This is our final regexp. To recap, we built a regexp by
· specifying the task in detail,
· breaking down the problem into smaller parts,
· translating the small parts into regexps,
· combining the regexps,
· and optimizing the final combined regexp.
These are also the typical steps involved in writing a computer program.
This makes perfect sense, because regular expressions are essentially pro-
grams written a little computer language that specifies patterns.
Using regular expressions in Perl
The last topic of Part 1 briefly covers how regexps are used in Perl pro-
grams. Where do they fit into Perl syntax?
We have already introduced the matching operator in its default "/regexp/"
and arbitrary delimiter "m!regexp!" forms. We have used the binding opera-
tor "=~" and its negation "!~" to test for string matches. Associated with
the matching operator, we have discussed the single line "//s", multi-line
"//m", case-insensitive "//i" and extended "//x" modifiers.
There are a few more things you might want to know about matching operators.
First, we pointed out earlier that variables in regexps are substituted
before the regexp is evaluated:
$pattern = âSeussâ;
while (<>) {
print if /$pattern/;
}
This will print any lines containing the word "Seuss". It is not as effi-
cient as it could be, however, because perl has to re-evaluate $pattern each
time through the loop. If $pattern wonât be changing over the lifetime of
the script, we can add the "//o" modifier, which directs perl to only per-
form variable substitutions once:
#!/usr/bin/perl
# Improved simple_grep
$regexp = shift;
while (<>) {
print if /$regexp/o; # a good deal faster
}
If you change $pattern after the first substitution happens, perl will
ignore it. If you donât want any substitutions at all, use the special
delimiter "mââ":
@pattern = (âSeussâ);
while (<>) {
print if mâ@patternâ; # matches literal â@patternâ, not âSeussâ
}
"mââ" acts like single quotes on a regexp; all other "m" delimiters act like
double quotes. If the regexp evaluates to the empty string, the regexp in
the last successful match is used instead. So we have
"dog" =~ /d/; # âdâ matches
"dogbert =~ //; # this matches the âdâ regexp used before
The final two modifiers "//g" and "//c" concern multiple matches. The modi-
fier "//g" stands for global matching and allows the matching operator to
match within a string as many times as possible. In scalar context, succes-
sive invocations against a string will have â"//g" jump from match to match,
keeping track of position in the string as it goes along. You can get or
set the position with the "pos()" function.
The use of "//g" is shown in the following example. Suppose we have a
string that consists of words separated by spaces. If we know how many
words there are in advance, we could extract the words using groupings:
$x = "cat dog house"; # 3 words
$x =~ /^\s*(\w+)\s+(\w+)\s+(\w+)\s*$/; # matches,
# $1 = âcatâ
# $2 = âdogâ
# $3 = âhouseâ
But what if we had an indeterminate number of words? This is the sort of
task "//g" was made for. To extract all words, form the simple regexp
"(\w+)" and loop over all matches with "/(\w+)/g":
while ($x =~ /(\w+)/g) {
print "Word is $1, ends at position ", pos $x, "\n";
}
prints
Word is cat, ends at position 3
Word is dog, ends at position 7
Word is house, ends at position 13
A failed match or changing the target string resets the position. If you
donât want the position reset after failure to match, add the "//c", as in
"/regexp/gc". The current position in the string is associated with the
string, not the regexp. This means that different strings have different
positions and their respective positions can be set or read independently.
In list context, "//g" returns a list of matched groupings, or if there are
no groupings, a list of matches to the whole regexp. So if we wanted just
the words, we could use
@words = ($x =~ /(\w+)/g); # matches,
# $word[0] = âcatâ
# $word[1] = âdogâ
# $word[2] = âhouseâ
Closely associated with the "//g" modifier is the "\G" anchor. The "\G"
anchor matches at the point where the previous "//g" match left off. "\G"
allows us to easily do context-sensitive matching:
$metric = 1; # use metric units
...
$x = <FILE>; # read in measurement
$x =~ /^([+-]?\d+)\s*/g; # get magnitude
$weight = $1;
if ($metric) { # error checking
print "Units error!" unless $x =~ /\Gkg\./g;
}
else {
print "Units error!" unless $x =~ /\Glbs\./g;
}
$x =~ /\G\s+(widgetâsprocket)/g; # continue processing
The combination of "//g" and "\G" allows us to process the string a bit at a
time and use arbitrary Perl logic to decide what to do next. Currently, the
"\G" anchor is only fully supported when used to anchor to the start of the
pattern.
"\G" is also invaluable in processing fixed length records with regexps.
Suppose we have a snippet of coding region DNA, encoded as base pair letters
"ATCGTTGAAT..." and we want to find all the stop codons "TGA". In a coding
region, codons are 3-letter sequences, so we can think of the DNA snippet as
a sequence of 3-letter records. The naive regexp
# expanded, this is "ATC GTT GAA TGC AAA TGA CAT GAC"
$dna = "ATCGTTGAATGCAAATGACATGAC";
$dna =~ /TGA/;
doesnât work; it may match a "TGA", but there is no guarantee that the match
is aligned with codon boundaries, e.g., the substring "GTT GAA" gives a
match. A better solution is
while ($dna =~ /(\w\w\w)*?TGA/g) { # note the minimal *?
print "Got a TGA stop codon at position ", pos $dna, "\n";
}
which prints
Got a TGA stop codon at position 18
Got a TGA stop codon at position 23
Position 18 is good, but position 23 is bogus. What happened?
