path: root/doc/qvaluevector.doc

/****************************************************************************
**
** QValueVector class documentation
**
** Copyright (C) 1992-2008 Trolltech ASA.  All rights reserved.
**
** This file is part of the TQt GUI Toolkit.
**
** This file may be used under the terms of the GNU General
** Public License versions 2.0 or 3.0 as published by the Free
** Software Foundation and appearing in the files LICENSE.GPL2
** and LICENSE.GPL3 included in the packaging of this file.
** Alternatively you may (at your option) use any later version
** of the GNU General Public License if such license has been
** publicly approved by Trolltech ASA (or its successors, if any)
** and the KDE Free TQt Foundation.
**
** Please review the following information to ensure GNU General
** Public Licensing requirements will be met:
** http://trolltech.com/products/qt/licenses/licensing/opensource/.
** If you are unsure which license is appropriate for your use, please
** review the following information:
** http://trolltech.com/products/qt/licenses/licensing/licensingoverview
** or contact the sales department at [email protected].
**
** This file may be used under the terms of the Q Public License as
** defined by Trolltech ASA and appearing in the file LICENSE.QPL
** included in the packaging of this file.  Licensees holding valid Qt
** Commercial licenses may use this file in accordance with the Qt
** Commercial License Agreement provided with the Software.
**
** This file is provided "AS IS" with NO WARRANTY OF ANY KIND,
** INCLUDING THE WARRANTIES OF DESIGN, MERCHANTABILITY AND FITNESS FOR
** A PARTICULAR PURPOSE. Trolltech reserves all rights not granted
** herein.
**
**********************************************************************/


/*****************************************************************************
  QValueVector documentation
 *****************************************************************************/

/*!
    \class QValueVector ntqvaluevector.h
    \brief The QValueVector class is a value-based template class that provides a dynamic array.

    \ingroup qtl
    \ingroup tools
    \ingroup shared
    \mainclass
    \reentrant

    QValueVector is a TQt implementation of an STL-like vector
    container. It can be used in your application if the standard \c
    vector is not available for your target platforms. QValueVector is
    part of the \link ntqtl.html TQt Template Library\endlink.

    QValueVector\<T\> defines a template instance to create a vector
    of values that all have the class T. QValueVector does not store
    pointers to the members of the vector; it holds a copy of every
    member. QValueVector is said to be value based; in contrast,
    QPtrList and QDict are pointer based.

    QValueVector contains and manages a collection of objects of type
    T and provides random access iterators that allow the contained
    objects to be addressed. QValueVector owns the contained
    elements. For more relaxed ownership semantics, see QPtrCollection
    and friends, which are pointer-based containers.

    QValueVector provides good performance if you append or remove
    elements from the end of the vector. If you insert or remove
    elements from anywhere but the end, performance is very bad. The
    reason for this is that elements must to be copied into new
    positions.

    Some classes cannot be used within a QValueVector: for example,
    all classes derived from QObject and thus all classes that
    implement widgets. Only values can be used in a QValueVector. To
    qualify as a value the class must provide:
    \list
    \i a copy constructor;
    \i an assignment operator;
    \i a default constructor, i.e., a constructor that does not take any arguments.
    \endlist

    Note that C++ defaults to field-by-field assignment operators and
    copy constructors if no explicit version is supplied. In many
    cases this is sufficient.

    QValueVector uses an STL-like syntax to manipulate and address the
    objects it contains. See \link ntqtl.html this document\endlink for
    more information.

