path: root/doc/re.rdoc

# -*- mode: rdoc; coding: utf-8; fill-column: 74; -*-

Regular expressions (<i>regexp</i>s) are patterns which describe the
contents of a string. They're used for testing whether a string contains a
given pattern, or extracting the portions that match. They are created
with the <tt>/</tt><i>pat</i><tt>/</tt> and
<tt>%r{</tt><i>pat</i><tt>}</tt> literals or the <tt>Regexp.new</tt>
constructor.

A regexp is usually delimited with forward slashes (<tt>/</tt>). For
example:

    /hay/ =~ 'haystack'   #=> 0
    /y/.match('haystack') #=> #<MatchData "y">

If a string contains the pattern it is said to <i>match</i>. A literal
string matches itself.

    # 'haystack' does not contain the pattern 'needle', so doesn't match.
    /needle/.match('haystack') #=> nil
    # 'haystack' does contain the pattern 'hay', so it matches
    /hay/.match('haystack')    #=> #<MatchData "hay">

Specifically, <tt>/st/</tt> requires that the string contains the letter
_s_ followed by the letter _t_, so it matches _haystack_, also.

== Metacharacters and Escapes

The following are <i>metacharacters</i> <tt>(</tt>, <tt>)</tt>,
<tt>[</tt>, <tt>]</tt>, <tt>{</tt>, <tt>}</tt>, <tt>.</tt>, <tt>?</tt>,
<tt>+</tt>, <tt>*</tt>. They have a specific meaning when appearing in a
pattern. To match them literally they must be backslash-escaped. To match
a backslash literally backslash-escape that: <tt>\\\\\\</tt>.

    /1 \+ 2 = 3\?/.match('Does 1 + 2 = 3?') #=> #<MatchData "1 + 2 = 3?">

Patterns behave like double-quoted strings so can contain the same
backslash escapes.

    /\s\u{6771 4eac 90fd}/.match("Go to 東京都")
        #=> #<MatchData " 東京都">

Arbitrary Ruby expressions can be embedded into patterns with the
<tt>#{...}</tt> construct.

    place = "東京都"
    /#{place}/.match("Go to 東京都")
        #=> #<MatchData "東京都">

== Character Classes

A <i>character class</i> is delimited with square brackets (<tt>[</tt>,
<tt>]</tt>) and lists characters that may appear at that point in the
match. <tt>/[ab]/</tt> means _a_ or _b_, as opposed to <tt>/ab/</tt> which
means _a_ followed by _b_.

    /W[aeiou]rd/.match("Word") #=> #<MatchData "Word">

Within a character class the hyphen (<tt>-</tt>) is a metacharacter
denoting an inclusive range of characters. <tt>[abcd]</tt> is equivalent
to <tt>[a-d]</tt>. A range can be followed by another range, so
<tt>[abcdwxyz]</tt> is equivalent to <tt>[a-dw-z]</tt>. The order in which
ranges or individual characters appear inside a character class is
irrelevant.

    /[0-9a-f]/.match('9f') #=> #<MatchData "9">
    /[9f]/.match('9f')     #=> #<MatchData "9">

If the first character of a character class is a caret (<tt>^</tt>) the
class is inverted: it matches any character _except_ those named.

    /[^a-eg-z]/.match('f') #=> #<MatchData "f">

A character class may contain another character class. By itself this
isn't useful because <tt>[a-z[0-9]]</tt> describes the same set as
<tt>[a-z0-9]</tt>. However, character classes also support the <tt>&&</tt>
operator which performs set intersection on its arguments. The two can be
combined as follows:

    /[a-w&&[^c-g]z]/ # ([a-w] AND ([^c-g] OR z))
    # This is equivalent to:
    /[abh-w]/

The following metacharacters also behave like character classes:

* <tt>/./</tt> - Any character except a newline.
* <tt>/./m</tt> - Any character (the +m+ modifier enables multiline mode)
* <tt>/\w/</tt> - A word character (<tt>[a-zA-Z0-9_]</tt>)
* <tt>/\W/</tt> - A non-word character (<tt>[^a-zA-Z0-9_]</tt>)
* <tt>/\d/</tt> - A digit character (<tt>[0-9]</tt>)
* <tt>/\D/</tt> - A non-digit character (<tt>[^0-9]</tt>)
* <tt>/\h/</tt> - A hexdigit character (<tt>[0-9a-fA-F]</tt>)
* <tt>/\H/</tt> - A non-hexdigit character (<tt>[^0-9a-fA-F]</tt>)
* <tt>/\s/</tt> - A whitespace character: <tt>/[ \t\r\n\f]/</tt>
* <tt>/\S/</tt> - A non-whitespace character: <tt>/[^ \t\r\n\f]/</tt>

