AuroraMiddleware/wxWidgets
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

interface revisions

Browse Source 
git-svn-id: https://svn.wxwidgets.org/svn/wx/wxWidgets/trunk@56444 c3d73ce0-8a6f-49c7-b76d-6d57e0e08775
This commit is contained in:
Francesco Montorsi 2008-10-19 15:14:14 +00:00
parent 89ff7d649f
commit 6a93e7940b
3 changed files with 269 additions and 228 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

70
interface/wx/list.h
View File

@ -7,26 +7,28 @@
/////////////////////////////////////////////////////////////////////////////
/**
@class wxList
The wxList<T> class provides linked list functionality. It has been rewritten
to be type safe and to provide the full API of the STL std::list container and
should be used like it. The exception is that wxList<T> actually stores
pointers and therefore its iterators return pointers and not references
to the actual objets in the list (see example below) and @e value_type
is defined as @e T*. wxList<T> destroys an object after removing it only
if wxList::DeleteContents has been called.
The wxList<T> class provides linked list functionality.
This class has been rewritten to be type safe and to provide the full API of
the STL std::list container and should be used like it.
The exception is that wxList<T> actually stores pointers and therefore its
iterators return pointers and not references to the actual objets in the list
(see example below) and @e value_type is defined as @e T*.
wxList<T> destroys an object after removing it only if wxList::DeleteContents
has been called.
wxList<T> is not a real template and it requires that you declare and define
each wxListT class in your program. This is done with @e WX_DECLARE_LIST
and @e WX_DEFINE_LIST macros (see example). We hope that we'll be able
to provide a proper template class providing both the STL std::list
and the old wxList API in the future.
and @e WX_DEFINE_LIST macros (see example). We hope that we'll be able to
provide a proper template class providing both the STL std::list and the old
wxList API in the future.
Please refer to the STL std::list documentation for further
information on how to use the class. Below we documented both
the supported STL and the legacy API that originated from the
old wxList class and which can still be used alternatively for
the the same class.
Please refer to the STL std::list documentation for further information on how
to use the class. Below we documented both the supported STL and the legacy API
that originated from the old wxList class and which can still be used alternatively
for the the same class.
Note that if you compile wxWidgets in STL mode (wxUSE_STL defined as 1)
then wxList<T> will actually derive from std::list and just add a legacy
@ -78,6 +80,11 @@
}
@endcode
For compatibility with previous versions wxList and wxStringList classes are
still defined, but their usage is deprecated and they will disappear in the
future versions completely.
The use of the latter is especially discouraged as it is not only unsafe but
is also much less efficient than wxArrayString class.
@library{wxbase}
@category{data}
@ -199,17 +206,17 @@ public:
wxList<T>::compatibility_iterator Item(size_t index) const;
/**
@note This function is deprecated, use Find() instead.
@deprecated This function is deprecated, use Find() instead.
*/
wxList<T>::compatibility_iterator Member(T* object) const;
/**
@note This function is deprecated, use Item() instead.
@deprecated This function is deprecated, use Item() instead.
*/
wxList<T>::compatibility_iterator Nth(int n) const;
/**
@note This function is deprecated, use wxList::GetCount instead.
@deprecated This function is deprecated, use wxList::GetCount instead.
Returns the number of elements in the list.
*/
int Number() const;
@ -351,20 +358,18 @@ public:
void remove(const_reference v);
/**
Returns a reverse iterator pointing to the end of the
reversed list.
Returns a reverse iterator pointing to the end of the reversed list.
*/
reverse_iterator rend();
/**
Returns a const reverse iterator pointing to the end of the
reversed list.
Returns a const reverse iterator pointing to the end of the reversed list.
*/
const_reverse_iterator rend() const;
/**
Resizes the list. If the the list is enlarges items with
the value @e v are appended to the list.
Resizes the list.
If the the list is enlarges items with the value @e v are appended to the list.
*/
void resize(size_type n);
@ -384,13 +389,14 @@ public:
/**
@class wxNode
wxNodeBase is the node structure used in linked lists (see
wxList) and derived classes. You should never use wxNodeBase
class directly, however, because it works with untyped (@c void *) data and
this is unsafe. Use wxNodeBase-derived classes which are automatically defined
by WX_DECLARE_LIST and WX_DEFINE_LIST macros instead as described in
wxList documentation (see example there). Also note that
although there is a class called wxNode, it is defined for backwards
wxNodeBase is the node structure used in linked lists (see wxList) and derived
classes. You should never use wxNodeBase class directly, however, because it
works with untyped (@c void *) data and this is unsafe.