The answer is that our regexp works well until we get past the last real
match. Then the regexp will fail to match a synchronized "TGA" and start
stepping ahead one character position at a time, not what we want. The
solution is to use "\G" to anchor the match to the codon alignment:
while ($dna =~ /\G(\w\w\w)*?TGA/g) {
print "Got a TGA stop codon at position ", pos $dna, "\n";
}
This prints
Got a TGA stop codon at position 18
which is the correct answer. This example illustrates that it is important
not only to match what is desired, but to reject what is not desired.
search and replace
Regular expressions also play a big role in search and replace operations in
Perl. Search and replace is accomplished with the "s///" operator. The
general form is "s/regexp/replacement/modifiers", with everything we know
about regexps and modifiers applying in this case as well. The "replace-
ment" is a Perl double quoted string that replaces in the string whatever is
matched with the "regexp". The operator "=~" is also used here to associate
a string with "s///". If matching against $_, the "$_ =~" can be dropped.
If there is a match, "s///" returns the number of substitutions made, other-
wise it returns false. Here are a few examples:
$x = "Time to feed the cat!";
$x =~ s/cat/hacker/; # $x contains "Time to feed the hacker!"
if ($x =~ s/^(Time.*hacker)!$/$1 now!/) {
$more_insistent = 1;
}
$y = "âquoted wordsâ";
$y =~ s/^â(.*)â$/$1/; # strip single quotes,
# $y contains "quoted words"
In the last example, the whole string was matched, but only the part inside
the single quotes was grouped. With the "s///" operator, the matched vari-
ables $1, $2, etc. are immediately available for use in the replacement
expression, so we use $1 to replace the quoted string with just what was
quoted. With the global modifier, "s///g" will search and replace all
occurrences of the regexp in the string:
$x = "I batted 4 for 4";
$x =~ s/4/four/; # doesnât do it all:
# $x contains "I batted four for 4"
$x = "I batted 4 for 4";
$x =~ s/4/four/g; # does it all:
# $x contains "I batted four for four"
If you prefer âregexâ over âregexpâ in this tutorial, you could use the fol-
lowing program to replace it:
% cat > simple_replace
#!/usr/bin/perl
$regexp = shift;
$replacement = shift;
while (<>) {
s/$regexp/$replacement/go;
print;
}
^D
% simple_replace regexp regex perlretut.pod
In "simple_replace" we used the "s///g" modifier to replace all occurrences
of the regexp on each line and the "s///o" modifier to compile the regexp
only once. As with "simple_grep", both the "print" and the "s/$reg-
exp/$replacement/go" use $_ implicitly.
A modifier available specifically to search and replace is the "s///e" eval-
uation modifier. "s///e" wraps an "eval{...}" around the replacement string
and the evaluated result is substituted for the matched substring. "s///e"
is useful if you need to do a bit of computation in the process of replacing
text. This example counts character frequencies in a line:
$x = "Bill the cat";
$x =~ s/(.)/$chars{$1}++;$1/eg; # final $1 replaces char with itself
print "frequency of â$_â is $chars{$_}\n"
foreach (sort {$chars{$b} <=> $chars{$a}} keys %chars);
This prints
frequency of â â is 2
frequency of âtâ is 2
frequency of âlâ is 2
frequency of âBâ is 1
frequency of âcâ is 1
frequency of âeâ is 1
frequency of âhâ is 1
frequency of âiâ is 1
frequency of âaâ is 1
As with the match "m//" operator, "s///" can use other delimiters, such as
"s!!!" and "s{}{}", and even "s{}//". If single quotes are used "sâââ",
then the regexp and replacement are treated as single quoted strings and
there are no substitutions. "s///" in list context returns the same thing
as in scalar context, i.e., the number of matches.
The split operator
The "split" function can also optionally use a matching operator "m//" to
split a string. "split /regexp/, string, limit" splits "string" into a list
of substrings and returns that list. The regexp is used to match the char-
acter sequence that the "string" is split with respect to. The "limit", if
present, constrains splitting into no more than "limit" number of strings.
For example, to split a string into words, use
$x = "Calvin and Hobbes";
@words = split /\s+/, $x; # $word[0] = âCalvinâ
# $word[1] = âandâ
# $word[2] = âHobbesâ
If the empty regexp "//" is used, the regexp always matches and the string
is split into individual characters. If the regexp has groupings, then list
produced contains the matched substrings from the groupings as well. For
instance,
$x = "/usr/bin/perl";
@dirs = split m!/!, $x; # $dirs[0] = ââ
# $dirs[1] = âusrâ
# $dirs[2] = âbinâ
# $dirs[3] = âperlâ
@parts = split m!(/)!, $x; # $parts[0] = ââ
# $parts[1] = â/â
# $parts[2] = âusrâ
# $parts[3] = â/â
# $parts[4] = âbinâ
# $parts[5] = â/â
# $parts[6] = âperlâ
Since the first character of $x matched the regexp, "split" prepended an
empty initial element to the list.
If you have read this far, congratulations! You now have all the basic tools
needed to use regular expressions to solve a wide range of text processing
problems. If this is your first time through the tutorial, why not stop
here and play around with regexps a while... Part 2 concerns the more eso-
teric aspects of regular expressions and those concepts certainly arenât
needed right at the start.
Part 2: Power tools
OK, you know the basics of regexps and you want to know more. If matching
regular expressions is analogous to a walk in the woods, then the tools dis-
cussed in Part 1 are analogous to topo maps and a compass, basic tools we
use all the time. Most of the tools in part 2 are analogous to flare guns
and satellite phones. They arenât used too often on a hike, but when we are
stuck, they can be invaluable.
What follows are the more advanced, less used, or sometimes esoteric capa-
bilities of perl regexps. In Part 2, we will assume you are comfortable
with the basics and concentrate on the new features.
More on characters, strings, and character classes
There are a number of escape sequences and character classes that we havenât
covered yet.
