regex/doc/match_result.qbk
John Maddock 132c39a42c Update regex_match requirements.
Fixes #4215.

[SVN r61976]
2010-05-15 09:15:20 +00:00

565 lines
26 KiB
Plaintext

[/
Copyright 2006-2007 John Maddock.
Distributed under the Boost Software License, Version 1.0.
(See accompanying file LICENSE_1_0.txt or copy at
http://www.boost.org/LICENSE_1_0.txt).
]
[section:match_results match_results]
[h4 Synopsis]
#include <boost/regex.hpp>
Regular expressions are different from many simple pattern-matching algorithms
in that as well as finding an overall match they can also produce
sub-expression matches: each sub-expression being delimited in the
pattern by a pair of parenthesis (...). There has to be some method for
reporting sub-expression matches back to the user: this is achieved this by
defining a class `match_results` that acts as an indexed collection of
sub-expression matches, each sub-expression match being contained in an
object of type [sub_match].
Template class `match_results` denotes a collection of character
sequences representing the result of a regular expression match. Objects of
type `match_results` are passed to the algorithms [regex_match] and [regex_search],
and are returned by the iterator [regex_iterator]. Storage for the
collection is allocated and freed as necessary by the member functions of
class `match_results`.
The template class `match_results` conforms to the requirements of a Sequence,
as specified in (lib.sequence.reqmts), except that only operations defined for
const-qualified Sequences are supported.
Class template `match_results` is most commonly used as one of the typedefs
`cmatch`, `wcmatch`, `smatch`, or `wsmatch`:
template <class BidirectionalIterator,
class Allocator = std::allocator<sub_match<BidirectionalIterator> >
class match_results;
typedef match_results<const char*> cmatch;
typedef match_results<const wchar_t*> wcmatch;
typedef match_results<string::const_iterator> smatch;
typedef match_results<wstring::const_iterator> wsmatch;
template <class BidirectionalIterator,
class Allocator = std::allocator<sub_match<BidirectionalIterator> >
class match_results
{
public:
typedef sub_match<BidirectionalIterator> value_type;
typedef const value_type& const_reference;
typedef const_reference reference;
typedef implementation defined const_iterator;
typedef const_iterator iterator;
typedef typename iterator_traits<BidirectionalIterator>::difference_type difference_type;
typedef typename Allocator::size_type size_type;
typedef Allocator allocator_type;
typedef typename iterator_traits<BidirectionalIterator>::value_type char_type;
typedef basic_string<char_type> string_type;
// construct/copy/destroy:
``[link boost_regex.match_results.construct explicit match_results]``(const Allocator& a = Allocator());
``[link boost_regex.match_results.copy_construct match_results]``(const match_results& m);
``[link boost_regex.match_results.assign match_results& operator=]``(const match_results& m);
~match_results();
// size:
size_type ``[link boost_regex.match_results.size size]``() const;
size_type ``[link boost_regex.match_results.max_size max_size]``() const;
bool ``[link boost_regex.match_results.empty empty]``() const;
// element access:
difference_type ``[link boost_regex.match_results.length length]``(int sub = 0) const;
difference_type ``[link boost_regex.match_results.length length]``(const char_type* sub) const;
template <class charT>
difference_type ``[link boost_regex.match_results.length length]``(const charT* sub) const;
template <class charT, class Traits, class A>
difference_type ``[link boost_regex.match_results.length length]``(const std::basic_string<charT, Traits, A>& sub) const;
difference_type ``[link boost_regex.match_results.position position]``(unsigned int sub = 0) const;
difference_type ``[link boost_regex.match_results.position position]``(const char_type* sub) const;
template <class charT>
difference_type ``[link boost_regex.match_results.position position]``(const charT* sub) const;
template <class charT, class Traits, class A>
difference_type ``[link boost_regex.match_results.position position]``(const std::basic_string<charT, Traits, A>& sub) const;
string_type ``[link boost_regex.match_results.str str]``(int sub = 0) const;
string_type ``[link boost_regex.match_results.str str]``(const char_type* sub)const;
template <class Traits, class A>
string_type ``[link boost_regex.match_results.str str]``(const std::basic_string<char_type, Traits, A>& sub)const;
template <class charT>
string_type ``[link boost_regex.match_results.str str]``(const charT* sub)const;
template <class charT, class Traits, class A>
string_type ``[link boost_regex.match_results.str str]``(const std::basic_string<charT, Traits, A>& sub)const;
const_reference ``[link boost_regex.match_results.subscript operator\[\]]``(int n) const;