POSIX <i>bracket expressions</i> are also similar to character classes.
They provide a portable alternative to the above, with the added benefit
that they encompass non-ASCII characters. For instance, <tt>/\d/</tt>
matches only the ASCII decimal digits (0-9); whereas <tt>/[[:digit:]]/</tt>
matches any character in the Unicode _Nd_ category.

* <tt>/[[:alnum:]]/</tt> - Alphabetic and numeric character
* <tt>/[[:alpha:]]/</tt> - Alphabetic character
* <tt>/[[:blank:]]/</tt> - Space or tab
* <tt>/[[:cntrl:]]/</tt> - Control character
* <tt>/[[:digit:]]/</tt> - Digit
* <tt>/[[:graph:]]/</tt> - Non-blank character (excludes spaces, control
  characters, and similar)
* <tt>/[[:lower:]]/</tt> - Lowercase alphabetical character
* <tt>/[[:print:]]/</tt> - Like [:graph:], but includes the space character
* <tt>/[[:punct:]]/</tt> - Punctuation character
* <tt>/[[:space:]]/</tt> - Whitespace character (<tt>[:blank:]</tt>, newline,
   carriage return, etc.)
* <tt>/[[:upper:]]/</tt> - Uppercase alphabetical
* <tt>/[[:xdigit:]]/</tt> - Digit allowed in a hexadecimal number (i.e.,
  0-9a-fA-F)

Ruby also supports the following non-POSIX character classes:

* <tt>/[[:word:]]/</tt> - A character in one of the following Unicode
  general categories _Letter_, _Mark_, _Number_,
  <i>Connector_Punctuation<i/i>
* <tt>/[[:ascii:]]/</tt> - A character in the ASCII character set

    # U+06F2 is "EXTENDED ARABIC-INDIC DIGIT TWO"
    /[[:digit:]]/.match("\u06F2")    #=> #<MatchData "\u{06F2}">
    /[[:upper:]][[:lower:]]/.match("Hello") #=> #<MatchData "He">
    /[[:xdigit:]][[:xdigit:]]/.match("A6")  #=> #<MatchData "A6">

== Repetition

The constructs described so far match a single character. They can be
followed by a repetition metacharacter to specify how many times they need
to occur. Such metacharacters are called <i>quantifiers</i>.

* <tt>*</tt> - Zero or more times
* <tt>+</tt> - One or more times
* <tt>?</tt> - Zero or one times (optional)
* <tt>{</tt><i>n</i><tt>}</tt> - Exactly <i>n</i> times
* <tt>{</tt><i>n</i><tt>,}</tt> - <i>n</i> or more times
* <tt>{,</tt><i>m</i><tt>}</tt> - <i>m</i> or less times
* <tt>{</tt><i>n</i><tt>,</tt><i>m</i><tt>}</tt> - At least <i>n</i> and
  at most <i>m</i> times

    # At least one uppercase character ('H'), at least one lowercase
    # character ('e'), two 'l' characters, then one 'o'
    "Hello".match(/[[:upper:]]+[[:lower:]]+l{2}o/) #=> #<MatchData "Hello">

Repetition is <i>greedy</i> by default: as many occurrences as possible
are matched while still allowing the overall match to succeed. By
contrast, <i>lazy</i> matching makes the minimal amount of matches
necessary for overall success. A greedy metacharacter can be made lazy by
following it with <tt>?</tt>.

    # Both patterns below match the string. The fist uses a greedy
    # quantifier so '.+' matches '<a><b>'; the second uses a lazy
    # quantifier so '.+?' matches '<a>'.
    /<.+>/.match("<a><b>")  #=> #<MatchData "<a><b>">
    /<.+?>/.match("<a><b>") #=> #<MatchData "<a>">

A quantifier followed by <tt>+</tt> matches <i>possessively</i>: once it
has matched it does not backtrack. They behave like greedy quantifiers,
but having matched they refuse to "give up" their match even if this
jeopardises the overall match.

