Class: RE2::Regexp

Inherits:
Object show all
Defined in:
ext/re2/re2.cc,
lib/re2/regexp.rb

Class Method Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(pattern) ⇒ RE2::Regexp #initialize(pattern, options) ⇒ RE2::Regexp

Returns a new RE2::Regexp object with a compiled version of pattern stored inside.

Overloads:

  • #initialize(pattern) ⇒ RE2::Regexp

    Returns a new RE2::Regexp object with a compiled version of pattern stored inside with the default options.

    Parameters:

    • pattern (String) —

      the pattern to compile

    Raises:

    • (TypeError) —

      if the given pattern can't be coerced to a String

    • (NoMemoryError) —

      if memory could not be allocated for the compiled pattern

  • #initialize(pattern, options) ⇒ RE2::Regexp

    Returns a new RE2::Regexp object with a compiled version of pattern stored inside with the specified options.

    Parameters:

    • pattern (String) —

      the pattern to compile

    • options (Hash) —

      the options with which to compile the pattern

    Options Hash (options):

    • :utf8 (Boolean) — default: true —

      text and pattern are UTF-8; otherwise Latin-1

    • :posix_syntax (Boolean) — default: false —

      restrict regexps to POSIX egrep syntax

    • :longest_match (Boolean) — default: false —

      search for longest match, not first match

    • :log_errors (Boolean) — default: true —

      log syntax and execution errors to ERROR

    • :max_mem (Integer) —

      approx. max memory footprint of RE2

    • :literal (Boolean) — default: false —

      interpret string as literal, not regexp

    • :never_nl (Boolean) — default: false —

      never match \n, even if it is in regexp

    • :case_sensitive (Boolean) — default: true —

      match is case-sensitive (regexp can override with (?i) unless in posix_syntax mode)

    • :perl_classes (Boolean) — default: false —

      allow Perl's \d \s \w \D \S \W when in posix_syntax mode

    • :word_boundary (Boolean) — default: false —

      allow \b \B (word boundary and not) when in posix_syntax mode

    • :one_line (Boolean) — default: false —

      ^ and $ only match beginning and end of text when in posix_syntax mode

    Raises:

    • (TypeError) —

      if the given pattern can't be coerced to a String

    • (NoMemoryError) —

      if memory could not be allocated for the compiled pattern



1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
 
# File 'ext/re2/re2.cc', line 1441

static VALUE re2_regexp_initialize(int argc, VALUE *argv, VALUE self) {
  VALUE pattern, options;
  re2_pattern *p;

  rb_scan_args(argc, argv, "11", &pattern, &options);

  /* Ensure pattern is a string. */
  StringValue(pattern);

  TypedData_Get_Struct(self, re2_pattern, &re2_regexp_data_type, p);

  RE2::Options re2_options;

  if (RTEST(options)) {
    parse_re2_options(&re2_options, options);
  }

  rb_check_frozen(self);

  p->pattern = new(std::nothrow) RE2(
      re2::StringPiece(RSTRING_PTR(pattern), RSTRING_LEN(pattern)), re2_options);

  if (p->pattern == nullptr) {
    rb_raise(rb_eNoMemError, "not enough memory to allocate RE2 object");
  }

  rb_obj_freeze(self);

  return self;
}

Class Method Details

.initialize(pattern) ⇒ RE2::Regexp .initialize(pattern, options) ⇒ RE2::Regexp

Returns a new RE2::Regexp object with a compiled version of pattern stored inside.

Overloads:

  • .initialize(pattern) ⇒ RE2::Regexp

    Returns a new RE2::Regexp object with a compiled version of pattern stored inside with the default options.

    Parameters:

    • pattern (String) —

      the pattern to compile

    Returns:

    Raises:

    • (TypeError) —

      if the given pattern can't be coerced to a String

    • (NoMemoryError) —

      if memory could not be allocated for the compiled pattern

  • .initialize(pattern, options) ⇒ RE2::Regexp

    Returns a new RE2::Regexp object with a compiled version of pattern stored inside with the specified options.

    Parameters:

    • pattern (String) —

      the pattern to compile

    • options (Hash) —

      the options with which to compile the pattern

    Options Hash (options):

    • :utf8 (Boolean) — default: true —

      text and pattern are UTF-8; otherwise Latin-1

    • :posix_syntax (Boolean) — default: false —

      restrict regexps to POSIX egrep syntax

    • :longest_match (Boolean) — default: false —

      search for longest match, not first match

    • :log_errors (Boolean) — default: true —

      log syntax and execution errors to ERROR

    • :max_mem (Integer) —

      approx. max memory footprint of RE2

    • :literal (Boolean) — default: false —

      interpret string as literal, not regexp

    • :never_nl (Boolean) — default: false —

      never match \n, even if it is in regexp

    • :case_sensitive (Boolean) — default: true —

      match is case-sensitive (regexp can override with (?i) unless in posix_syntax mode)

    • :perl_classes (Boolean) — default: false —

      allow Perl's \d \s \w \D \S \W when in posix_syntax mode

    • :word_boundary (Boolean) — default: false —

      allow \b \B (word boundary and not) when in posix_syntax mode

    • :one_line (Boolean) — default: false —

      ^ and $ only match beginning and end of text when in posix_syntax mode

    Returns:

    Raises:

    • (TypeError) —

      if the given pattern can't be coerced to a String

    • (NoMemoryError) —

      if memory could not be allocated for the compiled pattern

.escape(unquoted) ⇒ String

Returns a version of str with all potentially meaningful regexp characters escaped using `QuoteMeta`. The returned string, used as a regular expression, will exactly match the original string.

Examples:

RE2.escape("1.5-2.0?")         #=> "1\\.5\\-2\\.0\\?"
RE2.quote("1.5-2.0?")          #=> "1\\.5\\-2\\.0\\?"
RE2::Regexp.escape("1.5-2.0?") #=> "1\\.5\\-2\\.0\\?"
RE2::Regexp.quote("1.5-2.0?")  #=> "1\\.5\\-2\\.0\\?"