    Example:
    \code
    #include <ntqvaluevector.h>
    #include <ntqstring.h>
    #include <stdio.h>

    class Employee
    {
    public:
	Employee(): s(0) {}
	Employee( const QString& name, int salary )
	    : n( name ), s( salary )
	{ }

	QString name()   const	 	{ return n; }
	int	salary() const	 	{ return s; }
	void	setSalary( int salary )	{ s = salary; }
    private:
	QString n;
	int     s;
    };

    int main()
    {
	typedef QValueVector<Employee> EmployeeVector;
	EmployeeVector vec( 3 );  // vector of 3 Employees

	vec[0] = Employee( "Bill", 50000 );
	vec[1] = Employee( "Steve", 80000 );
	vec[2] = Employee( "Ron", 60000 );

	Employee joe( "Joe", 50000 );
	vec.push_back( joe );  // vector expands to accommodate 4 Employees
	joe.setSalary( 70000 );
	
	EmployeeVector::iterator it;
	for( it = vec.begin(); it != vec.end(); ++it )
	    printf( "%s earns %d\n", (*it).name().latin1(), (*it).salary() );

	return 0;
    }
    \endcode

    Program output:
    \code
	Bill earns 50000
	Steve earns 80000
	Ron earns 60000
	Joe earns 50000
    \endcode

    As you can see, the most recent change to Joe's salary did not
    affect the value in the vector because the vector created a copy
    of Joe's entry.

    Many TQt functions return const value vectors; to iterate over
    these you should make a copy and iterate over the copy.

    There are several ways to find items in the vector. The begin()
    and end() functions return iterators to the beginning and end of
    the vector. The advantage of getting an iterator is that you can
    move forward or backward from this position by
    incrementing/decrementing the iterator. The iterator returned by
    end() points to the element which is one past the last element in
    the container. The past-the-end iterator is still associated with
    the vector it belongs to, however it is \e not dereferenceable;
    operator*() will not return a well-defined value. If the vector is
    empty(), the iterator returned by begin() will equal the iterator
    returned by end().

    The fastest way to access an element of a vector is by using
    operator[]. This function provides random access and will return
    a reference to the element located at the specified index. Thus,
    you can access every element directly, in constant time, providing
    you know the location of the element. It is undefined to access
    an element that does not exist (your application will probably
    crash). For example:

    \code
    QValueVector<int> vec1;  // an empty vector
    vec1[10] = 4;  // WARNING: undefined, probably a crash

    QValueVector<QString> vec2(25); // initialize with 25 elements
    vec2[10] = "Dave";  // OK
    \endcode

    Whenever inserting, removing or referencing elements in a vector,
    always make sure you are referring to valid positions. For
    example:

    \code
    void func( QValueVector<int>& vec )
    {
	if ( vec.size() > 10 ) {
	    vec[9] = 99; // OK
	}
    };
    \endcode

    The iterators provided by vector are random access iterators,
    therefore you can use them with many generic algorithms, for
    example, algorithms provided by the STL or the \link ntqtl.html
    QTL\endlink. 

    Another way to find an element in the vector is by using the
    std::find() or \link ntqtl.html#qFind tqFind()\endlink algorithms.
    For example:

    \code
    QValueVector<int> vec;
    ...
    QValueVector<int>::const_iterator it = tqFind( vec.begin(), vec.end(), 3 );
    if ( it != vector.end() )
	// 'it' points to the found element
    \endcode

    It is safe to have multiple iterators on the vector at the same
    time. Since QValueVector manages memory dynamically, all iterators
    can become invalid if a memory reallocation occurs. For example,
    if some member of the vector is removed, iterators that point to
    the removed element and to all following elements become
    invalidated. Inserting into the middle of the vector will
    invalidate all iterators. For convenience, the function back()
    returns a reference to the last element in the vector, and front()
    returns a reference to the first element. If the vector is
    empty(), both back() and front() have undefined behavior (your
    application will crash or do unpredictable things). Use back() and
    front() with caution, for example:

    \code
    QValueVector<int> vec( 3 );
    vec.push_back( 1 );
    vec.push_back( 2 );
    vec.push_back( 3 );
    ...
    if ( !vec.empty() ) {
	// OK: modify the first element
	int& i = vec.front();
	i = 18;
    }
    ...
    QValueVector<double> dvec;
    double d = dvec.back(); // undefined behavior
    \endcode

    Because QValueVector manages memory dynamically, it is recommended
    that you contruct a vector with an initial size. Inserting and
    removing elements happens fastest when:
    \list
    \i Inserting or removing elements happens at the end() of the
    vector;
    \i The vector does not need to allocate additional memory.
    \endlist

    By creating a QValueVector with a sufficiently large initial size,
    there will be less memory allocations. Do not use an initial size
    that is too big, since it will still take time to construct all
    the empty entries, and the extra space will be wasted if it is
    never used.