== Capturing

Parentheses can be used for <i>capturing</i>. The text enclosed by the
<i>n</i><sup>th</sup> group of parentheses can be subsequently referred to
with <i>n</i>. Within a pattern use the <i>backreference</i>
<tt>\</tt><i>n</i>; outside of the pattern use
<tt>MatchData[</tt><i>n</i><tt>]</tt>.

    # 'at' is captured by the first group of parentheses, then referred to
    # later with \1
    /[csh](..) [csh]\1 in/.match("The cat sat in the hat")
        #=> #<MatchData "cat sat in" 1:"at">
    # Regexp#match returns a MatchData object which makes the captured
    # text available with its #[] method.
    /[csh](..) [csh]\1 in/.match("The cat sat in the hat")[1] #=> 'at'

Capture groups can be referred to by name when defined with the
<tt>(?<</tt><i>name</i><tt>>)</tt> or <tt>(?'</tt><i>name</i><tt>')</tt>
constructs.

    /\$(?<dollars>\d+)\.(?<cents>\d+)/.match("$3.67")
        => #<MatchData "$3.67" dollars:"3" cents:"67">
    /\$(?<dollars>\d+)\.(?<cents>\d+)/.match("$3.67")[:dollars] #=> "3"

Named groups can be backreferenced with <tt>\k<</tt><i>name</i><tt>></tt>,
where _name_ is the group name.

    /(?<vowel>[aeiou]).\k<vowel>.\k<vowel>/.match('ototomy')
        #=> #<MatchData "ototo" vowel:"o">

*Note*: A regexp can't use named backreferences and numbered
backreferences simultaneously.

When named capture groups are used with a literal regexp on the left-hand
side of an expression and the <tt>=~</tt> operator, the captured text is
also assigned to local variables with corresponding names.

    /\$(?<dollars>\d+)\.(?<cents>\d+)/ =~ "$3.67" #=> 0
    dollars #=> "3"

== Grouping

Parentheses also <i>group</i> the terms they enclose, allowing them to be
quantified as one <i>atomic</i> whole.

    # The pattern below matches a vowel followed by 2 word characters:
    # 'aen'
    /[aeiou]\w{2}/.match("Caenorhabditis elegans") #=> #<MatchData "aen">
    # Whereas the following pattern matches a vowel followed by a word
    # character, twice, i.e. <tt>[aeiou]\w[aeiou]\w</tt>: 'enor'.
    /([aeiou]\w){2}/.match("Caenorhabditis elegans")
        #=> #<MatchData "enor" 1:"or">

The <tt>(?:</tt>...<tt>)</tt> construct provides grouping without
capturing. That is, it combines the terms it contains into an atomic whole
without creating a backreference. This benefits performance at the slight
expense of readabilty.

    # The group of parentheses captures 'n' and the second 'ti'. The
    # second group is referred to later with the backreference \2
    /I(n)ves(ti)ga\2ons/.match("Investigations")
        #=> #<MatchData "Investigations" 1:"n" 2:"ti">
    # The first group of parentheses is now made non-capturing with '?:',
    # so it still matches 'n', but doesn't create the backreference. Thus,
    # the backreference \1 now refers to 'ti'.
    /I(?:n)ves(ti)ga\1ons/.match("Investigations")
        #=> #<MatchData "Investigations" 1:"ti">

=== Atomic Grouping

Grouping can be made <i>atomic</i> with
<tt>(?></tt><i>pat</i><tt>)</tt>. This causes the subexpression <i>pat</i>
to be matched independently of the rest of the expression such that what
it matches becomes fixed for the remainder of the match, unless the entire
subexpression must be abandoned and subsequently revisited. In this
way <i>pat</i> is treated as a non-divisible whole. Atomic grouping is
typically used to optimise patterns so as to prevent the regular
expression engine from backtracking needlesly.

    # The <tt>"</tt> in the pattern below matches the first character of
    # the string, then <tt>.*</tt> matches <i>Quote"</i>. This causes the
    # overall match to fail, so the text matched by <tt>.*</tt> is
    # backtracked by one position, which leaves the final character of the
    # string available to match <tt>"</tt>
          /".*"/.match('"Quote"')     #=> #<MatchData "\"Quote\"">
    # If <tt>.*</tt> is grouped atomically, it refuses to backtrack
    # <i>Quote"</i>, even though this means that the overall match fails
    /"(?>.*)"/.match('"Quote"') #=> nil

== Subexpression Calls

The <tt>\g<</tt><i>name</i><tt>></tt> syntax matches the previous
subexpression named _name_, which can be a group name or number, again.
This differs from backreferences in that it re-executes the group rather
than simply trying to re-match the same text.