There are several escape sequences that convert characters or strings
between upper and lower case. "\l" and "\u" convert the next character to
lower or upper case, respectively:
$x = "perl";
$string =~ /\u$x/; # matches âPerlâ in $string
$x = "M(rs?âs)\\."; # note the double backslash
$string =~ /\l$x/; # matches âmr.â, âmrs.â, and âms.â,
"\L" and "\U" converts a whole substring, delimited by "\L" or "\U" and
"\E", to lower or upper case:
$x = "This word is in lower case:\L SHOUT\E";
$x =~ /shout/; # matches
$x = "I STILL KEYPUNCH CARDS FOR MY 360"
$x =~ /\Ukeypunch/; # matches punch card string
If there is no "\E", case is converted until the end of the string. The reg-
exps "\L\u$word" or "\u\L$word" convert the first character of $word to
uppercase and the rest of the characters to lowercase.
Control characters can be escaped with "\c", so that a control-Z character
would be matched with "\cZ". The escape sequence "\Q"..."\E" quotes, or
protects most non-alphabetic characters. For instance,
$x = "\QThat !^*&%~& cat!";
$x =~ /\Q!^*&%~&\E/; # check for rough language
It does not protect "$" or "@", so that variables can still be substituted.
With the advent of 5.6.0, perl regexps can handle more than just the stan-
dard ASCII character set. Perl now supports Unicode, a standard for encod-
ing the character sets from many of the worldâs written languages. Unicode
does this by allowing characters to be more than one byte wide. Perl uses
the UTF-8 encoding, in which ASCII characters are still encoded as one byte,
but characters greater than "chr(127)" may be stored as two or more bytes.
What does this mean for regexps? Well, regexp users donât need to know much
about perlâs internal representation of strings. But they do need to know
1) how to represent Unicode characters in a regexp and 2) when a matching
operation will treat the string to be searched as a sequence of bytes (the
old way) or as a sequence of Unicode characters (the new way). The answer
to 1) is that Unicode characters greater than "chr(127)" may be represented
using the "\x{hex}" notation, with "hex" a hexadecimal integer:
/\x{263a}/; # match a Unicode smiley face :)
Unicode characters in the range of 128-255 use two hexadecimal digits with
braces: "\x{ab}". Note that this is different than "\xab", which is just a
hexadecimal byte with no Unicode significance.
NOTE: in Perl 5.6.0 it used to be that one needed to say "use utf8" to use
any Unicode features. This is no more the case: for almost all Unicode pro-
cessing, the explicit "utf8" pragma is not needed. (The only case where it
matters is if your Perl script is in Unicode and encoded in UTF-8, then an
explicit "use utf8" is needed.)
Figuring out the hexadecimal sequence of a Unicode character you want or
deciphering someone elseâs hexadecimal Unicode regexp is about as much fun
as programming in machine code. So another way to specify Unicode charac-
ters is to use the named character escape sequence "\N{name}". "name" is a
name for the Unicode character, as specified in the Unicode standard. For
instance, if we wanted to represent or match the astrological sign for the
planet Mercury, we could use
use charnames ":full"; # use named chars with Unicode full names
$x = "abc\N{MERCURY}def";
$x =~ /\N{MERCURY}/; # matches
One can also use short names or restrict names to a certain alphabet:
use charnames â:fullâ;
print "\N{GREEK SMALL LETTER SIGMA} is called sigma.\n";
use charnames ":short";
print "\N{greek:Sigma} is an upper-case sigma.\n";
use charnames qw(greek);
print "\N{sigma} is Greek sigma\n";
A list of full names is found in the file Names.txt in the
lib/perl5/5.X.X/unicore directory.
The answer to requirement 2), as of 5.6.0, is that if a regexp contains Uni-
code characters, the string is searched as a sequence of Unicode characters.
Otherwise, the string is searched as a sequence of bytes. If the string is
being searched as a sequence of Unicode characters, but matching a single
byte is required, we can use the "\C" escape sequence. "\C" is a character
class akin to "." except that it matches any byte 0-255. So
use charnames ":full"; # use named chars with Unicode full names
$x = "a";
$x =~ /\C/; # matches âaâ, eats one byte
$x = "";
$x =~ /\C/; # doesnât match, no bytes to match
$x = "\N{MERCURY}"; # two-byte Unicode character
$x =~ /\C/; # matches, but dangerous!
The last regexp matches, but is dangerous because the string character posi-
tion is no longer synchronized to the string byte position. This generates
the warning âMalformed UTF-8 characterâ. The "\C" is best used for matching
the binary data in strings with binary data intermixed with Unicode charac-
ters.
Let us now discuss the rest of the character classes. Just as with Unicode
characters, there are named Unicode character classes represented by the
"\p{name}" escape sequence. Closely associated is the "\P{name}" character
class, which is the negation of the "\p{name}" class. For example, to match
lower and uppercase characters,
use charnames ":full"; # use named chars with Unicode full names
$x = "BOB";
$x =~ /^\p{IsUpper}/; # matches, uppercase char class
$x =~ /^\P{IsUpper}/; # doesnât match, char class sans uppercase
$x =~ /^\p{IsLower}/; # doesnât match, lowercase char class
$x =~ /^\P{IsLower}/; # matches, char class sans lowercase
Here is the association between some Perl named classes and the traditional
Unicode classes:
Perl class name Unicode class name or regular expression
IsAlpha /^[LM]/
IsAlnum /^[LMN]/
IsASCII $code <= 127
IsCntrl /^C/
IsBlank $code =~ /^(0020â0009)$/ ââ /^Z[^lp]/
IsDigit Nd
IsGraph /^([LMNPS]âCo)/
IsLower Ll
IsPrint /^([LMNPS]âCoâZs)/
IsPunct /^P/
IsSpace /^Z/ ââ ($code =~ /^(0009â000Aâ000Bâ000Câ000D)$/
IsSpacePerl /^Z/ ââ ($code =~ /^(0009â000Aâ000Câ000Dâ0085â2028â2029)$/
IsUpper /^L[ut]/
IsWord /^[LMN]/ ââ $code eq "005F"
IsXDigit $code =~ /^00(3[0-9]â[46][1-6])$/
You can also use the official Unicode class names with the "\p" and "\P",
like "\p{L}" for Unicode âlettersâ, or "\p{Lu}" for uppercase letters, or
"\P{Nd}" for non-digits. If a "name" is just one letter, the braces can be
dropped. For instance, "\pM" is the character class of Unicode âmarksâ, for
example accent marks. For the full list see perlunicode.