Use wxNodeBase-derived classes which are automatically defined by WX_DECLARE_LIST
and WX_DEFINE_LIST macros instead as described in wxList documentation
(see example there).
Also note that although there is a class called wxNode, it is defined for backwards
compatibility only and usage of this class is strongly deprecated.
In the documentation below, the type @c T should be thought of as a
@ -400,7 +406,7 @@ public:
helps to think of it as if it were.
@library{wxbase}
@category{FIXME}
@category{data}
@see wxList<T>, wxHashTable
*/

334
interface/wx/log.h
View File

@ -51,6 +51,7 @@ public:
Called if the user closes the window interactively, will not be
called if it is destroyed for another reason (such as when program
exits).
Return @true from here to allow the frame to close, @false to
prevent this from happening.
@ -81,16 +82,13 @@ public:
/**
@class wxLogInterposerTemp
A special version of wxLogChain which uses itself as the
new log target. It forwards log messages to the previously installed one in
addition to
processing them itself. Unlike wxLogInterposer, it doesn't
delete the old target which means it can be used to temporarily redirect log
output.
A special version of wxLogChain which uses itself as the new log target.
It forwards log messages to the previously installed one in addition to
processing them itself. Unlike wxLogInterposer, it doesn't delete the old
target which means it can be used to temporarily redirect log output.
As per wxLogInterposer, this class must be derived from to implement
wxLog::DoLog
and/or wxLog::DoLogString methods.
wxLog::DoLog and/or wxLog::DoLogString methods.
@library{wxbase}
@category{logging}
@ -101,6 +99,7 @@ public:
/**
The default constructor installs this object as the current active log target.
*/
wxLogInterposerTemp();
};
@ -159,8 +158,7 @@ public:
/**
Returns @true if the messages are passed to the previously active log
target (default) or @false if PassMessages()
had been called.
target (default) or @false if PassMessages() had been called.
*/
bool IsPassingMessages() const;
@ -173,9 +171,10 @@ public:
void PassMessages(bool passMessages);
/**
Sets another log target to use (may be @NULL). The log target specified
in the wxLogChain(wxLog*) constructor or in a previous call to
this function is deleted.
Sets another log target to use (may be @NULL).
The log target specified in the wxLogChain(wxLog*) constructor or in a
previous call to this function is deleted.
This doesn't change the old log target value (the one the messages are
forwarded to) which still remains the same as was active when wxLogChain
object was created.
@ -436,8 +435,7 @@ public:
by the new lines.
All the messages collected so far can be shown to the user (and the current
buffer cleared) by calling the overloaded wxLogBuffer::Flush
method.
buffer cleared) by calling the overloaded wxLogBuffer::Flush method.
@library{wxbase}
@category{logging}
@ -455,8 +453,8 @@ public:
/**
Returns the current buffer contains. Messages from different log function calls
are separated with the new lines in the buffer.
The buffer can be cleared by Flush() which will
also show the current contents to the user.
The buffer can be cleared by Flush() which will also show the current
contents to the user.
*/
const wxString GetBuffer();
};
@ -466,17 +464,15 @@ public:
/**
@class wxLogInterposer
A special version of wxLogChain which uses itself as the
new log target. It forwards log messages to the previously installed one in
addition to
A special version of wxLogChain which uses itself as the new log target.
It forwards log messages to the previously installed one in addition to
processing them itself.
Unlike wxLogChain which is usually used directly as is,
this class must be derived from to implement wxLog::DoLog
and/or wxLog::DoLogString methods.
Unlike wxLogChain which is usually used directly as is, this class must be
derived from to implement wxLog::DoLog and/or wxLog::DoLogString methods.
wxLogInterposer destroys the previous log target in its destructor. If you
don't want this to happen, use wxLogInterposerTemp instead.
wxLogInterposer destroys the previous log target in its destructor.
If you don't want this to happen, use wxLogInterposerTemp instead.
@library{wxbase}
@category{logging}
@ -487,6 +483,7 @@ public:
/**
The default constructor installs this object as the current active log target.
*/
wxLogInterposer();
};
@ -530,116 +527,132 @@ public:
Otherwise, it is completely hidden behind the @e wxLogXXX() functions and
you may not even know about its existence.