    # Matches a <i>(</i> character and assigns it to the <tt>paren</tt>
    # group, tries to call that the <tt>paren</tt> sub-expression again
    # but fails, then matches a literal <i>)</i>.
    /\A(?<paren>\(\g<paren>*\))*\z/ =~ '()'


    /\A(?<paren>\(\g<paren>*\))*\z/ =~ '(())' #=> 0
    # ^1
    #      ^2
    #           ^3
    #                 ^4
    #      ^5
    #           ^6
    #                      ^7
    #                       ^8
    #                       ^9
    #                           ^10

1.  Matches at the beginning of the string, i.e. before the first
    character.
2.  Enters a named capture group called <tt>paren</tt>
3.  Matches a literal <i>(</i>, the first character in the string
4.  Calls the <tt>paren</tt> group again, i.e. recurses back to the
    second step
5.  Re-enters the <tt>paren</tt> group
6.  Matches a literal <i>(</i>, the second character in the
    string
7.  Try to call <tt>paren</tt> a third time, but fail because
    doing so would prevent an overall successful match
8.  Match a literal <i>)</i>, the third character in the string.
    Marks the end of the second recursive call
9.  Match a literal <i>)</i>, the fourth character in the string
10. Match the end of the string

== Alternation

The vertical bar metacharacter (<tt>|</tt>) combines two expressions into
a single one that matches either of the expressions. Each expression is an
<i>alternative</i>.

    /\w(and|or)\w/.match("Feliformia") #=> #<MatchData "form" 1:"or">
    /\w(and|or)\w/.match("furandi")    #=> #<MatchData "randi" 1:"and">
    /\w(and|or)\w/.match("dissemblance") #=> nil

== Character Properties

The <tt>\p{}</tt> construct matches characters with the named property,
much like POSIX bracket classes.

* <tt>/\p{Alnum}/</tt> - Alphabetic and numeric character
* <tt>/\p{Alpha}/</tt> - Alphabetic character
* <tt>/\p{Blank}/</tt> - Space or tab
* <tt>/\p{Cntrl}/</tt> - Control character
* <tt>/\p{Digit}/</tt> - Digit
* <tt>/\p{Graph}/</tt> - Non-blank character (excludes spaces, control
  characters, and similar)
* <tt>/\p{Lower}/</tt> - Lowercase alphabetical character
* <tt>/\p{Print}/</tt> - Like <tt>\p{Graph}</tt>, but includes the space character
* <tt>/\p{Punct}/</tt> - Punctuation character
* <tt>/\p{Space}/</tt> - Whitespace character (<tt>[:blank:]</tt>, newline,
  carriage return, etc.)
* <tt>/\p{Upper}/</tt> - Uppercase alphabetical
* <tt>/\p{XDigit}/</tt> - Digit allowed in a hexadecimal number (i.e., 0-9a-fA-F)
* <tt>/\p{Word}/</tt> - A member of one of the following Unicode general
  category <i>Letter</i>, <i>Mark</i>, <i>Number</i>,
  <i>Connector\_Punctuation</i>
* <tt>/\p{ASCII}/</tt> - A character in the ASCII character set
* <tt>/\p{Any}/</tt> - Any Unicode character (including unassigned
  characters)
* <tt>/\p{Assigned}/</tt> - An assigned character

A Unicode character's <i>General Category</i> value can also be matched
with <tt>\p{</tt><i>Ab</i><tt>}</tt> where <i>Ab</i> is the category's
abbreviation as described below:

* <tt>/\p{L}/</tt> - 'Letter'
* <tt>/\p{Ll}/</tt> - 'Letter: Lowercase'
* <tt>/\p{Lm}/</tt> - 'Letter: Mark'
* <tt>/\p{Lo}/</tt> - 'Letter: Other'
* <tt>/\p{Lt}/</tt> - 'Letter: Titlecase'
* <tt>/\p{Lu}/</tt> - 'Letter: Uppercase
* <tt>/\p{Lo}/</tt> - 'Letter: Other'
* <tt>/\p{M}/</tt> - 'Mark'
* <tt>/\p{Mn}/</tt> - 'Mark: Nonspacing'
* <tt>/\p{Mc}/</tt> - 'Mark: Spacing Combining'
* <tt>/\p{Me}/</tt> - 'Mark: Enclosing'
* <tt>/\p{N}/</tt> - 'Number'
* <tt>/\p{Nd}/</tt> - 'Number: Decimal Digit'
* <tt>/\p{Nl}/</tt> - 'Number: Letter'
* <tt>/\p{No}/</tt> - 'Number: Other'
* <tt>/\p{P}/</tt> - 'Punctuation'
* <tt>/\p{Pc}/</tt> - 'Punctuation: Connector'
* <tt>/\p{Pd}/</tt> - 'Punctuation: Dash'
* <tt>/\p{Ps}/</tt> - 'Punctuation: Open'
* <tt>/\p{Pe}/</tt> - 'Punctuation: Close'
* <tt>/\p{Pi}/</tt> - 'Punctuation: Initial Quote'
* <tt>/\p{Pf}/</tt> - 'Punctuation: Final Quote'
* <tt>/\p{Po}/</tt> - 'Punctuation: Other'
* <tt>/\p{S}/</tt> - 'Symbol'
* <tt>/\p{Sm}/</tt> - 'Symbol: Math'
* <tt>/\p{Sc}/</tt> - 'Symbol: Currency'
* <tt>/\p{Sc}/</tt> - 'Symbol: Currency'
* <tt>/\p{Sk}/</tt> - 'Symbol: Modifier'
* <tt>/\p{So}/</tt> - 'Symbol: Other'
* <tt>/\p{Z}/</tt> - 'Separator'
* <tt>/\p{Zs}/</tt> - 'Separator: Space'
* <tt>/\p{Zl}/</tt> - 'Separator: Line'
* <tt>/\p{Zp}/</tt> - 'Separator: Paragraph'
* <tt>/\p{C}/</tt> - 'Other'
* <tt>/\p{Cc}/</tt> - 'Other: Control'
* <tt>/\p{Cf}/</tt> - 'Other: Format'
* <tt>/\p{Cn}/</tt> - 'Other: Not Assigned'
* <tt>/\p{Co}/</tt> - 'Other: Private Use'
* <tt>/\p{Cs}/</tt> - 'Other: Surrogate'

Lastly, <tt>\p{}</tt> matches a character's Unicode <i>script</i>. The
following scripts are supported: <i>Arabic</i>, <i>Armenian</i>,
<i>Balinese</i>, <i>Bengali</i>, <i>Bopomofo</i>, <i>Braille</i>,
<i>Buginese</i>, <i>Buhid</i>, <i>Canadian_Aboriginal</i>, <i>Carian</i>,
<i>Cham</i>, <i>Cherokee</i>, <i>Common</i>, <i>Coptic</i>,
<i>Cuneiform</i>, <i>Cypriot</i>, <i>Cyrillic</i>, <i>Deseret</i>,
<i>Devanagari</i>, <i>Ethiopic</i>, <i>Georgian</i>, <i>Glagolitic</i>,
<i>Gothic</i>, <i>Greek</i>, <i>Gujarati</i>, <i>Gurmukhi</i>, <i>Han</i>,
<i>Hangul</i>, <i>Hanunoo</i>, <i>Hebrew</i>, <i>Hiragana</i>,
<i>Inherited</i>, <i>Kannada</i>, <i>Katakana</i>, <i>Kayah_Li</i>,
<i>Kharoshthi</i>, <i>Khmer</i>, <i>Lao</i>, <i>Latin</i>, <i>Lepcha</i>,
<i>Limbu</i>, <i>Linear_B</i>, <i>Lycian</i>, <i>Lydian</i>,
<i>Malayalam</i>, <i>Mongolian</i>, <i>Myanmar</i>, <i>New_Tai_Lue</i>,
<i>Nko</i>, <i>Ogham</i>, <i>Ol_Chiki</i>, <i>Old_Italic</i>,
<i>Old_Persian</i>, <i>Oriya</i>, <i>Osmanya</i>, <i>Phags_Pa</i>,
<i>Phoenician</i>, <i>Rejang</i>, <i>Runic</i>, <i>Saurashtra</i>,
<i>Shavian</i>, <i>Sinhala</i>, <i>Sundanese</i>, <i>Syloti_Nagri</i>,
<i>Syriac</i>, <i>Tagalog</i>, <i>Tagbanwa</i>, <i>Tai_Le</i>,
<i>Tamil</i>, <i>Telugu</i>, <i>Thaana</i>, <i>Thai</i>, <i>Tibetan</i>,
<i>Tifinagh</i>, <i>Ugaritic</i>, <i>Vai</i>, and <i>Yi</i>.