The Unicode has also been separated into various sets of characters which
you can test with "\p{In...}" (in) and "\P{In...}" (not in), for example
"\p{Latin}", "\p{Greek}", or "\P{Katakana}". For the full list see perluni-
code.
"\X" is an abbreviation for a character class sequence that includes the
Unicode âcombining character sequencesâ. A âcombining character sequenceâ
is a base character followed by any number of combining characters. An
example of a combining character is an accent. Using the Unicode full
names, e.g., "A + COMBINING RING" is a combining character sequence with
base character "A" and combining character "COMBINING RING" , which trans-
lates in Danish to A with the circle atop it, as in the word Angstrom. "\X"
is equivalent to "\PM\pM*}", i.e., a non-mark followed by one or more marks.
For the full and latest information about Unicode see the latest Unicode
standard, or the Unicode Consortiumâs website http://www.unicode.org/
As if all those classes werenât enough, Perl also defines POSIX style char-
acter classes. These have the form "[:name:]", with "name" the name of the
POSIX class. The POSIX classes are "alpha", "alnum", "ascii", "cntrl",
"digit", "graph", "lower", "print", "punct", "space", "upper", and "xdigit",
and two extensions, "word" (a Perl extension to match "\w"), and "blank" (a
GNU extension). If "utf8" is being used, then these classes are defined the
same as their corresponding perl Unicode classes: "[:upper:]" is the same as
"\p{IsUpper}", etc. The POSIX character classes, however, donât require
using "utf8". The "[:digit:]", "[:word:]", and "[:space:]" correspond to
the familiar "\d", "\w", and "\s" character classes. To negate a POSIX
class, put a "^" in front of the name, so that, e.g., "[:^digit:]" corre-
sponds to "\D" and under "utf8", "\P{IsDigit}". The Unicode and POSIX char-
acter classes can be used just like "\d", with the exception that POSIX
character classes can only be used inside of a character class:
/\s+[abc[:digit:]xyz]\s*/; # match a,b,c,x,y,z, or a digit
/^=item\sdigit:/; # match â=itemâ,
# followed by a space and a digit
use charnames ":full";
/\s+[abc\p{IsDigit}xyz]\s+/; # match a,b,c,x,y,z, or a digit
/^=item\s\p{IsDigit}/; # match â=itemâ,
# followed by a space and a digit
Whew! That is all the rest of the characters and character classes.
Compiling and saving regular expressions
In Part 1 we discussed the "//o" modifier, which compiles a regexp just
once. This suggests that a compiled regexp is some data structure that can
be stored once and used again and again. The regexp quote "qr//" does
exactly that: "qr/string/" compiles the "string" as a regexp and transforms
the result into a form that can be assigned to a variable:
$reg = qr/foo+bar?/; # reg contains a compiled regexp
Then $reg can be used as a regexp:
$x = "fooooba";
$x =~ $reg; # matches, just like /foo+bar?/
$x =~ /$reg/; # same thing, alternate form
$reg can also be interpolated into a larger regexp:
$x =~ /(abc)?$reg/; # still matches
As with the matching operator, the regexp quote can use different delim-
iters, e.g., "qr!!", "qr{}" and "qr~~". The single quote delimiters "qrââ"
prevent any interpolation from taking place.
Pre-compiled regexps are useful for creating dynamic matches that donât need
to be recompiled each time they are encountered. Using pre-compiled reg-
exps, "simple_grep" program can be expanded into a program that matches mul-
tiple patterns:
% cat > multi_grep
#!/usr/bin/perl
# multi_grep - match any of <number> regexps
# usage: multi_grep <number> regexp1 regexp2 ... file1 file2 ...
$number = shift;
$regexp[$_] = shift foreach (0..$number-1);
@compiled = map qr/$_/, @regexp;
while ($line = <>) {
foreach $pattern (@compiled) {
if ($line =~ /$pattern/) {
print $line;
last; # we matched, so move onto the next line
}
}
}
^D
% multi_grep 2 last for multi_grep
$regexp[$_] = shift foreach (0..$number-1);
foreach $pattern (@compiled) {
last;
Storing pre-compiled regexps in an array @compiled allows us to simply loop
through the regexps without any recompilation, thus gaining flexibility
without sacrificing speed.
Embedding comments and modifiers in a regular expression
Starting with this section, we will be discussing Perlâs set of extended
patterns. These are extensions to the traditional regular expression syntax
that provide powerful new tools for pattern matching. We have already seen
extensions in the form of the minimal matching constructs "??", "*?", "+?",
"{n,m}?", and "{n,}?". The rest of the extensions below have the form
"(?char...)", where the "char" is a character that determines the type of
extension.
The first extension is an embedded comment "(?#text)". This embeds a com-
ment into the regular expression without affecting its meaning. The comment
should not have any closing parentheses in the text. An example is
/(?# Match an integer:)[+-]?\d+/;
This style of commenting has been largely superseded by the raw, freeform
commenting that is allowed with the "//x" modifier.