const_reference ``[link boost_regex.match_results.subscript operator\[\]]``(const char_type* n) const;
template <class Traits, class A>
const_reference ``[link boost_regex.match_results.subscript operator\[\]]``(const std::basic_string<char_type, Traits, A>& n) const;
template <class charT>
const_reference ``[link boost_regex.match_results.subscript operator\[\]]``(const charT* n) const;
template <class charT, class Traits, class A>
const_reference ``[link boost_regex.match_results.subscript operator\[\]]``(const std::basic_string<charT, Traits, A>& n) const;
const_reference ``[link boost_regex.match_results.prefix prefix]``() const;
const_reference ``[link boost_regex.match_results.suffix suffix]``() const;
const_iterator ``[link boost_regex.match_results.begin begin]``() const;
const_iterator ``[link boost_regex.match_results.end end]``() const;
// format:
template <class OutputIterator, class Formatter>
OutputIterator ``[link boost_regex.match_results.format format]``(OutputIterator out,
Formatter fmt,
match_flag_type flags = format_default) const;
template <class Formatter>
string_type ``[link boost_regex.match_results.format2 format]``(Formatter fmt,
match_flag_type flags = format_default) const;
allocator_type ``[link boost_regex.match_results.get_allocator get_allocator]``() const;
void ``[link boost_regex.match_results.swap swap]``(match_results& that);
#ifdef BOOST_REGEX_MATCH_EXTRA
typedef typename value_type::capture_sequence_type capture_sequence_type;
const capture_sequence_type& ``[link boost_regex.match_results.captures captures]``(std::size_t i)const;
#endif
};
template <class BidirectionalIterator, class Allocator>
bool ``[link boost_regex.match_results.op_eq operator ==]`` (const match_results<BidirectionalIterator, Allocator>& m1,
const match_results<BidirectionalIterator, Allocator>& m2);
template <class BidirectionalIterator, class Allocator>
bool ``[link boost_regex.match_results.op_ne operator !=]`` (const match_results<BidirectionalIterator, Allocator>& m1,
const match_results<BidirectionalIterator, Allocator>& m2);
template <class charT, class traits, class BidirectionalIterator, class Allocator>
basic_ostream<charT, traits>&
``[link boost_regex.match_results.op_stream operator <<]`` (basic_ostream<charT, traits>& os,
const match_results<BidirectionalIterator, Allocator>& m);
template <class BidirectionalIterator, class Allocator>
void ``[link boost_regex.match_results.op_swap swap]``(match_results<BidirectionalIterator, Allocator>& m1,
match_results<BidirectionalIterator, Allocator>& m2);
[h4 Description]
In all `match_results` constructors, a copy of the Allocator argument is used
for any memory allocation performed by the constructor or member functions
during the lifetime of the object.
[#boost_regex.match_results.construct]
match_results(const Allocator& a = Allocator());
[*Effects]: Constructs an object of class `match_results`. The postconditions of
this function are indicated in the table:
[table
[[Element][Value]]
[[empty()][true]]
[[size()][0]]
[[str()][basic_string<charT>()]]
]
[#boost_regex.match_results.copy_construct]
match_results(const match_results& m);
[*Effects]: Constructs an object of class match_results, as a copy of m.
[#boost_regex.match_results.assign]
match_results& operator=(const match_results& m);
[*Effects]: Assigns m to *this. The postconditions of this function are
indicated in the table:
[table
[[Element][Value]]
[[empty()][m.empty().]]
[[size()][m.size().]]
[[str(n)][m.str(n) for all integers n < m.size().]]
[[prefix()][m.prefix().]]
[[suffix()][m.suffix().]]
[[(*this)\[n\]][m\[n\] for all integers n < m.size().]]
[[length(n)][m.length(n) for all integers n < m.size().]]
[[position(n)][m.position(n) for all integers n < m.size().]]
]
[#boost_regex.match_results.size]
size_type size()const;
[*Effects]: Returns the number of [sub_match] elements stored in *this; that is
the number of marked sub-expressions in the regular expression that was
matched plus one.
[#boost_regex.match_results.max_size]
size_type max_size()const;
[*Effects]: Returns the maximum number of [sub_match] elements that can be
stored in *this.
[#boost_regex.match_results.empty]
bool empty()const;
[*Effects]: Returns size() == 0.
[#boost_regex.match_results.length]
difference_type length(int sub = 0)const;
difference_type length(const char_type* sub)const;
template <class charT>
difference_type length(const charT* sub)const;
template <class charT, class Traits, class A>
difference_type length(const std::basic_string<charT, Traits, A>&)const;
[*Requires]: that the match_results object has been initialized as a result of a
successful call to [regex_search] or [regex_match] or was returned from a [regex_iterator], and
that the underlying iterators have not been subsequently invalidated. Will raise a
`std::logic_error` if the match_results object was not initialized.