Parameters:

  • unquoted (String) —

    the unquoted string

Returns:

  • (String) —

    the escaped string

Raises:

  • (TypeError) —

    if the given unquoted string cannot be coerced to a String



2457
2458
2459
2460
2461
2462
2463
2464
 
# File 'ext/re2/re2.cc', line 2457

static VALUE re2_escape(VALUE, VALUE unquoted) {
  StringValue(unquoted);

  std::string quoted_string = RE2::QuoteMeta(
      re2::StringPiece(RSTRING_PTR(unquoted), RSTRING_LEN(unquoted)));

  return rb_str_new(quoted_string.data(), quoted_string.size());
}

.match_has_endpos_argument? ⇒ Boolean

Returns whether the underlying RE2 version supports passing an endpos argument to Match. If not, #match will raise an error if attempting to pass an endpos.

Returns:

  • (Boolean) —

    whether the underlying Match has an endpos argument



2210
2211
2212
2213
2214
2215
2216
 
# File 'ext/re2/re2.cc', line 2210

static VALUE re2_regexp_match_has_endpos_argument_p(VALUE) {
#ifdef HAVE_ENDPOS_ARGUMENT
  return Qtrue;
#else
  return Qfalse;
#endif
}

.quote(unquoted) ⇒ String

Returns a version of str with all potentially meaningful regexp characters escaped using `QuoteMeta`. The returned string, used as a regular expression, will exactly match the original string.

Examples:

RE2.escape("1.5-2.0?")         #=> "1\\.5\\-2\\.0\\?"
RE2.quote("1.5-2.0?")          #=> "1\\.5\\-2\\.0\\?"
RE2::Regexp.escape("1.5-2.0?") #=> "1\\.5\\-2\\.0\\?"
RE2::Regexp.quote("1.5-2.0?")  #=> "1\\.5\\-2\\.0\\?"

Parameters:

  • unquoted (String) —

    the unquoted string

Returns:

  • (String) —

    the escaped string

Raises:

  • (TypeError) —

    if the given unquoted string cannot be coerced to a String



2457
2458
2459
2460
2461
2462
2463
2464
 
# File 'ext/re2/re2.cc', line 2457

static VALUE re2_escape(VALUE, VALUE unquoted) {
  StringValue(unquoted);

  std::string quoted_string = RE2::QuoteMeta(
      re2::StringPiece(RSTRING_PTR(unquoted), RSTRING_LEN(unquoted)));

  return rb_str_new(quoted_string.data(), quoted_string.size());
}

Instance Method Details

#==(other) ⇒ Boolean Also known as: eql?

Return whether this regular expression has the same pattern and options as another RE2::Regexp.

Examples:

RE2::Regexp.new('abc') == RE2::Regexp.new('abc')                        #=> true
RE2::Regexp.new('abc') == RE2::Regexp.new('abc', case_sensitive: false) #=> false

Parameters:

  • other (Object) —

    the object to compare against

Returns:

  • (Boolean) —

    whether the two regular expressions are equal



101
102
103
 
# File 'lib/re2/regexp.rb', line 101

def ==(other)
  other.is_a?(RE2::Regexp) && source == other.source && options == other.options
end

#===(text) ⇒ Boolean

Match the pattern against any substring of the given text, returning false if it is not a String, Symbol, or object responding to to_str so that an RE2::Regexp can be used in a case statement like Ruby's own Regexp#===.

Examples:

case "bob 123"
when RE2::Regexp.new('\d+')
  "numbers"
end

Parameters:

  • text (Object) —

    the text to search; anything other than a String, Symbol, or object responding to to_str returns false

Returns:

  • (Boolean) —

    whether the match was successful



85
86
87
88
89
90
91
 
# File 'lib/re2/regexp.rb', line 85

def ===(text)
  text = text.to_s if text.is_a?(Symbol)
  text = ::String.try_convert(text)
  return false unless text

  match?(text)
end

#=~(text) ⇒ Boolean

Returns true if the pattern matches any substring of the given text using `PartialMatch`.

Parameters:

  • text (String, Symbol, nil) —

    the text to search; nil returns false

Returns:

  • (Boolean) —

    whether the match was successful

Raises:

  • (TypeError) —

    if text cannot be coerced to a String



2113
2114
2115
2116
2117
2118
2119
2120
2121
2122
2123
2124
2125
2126
2127
2128
2129
 
# File 'ext/re2/re2.cc', line 2113

static VALUE re2_regexp_match_p(const VALUE self, VALUE text) {
  if (NIL_P(text)) {
    return Qfalse;
  }
  if (SYMBOL_P(text)) {
    text = rb_sym2str(text);
  }
  StringValue(text);
  text = rb_str_new_frozen(text);

  re2_pattern *p = unwrap_re2_regexp(self);
  bool matched = re2_match_without_gvl(
      p->pattern, text, 0, RSTRING_LEN(text), RE2::UNANCHORED, 0, 0);
  RB_GC_GUARD(text);

  return BOOL2RUBY(matched);
}

#case_insensitive? ⇒ Boolean

Returns whether or not the regular expression was compiled with the case_sensitive option set to false.

Examples:

re2 = RE2::Regexp.new("woo?", case_sensitive: true)
re2.case_insensitive? #=> false
re2.casefold?         #=> false

Returns:

  • (Boolean) —

    the inverse of the case_sensitive option



1683
1684
1685
 
# File 'ext/re2/re2.cc', line 1683

static VALUE re2_regexp_case_insensitive(const VALUE self) {
  return BOOL2RUBY(re2_regexp_case_sensitive(self) != Qtrue);
}

#case_sensitive? ⇒ Boolean

Returns whether or not the regular expression was compiled with the case_sensitive option set to true.

Examples:

re2 = RE2::Regexp.new("woo?", case_sensitive: true)
re2.case_sensitive? #=> true

Returns:

  • (Boolean) —

    the case_sensitive option



1667
1668
1669
1670
1671
 
# File 'ext/re2/re2.cc', line 1667

static VALUE re2_regexp_case_sensitive(const VALUE self) {
  re2_pattern *p = unwrap_re2_regexp(self);

  return BOOL2RUBY(p->pattern->options().case_sensitive());
}

#casefold? ⇒ Boolean

Returns whether or not the regular expression was compiled with the case_sensitive option set to false.