    Because QValueVector is value-based there is no need to be careful
    about deleting elements in the vector. The vector holds its own
    copies and will free them if the corresponding member or the
    vector itself is deleted. You can force the vector to free all of
    its items with clear().

    QValueVector is shared implicitly, which means it can be copied in
    constant time. If multiple QValueVector instances share the same
    data and one needs to modify its contents, this modifying instance
    makes a copy and modifies its private copy; it thus does not
    affect the other instances. This is often called "copy on write".
    If a QValueVector is being used in a multi-threaded program, you
    must protect all access to the vector. See QMutex.

    There are several ways to insert elements into the vector. The
    push_back() function insert elements into the end of the vector,
    and is usually fastest. The insert() function can be used to add
    elements at specific positions within the vector.

    Items can be also be removed from the vector in several ways.
    There are several variants of the erase() function which removes a
    specific element, or range of elements, from the vector.

    Vectors can be also sorted with various STL algorithms , or it can
    be sorted using the \link ntqtl.html TQt Template Library\endlink.
    For example with qHeapSort():

    Example:
    \code
    QValueVector<int> v( 4 );
    v.push_back( 5 );
    v.push_back( 8 );
    v.push_back( 3 );
    v.push_back( 4 );
    qHeapSort( v );
    \endcode

    QValueVector stores its elements in contiguous memory. This means
    that you can use a QValueVector in any situation that requires an
    array. 
*/

/*! \enum QValueVector::value_type
	The type of the object stored in the vector. */
/*! \enum QValueVector::ValueType
	The type of the object stored in the vector. */
/*! \enum QValueVector::pointer
	The pointer to T type. */
/*! \enum QValueVector::const_pointer
	The const pointer to T type. */
/*! \enum QValueVector::iterator
	The vector's iterator type. */
/*! \enum QValueVector::const_iterator
	The vector's const iterator type. */
/*! \enum QValueVector::Iterator
	The vector's iterator type. */
/*! \enum QValueVector::ConstIterator
	The vector's const iterator type. */
/*! \enum QValueVector::reference
	The reference to T type. */
/*! \enum QValueVector::const_reference
	The const reference to T type. */
/*! \enum QValueVector::size_type
	An unsigned integral type, used to represent various sizes. */
/*! \enum QValueVector::difference_type
	A signed integral type used to represent the distance between two iterators. */


/*!
    \fn QValueVector::QValueVector()

    Constructs an empty vector without any elements. To create a
    vector which reserves an initial amount of space for elements, use
    \c QValueVector(size_type n). 
*/

/*!
    \fn QValueVector::QValueVector( const QValueVector<T>& v )

    Constructs a copy of \a v.

    This operation costs O(1) time because QValueVector is implicitly
    shared.

    The first modification to the vector does takes O(n) time, because
    the elements must be copied.
*/

/*!
    \fn QValueVector::QValueVector( std::vector<T>& v )

    Constructs a copy of \a v.
*/

/*!
    \fn QValueVector::QValueVector( const std::vector<T>& v )

    This operation costs O(n) time because \a v is copied.
*/

/*!
    \fn QValueVector::QValueVector( size_type n, const T& val )

    Constructs a vector with an initial size of \a n elements. Each
    element is initialized with the value of \a val.
*/

/*!
    \fn QValueVector::~QValueVector()

    Destroys the vector, destroying all elements and freeing the
    allocated memory. References to the values in the vector and all
    iterators of this vector become invalidated. Note that it is
    impossible for an iterator to check whether or not it is valid:
    QValueVector is tuned for performance, not for error checking.
*/

/*!
    \fn QValueVector<T>& QValueVector::operator=( const QValueVector<T>& v )

    Assigns \a v to this vector and returns a reference to this vector.