[*Effects]: Returns the length of sub-expression /sub/, that is to say:
`(*this)[sub].length()`.
The overloads that accept a string refer to a named sub-expression /n/.
In the event that there is no such named sub-expression then returns zero.
The template overloads of this function, allow the string and\/or character type
to be different from the character type of the underlying sequence and\/or regular expression:
in this case the characters will be widened to the underlying character type of the original regular expression.
A compiler error will occur if the argument passes a wider character type than the underlying sequence.
These overloads allow a normal narrow character C string literal to be used as an argument, even when
the underlying character type of the expression being matched may be something more exotic such as a
Unicode character type.
[#boost_regex.match_results.position]
difference_type position(unsigned int sub = 0)const;
difference_type position(const char_type* sub)const;
template <class charT>
difference_type position(const charT* sub)const;
template <class charT, class Traits, class A>
difference_type position(const std::basic_string<charT, Traits, A>&)const;
[*Requires]: that the match_results object has been initialized as a result of a
successful call to [regex_search] or [regex_match] or was returned from a [regex_iterator], and
that the underlying iterators have not been subsequently invalidated. Will raise a
`std::logic_error` if the match_results object was not initialized.
[*Effects]: Returns the starting location of sub-expression /sub/, or -1 if /sub/ was
not matched. Note that if this represents a partial match , then `position()`
will return the location of the partial match even though `(*this)[0].matched` is false.
The overloads that accept a string refer to a named sub-expression /n/.
In the event that there is no such named sub-expression then returns -1.
The template overloads of this function, allow the string and\/or character type
to be different from the character type of the underlying sequence and\/or regular expression:
in this case the characters will be widened to the underlying character type of the original regular expression.
A compiler error will occur if the argument passes a wider character type than the underlying sequence.
These overloads allow a normal narrow character C string literal to be used as an argument, even when
the underlying character type of the expression being matched may be something more exotic such as a
Unicode character type.
[#boost_regex.match_results.str]
string_type str(int sub = 0)const;
string_type str(const char_type* sub)const;
template <class Traits, class A>
string_type str(const std::basic_string<char_type, Traits, A>& sub)const;
template <class charT>
string_type str(const charT* sub)const;
template <class charT, class Traits, class A>
string_type str(const std::basic_string<charT, Traits, A>& sub)const;
[*Requires]: that the match_results object has been initialized as a result of a
successful call to [regex_search] or [regex_match] or was returned from a [regex_iterator], and
that the underlying iterators have not been subsequently invalidated. Will raise a
`std::logic_error` if the match_results object was not initialized.
[*Effects]: Returns sub-expression /sub/ as a string: `string_type((*this)[sub])`.
The overloads that accept a string, return the string that matched the named sub-expression /n/.
In the event that there is no such named sub-expression then returns an empty string.
The template overloads of this function, allow the string and\/or character type
to be different from the character type of the underlying sequence and\/or regular expression:
in this case the characters will be widened to the underlying character type of the original regular expression.
A compiler error will occur if the argument passes a wider character type than the underlying sequence.
These overloads allow a normal narrow character C string literal to be used as an argument, even when
the underlying character type of the expression being matched may be something more exotic such as a
Unicode character type.
[#boost_regex.match_results.subscript]
const_reference operator[](int n) const;
const_reference operator[](const char_type* n) const;
template <class Traits, class A>
const_reference operator[](const std::basic_string<char_type, Traits, A>& n) const;
template <class charT>
const_reference operator[](const charT* n) const;
template <class charT, class Traits, class A>
const_reference operator[](const std::basic_string<charT, Traits, A>& n) const;
[*Requires]: that the match_results object has been initialized as a result of a
successful call to [regex_search] or [regex_match] or was returned from a [regex_iterator], and
that the underlying iterators have not been subsequently invalidated. Will raise a
`std::logic_error` if the match_results object was not initialized.
[*Effects]: Returns a reference to the [sub_match] object representing the character
sequence that matched marked sub-expression /n/. If `n == 0` then returns a
reference to a [sub_match] object representing the character sequence that
matched the whole regular expression. If /n/ is out of range, or if /n/ is an
unmatched sub-expression, then returns a [sub_match] object whose matched
member is false.
The overloads that accept a string, return a reference to the [sub_match]
object representing the character sequence that matched the named sub-expression /n/.