Examples:

re2 = RE2::Regexp.new("woo?", case_sensitive: true)
re2.case_insensitive? #=> false
re2.casefold?         #=> false

Returns:

  • (Boolean) —

    the inverse of the case_sensitive option



1683
1684
1685
 
# File 'ext/re2/re2.cc', line 1683

static VALUE re2_regexp_case_insensitive(const VALUE self) {
  return BOOL2RUBY(re2_regexp_case_sensitive(self) != Qtrue);
}

#error ⇒ String?

If the RE2::Regexp could not be created properly, returns an error string otherwise returns nil.

Returns:

  • (String, nil) —

    the error string or nil



1738
1739
1740
1741
1742
1743
1744
1745
1746
 
# File 'ext/re2/re2.cc', line 1738

static VALUE re2_regexp_error(const VALUE self) {
  re2_pattern *p = unwrap_re2_regexp(self);

  if (p->pattern->ok()) {
    return Qnil;
  } else {
    return rb_str_new(p->pattern->error().data(), p->pattern->error().size());
  }
}

#error_arg ⇒ String?

If the RE2::Regexp could not be created properly, returns the offending portion of the regexp otherwise returns nil.

Note RE2 only supports UTF-8 and ISO-8859-1 encoding so strings will be returned in UTF-8 by default or ISO-8859-1 if the :utf8 option for the RE2::Regexp is set to false (any other encoding's behaviour is undefined).

Returns:

  • (String, nil) —

    the offending portion of the regexp or nil



1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
 
# File 'ext/re2/re2.cc', line 1758

static VALUE re2_regexp_error_arg(const VALUE self) {
  re2_pattern *p = unwrap_re2_regexp(self);

  if (p->pattern->ok()) {
    return Qnil;
  } else {
    return encoded_str_new(p->pattern->error_arg().data(),
        p->pattern->error_arg().size(),
        p->pattern->options().encoding());
  }
}

#full_match(text, options = {}) ⇒ RE2::MatchData, ...

Match the pattern against the given text exactly and return a MatchData instance with the specified number of submatches (defaults to the total number of capturing groups) or a boolean (if no submatches are required).

The number of submatches has a significant impact on performance: requesting one submatch is much faster than requesting more than one and requesting zero submatches is faster still.

Examples:

r = RE2::Regexp.new('w(o)(o)')
r.full_match('woo')                #=> #<RE2::MatchData "woo" 1:"o" 2:"o">
r.full_match('woot')               #=> nil
r.full_match('woo', submatches: 1) #=> #<RE2::MatchData "woo" 1:"o">
r.full_match('woo', submatches: 0) #=> true

Parameters:

  • text (String) —

    the text to search

  • options (Hash) (defaults to: {}) —

    the options with which to perform the match

Options Hash (options):

  • :submatches (Integer) —

    how many submatches to extract (0 is fastest), defaults to the total number of capturing groups

Returns:

  • (RE2::MatchData, nil) —

    if extracting any submatches

  • (Boolean) —

    if not extracting any submatches

Raises:

  • (ArgumentError) —

    if given a negative number of submatches

  • (NoMemoryError) —

    if there was not enough memory to allocate the matches

  • (TypeError) —

    if given non-numeric submatches or non-hash options



68
69
70
 
# File 'lib/re2/regexp.rb', line 68

def full_match(text, options = {})
  match(text, Hash(options).merge(anchor: :anchor_both))
end

#full_match?(text) ⇒ Boolean

Returns true if the pattern matches the given text using `FullMatch`.

Parameters:

  • text (String, Symbol, nil) —

    the text to search; nil returns false

Returns:

  • (Boolean) —

    whether the match was successful

Raises:

  • (TypeError) —

    if text cannot be coerced to a String



2140
2141
2142
2143
2144
2145
2146
2147
2148
2149
2150
2151
2152
2153
2154
2155
2156
 
# File 'ext/re2/re2.cc', line 2140

static VALUE re2_regexp_full_match_p(const VALUE self, VALUE text) {
  if (NIL_P(text)) {
    return Qfalse;
  }
  if (SYMBOL_P(text)) {
    text = rb_sym2str(text);
  }
  StringValue(text);
  text = rb_str_new_frozen(text);

  re2_pattern *p = unwrap_re2_regexp(self);
  bool matched = re2_match_without_gvl(
      p->pattern, text, 0, RSTRING_LEN(text), RE2::ANCHOR_BOTH, 0, 0);
  RB_GC_GUARD(text);

  return BOOL2RUBY(matched);
}

#hash ⇒ Integer

Return a hash based on the pattern and options so that equal RE2::Regexps can be used as Hash keys.

Returns:

  • (Integer) —

    the hash of the regular expression



111
112
113
 
# File 'lib/re2/regexp.rb', line 111

def hash
  [source, options].hash
end

#initialize_copy(other) ⇒ Object 



1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
 
# File 'ext/re2/re2.cc', line 1472

static VALUE re2_regexp_initialize_copy(VALUE self, VALUE other) {
  re2_pattern *self_p;
  re2_pattern *other_p = unwrap_re2_regexp(other);

  TypedData_Get_Struct(self, re2_pattern, &re2_regexp_data_type, self_p);

  rb_check_frozen(self);

  if (self_p->pattern) {
    delete self_p->pattern;
    self_p->pattern = nullptr;
  }

  self_p->pattern = new(std::nothrow) RE2(other_p->pattern->pattern(),
                                          other_p->pattern->options());
  if (self_p->pattern == nullptr) {
    rb_raise(rb_eNoMemError, "not enough memory to allocate RE2 object");
  }

  rb_obj_freeze(self);

  return self;
}

#inspect ⇒ String

Returns a printable version of the regular expression.

Note RE2 only supports UTF-8 and ISO-8859-1 encoding so strings will be returned in UTF-8 by default or ISO-8859-1 if the :utf8 option for the RE2::Regexp is set to false (any other encoding's behaviour is undefined).