The modifiers "//i", "//m", "//s", and "//x" can also embedded in a regexp
using "(?i)", "(?m)", "(?s)", and "(?x)". For instance,
/(?i)yes/; # match âyesâ case insensitively
/yes/i; # same thing
/(?x)( # freeform version of an integer regexp
[+-]? # match an optional sign
\d+ # match a sequence of digits
)
/x;
Embedded modifiers can have two important advantages over the usual modi-
fiers. Embedded modifiers allow a custom set of modifiers to each regexp
pattern. This is great for matching an array of regexps that must have dif-
ferent modifiers:
$pattern[0] = â(?i)doctorâ;
$pattern[1] = âJohnsonâ;
...
while (<>) {
foreach $patt (@pattern) {
print if /$patt/;
}
}
The second advantage is that embedded modifiers only affect the regexp
inside the group the embedded modifier is contained in. So grouping can be
used to localize the modifierâs effects:
/Answer: ((?i)yes)/; # matches âAnswer: yesâ, âAnswer: YESâ, etc.
Embedded modifiers can also turn off any modifiers already present by using,
e.g., "(?-i)". Modifiers can also be combined into a single expression,
e.g., "(?s-i)" turns on single line mode and turns off case insensitivity.
Non-capturing groupings
We noted in Part 1 that groupings "()" had two distinct functions: 1) group
regexp elements together as a single unit, and 2) extract, or capture, sub-
strings that matched the regexp in the grouping. Non-capturing groupings,
denoted by "(?:regexp)", allow the regexp to be treated as a single unit,
but donât extract substrings or set matching variables $1, etc. Both cap-
turing and non-capturing groupings are allowed to co-exist in the same reg-
exp. Because there is no extraction, non-capturing groupings are faster
than capturing groupings. Non-capturing groupings are also handy for choos-
ing exactly which parts of a regexp are to be extracted to matching vari-
ables:
# match a number, $1-$4 are set, but we only want $1
/([+-]?\ *(\d+(\.\d*)?â\.\d+)([eE][+-]?\d+)?)/;
# match a number faster , only $1 is set
/([+-]?\ *(?:\d+(?:\.\d*)?â\.\d+)(?:[eE][+-]?\d+)?)/;
# match a number, get $1 = whole number, $2 = exponent
/([+-]?\ *(?:\d+(?:\.\d*)?â\.\d+)(?:[eE]([+-]?\d+))?)/;
Non-capturing groupings are also useful for removing nuisance elements gath-
ered from a split operation:
$x = â12a34b5â;
@num = split /(aâb)/, $x; # @num = (â12â,âaâ,â34â,âbâ,â5â)
@num = split /(?:aâb)/, $x; # @num = (â12â,â34â,â5â)
Non-capturing groupings may also have embedded modifiers: "(?i-m:regexp)" is
a non-capturing grouping that matches "regexp" case insensitively and turns
off multi-line mode.
Looking ahead and looking behind
This section concerns the lookahead and lookbehind assertions. First, a
little background.
In Perl regular expressions, most regexp elements âeat upâ a certain amount
of string when they match. For instance, the regexp element "[abc}]" eats
up one character of the string when it matches, in the sense that perl moves
to the next character position in the string after the match. There are
some elements, however, that donât eat up characters (advance the character
position) if they match. The examples we have seen so far are the anchors.
The anchor "^" matches the beginning of the line, but doesnât eat any char-
acters. Similarly, the word boundary anchor "\b" matches, e.g., if the
character to the left is a word character and the character to the right is
a non-word character, but it doesnât eat up any characters itself. Anchors
are examples of âzero-width assertionsâ. Zero-width, because they consume
no characters, and assertions, because they test some property of the
string. In the context of our walk in the woods analogy to regexp matching,
most regexp elements move us along a trail, but anchors have us stop a
moment and check our surroundings. If the local environment checks out, we
can proceed forward. But if the local environment doesnât satisfy us, we
must backtrack.
Checking the environment entails either looking ahead on the trail, looking
behind, or both. "^" looks behind, to see that there are no characters
before. "$" looks ahead, to see that there are no characters after. "\b"
looks both ahead and behind, to see if the characters on either side differ
in their âwordâ-ness.
The lookahead and lookbehind assertions are generalizations of the anchor
concept. Lookahead and lookbehind are zero-width assertions that let us
specify which characters we want to test for. The lookahead assertion is
denoted by "(?=regexp)" and the lookbehind assertion is denoted by
"(?<=fixed-regexp)". Some examples are
$x = "I catch the housecat âTom-catâ with catnip";
$x =~ /cat(?=\s+)/; # matches âcatâ in âhousecatâ
@catwords = ($x =~ /(?<=\s)cat\w+/g); # matches,
# $catwords[0] = âcatchâ
# $catwords[1] = âcatnipâ
$x =~ /\bcat\b/; # matches âcatâ in âTom-catâ
$x =~ /(?<=\s)cat(?=\s)/; # doesnât match; no isolated âcatâ in
# middle of $x
Note that the parentheses in "(?=regexp)" and "(?<=regexp)" are non-captur-
ing, since these are zero-width assertions. Thus in the second regexp, the
substrings captured are those of the whole regexp itself. Lookahead
"(?=regexp)" can match arbitrary regexps, but lookbehind "(?<=fixed-regexp)"
only works for regexps of fixed width, i.e., a fixed number of characters
long. Thus "(?<=(abâbc))" is fine, but "(?<=(ab)*)" is not. The negated
versions of the lookahead and lookbehind assertions are denoted by "(?!reg-
exp)" and "(?<!fixed-regexp)" respectively. They evaluate true if the reg-
exps do not match:
$x = "foobar";
$x =~ /foo(?!bar)/; # doesnât match, âbarâ follows âfooâ
$x =~ /foo(?!baz)/; # matches, âbazâ doesnât follow âfooâ
$x =~ /(?<!\s)foo/; # matches, there is no \s before âfooâ
The "\C" is unsupported in lookbehind, because the already treacherous defi-
nition of "\C" would become even more so when going backwards.
Using independent subexpressions to prevent backtracking
The last few extended patterns in this tutorial are experimental as of
5.6.0. Play with them, use them in some code, but donât rely on them just
yet for production code.
Independent subexpressions are regular expressions, in the context of a
larger regular expression, that function independently of the larger regular
expression. That is, they consume as much or as little of the string as
they wish without regard for the ability of the larger regexp to match.