In the event that there is no such named sub-expression then returns a [sub_match] object whose matched
member is false.
The template overloads of this function, allow the string and\/or character type
to be different from the character type of the underlying sequence and\/or regular expression:
in this case the characters will be widened to the underlying character type of the original regular expression.
A compiler error will occur if the argument passes a wider character type than the underlying sequence.
These overloads allow a normal narrow character C string literal to be used as an argument, even when
the underlying character type of the expression being matched may be something more exotic such as a
Unicode character type.
[#boost_regex.match_results.prefix]
const_reference prefix()const;
[*Requires]: that the match_results object has been initialized as a result of a
successful call to [regex_search] or [regex_match] or was returned from a [regex_iterator], and
that the underlying iterators have not been subsequently invalidated. Will raise a
`std::logic_error` if the match_results object was not initialized.
[*Effects]: Returns a reference to the [sub_match] object representing the
character sequence from the start of the string being matched or searched, to the
start of the match found.
[#boost_regex.match_results.suffix]
const_reference suffix()const;
[*Requires]: that the match_results object has been initialized as a result of a
successful call to [regex_search] or [regex_match] or was returned from a [regex_iterator], and
that the underlying iterators have not been subsequently invalidated. Will raise a
`std::logic_error` if the match_results object was not initialized.
[*Effects]: Returns a reference to the [sub_match] object representing the
character sequence from the end of the match found to the end of the
string being matched or searched.
[#boost_regex.match_results.begin]
const_iterator begin()const;
[*Effects]: Returns a starting iterator that enumerates over all the marked
sub-expression matches stored in *this.
[#boost_regex.match_results.end]
const_iterator end()const;
[*Effects]: Returns a terminating iterator that enumerates over all the
marked sub-expression matches stored in *this.
[#boost_regex.match_results_format]
[#boost_regex.match_results.format]
template <class OutputIterator, class Formatter>
OutputIterator format(OutputIterator out,
Formatter fmt,
match_flag_type flags = format_default);
[*Requires]: The type `OutputIterator` conforms to the Output Iterator requirements
(C++ std 24.1.2).
The type `Formatter` must be either a pointer to a null-terminated string
of type `char_type[]`, or be a container of `char_type`'s (for example
`std::basic_string<char_type>`) or be a unary, binary or ternary functor
that computes the replacement string from a function call: either
`fmt(*this)` which must return a container of `char_type`'s to be used as the
replacement text, or either `fmt(*this, out)` or `fmt(*this, out, flags)`, both of
which write the replacement text to `*out`, and then return the new
OutputIterator position. Note that if the formatter is a functor, then it is
['passed by value]: users that want to pass function objects with internal state
might want to use [@../../../../doc/html/ref.html Boost.Ref] to wrap the object so
that it's passed by reference.
[*Requires]: that the match_results object has been initialized as a result of a
successful call to [regex_search] or [regex_match] or was returned from a [regex_iterator], and
that the underlying iterators have not been subsequently invalidated. Will raise a
`std::logic_error` if the match_results object was not initialized.
[*Effects]: If `fmt` is either a null-terminated string, or a
container of `char_type`'s, then copies the character sequence `[fmt.begin(), fmt.end())` to
`OutputIterator` /out/. For each format specifier or escape sequence in
/fmt/, replace that sequence with either the character(s) it represents,
or the sequence of characters within `*this` to which it refers.
The bitmasks specified in flags determines what format specifiers or
escape sequences are recognized, by default this is the format used by
ECMA-262, ECMAScript Language Specification, Chapter 15 part
5.4.11 String.prototype.replace.
If `fmt` is a function object, then depending on the number of arguments
the function object accepts, it will either:
* Call `fmt(*this)` and copy the string returned to `OutputIterator`
/out/.
* Call `fmt(*this, out)`.
* Call `fmt(*this, out, flags)`.
In all cases the new position of the `OutputIterator` is returned.
See the [link boost_regex.format format syntax guide for more information].
[*Returns]: out.
[#boost_regex.match_results.format2]
template <class Formatter>
string_type format(Formatter fmt,
match_flag_type flags = format_default);
[*Requires]
The type `Formatter` must be either a pointer to a null-terminated string
of type `char_type[]`, or be a container of `char_type`'s (for example
`std::basic_string<char_type>`) or be a unary, binary or ternary functor
that computes the replacement string from a function call: either
`fmt(*this)` which must return a container of `char_type`'s to be used as the
replacement text, or either `fmt(*this, out)` or `fmt(*this, out, flags)`, both of
which write the replacement text to `*out`, and then return the new
OutputIterator position.