Examples:

re2 = RE2::Regexp.new("woo?")
re2.inspect #=> "#<RE2::Regexp /woo?/>"

Returns:

  • (String) —

    a printable version of the regular expression



1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
 
# File 'ext/re2/re2.cc', line 1509

static VALUE re2_regexp_inspect(const VALUE self) {
  re2_pattern *p = unwrap_re2_regexp(self);

  std::ostringstream output;

  output << "#<RE2::Regexp /" << p->pattern->pattern() << "/>";

  return encoded_str_new(output.str().data(), output.str().length(),
      p->pattern->options().encoding());
}

#literal? ⇒ Boolean

Returns whether or not the regular expression was compiled with the literal option set to true.

Examples:

re2 = RE2::Regexp.new("woo?", literal: true)
re2.literal? #=> true

Returns:

  • (Boolean) —

    the literal option



1637
1638
1639
1640
1641
 
# File 'ext/re2/re2.cc', line 1637

static VALUE re2_regexp_literal(const VALUE self) {
  re2_pattern *p = unwrap_re2_regexp(self);

  return BOOL2RUBY(p->pattern->options().literal());
}

#log_errors? ⇒ Boolean

Returns whether or not the regular expression was compiled with the log_errors option set to true.

Examples:

re2 = RE2::Regexp.new("woo?", log_errors: true)
re2.log_errors? #=> true

Returns:

  • (Boolean) —

    the log_errors option



1608
1609
1610
1611
1612
 
# File 'ext/re2/re2.cc', line 1608

static VALUE re2_regexp_log_errors(const VALUE self) {
  re2_pattern *p = unwrap_re2_regexp(self);

  return BOOL2RUBY(p->pattern->options().log_errors());
}

#longest_match? ⇒ Boolean

Returns whether or not the regular expression was compiled with the longest_match option set to true.

Examples:

re2 = RE2::Regexp.new("woo?", longest_match: true)
re2.longest_match? #=> true

Returns:

  • (Boolean) —

    the longest_match option



1593
1594
1595
1596
1597
 
# File 'ext/re2/re2.cc', line 1593

static VALUE re2_regexp_longest_match(const VALUE self) {
  re2_pattern *p = unwrap_re2_regexp(self);

  return BOOL2RUBY(p->pattern->options().longest_match());
}

#match(text) ⇒ RE2::MatchData, ... #match(text, options) ⇒ RE2::MatchData, ... #match(text, submatches) ⇒ RE2::MatchData, ...

General matching: match the pattern against the given text using `Match` and return a MatchData instance with the specified number of submatches (defaults to the total number of capturing groups) or a boolean (if no submatches are required).

The number of submatches has a significant impact on performance: requesting one submatch is much faster than requesting more than one and requesting zero submatches is faster still.

Overloads:

  • #match(text) ⇒ RE2::MatchData, ...

    Returns a MatchData containing the matching pattern and all submatches resulting from looking for the regexp in text if the pattern contains capturing groups.

    Returns either true or false indicating whether a successful match was made if the pattern contains no capturing groups.

    Examples:

    Matching with capturing groups

    r = RE2::Regexp.new('w(o)(o)')
r.match('woo') #=> #<RE2::MatchData "woo" 1:"o" 2:"o">

    Matching without capturing groups

    r = RE2::Regexp.new('woo')
r.match('woo') #=> true

    Parameters:

    • text (String, Symbol, nil) —

      the text to search; nil returns nil

    Returns:

    • (RE2::MatchData, nil) —

      if the pattern contains capturing groups

    • (Boolean) —

      if the pattern does not contain capturing groups

    Raises:

    • (NoMemoryError) —

      if there was not enough memory to allocate the submatches

    • (TypeError) —

      if given text that cannot be coerced to a String

  • #match(text, options) ⇒ RE2::MatchData, ...

    See match(text) but with customisable offsets for starting and ending matches, optional anchoring to the start or both ends of the text and a specific number of submatches to extract (padded with nils if necessary).

    Examples:

    Matching with capturing groups

    r = RE2::Regexp.new('w(o)(o)')
r.match('woo', submatches: 1) #=> #<RE2::MatchData "woo" 1:"o">
r.match('woo', submatches: 3) #=> #<RE2::MatchData "woo" 1:"o" 2:"o" 3:nil>
r.match('woot', anchor: :anchor_both, submatches: 0)
#=> false
r.match('woot', anchor: :anchor_start, submatches: 0)
#=> true

    Matching without capturing groups

    r = RE2::Regexp.new('wo+')
r.match('woot', anchor: :anchor_both)  #=> false
r.match('woot', anchor: :anchor_start) #=> true

    Parameters:

    • text (String, Symbol, nil) —

      the text to search; nil returns nil

    • options (Hash) —

      the options with which to perform the match

    Options Hash (options):

    • :startpos (Integer) — default: 0 —

      offset at which to start matching

    • :endpos (Integer) —

      offset at which to stop matching, defaults to the text length

    • :anchor (Symbol) — default: :unanchored —

      one of :unanchored, :anchor_start, :anchor_both to anchor the match

    • :submatches (Integer) —

      how many submatches to extract (0 is fastest), defaults to the number of capturing groups

    Returns:

    • (RE2::MatchData, nil) —

      if extracting any submatches

    • (Boolean) —

      if not extracting any submatches

    Raises:

    • (ArgumentError) —

      if given a negative number of submatches, invalid anchor or invalid startpos, endpos pair

    • (NoMemoryError) —

      if there was not enough memory to allocate the matches

    • (TypeError) —

      if given non-String text, non-numeric number of submatches, non-symbol anchor or non-hash options

    • (RE2::Regexp::UnsupportedError) —

      if given an endpos argument on a version of RE2 that does not support it

  • #match(text, submatches) ⇒ RE2::MatchData, ...
    Deprecated.

    Legacy syntax for matching against text with a specific number of submatches to extract. Use match(text, submatches: n) instead.