Independent subexpressions are represented by "(?>regexp)". We can illus-
trate their behavior by first considering an ordinary regexp:
$x = "ab";
$x =~ /a*ab/; # matches
This obviously matches, but in the process of matching, the subexpression
"a*" first grabbed the "a". Doing so, however, wouldnât allow the whole
regexp to match, so after backtracking, "a*" eventually gave back the "a"
and matched the empty string. Here, what "a*" matched was dependent on what
the rest of the regexp matched.
Contrast that with an independent subexpression:
$x =~ /(?>a*)ab/; # doesnât match!
The independent subexpression "(?>a*)" doesnât care about the rest of the
regexp, so it sees an "a" and grabs it. Then the rest of the regexp "ab"
cannot match. Because "(?>a*)" is independent, there is no backtracking and
the independent subexpression does not give up its "a". Thus the match of
the regexp as a whole fails. A similar behavior occurs with completely
independent regexps:
$x = "ab";
$x =~ /a*/g; # matches, eats an âaâ
$x =~ /\Gab/g; # doesnât match, no âaâ available
Here "//g" and "\G" create a âtag teamâ handoff of the string from one reg-
exp to the other. Regexps with an independent subexpression are much like
this, with a handoff of the string to the independent subexpression, and a
handoff of the string back to the enclosing regexp.
The ability of an independent subexpression to prevent backtracking can be
quite useful. Suppose we want to match a non-empty string enclosed in
parentheses up to two levels deep. Then the following regexp matches:
$x = "abc(de(fg)h"; # unbalanced parentheses
$x =~ /\( ( [^()]+ â \([^()]*\) )+ \)/x;
The regexp matches an open parenthesis, one or more copies of an alterna-
tion, and a close parenthesis. The alternation is two-way, with the first
alternative "[^()]+" matching a substring with no parentheses and the second
alternative "\([^()]*\)" matching a substring delimited by parentheses.
The problem with this regexp is that it is pathological: it has nested inde-
terminate quantifiers of the form "(a+âb)+". We discussed in Part 1 how
nested quantifiers like this could take an exponentially long time to exe-
cute if there was no match possible. To prevent the exponential blowup, we
need to prevent useless backtracking at some point. This can be done by
enclosing the inner quantifier as an independent subexpression:
$x =~ /\( ( (?>[^()]+) â \([^()]*\) )+ \)/x;
Here, "(?>[^()]+)" breaks the degeneracy of string partitioning by gobbling
up as much of the string as possible and keeping it. Then match failures
fail much more quickly.
Conditional expressions
A conditional expression is a form of if-then-else statement that allows
one to choose which patterns are to be matched, based on some condition.
There are two types of conditional expression: "(?(condition)yes-regexp)"
and "(?(condition)yes-regexpâno-regexp)". "(?(condition)yes-regexp)" is
like an âif () {}â statement in Perl. If the "condition" is true, the
"yes-regexp" will be matched. If the "condition" is false, the "yes-regexp"
will be skipped and perl will move onto the next regexp element. The second
form is like an âif () {} else {}â statement in Perl. If the "condition"
is true, the "yes-regexp" will be matched, otherwise the "no-regexp" will be
matched.
The "condition" can have two forms. The first form is simply an integer in
parentheses "(integer)". It is true if the corresponding backreference
"\integer" matched earlier in the regexp. The second form is a bare zero
width assertion "(?...)", either a lookahead, a lookbehind, or a code asser-
tion (discussed in the next section).
The integer form of the "condition" allows us to choose, with more flexibil-
ity, what to match based on what matched earlier in the regexp. This
searches for words of the form "$x$x" or "$x$y$y$x":
% simple_grep â^(\w+)(\w+)?(?(2)\2\1â\1)$â /usr/dict/words
beriberi
coco
couscous
deed
...
toot
toto
tutu
The lookbehind "condition" allows, along with backreferences, an earlier
part of the match to influence a later part of the match. For instance,
/[ATGC]+(?(?<=AA)GâC)$/;
matches a DNA sequence such that it either ends in "AAG", or some other base
pair combination and "C". Note that the form is "(?(?<=AA)GâC)" and not
"(?((?<=AA))GâC)"; for the lookahead, lookbehind or code assertions, the
parentheses around the conditional are not needed.
A bit of magic: executing Perl code in a regular expression
Normally, regexps are a part of Perl expressions. Code evaluation expres-
sions turn that around by allowing arbitrary Perl code to be a part of a
regexp. A code evaluation expression is denoted "(?{code})", with "code" a
string of Perl statements.
Code expressions are zero-width assertions, and the value they return
depends on their environment. There are two possibilities: either the code
expression is used as a conditional in a conditional expression "(?(condi-
tion)...)", or it is not. If the code expression is a conditional, the code
is evaluated and the result (i.e., the result of the last statement) is used
to determine truth or falsehood. If the code expression is not used as a
conditional, the assertion always evaluates true and the result is put into
the special variable $^R. The variable $^R can then be used in code expres-
sions later in the regexp. Here are some silly examples:
$x = "abcdef";
$x =~ /abc(?{print "Hi Mom!";})def/; # matches,
# prints âHi Mom!â
$x =~ /aaa(?{print "Hi Mom!";})def/; # doesnât match,
# no âHi Mom!â
Pay careful attention to the next example:
$x =~ /abc(?{print "Hi Mom!";})ddd/; # doesnât match,
# no âHi Mom!â
# but why not?
At first glance, youâd think that it shouldnât print, because obviously the
"ddd" isnât going to match the target string. But look at this example:
$x =~ /abc(?{print "Hi Mom!";})[d]dd/; # doesnât match,
# but _does_ print
Hmm. What happened here? If youâve been following along, you know that the
above pattern should be effectively the same as the last one -- enclosing
the d in a character class isnât going to change what it matches. So why
does the first not print while the second one does?