@section overview_wxLog_deriving Deriving your own log target
There are two functions which must be implemented by any derived class to
actually process the log messages: DoLog() and
DoLogString(). The second function receives a string
which just has to be output in some way and the easiest way to write a new log
target is to override just this function in the derived class. If more control
over the output format is needed, then the first function must be overridden
which allows to construct custom messages depending on the log level or even
do completely different things depending on the message severity (for example,
throw away all messages except warnings and errors, show warnings on the
screen and forward the error messages to the user's (or programmer's) cell
phone - maybe depending on whether the timestamp tells us if it is day or
night in the current time zone).
There also functions to support message buffering. Why are they needed?
Some of wxLog implementations, most notably the standard wxLogGui class,
buffer the messages (for example, to avoid showing the user a zillion of modal
message boxes one after another -- which would be really annoying).
Flush() shows them all and clears the buffer contents.
This function doesn't do anything if the buffer is already empty.
See also:
@li Flush()
@li FlushActive()
@section log_derivingyours Deriving your own log target
@section overview_wxLog_Trace_Masks Using trace masks
There are two functions which must be implemented by any derived class to
actually process the log messages: DoLog() and DoLogString().
The second function receives a string which just has to be output in some way
and the easiest way to write a new log target is to override just this function
in the derived class.
The functions below allow some limited customization of wxLog behaviour
without writing a new log target class (which, aside from being a matter of
several minutes, allows you to do anything you want).
The verbose messages are the trace messages which are not disabled in the
release mode and are generated by wxLogVerbose(). They
are not normally shown to the user because they present little interest, but
may be activated, for example, in order to help the user find some program
problem.
As for the (real) trace messages, their handling depends on the settings of
the (application global) @e trace mask which can either be specified using
SetTraceMask(), GetTraceMask() and wxLogTrace() which takes an integer mask
or using AddTraceMask() for string trace masks.
The difference between bit-wise and string trace masks is that a message using
integer trace mask will only be logged if all bits of the mask are set in the
current mask while a message using string mask will be logged simply if the
mask had been added before to the list of allowed ones.
For example,
If more control over the output format is needed, then the first function must
be overridden which allows to construct custom messages depending on the log level
or even do completely different things depending on the message severity
(for example, throw away all messages except warnings and errors, show warnings
on the screen and forward the error messages to the user's (or programmer's) cell
phone - maybe depending on whether the timestamp tells us if it is day or
night in the current time zone).
@code
wxLogTrace( wxTraceRefCount|wxTraceOleCalls, "Active object ref count: %d", nRef );
@endcode
There also functions to support message buffering. Why are they needed?
Some of wxLog implementations, most notably the standard wxLogGui class,
buffer the messages (for example, to avoid showing the user a zillion of modal
message boxes one after another -- which would be really annoying).
will do something only if the current trace mask contains both
@c wxTraceRefCount and @c wxTraceOle, but
Flush() shows them all and clears the buffer contents.
This function doesn't do anything if the buffer is already empty.
@code
wxLogTrace( wxTRACE_OleCalls, "IFoo::Bar() called" );
@endcode
See also:
@li Flush()
@li FlushActive()
will log the message if it was preceded by
@code
wxLog::AddTraceMask( wxTRACE_OleCalls);
@endcode
@section log_tracemasks Using trace masks
Using string masks is simpler and allows you to easily add custom ones, so this is
the preferred way of working with trace messages. The integer trace mask is
kept for compatibility and for additional (but very rarely needed) flexibility
only.
The standard trace masks are given in wxLogTrace() documentation.
Finally, the @e wxLog::DoLog() function automatically prepends a time stamp
to all the messages. The format of the time stamp may be changed: it can be
any string with % specifications fully described in the documentation of the
standard @e strftime() function. For example, the default format is
"[%d/%b/%y %H:%M:%S] " which gives something like "[17/Sep/98 22:10:16] "
(without quotes) for the current date. Setting an empty string as the time
format disables timestamping of the messages completely.
See also
@li AddTraceMask()
@li RemoveTraceMask()
@li ClearTraceMasks()
@li GetTraceMasks()
@li IsAllowedTraceMask()
@li SetVerbose()
@li GetVerbose()
@li SetTimestamp()
@li GetTimestamp()
@li SetTraceMask()
@li GetTraceMask()
@li SetRepetitionCounting()
@li GetRepetitionCounting()
The functions below allow some limited customization of wxLog behaviour
without writing a new log target class (which, aside from being a matter of
several minutes, allows you to do anything you want).
