I am tying to generalize calling NSLog and avoid having to comment-out the calls when I am finished debugging.
I have:
#define USE_ADLog 0
#define USE_RPLog 0
#define USE_DLLog 1
void ConsoleADLog(NSString *message, ...);
void ConsoleRPLog(NSString *message, ...);
void ConsoleDLLog(NSString *message, ...);
and, for example:
void ConsoleADLog(NSString *message, ...) {
#if (USE_ADLog)
va_list optionalArgs;
va_start(optionalArgs, message); // after the parm = message
va_end(optionalArgs);
NSLog(message, optionalArgs);
#endif
}
So far, so good ... however, as soon as I call, for example,:
ConsoleDLLog(#"parm1 = %#, parm2 = %#", parm1, parm2);
which call is inside a secondary thread, I bomb. I thought?? that va_start, va_end were thread safe.
... or is the problem %# ... I know %f works ???
Apparently not!, so how do I make them thread safe ... plain ole
NSLog(#"whatever %#", whateverParm)
works, but not the function above.
Thanks,
this is why many variadic functions include variants which accept va_lists.
see NSLogv.
#if DEBUG == 0
#define DebugLog(...)
#elif DEBUG == 1
#define DebugLog(...) NSLog(__VA_ARGS__)
#endif
Then invoke it using DebugLog(#"Uh oh: %#", someArgument);. If DEBUG is set to 1, the preprocessor will emit NSLog(#"Uh oh: %#", someArgument);. If it's set to 0, it will not emit anything.
Related
How do I wait for a key in GNU EFI?
I intend it to wait for a single key, then continue execution.
My code:
#include <efi.h>
#include <efilib.h>
#include <stdlib.h>
EFI_STATUS efi_main (EFI_HANDLE ImageHandle, EFI_SYSTEM_TABLE *SystemTable) {
EFI_STATUS Status;
ST = SystemTable;
Status = uefi_call_wrapper(ST->ConOut->ClearScreen, 1, ST->ConOut);
if (EFI_ERROR(Status)){
...
return Status;
}
...
Status = ST->ConIn->Reset(ST->ConIn,1!=1);
if (EFI_ERROR(Status)){
...
return Status;
}
// wait for key here
return EFI_SUCCESS;
}
You can get some ideas from the EDK2 implementation of UEFI Shell.
The basic principle is that you need to set up an event waiting for a keypress.
See also section 12.3 Simple Text Input Protocol in UEFI specification v2.8.
TOP OUTSIDE OF efi_main(...){...} YOU SHOULD ADD EFI_INPUT_KEY Key;
INSIDE OF efi_main AFTER,
#if defined(_GNU_EFI)
InitializeLib(ImageHandle, SystemTable);
#endif
ADD UINTN KeyEvent = 0;
OR WHATEVER YOUR WAY IS...
Print(L"PRESS ANY KEY OR PLEASE PRESS ESC TO EXIT.%N\n");
uefi_call_wrapper(SystemTable->ConOut->OutputString, 1, SystemTable->ConOut, L"START KEY READ\n"); // YES WE COULD DO SIMPLY Print(...);
SystemTable->ConIn->Reset(SystemTable->ConIn, FALSE);
//NOW WE SHOULD READ SOME KEYS
//YOU CAN ADD ANY OTHER OPTION HERE, WHEN CAPSLOCK ON THAT WILL PRINT FIRST CHAR CAPITALIZED BUT NOT REST OF CHARACTERS SINCE WE RESET...
//IF YOU WANT MORE KEYS SIMPLY IN VISUAL STUDIO TYPE SCAN_ AND YOU WILL SEE OTHER KEYS
/*...OTHER...*/
while ((UINTN)Key.ScanCode != SCAN_ESC)
{
SystemTable->BootServices->WaitForEvent(1, &SystemTable->ConIn->WaitForKey, &KeyEvent);
SystemTable->ConIn->ReadKeyStroke(SystemTable->ConIn, &Key);
SystemTable->ConIn->Reset(SystemTable->ConIn, FALSE);
Print(L"%c", Key.UnicodeChar);
}
/*...OTHER...*/
SystemTable->RuntimeServices->ResetSystem(EfiResetShutdown, EFI_SUCCESS, 0, NULL);
return EFI_SUCCESS;
So I'm working on a Gtk/X11/Linux app that does screen capture to .gif and one of the methods of stopping the capture is a key press (Esc, Space or End). You can also use a timeout. However to implement the key press to end capture I have to be able to grab the key such that I can get an event even though my window doesn't have focus (it's actually invisible during capture). I believe XGrabKey is the right X11 function for this task:
Window w = Gtk::gdk_x11_drawable_get_xid(Gtk::gtk_widget_get_window(Handle()));
KeyCode kc = XKeysymToKeycode(Gtk::gdk_display, HotKeyCode);
int r = XGrabKey( Gtk::gdk_display,
kc,
0 /* modifiers */,
w /* grab_window */,
TRUE /* owner_events */,
GrabModeAsync /* pointer_mode */,
GrabModeAsync /* keyboard_mode */);
printf("XGrabKey(%p, 0x%x/%x)=%i\n", w, HotKeyCode, kc, r);
Where 'HotKeyCode' is say XK_Escape or something e.g.:
XGrabKey(0x3e00003, 0xff1b/9)=1
XGrabKey is returning '1' or BadRequest. What am I doing wrong here?