The answer lies in the optimizations the REx engine makes. In the first
case, all the engine sees are plain old characters (aside from the "?{}"
construct). Itâs smart enough to realize that the string âdddâ doesnât occur
in our target string before actually running the pattern through. But in the
second case, weâve tricked it into thinking that our pattern is more compli-
cated than it is. It takes a look, sees our character class, and decides
that it will have to actually run the pattern to determine whether or not it
matches, and in the process of running it hits the print statement before it
discovers that we donât have a match.
To take a closer look at how the engine does optimizations, see the section
"Pragmas and debugging" below.
More fun with "?{}":
$x =~ /(?{print "Hi Mom!";})/; # matches,
# prints âHi Mom!â
$x =~ /(?{$c = 1;})(?{print "$c";})/; # matches,
# prints â1â
$x =~ /(?{$c = 1;})(?{print "$^R";})/; # matches,
# prints â1â
The bit of magic mentioned in the section title occurs when the regexp back-
tracks in the process of searching for a match. If the regexp backtracks
over a code expression and if the variables used within are localized using
"local", the changes in the variables produced by the code expression are
undone! Thus, if we wanted to count how many times a character got matched
inside a group, we could use, e.g.,
$x = "aaaa";
$count = 0; # initialize âaâ count
$c = "bob"; # test if $c gets clobbered
$x =~ /(?{local $c = 0;}) # initialize count
( a # match âaâ
(?{local $c = $c + 1;}) # increment count
)* # do this any number of times,
aa # but match âaaâ at the end
(?{$count = $c;}) # copy local $c var into $count
/x;
print "âaâ count is $count, \$c variable is â$câ\n";
This prints
âaâ count is 2, $c variable is âbobâ
If we replace the " (?{local $c = $c + 1;})" with " (?{$c = $c + 1;})" ,
the variable changes are not undone during backtracking, and we get
âaâ count is 4, $c variable is âbobâ
Note that only localized variable changes are undone. Other side effects of
code expression execution are permanent. Thus
$x = "aaaa";
$x =~ /(a(?{print "Yow\n";}))*aa/;
produces
Yow
Yow
Yow
Yow
The result $^R is automatically localized, so that it will behave properly
in the presence of backtracking.
This example uses a code expression in a conditional to match the article
âtheâ in either English or German:
$lang = âDEâ; # use German
...
$text = "das";
print "matched\n"
if $text =~ /(?(?{
$lang eq âENâ; # is the language English?
})
the â # if so, then match âtheâ
(dieâdasâder) # else, match âdieâdasâderâ
)
/xi;
Note that the syntax here is "(?(?{...})yes-regexpâno-regexp)", not
"(?((?{...}))yes-regexpâno-regexp)". In other words, in the case of a code
expression, we donât need the extra parentheses around the conditional.
If you try to use code expressions with interpolating variables, perl may
surprise you:
$bar = 5;
$pat = â(?{ 1 })â;
/foo(?{ $bar })bar/; # compiles ok, $bar not interpolated
/foo(?{ 1 })$bar/; # compile error!
/foo${pat}bar/; # compile error!
$pat = qr/(?{ $foo = 1 })/; # precompile code regexp
/foo${pat}bar/; # compiles ok
If a regexp has (1) code expressions and interpolating variables, or (2) a
variable that interpolates a code expression, perl treats the regexp as an
error. If the code expression is precompiled into a variable, however,
interpolating is ok. The question is, why is this an error?
The reason is that variable interpolation and code expressions together pose
a security risk. The combination is dangerous because many programmers who
write search engines often take user input and plug it directly into a reg-
exp:
$regexp = <>; # read user-supplied regexp
$chomp $regexp; # get rid of possible newline
$text =~ /$regexp/; # search $text for the $regexp
If the $regexp variable contains a code expression, the user could then exe-
cute arbitrary Perl code. For instance, some joker could search for "sys-
tem(ârm -rf *â);" to erase your files. In this sense, the combination of
interpolation and code expressions taints your regexp. So by default, using
both interpolation and code expressions in the same regexp is not allowed.
If youâre not concerned about malicious users, it is possible to bypass this
security check by invoking "use re âevalâ" :
use re âevalâ; # throw caution out the door
$bar = 5;
$pat = â(?{ 1 })â;
/foo(?{ 1 })$bar/; # compiles ok
/foo${pat}bar/; # compiles ok
Another form of code expression is the pattern code expression . The pat-
tern code expression is like a regular code expression, except that the
result of the code evaluation is treated as a regular expression and matched
immediately. A simple example is
$length = 5;
$char = âaâ;
$x = âaaaaabbâ;
$x =~ /(??{$char x $length})/x; # matches, there are 5 of âaâ
This final example contains both ordinary and pattern code expressions. It
detects if a binary string 1101010010001... has a Fibonacci spacing
0,1,1,2,3,5,... of the 1âs:
$s0 = 0; $s1 = 1; # initial conditions
$x = "1101010010001000001";
print "It is a Fibonacci sequence\n"
if $x =~ /^1 # match an initial â1â
(
(??{â0â x $s0}) # match $s0 of â0â
1 # and then a â1â
(?{
$largest = $s0; # largest seq so far
$s2 = $s1 + $s0; # compute next term
$s0 = $s1; # in Fibonacci sequence
$s1 = $s2;
})
)+ # repeat as needed
$ # that is all there is
/x;
print "Largest sequence matched was $largest\n";
This prints
It is a Fibonacci sequence
Largest sequence matched was 5
Ha! Try that with your garden variety regexp package...
Note that the variables $s0 and $s1 are not substituted when the regexp is
compiled, as happens for ordinary variables outside a code expression.
Rather, the code expressions are evaluated when perl encounters them during
the search for a match.