    # Unicode codepoint U+06E9 is named "ARABIC PLACE OF SAJDAH" and
    # belongs to the Arabic script.
    /\p{Arabic}/.match("\u06E9") #=> #<MatchData "\u06E9">

All character properties can be inverted by prefixing their name with a
caret (<tt>^</tt>).

    # Letter 'A' is not in the Unicode Ll (Letter; Lowercase) category, so
    # this match succeeds
    /\p{^Ll}/.match("A") #=> #<MatchData "A">

== Anchors

Anchors are metacharacter that match the zero-width positions between
characters, <i>anchoring</i> the match to a specific position.

* <tt>^</tt> - Matches beginning of line
* <tt>$</tt> - Matches end of line
* <tt>\A</tt> - Matches beginning of string.
* <tt>\Z</tt> - Matches end of string. If string ends with a newline,
  it matches just before newline
* <tt>\z</tt> - Matches end of string
* <tt>\G</tt> - Matches point where last match finished
* <tt>\b</tt> - Matches word boundaries when outside brackets; backspace
   (0x08) inside brackets
* <tt>\B</tt> - Matches non-word boundaries
* <tt>(?=</tt><i>pat</i><tt>)</tt> - <i>Positive lookahead</i> assertion:
  ensures that the following characters match <i>pat</i>, but doesn't
  include those characters in the matched text
* <tt>(?!</tt><i>pat</i><tt>)</tt> - <i>Negative lookahead</i> assertion:
  ensures that the following characters do not match <i>pat</i>, but
  doesn't include those characters in the matched text
* <tt>(?<=</tt><i>pat</i><tt>)</tt> - <i>Positive lookbehind</i>
  assertion: ensures that the preceding characters match <i>pat</i>, but
  doesn't include those characters in the matched text
* <tt>(?<!</tt><i>pat</i><tt>)</tt> - <i>Negative lookbehind</i>
  assertion: ensures that the preceding characters do not match
  <i>pat</i>, but doesn't include those characters in the matched text

    # If a pattern isn't anchored it can begin at any point in the string
    /real/.match("surrealist") #=> #<MatchData "real">
    # Anchoring the pattern to the beginning of the string forces the
    # match to start there. 'real' doesn't occur at the beginning of the
    # string, so now the match fails
    /\Areal/.match("surrealist") #=> nil
    # The match below fails because although 'Demand' contains 'and', the
    pattern does not occur at a word boundary.
    /\band/.match("Demand")
    # Whereas in the following example 'and' has been anchored to a
    # non-word boundary so instead of matching the first 'and' it matches
    # from the fourth letter of 'demand' instead
    /\Band.+/.match("Supply and demand curve") #=> #<MatchData "and curve">
    # The pattern below uses positive lookahead and positive lookbehind to
    # match text appearing in <b></b> tags without including the tags in the
    # match
    /(?<=<b>)\w+(?=<\/b>)/.match("Fortune favours the <b>bold</b>")
        #=> #<MatchData "bold">

== Options

The end delimiter for a regexp can be followed by one or more single-letter
options which control how the pattern can match.

* <tt>/pat/i</tt> - Ignore case
* <tt>/pat/m</tt> - Treat a newline as a character matched by <tt>.</tt>
* <tt>/pat/x</tt> - Ignore whitespace and comments in the pattern
* <tt>/pat/o</tt> - Perform <tt>#{}</tt> interpolation only once

<tt>i</tt>, <tt>m</tt>, and <tt>x</tt> can also be applied on the
subexpression level with the
<tt>(?</tt><i>on</i><tt>-</tt><i>off</i><tt>)</tt> construct, which
enables options <i>on</i>, and disables options <i>off</i> for the
expression enclosed by the parentheses.

    /a(?i:b)c/.match('aBc') #=> #<MatchData "aBc">
    /a(?i:b)c/.match('abc') #=> #<MatchData "abc">

== Free-Spacing Mode and Comments

As mentioned above, the <tt>x</tt> option enables <i>free-spacing</i>
mode. Literal white space inside the pattern is ignored, and the
octothorpe (<tt>#</tt>) character introduces a comment until the end of
the line. This allows the components of the pattern to be organised in a
potentially more readable fashion.