    Examples:

    r = RE2::Regexp.new('w(o)(o)')
r.match('woo', 0) #=> true
r.match('woo', 1) #=> #<RE2::MatchData "woo" 1:"o">
r.match('woo', 2) #=> #<RE2::MatchData "woo" 1:"o" 2:"o">

    Parameters:

    • text (String, Symbol, nil) —

      the text to search; nil returns nil

    • submatches (Integer) —

      the number of submatches to extract

    Returns:

    • (RE2::MatchData, nil) —

      if extracting any submatches

    • (Boolean) —

      if not extracting any submatches

    Raises:

    • (NoMemoryError) —

      if there was not enough memory to allocate the submatches

    • (TypeError) —

      if given non-numeric number of submatches



1949
1950
1951
1952
1953
1954
1955
1956
1957
1958
1959
1960
1961
1962
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
2051
2052
2053
2054
2055
2056
2057
2058
2059
2060
2061
2062
2063
2064
2065
2066
2067
2068
2069
2070
2071
2072
2073
2074
2075
2076
2077
2078
2079
2080
2081
2082
2083
2084
2085
2086
2087
2088
2089
2090
2091
2092
2093
2094
2095
2096
2097
2098
2099
2100
2101
2102
 
# File 'ext/re2/re2.cc', line 1949

static VALUE re2_regexp_match(int argc, VALUE *argv, const VALUE self) {
  re2_pattern *p;
  re2_matchdata *m;
  VALUE text, options;

  rb_scan_args(argc, argv, "11", &text, &options);

  if (NIL_P(text)) {
    return Qnil;
  }
  if (SYMBOL_P(text)) {
    text = rb_sym2str(text);
  }

  /* Coerce and freeze text to prevent mutation. */
  StringValue(text);
  text = rb_str_new_frozen(text);

  p = unwrap_re2_regexp(self);

  int n;
  size_t startpos = 0;
  size_t endpos = RSTRING_LEN(text);
  RE2::Anchor anchor = RE2::UNANCHORED;

  if (RTEST(options)) {
    if (RB_INTEGER_TYPE_P(options)) {
      n = NUM2INT(options);

      if (n < 0) {
        rb_raise(rb_eArgError, "number of matches should be >= 0");
      }
    } else {
      if (TYPE(options) != T_HASH) {
        options = rb_Hash(options);
      }

      VALUE endpos_option = rb_hash_aref(options, ID2SYM(id_endpos));
      if (!NIL_P(endpos_option)) {
#ifdef HAVE_ENDPOS_ARGUMENT
        ssize_t endpos_value = NUM2SSIZET(endpos_option);

        if (endpos_value < 0) {
          rb_raise(rb_eArgError, "endpos should be >= 0");
        }

        endpos = static_cast<size_t>(endpos_value);
#else
        rb_raise(re2_eRegexpUnsupportedError, "current version of RE2::Match() does not support endpos argument");
#endif
      }

      VALUE anchor_option = rb_hash_aref(options, ID2SYM(id_anchor));
      if (!NIL_P(anchor_option)) {
        Check_Type(anchor_option, T_SYMBOL);

        ID id_anchor_option = SYM2ID(anchor_option);
        if (id_anchor_option == id_unanchored) {
          anchor = RE2::UNANCHORED;
        } else if (id_anchor_option == id_anchor_start) {
          anchor = RE2::ANCHOR_START;
        } else if (id_anchor_option == id_anchor_both) {
          anchor = RE2::ANCHOR_BOTH;
        } else {
          rb_raise(rb_eArgError, "anchor should be one of: :unanchored, :anchor_start, :anchor_both");
        }
      }

      VALUE submatches_option = rb_hash_aref(options, ID2SYM(id_submatches));
      if (!NIL_P(submatches_option)) {
        n = NUM2INT(submatches_option);

        if (n < 0) {
          rb_raise(rb_eArgError, "number of matches should be >= 0");
        }
      } else {
        if (!p->pattern->ok()) {
          return Qnil;
        }

        n = p->pattern->NumberOfCapturingGroups();
      }

      VALUE startpos_option = rb_hash_aref(options, ID2SYM(id_startpos));
      if (!NIL_P(startpos_option)) {
        ssize_t startpos_value = NUM2SSIZET(startpos_option);

        if (startpos_value < 0) {
          rb_raise(rb_eArgError, "startpos should be >= 0");
        }

        startpos = static_cast<size_t>(startpos_value);
      }
    }
  } else {
    if (!p->pattern->ok()) {
      return Qnil;
    }

    n = p->pattern->NumberOfCapturingGroups();
  }

  if (startpos > endpos) {
    rb_raise(rb_eArgError, "startpos should be <= endpos");
  }

#ifndef HAVE_ENDPOS_ARGUMENT
  /* Old RE2's Match() takes int startpos. Reject values that would overflow. */
  if (startpos > INT_MAX) {
    rb_raise(rb_eRangeError, "startpos should be <= %d", INT_MAX);
  }
#endif

  if (n == 0) {
    bool matched = re2_match_without_gvl(
        p->pattern, text, startpos, endpos, anchor, 0, 0);
    RB_GC_GUARD(text);

    return BOOL2RUBY(matched);
  } else {
    if (n == INT_MAX) {
      rb_raise(rb_eRangeError, "number of matches should be < %d", INT_MAX);
    }

    /* Because match returns the whole match as well. */
    n += 1;

    re2::StringPiece *matches = new(std::nothrow) re2::StringPiece[n];
    if (matches == nullptr) {
      rb_raise(rb_eNoMemError,
               "not enough memory to allocate StringPieces for matches");
    }

    bool matched = re2_match_without_gvl(
        p->pattern, text, startpos, endpos, anchor, matches, n);
    RB_GC_GUARD(text);

    if (matched) {
      VALUE matchdata = rb_class_new_instance(0, 0, re2_cMatchData);
      TypedData_Get_Struct(matchdata, re2_matchdata, &re2_matchdata_data_type, m);

      RB_OBJ_WRITE(matchdata, &m->regexp, self);
      RB_OBJ_WRITE(matchdata, &m->text, text);
      m->matches = matches;
      m->number_of_matches = n;

      return matchdata;
    } else {
      delete[] matches;

      return Qnil;
    }
  }
}

#match?(text) ⇒ Boolean

Returns true if the pattern matches any substring of the given text using `PartialMatch`.