The regexp without the "//x" modifier is
/^1((??{â0âx$s0})1(?{$largest=$s0;$s2=$s1+$s0$s0=$s1;$s1=$s2;}))+$/;
and is a great start on an Obfuscated Perl entry :-) When working with code
and conditional expressions, the extended form of regexps is almost neces-
sary in creating and debugging regexps.
Pragmas and debugging
Speaking of debugging, there are several pragmas available to control and
debug regexps in Perl. We have already encountered one pragma in the previ-
ous section, "use re âevalâ;" , that allows variable interpolation and code
expressions to coexist in a regexp. The other pragmas are
use re âtaintâ;
$tainted = <>;
@parts = ($tainted =~ /(\w+)\s+(\w+)/; # @parts is now tainted
The "taint" pragma causes any substrings from a match with a tainted vari-
able to be tainted as well. This is not normally the case, as regexps are
often used to extract the safe bits from a tainted variable. Use "taint"
when you are not extracting safe bits, but are performing some other pro-
cessing. Both "taint" and "eval" pragmas are lexically scoped, which means
they are in effect only until the end of the block enclosing the pragmas.
use re âdebugâ;
/^(.*)$/s; # output debugging info
use re âdebugcolorâ;
/^(.*)$/s; # output debugging info in living color
The global "debug" and "debugcolor" pragmas allow one to get detailed debug-
ging info about regexp compilation and execution. "debugcolor" is the same
as debug, except the debugging information is displayed in color on termi-
nals that can display termcap color sequences. Here is example output:
% perl -e âuse re "debug"; "abc" =~ /a*b+c/;â
Compiling REx âa*b+câ
size 9 first at 1
1: STAR(4)
2: EXACT <a>(0)
4: PLUS(7)
5: EXACT <b>(0)
7: EXACT <c>(9)
9: END(0)
floating âbcâ at 0..2147483647 (checking floating) minlen 2
Guessing start of match, REx âa*b+câ against âabcâ...
Found floating substr âbcâ at offset 1...
Guessed: match at offset 0
Matching REx âa*b+câ against âabcâ
Setting an EVAL scope, savestack=3
0 <> <abc> â 1: STAR
EXACT <a> can match 1 times out of 32767...
Setting an EVAL scope, savestack=3
1 <a> <bc> â 4: PLUS
EXACT <b> can match 1 times out of 32767...
Setting an EVAL scope, savestack=3
2 <ab> <c> â 7: EXACT <c>
3 <abc> <> â 9: END
Match successful!
Freeing REx: âa*b+câ
If you have gotten this far into the tutorial, you can probably guess what
the different parts of the debugging output tell you. The first part
Compiling REx âa*b+câ
size 9 first at 1
1: STAR(4)
2: EXACT <a>(0)
4: PLUS(7)
5: EXACT <b>(0)
7: EXACT <c>(9)
9: END(0)
describes the compilation stage. STAR(4) means that there is a starred
object, in this case âaâ, and if it matches, goto line 4, i.e., PLUS(7).
The middle lines describe some heuristics and optimizations performed before
a match:
floating âbcâ at 0..2147483647 (checking floating) minlen 2
Guessing start of match, REx âa*b+câ against âabcâ...
Found floating substr âbcâ at offset 1...
Guessed: match at offset 0
Then the match is executed and the remaining lines describe the process:
Matching REx âa*b+câ against âabcâ
Setting an EVAL scope, savestack=3
0 <> <abc> â 1: STAR
EXACT <a> can match 1 times out of 32767...
Setting an EVAL scope, savestack=3
1 <a> <bc> â 4: PLUS
EXACT <b> can match 1 times out of 32767...
Setting an EVAL scope, savestack=3
2 <ab> <c> â 7: EXACT <c>
3 <abc> <> â 9: END
Match successful!
Freeing REx: âa*b+câ
Each step is of the form "n <x> <y>" , with "<x>" the part of the string
matched and "<y>" the part not yet matched. The "â 1: STAR" says that perl
is at line number 1 n the compilation list above. See "Debugging regular
expressions" in perldebguts for much more detail.
An alternative method of debugging regexps is to embed "print" statements
within the regexp. This provides a blow-by-blow account of the backtracking
in an alternation:
"that this" =~ m@(?{print "Start at position ", pos, "\n";})
t(?{print "t1\n";})
h(?{print "h1\n";})
i(?{print "i1\n";})
s(?{print "s1\n";})
â
t(?{print "t2\n";})
h(?{print "h2\n";})
a(?{print "a2\n";})
t(?{print "t2\n";})
(?{print "Done at position ", pos, "\n";})
@x;
prints
Start at position 0
t1
h1
t2
h2
a2
t2
Done at position 4
BUGS
Code expressions, conditional expressions, and independent expressions are experimental. Don't use them in production code. Yet.
SEE ALSO
This is just a tutorial. For the full story on perl regular expressions, see the perlre regular expressions reference page.
For more information on the matching "m//" and substitution "s///" operators, see "Regexp Quote-Like Operators" in perlop. For information on the "split" operation, see "split" in perlfunc.
For an excellent all-around resource on the care and feeding of regular expressions, see the book Mastering Regular Expressions by Jeffrey Friedl (published by O'Reilly, ISBN 1556592-257-3).
AUTHOR AND COPYRIGHT
Copyright (c) 2000 Mark Kvale All rights reserved.
This document may be distributed under the same terms as Perl itself.
Acknowledgments
The inspiration for the stop codon DNA example came from the ZIP code example in chapter 7 of Mastering Regular Expressions.
The author would like to thank Jeff Pinyan, Andrew Johnson, Peter Haworth, Ronald J Kimball, and Joe Smith for all their helpful comments.
Edit Log
- 2006-07-15 Transcribed from perlretut manpage for Perl v5.8.6 dated 2004-11-05 (as distributed in Fedora Core 4)