    All iterators of the current vector become invalidated by this
    operation. The cost of such an assignment is O(1) since
    QValueVector is implicitly shared.
*/

/*!
    \fn QValueVector<T>& QValueVector::operator=( const std::vector<T>& v )

    \overload

    Assigns \a v to this vector and returns a reference to this vector.

    All iterators of the current vector become invalidated by this
    operation. The cost of this assignment is O(n) since \a v is
    copied.
*/

/*!
    \fn size_type QValueVector::size() const

    Returns the number of elements in the vector.

    This function is provided for STL compatibility. It is equivalent
    to count().

  \sa empty()
*/

/*!
    \fn size_type QValueVector::count() const

    Returns the number of items in the vector.

    \sa isEmpty()
*/

/*!
    \fn bool QValueVector::empty() const

    Returns TRUE if the vector is empty; otherwise returns FALSE.
    Equivalent to size()==0, only faster.

    This function is provided for STL compatibility. It is equivalent
    to isEmpty().

    \sa size()
*/

/*!
    \fn bool QValueVector::isEmpty() const

    Returns TRUE if the vector is empty; returns FALSE otherwise.

    \sa count()
*/

/*!
    \fn size_type QValueVector::capacity() const

    Returns the maximum number of elements that can be stored in the
    vector without forcing memory reallocation. If memory reallocation
    takes place, some or all iterators may become invalidated.
*/

/*!
    \fn iterator QValueVector::begin()

    Returns an iterator pointing to the beginning of the vector. If
    the vector is empty(), the returned iterator will equal end().
*/

/*!
    \fn const_iterator QValueVector::begin() const

    \overload

    Returns a const iterator pointing to the beginning of the vector.
    If the vector is empty(), the returned iterator will equal end().
*/

/*!
    \fn const_iterator QValueVector::constBegin() const

    Returns a const iterator pointing to the beginning of the vector.
    If the vector is empty(), the returned iterator will equal end().

    \sa constEnd();
*/

/*!
    \fn iterator QValueVector::end()

    Returns an iterator pointing behind the last element of the
    vector.
*/

/*!
    \fn const_iterator QValueVector::end() const

    \overload

    Returns a const iterator pointing behind the last element of the
    vector.
*/

/*!
    \fn const_iterator QValueVector::constEnd() const

    Returns a const iterator pointing behind the last element of the
    vector.

    \sa constBegin()
*/

/*!
    \fn reference QValueVector::at( size_type i , bool* ok )

    Returns a reference to the element with index \a i. If \a ok is
    non-null, and the index \a i is out of range, *\a ok is set to
    FALSE and the returned reference is undefined. If the index \a i
    is within the range of the vector, and \a ok is non-null, *\a ok
    is set to TRUE and the returned reference is well defined.
*/

/*!
    \fn const_reference QValueVector::at( size_type i , bool* ok ) const

    \overload

    Returns a const reference to the element with index \a i. If \a ok
    is non-null, and the index \a i is out of range, *\a ok is set to
    FALSE and the returned reference is undefined. If the index \a i
    is within the range of the vector, and \a ok is non-null, *\a ok
    is set to TRUE and the returned reference is well defined.
*/

/*!
    \fn reference QValueVector::operator[]( size_type i )

    Returns a reference to the element at index \a i. If \a i is out
    of range, this function has undefined behavior.

    \sa at()
*/

/*!
    \fn const_reference QValueVector::operator[]( size_type i ) const

    \overload

    Returns a const reference to the element at index \a i. If \a i is
    out of range, this function has undefined behavior.

    \sa at()
*/

/*!
    \fn reference QValueVector::front()

    Returns a reference to the first element in the vector. If there
    is no first element, this function has undefined behavior. 