The verbose messages are the trace messages which are not disabled in the
release mode and are generated by wxLogVerbose().
They are not normally shown to the user because they present little interest,
but may be activated, for example, in order to help the user find some program
problem.
@note Timestamping is disabled for Visual C++ users in debug builds by
default because otherwise it would be impossible to directly go to the line
from which the log message was generated by simply clicking in the debugger
window on the corresponding error message. If you wish to enable it, please
use SetTimestamp() explicitly.
As for the (real) trace messages, their handling depends on the settings of
the (application global) @e trace mask which can either be specified using
SetTraceMask(), GetTraceMask() and wxLogTrace() which takes an integer mask
or using AddTraceMask() for string trace masks.
@section overview_wxLog_Target Manipulating the log target
The difference between bit-wise and string trace masks is that a message using
integer trace mask will only be logged if all bits of the mask are set in the
current mask while a message using string mask will be logged simply if the
mask had been added before to the list of allowed ones.
For example,
@code
wxLogTrace( wxTraceRefCount|wxTraceOleCalls, "Active object ref count: %d", nRef );
@endcode
will do something only if the current trace mask contains both @c wxTraceRefCount
and @c wxTraceOle, but:
@code
wxLogTrace( wxTRACE_OleCalls, "IFoo::Bar() called" );
@endcode
will log the message if it was preceded by:
@code
wxLog::AddTraceMask( wxTRACE_OleCalls);
@endcode
Using string masks is simpler and allows you to easily add custom ones, so this
is the preferred way of working with trace messages. The integer trace mask is
kept for compatibility and for additional (but very rarely needed) flexibility
only.
The standard trace masks are given in wxLogTrace() documentation.
Finally, the @e wxLog::DoLog() function automatically prepends a time stamp
to all the messages. The format of the time stamp may be changed: it can be
any string with % specifications fully described in the documentation of the
standard @e strftime() function. For example, the default format is
"[%d/%b/%y %H:%M:%S] " which gives something like "[17/Sep/98 22:10:16] "
(without quotes) for the current date. Setting an empty string as the time
format disables timestamping of the messages completely.
See also
@li AddTraceMask()
@li RemoveTraceMask()
@li ClearTraceMasks()
@li GetTraceMasks()
@li IsAllowedTraceMask()
@li SetVerbose()
@li GetVerbose()
@li SetTimestamp()
@li GetTimestamp()
@li SetTraceMask()
@li GetTraceMask()
@li SetRepetitionCounting()
@li GetRepetitionCounting()
@note Timestamping is disabled for Visual C++ users in debug builds by
default because otherwise it would be impossible to directly go to the line
from which the log message was generated by simply clicking in the debugger
window on the corresponding error message. If you wish to enable it, please
use SetTimestamp() explicitly.
@section log_target Manipulating the log target
The functions in this section work with and manipulate the active log
target. The OnLog() is called by the @e wxLogXXX() functions
and invokes the DoLog() of the active log target if any.
Get/Set methods are used to install/query the current active target and,
finally, DontCreateOnDemand() disables the automatic creation of a standard
log target if none actually exists. It is only useful when the application
is terminating and shouldn't be used in other situations because it may
easily lead to a loss of messages.
See also:
@li OnLog()
@li GetActiveTarget()
@li SetActiveTarget()
@li DontCreateOnDemand()
@li Suspend()
@li Resume()
The functions in this section work with and manipulate the active log
target. The OnLog() is called by the @e wxLogXXX() functions
and invokes the DoLog() of the active log target if any.
Get/Set methods are used to install/query the current active target and,
finally, DontCreateOnDemand() disables the automatic creation of a standard
log target if none actually exists. It is only useful when the application
is terminating and shouldn't be used in other situations because it may
easily lead to a loss of messages. See also
@li OnLog()
@li GetActiveTarget()
@li SetActiveTarget()
@li DontCreateOnDemand()
@li Suspend()
@li Resume()
@library{wxcore}
@category{logging}
@ -650,16 +663,14 @@ class wxLog
{
public:
/**
Add the @a mask to the list of allowed masks for
wxLogTrace().
Add the @a mask to the list of allowed masks for wxLogTrace().
@see RemoveTraceMask(), GetTraceMasks()
*/
static void AddTraceMask(const wxString& mask);
/**
Removes all trace masks previously set with
AddTraceMask().
Removes all trace masks previously set with AddTraceMask().