    # A contrived pattern to match a number with optional decimal places
    float_pat = /\A
        [[:digit:]]+ # 1 or more digits before the decimal point
        (\.          # Decimal point
            [[:digit:]]+ # 1 or more digits after the decimal point
        )? # The decimal point and following digits are optional
    \Z/x
    float_pat.match('3.14') #=> #<MatchData "3.14" 1:".14">

*Note*: To match whitespace in an <tt>x</tt> pattern use an escape such as
<tt>\s</tt> or <tt>\p{Space}</tt>.

Comments can be included in a non-<tt>x</tt> pattern with the
<tt>(?#</tt><i>comment</i><tt>)</tt> construct, where <i>comment</i> is
arbitrary text ignored by the regexp engine.

== Encoding

Regular expressions are assumed to use the source encoding. This can be
overridden with one of the following modifiers.

* <tt>/</tt><i>pat</i><tt>/u</tt> - UTF-8
* <tt>/</tt><i>pat</i><tt>/e</tt> - EUC-JP
* <tt>/</tt><i>pat</i><tt>/s</tt> - Windows-31J
* <tt>/</tt><i>pat</i><tt>/n</tt> - ASCII-8BIT

A regexp can be matched against a string when they either share an
encoding, or the regexp's encoding is _US-ASCII_ and the string's encoding
is ASCII-compatible.

If a match between incompatible encodings is attempted an
<tt>Encoding::CompatibilityError</tt> exception is raised.

The <tt>Regexp#fixed_encoding?</tt> predicate indicates whether the regexp
has a <i>fixed</i> encoding, that is one incompatible with ASCII. A
regexp's encoding can be explicitly fixed by supplying
<tt>Regexp::FIXEDENCODING</tt> as the second argument of
<tt>Regexp.new</tt>:

    r = Regexp.new("a".force_encoding("iso-8859-1"),Regexp::FIXEDENCODING)
    r =~"a\u3042"
       #=> Encoding::CompatibilityError: incompatible encoding regexp match
            (ISO-8859-1 regexp with UTF-8 string)

== Performance

Certain pathological combinations of constructs can lead to abysmally bad
performance.

Consider a string of 25 <i>a</i>s, a <i>d</i>, 4 <i>a</i>s, and a
<i>c</i>.

    s = 'a' * 25 + 'd' 'a' * 4 + 'c'
        #=> "aaaaaaaaaaaaaaaaaaaaaaaaadadadadac"

The following patterns match instantly as you would expect:

    /(b|a)/ =~ s #=> 0
    /(b|a+)/ =~ s #=> 0
    /(b|a+)*\/ =~ s #=> 0

However, the following pattern takes appreciably longer:

    /(b|a+)*c/ =~ s #=> 32

This happens because an atom in the regexp is quantified by both an
immediate <tt>+</tt> and an enclosing <tt>*</tt> with nothing to
differentiate which is in control of any particular character. The
nondeterminism that results produces super-linear performance. (Consult
<i>Mastering Regular Expressions</i> (3rd ed.), pp 222, by
<i>Jeffery Friedl</i>, for an in-depth analysis). This particular case
can be fixed by use of atomic grouping, which prevents the unnecessary
backtracking:

    (start = Time.now) && /(b|a+)*c/ =~ s && (Time.now - start)
       #=> 24.702736882
    (start = Time.now) && /(?>b|a+)*c/ =~ s && (Time.now - start)
       #=> 0.000166571

A similar case is typified by the following example, which takes
approximately 60 seconds to execute for me:

    # Match a string of 29 <i>a</i>s against a pattern of 29 optional
    # <i>a</i>s followed by 29 mandatory <i>a</i>s.
    Regexp.new('a?' * 29 + 'a' * 29) =~ 'a' * 29

The 29 optional <i>a</i>s match the string, but this prevents the 29
mandatory <i>a</i>s that follow from matching. Ruby must then backtrack
repeatedly so as to satisfy as many of the optional matches as it can
while still matching the mandatory 29. It is plain to us that none of the
optional matches can succeed, but this fact unfortunately eludes Ruby.

One approach for improving performance is to anchor the match to the
beginning of the string, thus significantly reducing the amount of
backtracking needed.

    Regexp.new('\A' 'a?' * 29 + 'a' * 29).match('a' * 29)
        #=> #<MatchData "aaaaaaaaaaaaaaaaaaaaaaaaaaaaa">