    \sa empty() back()
*/

/*!
    \fn const_reference QValueVector::front() const

    \overload

    Returns a const reference to the first element in the vector. If
    there is no first element, this function has undefined behavior. 

    \sa empty() back()
*/

/*!
    \fn reference QValueVector::first()

    Returns a reference to the first item in the vector. If there is
    no first item, this function has undefined behavior.

    \sa empty() last()
*/

/*!
    \fn const_reference QValueVector::first() const

    \overload
*/

/*!
    \fn reference QValueVector::back()

    Returns a reference to the last element in the vector. If there is
    no last element, this function has undefined behavior. 

    \sa empty() front()
*/

/*!
    \fn const_reference QValueVector::back() const

    \overload

    Returns a const reference to the last element in the vector. If
    there is no last element, this function has undefined behavior. 

    \sa empty() front()
*/

/*!
    \fn reference QValueVector::last()

    Returns a reference to the last item in the vector. If there is no
    last item, this function has undefined behavior.

    \sa empty() first()
*/

/*!
    \fn const_reference QValueVector::last() const

    \overload
*/

/*!
    \fn void QValueVector::push_back( const T& x )

    Appends a copy of \a x to the end of the vector. This is the
    fastest way to add new elements.

    This function is provided for STL compatibility. It is equivalent
    to append().

    \sa insert()
*/

/*!
    \fn void QValueVector::append( const T& x )

    Appends a copy of \a x to the end of the vector.

    \sa push_back() insert()
*/

/*!
    \fn void QValueVector::pop_back()

    Removes the last item from the vector.

    This function is provided for STL compatibility.
*/

/*!
    \fn iterator QValueVector::insert( iterator pos, const T& x )

    Inserts a copy of \a x at the position immediately before \a pos.

    \sa push_back()
*/

/*! \fn void QValueVector::detachInternal()
    
    \internal
*/

/*!
    \fn iterator QValueVector::insert( iterator pos, size_type n, const T& x )

    \overload

    Inserts \a n copies of \a x immediately before position x.

    \sa push_back()
*/

/*!
    \fn void QValueVector::reserve( size_type n )

    Increases the vector's capacity. If \a n is less than or equal to
    capacity(), nothing happens. Otherwise, additional memory is
    allocated so that capacity() will be increased to a value greater
    than or equal to \a n. All iterators will then become invalidated.
    Note that the vector's size() and the values of existing elements
    remain unchanged.
*/

/*!
    \fn void QValueVector::resize( size_type n, const T& val = T() )

    Changes the size of the vector to \a n. If \a n is greater than
    the current size(), elements are added to the end and initialized
    with the value of \a val. If \a n is less than size(), elements
    are removed from the end. If \a n is equal to size() nothing
    happens.
*/

/*!
    \fn void QValueVector::clear()

    Removes all the elements from the vector.
*/

/*!
    \fn iterator QValueVector::erase( iterator pos )

    Removes the element at position \a pos and returns the position of
    the next element.
*/

/*!
    \fn iterator QValueVector::erase( iterator first, iterator last )

    \overload

    Removes all elements from \a first up to but not including \a last
    and returns the position of the next element.
*/

/*!
    \fn bool QValueVector::operator==( const QValueVector<T>& x ) const
    
    Returns TRUE if each element in this vector equals each
    corresponding element in \a x; otherwise returns FALSE.
*/

/*!
    \fn bool QValueVector::operator==( const QValueVector<T>& x ) 

    \overload

    Returns TRUE if each element in this vector equals each
    corresponding element in \a x; otherwise returns FALSE.
*/


/*!
    \fn void QValueVector::detach()

    \internal

    If the vector does not share its data with another QValueVector
    instance, nothing happens. Otherwise the function creates a new
    copy of this data and detaches from the shared one. This function
    is called whenever the vector is modified. The implicit sharing
    mechanism is implemented this way.
*/