@see RemoveTraceMask()
*/
@ -667,7 +678,8 @@ public:
/**
Disables time stamping of the log messages.
This function is new since wxWidgets version 2.9
@since 2.9.0
*/
static void SetTimestamp(const wxString& format);
@ -675,6 +687,7 @@ public:
Called to process the message of the specified severity. @a msg is the text
of the message as specified in the call of @e wxLogXXX() function which
generated it and @a timestamp is the moment when the message was generated.
The base class version prepends the timestamp to the message, adds a prefix
corresponding to the log level and then calls
DoLogString() with the resulting string.
@ -685,6 +698,7 @@ public:
/**
Called to log the specified string. The timestamp is already included in the
string but still passed to this function.
A simple implementation may just send the string to @c stdout or, better,
@c stderr.
*/
@ -694,14 +708,14 @@ public:
Instructs wxLog to not create new log targets on the fly if there is none
currently. (Almost) for internal use only: it is supposed to be called by the
application shutdown code.
Note that this function also calls
ClearTraceMasks().
Note that this function also calls ClearTraceMasks().
*/
static void DontCreateOnDemand();
/**
Shows all the messages currently in buffer and clears it. If the buffer
is already empty, nothing happens.
Shows all the messages currently in buffer and clears it.
If the buffer is already empty, nothing happens.
*/
virtual void Flush();
@ -733,8 +747,7 @@ public:
static const wxString GetTimestamp();
/**
Returns the current trace mask, see Customization() section
for details.
Returns the current trace mask, see Customization() section for details.
*/
static wxTraceMask GetTraceMask();
@ -751,38 +764,12 @@ public:
static bool GetVerbose();
/**
Returns @true if the @a mask is one of allowed masks for
wxLogTrace().
Returns @true if the @a mask is one of allowed masks for wxLogTrace().
See also: AddTraceMask(), RemoveTraceMask()
*/
static bool IsAllowedTraceMask(const wxString& mask);
/**
There are two functions which must be implemented by any derived class to
actually process the log messages: DoLog() and
DoLogString(). The second function receives a string
which just has to be output in some way and the easiest way to write a new log
target is to override just this function in the derived class. If more control
over the output format is needed, then the first function must be overridden
which allows you to construct custom messages depending on the log level or even
do completely different things depending on the message severity (for example,
throw away all messages except warnings and errors, show warnings on the
screen and forward the error messages to the user's (or programmer's) cell
phone - maybe depending on whether the timestamp tells us if it is day or
night in the current time zone).
There also functions to support message buffering. Why are they needed?
Some of wxLog implementations, most notably the standard wxLogGui class,
buffer the messages (for example, to avoid showing the user a zillion of modal
message boxes one after another -- which would be really annoying).
Flush() shows them all and clears the buffer contents.
This function doesn't do anything if the buffer is already empty.
Flush()
FlushActive()
*/
/**
Forwards the message at specified level to the @e DoLog() function of the
active log target if there is any, does nothing otherwise.
@ -792,7 +779,8 @@ public:
/**
Remove the @a mask from the list of allowed masks for
wxLogTrace().
See also: AddTraceMask()
@see AddTraceMask()
*/
static void RemoveTraceMask(const wxString& mask);
@ -834,8 +822,7 @@ public:
static void SetTimestamp(const wxString& format);
/**
Sets the trace mask, see Customization()
section for details.
Sets the trace mask, see @ref log_derivingyours section for details.
*/
static void SetTraceMask(wxTraceMask mask);
@ -846,14 +833,16 @@ public:
static void SetVerbose(bool verbose = true);
/**
Suspends the logging until Resume() is called. Note that
the latter must be called the same number of times as the former to undo it,
i.e. if you call Suspend() twice you must call Resume() twice as well.
Suspends the logging until Resume() is called.
Note that the latter must be called the same number of times as the former
to undo it, i.e. if you call Suspend() twice you must call Resume() twice as well.
Note that suspending the logging means that the log sink won't be be flushed
periodically, it doesn't have any effect if the current log target does the
logging immediately without waiting for Flush() to be
called (the standard GUI log target only shows the log dialog when it is
flushed, so Suspend() works as expected with it).
logging immediately without waiting for Flush() to be called (the standard
GUI log target only shows the log dialog when it is flushed, so Suspend()
works as expected with it).