Parameters:

  • text (String, Symbol, nil) —

    the text to search; nil returns false

Returns:

  • (Boolean) —

    whether the match was successful

Raises:

  • (TypeError) —

    if text cannot be coerced to a String



2113
2114
2115
2116
2117
2118
2119
2120
2121
2122
2123
2124
2125
2126
2127
2128
2129
 
# File 'ext/re2/re2.cc', line 2113

static VALUE re2_regexp_match_p(const VALUE self, VALUE text) {
  if (NIL_P(text)) {
    return Qfalse;
  }
  if (SYMBOL_P(text)) {
    text = rb_sym2str(text);
  }
  StringValue(text);
  text = rb_str_new_frozen(text);

  re2_pattern *p = unwrap_re2_regexp(self);
  bool matched = re2_match_without_gvl(
      p->pattern, text, 0, RSTRING_LEN(text), RE2::UNANCHORED, 0, 0);
  RB_GC_GUARD(text);

  return BOOL2RUBY(matched);
}

#max_mem ⇒ Integer

Returns the max_mem setting for the regular expression.

Examples:

re2 = RE2::Regexp.new("woo?", max_mem: 1024)
re2.max_mem #=> 1024

Returns:

  • (Integer) —

    the max_mem option



1622
1623
1624
1625
1626
 
# File 'ext/re2/re2.cc', line 1622

static VALUE re2_regexp_max_mem(const VALUE self) {
  re2_pattern *p = unwrap_re2_regexp(self);

  return INT2FIX(p->pattern->options().max_mem());
}

#named_captures ⇒ Hash

Returns a hash of names to capturing indices of groups.

Note RE2 stores capturing group names as UTF-8, even if the RE2::Regexp sets the :utf8 option to false so strings will always be returned in UTF-8.

Returns:

  • (Hash) —

    a hash of names to capturing indices



1853
1854
1855
1856
1857
1858
1859
1860
1861
1862
1863
1864
1865
 
# File 'ext/re2/re2.cc', line 1853

static VALUE re2_regexp_named_capturing_groups(const VALUE self) {
  re2_pattern *p = unwrap_re2_regexp(self);
  const auto& groups = p->pattern->NamedCapturingGroups();
  VALUE capturing_groups = rb_hash_new();

  for (const auto& group : groups) {
    rb_hash_aset(capturing_groups,
        rb_utf8_str_new(group.first.data(), group.first.size()),
        INT2FIX(group.second));
  }

  return capturing_groups;
}

#named_capturing_groups ⇒ Hash

Returns a hash of names to capturing indices of groups.

Note RE2 stores capturing group names as UTF-8, even if the RE2::Regexp sets the :utf8 option to false so strings will always be returned in UTF-8.

Returns:

  • (Hash) —

    a hash of names to capturing indices



1853
1854
1855
1856
1857
1858
1859
1860
1861
1862
1863
1864
1865
 
# File 'ext/re2/re2.cc', line 1853

static VALUE re2_regexp_named_capturing_groups(const VALUE self) {
  re2_pattern *p = unwrap_re2_regexp(self);
  const auto& groups = p->pattern->NamedCapturingGroups();
  VALUE capturing_groups = rb_hash_new();

  for (const auto& group : groups) {
    rb_hash_aset(capturing_groups,
        rb_utf8_str_new(group.first.data(), group.first.size()),
        INT2FIX(group.second));
  }

  return capturing_groups;
}

#names ⇒ Array<String>

Returns an array of names of all named capturing groups. Names are returned in alphabetical order rather than definition order, as RE2 stores named groups internally in a sorted map.

Note RE2 stores capturing group names as UTF-8, even if the RE2::Regexp sets the :utf8 option to false so strings will always be returned in UTF-8.

Examples:

RE2::Regexp.new('(?P<a>\d+) (?P<b>\w+)').names #=> ["a", "b"]

Returns:

  • (Array<String>) —

    an array of names of named capturing groups



431
432
433
434
435
436
437
438
439
440
441
442
 
# File 'ext/re2/re2.cc', line 431

static VALUE re2_regexp_names(const VALUE self) {
  re2_pattern *p = unwrap_re2_regexp(self);

  const auto& groups = p->pattern->NamedCapturingGroups();
  VALUE names = rb_ary_new2(groups.size());

  for (const auto& group : groups) {
    rb_ary_push(names, rb_utf8_str_new(group.first.data(), group.first.size()));
  }

  return names;
}

#never_nl? ⇒ Boolean

Returns whether or not the regular expression was compiled with the never_nl option set to true.

Examples:

re2 = RE2::Regexp.new("woo?", never_nl: true)
re2.never_nl? #=> true

Returns:

  • (Boolean) —

    the never_nl option



1652
1653
1654
1655
1656
 
# File 'ext/re2/re2.cc', line 1652

static VALUE re2_regexp_never_nl(const VALUE self) {
  re2_pattern *p = unwrap_re2_regexp(self);

  return BOOL2RUBY(p->pattern->options().never_nl());
}

#number_of_capturing_groups ⇒ Integer

Returns the number of capturing subpatterns, or -1 if the regexp wasn't valid on construction. The overall match ($0) does not count: if the regexp is "(a)(b)", returns 2.

Returns:

  • (Integer) —

    the number of capturing subpatterns



1838
1839
1840
1841
1842
 
# File 'ext/re2/re2.cc', line 1838

static VALUE re2_regexp_number_of_capturing_groups(const VALUE self) {
  re2_pattern *p = unwrap_re2_regexp(self);

  return INT2FIX(p->pattern->NumberOfCapturingGroups());
}

#ok? ⇒ Boolean

Returns whether or not the regular expression was compiled successfully.

Examples:

re2 = RE2::Regexp.new("woo?")
re2.ok? #=> true

Returns:

  • (Boolean) —

    whether or not compilation was successful



1548
1549
1550
1551
1552
 
# File 'ext/re2/re2.cc', line 1548

static VALUE re2_regexp_ok(const VALUE self) {
  re2_pattern *p = unwrap_re2_regexp(self);

  return BOOL2RUBY(p->pattern->ok());
}

#one_line? ⇒ Boolean

Returns whether or not the regular expression was compiled with the one_line option set to true.