[*Requires]: that the match_results object has been initialized as a result of a
successful call to [regex_search] or [regex_match] or was returned from a [regex_iterator], and
that the underlying iterators have not been subsequently invalidated. Will raise a
`std::logic_error` if the match_results object was not initialized.
[*Effects]:
If `fmt` is either a null-terminated string, or a
container of `char_type`'s, then copies the string /fmt/: For each format specifier or
escape sequence in /fmt/, replace that sequence with either the
character(s) it represents, or the sequence of characters within `*this` to
which it refers. The bitmasks specified in flags determines what format
specifiers or escape sequences are recognized, by default this is the format
used by ECMA-262, ECMAScript Language Specification, Chapter 15 part
5.4.11 String.prototype.replace.
If `fmt` is a function object, then depending on the number of arguments
the function object accepts, it will either:
* Call `fmt(*this)` and return the result.
* Call `fmt(*this, unspecified-output-iterator)`, where `unspecified-output-iterator`
is an unspecified OutputIterator type used to copy the output to the string result.
* Call `fmt(*this, unspecified-output-iterator, flags)`, where `unspecified-output-iterator`
is an unspecified OutputIterator type used to copy the output to the string result.
See the [link boost_regex.format format syntax guide for more information].
[#boost_regex.match_results.get_allocator]
allocator_type get_allocator()const;
[*Effects]: Returns a copy of the Allocator that was passed to the object's constructor.
[#boost_regex.match_results.swap]
void swap(match_results& that);
[*Effects]: Swaps the contents of the two sequences.
[*Postcondition]: *this contains the sequence of matched sub-expressions that were in that, that contains the sequence of matched sub-expressions that were in *this.
[*Complexity]: constant time.
[#boost_regex.match_results.capture_type]
typedef typename value_type::capture_sequence_type capture_sequence_type;
Defines an implementation-specific type that satisfies the requirements of
a standard library Sequence (21.1.1 including the optional Table 68 operations),
whose value_type is a `sub_match<BidirectionalIterator>`. This type happens to be
`std::vector<sub_match<BidirectionalIterator> >`, but you shouldn't actually
rely on that.
[#boost_regex.match_results.captures]
const capture_sequence_type& captures(std::size_t i)const;
[*Requires]: that the match_results object has been initialized as a result of a
successful call to [regex_search] or [regex_match] or was returned from a [regex_iterator], and
that the underlying iterators have not been subsequently invalidated. Will raise a
`std::logic_error` if the match_results object was not initialized.
[*Effects]: returns a sequence containing all the captures obtained for sub-expression i.
[*Returns]: `(*this)[i].captures();`
[*Preconditions]: the library must be built and used with BOOST_REGEX_MATCH_EXTRA defined,
and you must pass the flag match_extra to the regex matching functions
([regex_match], [regex_search], [regex_iterator] or [regex_token_iterator]) in
order for this member function to be defined and return useful information.
[*Rationale]: Enabling this feature has several consequences:
* sub_match occupies more memory resulting in complex expressions running out of memory or stack space more quickly during matching.
* The matching algorithms are less efficient at handling some features (independent sub-expressions for example), even when match_extra is not used.
* The matching algorithms are much less efficient (i.e. slower), when match_extra is used. Mostly this is down to the extra memory allocations that have to take place.
[#boost_regex.match_results.op_eq]
template <class BidirectionalIterator, class Allocator>
bool operator == (const match_results<BidirectionalIterator, Allocator>& m1,
const match_results<BidirectionalIterator, Allocator>& m2);
[*Effects]: Compares the two sequences for equality.
[#boost_regex.match_results.op_ne]
template <class BidirectionalIterator, class Allocator>
bool operator != (const match_results<BidirectionalIterator, Allocator>& m1,
const match_results<BidirectionalIterator, Allocator>& m2);
[*Effects]: Compares the two sequences for inequality.
[#boost_regex.match_results.op_stream]
template <class charT, class traits, class BidirectionalIterator, class Allocator>
basic_ostream<charT, traits>&
operator << (basic_ostream<charT, traits>& os,
const match_results<BidirectionalIterator, Allocator>& m);
[*Effects]: Writes the contents of /m/ to the stream /os/ as if by calling
`os << m.str()`; Returns /os/.
[#boost_regex.match_results.op_swap]
template <class BidirectionalIterator, class Allocator>
void swap(match_results<BidirectionalIterator, Allocator>& m1,
match_results<BidirectionalIterator, Allocator>& m2);
[*Effects]: Swaps the contents of the two sequences.
[endsect]