@see Resume(), wxLogNull
*/
@ -876,7 +865,7 @@ public:
For instance, the example of the overview:
@code
wxFile file;
wxFile file;
// wxFile.Open() normally complains if file can't be opened, we don't want it
{
@ -891,7 +880,7 @@ public:
would be better written as:
@code
wxFile file;
wxFile file;
// don't try to open file if it doesn't exist, we are prepared to deal with
// this ourselves - but all other errors are not expected
@ -921,6 +910,7 @@ public:
/**
Resumes logging.
*/
~wxLogNull();
};

93
interface/wx/longlong.h
View File

@ -29,10 +29,9 @@
If a native (i.e. supported directly by the compiler) 64 bit integer type was
found to exist, @e wxLongLong_t macro will be defined to correspond to it.
Also, in this case only, two additional macros will be defined:
wxLongLongFmtSpec() for printing 64 bit integers
using the standard @c printf() function (but see also
wxLongLong::ToString for a more portable solution) and
wxLL() for defining 64 bit integer compile-time constants.
- wxLongLongFmtSpec() for printing 64 bit integers using the standard @c printf()
function (but see also wxLongLong::ToString for a more portable solution);
- wxLL() for defining 64 bit integer compile-time constants.
@library{wxbase}
@category{data}
@ -40,6 +39,16 @@
class wxLongLong
{
public:
/**
Default constructor initializes the object to 0.
*/
wxLongLong();
/**
Constructor from native long long (only for compilers supporting it).
*/
wxLongLong(wxLongLong_t ll);
/**
Constructor from 2 longs: the high and low part are combined into one
wxLongLong.
@ -49,16 +58,17 @@ public:
//@{
/**
Returns an absolute value of wxLongLong - either making a copy (const version)
or modifying it in place (the second one). Not in wxULongLong.
or modifying it in place (the second one). Not in wxULongLong.
*/
wxLongLong Abs();
const wxLongLong& Abs();
wxLongLong Abs() const;
wxLongLong& Abs();
//@}
/**
This allows to convert a double value to wxLongLong type. Such conversion is
not always possible in which case the result will be silently truncated in a
platform-dependent way. Not in wxULongLong.
This allows to convert a double value to wxLongLong type.
Such conversion is not always possible in which case the result will be
silently truncated in a platform-dependent way. Not in wxULongLong.
*/
wxLongLong Assign(double d);
@ -73,7 +83,7 @@ public:
unsigned long GetLo() const;
/**
Convert to native long long (only for compilers supporting it)
Convert to native long long (only for compilers supporting it).
*/
wxLongLong_t GetValue() const;
@ -94,48 +104,83 @@ public:
*/
wxString ToString() const;
/**
Adds 2 wxLongLongs together and returns the result.
*/
wxLongLong operator+(const wxLongLong& ll) const;
//@{
/**
Pre/post increment operator.
*/
wxLongLong operator++();
wxLongLong operator++(int );
//@}
/**
Add another wxLongLong to this one.
*/
wxLongLong operator+(const wxLongLong& ll);
wxLongLong& operator+(const wxLongLong& ll);
/**
Subtracts 2 wxLongLongs and returns the result.
*/
wxLongLong operator-(const wxLongLong& ll) const;
/**
Subtracts another wxLongLong from this one.
*/
wxLongLong& operator-(const wxLongLong& ll);
//@{
/**
Pre/post increment operator.
*/
wxLongLong operator++();
wxLongLong operator++(int);
//@}
//@{
/**
Pre/post decrement operator.
*/
wxLongLong operator--();
wxLongLong operator--(int );
wxLongLong operator--(int);
//@}
/**
Subtracts another wxLongLong from this one.
Returns the value of this wxLongLong with opposite sign. Not in wxULongLong.
*/
wxLongLong operator-(const wxLongLong& ll);
wxLongLong operator-() const;
/**
Assignment operator from unsigned long long. The sign bit will be copied too.
@since 2.7.0
*/
wxLongLong& operator operator=(const wxULongLong& ll);
wxLongLong& operator=(const wxULongLong& ll);
/**
Assignment operator from native long long (only for compilers supporting it).
*/
wxLongLong& operator=(wxLongLong_t ll);
/**
Assignment operator from native unsigned long long (only for compilers supporting it).
@since 2.7.0
*/
wxLongLong& operator=(wxULongLong_t ll);
/**
Assignment operator from long.
@since 2.7.0
*/
wxLongLong& operator=(long l);
/**
Assignment operator from unsigned long.
@since 2.7.0
*/
wxLongLong& operator=(unsigned long l);
};