Examples:

re2 = RE2::Regexp.new("woo?", one_line: true)
re2.one_line? #=> true

Returns:

  • (Boolean) —

    the one_line option



1726
1727
1728
1729
1730
 
# File 'ext/re2/re2.cc', line 1726

static VALUE re2_regexp_one_line(const VALUE self) {
  re2_pattern *p = unwrap_re2_regexp(self);

  return BOOL2RUBY(p->pattern->options().one_line());
}

#options ⇒ Hash

Returns a hash of the options currently set for the RE2::Regexp.

Returns:

  • (Hash) —

    the options



1788
1789
1790
1791
1792
1793
1794
1795
1796
1797
1798
1799
1800
1801
1802
1803
1804
1805
1806
1807
1808
1809
1810
1811
1812
1813
1814
1815
1816
1817
1818
1819
1820
1821
1822
1823
1824
1825
1826
1827
1828
1829
 
# File 'ext/re2/re2.cc', line 1788

static VALUE re2_regexp_options(const VALUE self) {
  re2_pattern *p = unwrap_re2_regexp(self);
  VALUE options = rb_hash_new();

  rb_hash_aset(options, ID2SYM(id_utf8),
      BOOL2RUBY(p->pattern->options().encoding() == RE2::Options::EncodingUTF8));

  rb_hash_aset(options, ID2SYM(id_posix_syntax),
      BOOL2RUBY(p->pattern->options().posix_syntax()));

  rb_hash_aset(options, ID2SYM(id_longest_match),
      BOOL2RUBY(p->pattern->options().longest_match()));

  rb_hash_aset(options, ID2SYM(id_log_errors),
      BOOL2RUBY(p->pattern->options().log_errors()));

  rb_hash_aset(options, ID2SYM(id_max_mem),
      INT2FIX(p->pattern->options().max_mem()));

  rb_hash_aset(options, ID2SYM(id_literal),
      BOOL2RUBY(p->pattern->options().literal()));

  rb_hash_aset(options, ID2SYM(id_never_nl),
      BOOL2RUBY(p->pattern->options().never_nl()));

  rb_hash_aset(options, ID2SYM(id_case_sensitive),
      BOOL2RUBY(p->pattern->options().case_sensitive()));

  rb_hash_aset(options, ID2SYM(id_perl_classes),
      BOOL2RUBY(p->pattern->options().perl_classes()));

  rb_hash_aset(options, ID2SYM(id_word_boundary),
      BOOL2RUBY(p->pattern->options().word_boundary()));

  rb_hash_aset(options, ID2SYM(id_one_line),
      BOOL2RUBY(p->pattern->options().one_line()));

  /* This is a read-only hash after all... */
  rb_obj_freeze(options);

  return options;
}

#partial_match(text, options = {}) ⇒ RE2::MatchData, ...

Match the pattern against any substring of the given text and return a MatchData instance with the specified number of submatches (defaults to the total number of capturing groups) or a boolean (if no submatches are required).

The number of submatches has a significant impact on performance: requesting one submatch is much faster than requesting more than one and requesting zero submatches is faster still.

Examples:

r = RE2::Regexp.new('w(o)(o)')
r.partial_match('woot')                #=> #<RE2::MatchData "woo" 1:"o" 2:"o">
r.partial_match('nope')                #=> nil
r.partial_match('woot', submatches: 1) #=> #<RE2::MatchData "woo" 1:"o">
r.partial_match('woot', submatches: 0) #=> true

Parameters:

  • text (String) —

    the text to search

  • options (Hash) (defaults to: {}) —

    the options with which to perform the match

Options Hash (options):

  • :submatches (Integer) —

    how many submatches to extract (0 is fastest), defaults to the total number of capturing groups

Returns:

  • (RE2::MatchData, nil) —

    if extracting any submatches

  • (Boolean) —

    if not extracting any submatches

Raises:

  • (ArgumentError) —

    if given a negative number of submatches

  • (NoMemoryError) —

    if there was not enough memory to allocate the matches

  • (TypeError) —

    if given non-numeric submatches or non-hash options



39
40
41
 
# File 'lib/re2/regexp.rb', line 39

def partial_match(text, options = {})
  match(text, Hash(options).merge(anchor: :unanchored))
end

#partial_match?(text) ⇒ Boolean

Returns true if the pattern matches any substring of the given text using `PartialMatch`.

Parameters:

  • text (String, Symbol, nil) —

    the text to search; nil returns false

Returns:

  • (Boolean) —

    whether the match was successful

Raises:

  • (TypeError) —

    if text cannot be coerced to a String



2113
2114
2115
2116
2117
2118
2119
2120
2121
2122
2123
2124
2125
2126
2127
2128
2129
 
# File 'ext/re2/re2.cc', line 2113

static VALUE re2_regexp_match_p(const VALUE self, VALUE text) {
  if (NIL_P(text)) {
    return Qfalse;
  }
  if (SYMBOL_P(text)) {
    text = rb_sym2str(text);
  }
  StringValue(text);
  text = rb_str_new_frozen(text);

  re2_pattern *p = unwrap_re2_regexp(self);
  bool matched = re2_match_without_gvl(
      p->pattern, text, 0, RSTRING_LEN(text), RE2::UNANCHORED, 0, 0);
  RB_GC_GUARD(text);

  return BOOL2RUBY(matched);
}

#pattern ⇒ String

Returns a string version of the regular expression.

Note RE2 only supports UTF-8 and ISO-8859-1 encoding so strings will be returned in UTF-8 by default or ISO-8859-1 if the :utf8 option for the RE2::Regexp is set to false (any other encoding's behaviour is undefined).

Examples:

re2 = RE2::Regexp.new("woo?")
re2.to_s #=> "woo?"

Returns:

  • (String) —

    a string version of the regular expression



1532
1533
1534
1535
1536
1537
1538
 
# File 'ext/re2/re2.cc', line 1532

static VALUE re2_regexp_to_s(const VALUE self) {
  re2_pattern *p = unwrap_re2_regexp(self);

  return encoded_str_new(p->pattern->pattern().data(),
      p->pattern->pattern().size(),
      p->pattern->options().encoding());
}

#perl_classes? ⇒ Boolean

Returns whether or not the regular expression was compiled with the perl_classes option set to true.

Examples:

re2 = RE2::Regexp.new("woo?", perl_classes: true)
re2.perl_classes? #=> true

Returns:

  • (Boolean) —

    the perl_classes option



1696
1697
1698
1699
1700
 
# File 'ext/re2/re2.cc', line 1696

static VALUE re2_regexp_perl_classes(const VALUE self) {
  re2_pattern *p = unwrap_re2_regexp(self);

  return BOOL2RUBY(p->pattern->options().perl_classes());
}

#posix_syntax? ⇒ Boolean

Returns whether or not the regular expression was compiled with the posix_syntax option set to true.

Examples:

re2 = RE2::Regexp.new("woo?", posix_syntax: true)
re2.posix_syntax? #=> true

Returns:

  • (Boolean) —

    the posix_syntax option



1578
1579
1580
1581
1582
 
# File 'ext/re2/re2.cc', line 1578

static VALUE re2_regexp_posix_syntax(const VALUE self) {
  re2_pattern *p = unwrap_re2_regexp(self);

  return BOOL2RUBY(p->pattern->options().posix_syntax());
}

#program_size ⇒ Integer

Returns the program size, a very approximate measure of a regexp's "cost". Larger numbers are more expensive than smaller numbers.

Returns:

  • (Integer) —

    the regexp "cost"



1777
1778
1779
1780
1781
 
# File 'ext/re2/re2.cc', line 1777

static VALUE re2_regexp_program_size(const VALUE self) {
  re2_pattern *p = unwrap_re2_regexp(self);

  return INT2FIX(p->pattern->ProgramSize());
}

#scan(text) ⇒ RE2::Scanner

Returns a Scanner for scanning the given text incrementally with `FindAndConsume`.

Examples:

c = RE2::Regexp.new('(\w+)').scan("Foo bar baz")
#=> #<RE2::Scanner:0x0000000000000001>

Parameters:

  • text (text) —

    the text to scan incrementally

Returns:

Raises:

  • (TypeError) —

    if text cannot be coerced to a String



2170
2171
2172
2173
2174
2175
2176
2177
2178
2179
2180
2181
2182
2183
2184
2185
2186
2187
2188
2189
2190
2191
2192
2193
2194
2195
2196
2197
 
# File 'ext/re2/re2.cc', line 2170

static VALUE re2_regexp_scan(const VALUE self, VALUE text) {
  StringValue(text);
  text = rb_str_new_frozen(text);

  re2_pattern *p = unwrap_re2_regexp(self);
  re2_scanner *c;
  VALUE scanner = rb_class_new_instance(0, 0, re2_cScanner);
  TypedData_Get_Struct(scanner, re2_scanner, &re2_scanner_data_type, c);

  RB_OBJ_WRITE(scanner, &c->regexp, self);
  RB_OBJ_WRITE(scanner, &c->text, text);
  c->input = new(std::nothrow) re2::StringPiece(
      RSTRING_PTR(c->text), RSTRING_LEN(c->text));
  if (c->input == nullptr) {
    rb_raise(rb_eNoMemError,
             "not enough memory to allocate StringPiece for input");
  }

  if (p->pattern->ok()) {
    c->number_of_capturing_groups = p->pattern->NumberOfCapturingGroups();
  } else {
    c->number_of_capturing_groups = 0;
  }

  c->eof = false;

  return scanner;
}

#source ⇒ String

Returns a string version of the regular expression.

Note RE2 only supports UTF-8 and ISO-8859-1 encoding so strings will be returned in UTF-8 by default or ISO-8859-1 if the :utf8 option for the RE2::Regexp is set to false (any other encoding's behaviour is undefined).

Examples:

re2 = RE2::Regexp.new("woo?")
re2.to_s #=> "woo?"

Returns:

  • (String) —

    a string version of the regular expression



1532
1533
1534
1535
1536
1537
1538
 
# File 'ext/re2/re2.cc', line 1532

static VALUE re2_regexp_to_s(const VALUE self) {
  re2_pattern *p = unwrap_re2_regexp(self);

  return encoded_str_new(p->pattern->pattern().data(),
      p->pattern->pattern().size(),
      p->pattern->options().encoding());
}

#to_s ⇒ String

Returns a string version of the regular expression.

Note RE2 only supports UTF-8 and ISO-8859-1 encoding so strings will be returned in UTF-8 by default or ISO-8859-1 if the :utf8 option for the RE2::Regexp is set to false (any other encoding's behaviour is undefined).

Examples:

re2 = RE2::Regexp.new("woo?")
re2.to_s #=> "woo?"

Returns:

  • (String) —

    a string version of the regular expression



1532
1533
1534
1535
1536
1537
1538
 
# File 'ext/re2/re2.cc', line 1532

static VALUE re2_regexp_to_s(const VALUE self) {
  re2_pattern *p = unwrap_re2_regexp(self);

  return encoded_str_new(p->pattern->pattern().data(),
      p->pattern->pattern().size(),
      p->pattern->options().encoding());
}

#utf8? ⇒ Boolean

Returns whether or not the regular expression was compiled with the utf8 option set to true.

Examples:

re2 = RE2::Regexp.new("woo?", utf8: true)
re2.utf8? #=> true

Returns:

  • (Boolean) —

    the utf8 option



1563
1564
1565
1566
1567
 
# File 'ext/re2/re2.cc', line 1563

static VALUE re2_regexp_utf8(const VALUE self) {
  re2_pattern *p = unwrap_re2_regexp(self);

  return BOOL2RUBY(p->pattern->options().encoding() == RE2::Options::EncodingUTF8);
}

#word_boundary? ⇒ Boolean

Returns whether or not the regular expression was compiled with the word_boundary option set to true.

Examples:

re2 = RE2::Regexp.new("woo?", word_boundary: true)
re2.word_boundary? #=> true

Returns:

  • (Boolean) —

    the word_boundary option



1711
1712
1713
1714
1715
 
# File 'ext/re2/re2.cc', line 1711

static VALUE re2_regexp_word_boundary(const VALUE self) {
  re2_pattern *p = unwrap_re2_regexp(self);

  return BOOL2RUBY(p->pattern->options().word_boundary());
}