FYI the actual Xorg Xserver code in question appears to be here.
Edit: The latest incarnation of the code is:
int x_err_callback(Display *d, XErrorEvent *e)
{
char msg[256];
XGetErrorText(d, e->error_code, msg, sizeof(msg));
printf("X11Error %d (%s): request %d.%d\n",
e->error_code, msg, e->request_code,
e->minor_code);
return 0;
}
Gtk::GdkFilterReturn key_filter(Gtk::GdkXEvent *gdk_xevent,
Gtk::GdkEvent *event,
Gtk::gpointer data)
{
XKeyEvent *xevent = gdk_xevent;
if (xevent->type == KeyPress)
{
int key = ((XKeyEvent *)gdk_xevent)->keycode;
int keysym = XKeycodeToKeysym(Gtk::gdk_display, key, 0);
printf("caught keysym %i\n", keysym);
switch (keysym)
{
case 1: // your_keysym
// your key handler code
break;
}
}
return Gtk::GDK_FILTER_CONTINUE;
}
Gtk::GdkWindow *Root = Gtk::gdk_get_default_root_window();
KeyCode kc = XKeysymToKeycode(Gtk::gdk_display, HotKeyCode);
XSetErrorHandler(x_err_callback);
int r = XGrabKey( Gtk::gdk_display,
kc,
AnyModifier /* modifiers */,
GDK_WINDOW_XWINDOW(Root) /* grab_window */,
TRUE /* owner_events */,
GrabModeAsync /* pointer_mode */,
GrabModeSync /* keyboard_mode */);
Gtk::gdk_window_set_events(Root,
(Gtk::GdkEventMask)
(Gtk::GDK_KEY_PRESS_MASK |
Gtk::GDK_KEY_RELEASE_MASK));
Gtk::gdk_window_add_filter(NULL, key_filter, this);
AnyModifier actually results in an error. '0' doesn't. I know about the NumLock issue...
A return value of 1 does not mean that a BadRequest error occured. Xlib handles errors via an error handler, and the function will always return 1, if it returns at all.
Your code does not work because you have to do the XGrabKey on the root window (GetDefaultRootWindow(Gtk::gdk_display)). Here's a pure Xlib demo:
#include <X11/Xlib.h>
#include <X11/keysym.h>
#include <stdio.h>
int main() {
Display *d = XOpenDisplay(0);
Window root = DefaultRootWindow(d);
int keycode = XKeysymToKeycode(d, XK_BackSpace);
int rv = XGrabKey(d, keycode, AnyModifier, root, 1, GrabModeAsync, GrabModeAsync);
printf("XGrabKey returned %d\n", rv);
XEvent evt;
while(1) {
XNextEvent(d, &evt);
printf("Got event %d\n", evt.type);
}
}
To then capture the X11 events from GTK use gdk_window_add_filter on a NULL or on the root window and a GdkFilterFunc that processes the events associated with your global hotkey:
#include <X11/Xlib.h>
#include <X11/keysym.h>
#include <gtk/gtk.h>
#include <gdk/gdkx.h>
#include <stdio.h>
GdkFilterReturn filter(GdkXEvent *xevent, GdkEvent *event, gpointer data) {
XKeyEvent *ev = (XKeyEvent *)xevent;
if(ev->type == 2) {
printf("Backspace hit.\n");
}
return GDK_FILTER_CONTINUE;
}
int main(int argc, char *argv[]) {
gtk_init(&argc, &argv);
GdkScreen *scr = gdk_screen_get_default();
GdkWindow *groot = gdk_screen_get_root_window(scr);
gdk_window_set_events(groot, GDK_KEY_PRESS_MASK);
gdk_window_add_filter(groot, filter, NULL);
Display *d = gdk_x11_get_default_xdisplay();
Window root = GDK_WINDOW_XID(groot);
int keycode = XKeysymToKeycode(d, XK_BackSpace);
XGrabKey(d, keycode, AnyModifier, root, 1, GrabModeAsync, GrabModeAsync);
gtk_main();
}
As a side note, a modifier mask of 0 means that no modifiers must be enabled, even those that would not modify the meaning of a key. A grab on the letter "A" with a 0 modifier would not match NumLock + A. That's why I used AnyModifer.
I'm working on a university project that involves a lot of programming in C, especially with Portaudio & ALSA. At the moment i'm trying to make a callback function to pass audio through, standard input/output job. I was wondering if anybody could tell me how to print the floats from my inputBuffer to display in real time in the terminal? Here is the internal structure of my callback function so far.
Thanks very much for your help in advance!
#define SAMPLE_RATE (44100)
#define PA_SAMPLE_TYPE paFloat32
#define FRAMES_PER_BUFFER (64)
typedef float SAMPLE;
static int audio_callback( const void *inputBuffer, void *outputBuffer,
unsigned long framesPerBuffer,
const PaStreamCallbackTimeInfo* timeInfo,
PaStreamCallbackFlags statusFlags,
void *userData )
{
SAMPLE *out = (SAMPLE*)outputBuffer;
const SAMPLE *in = (const SAMPLE*)inputBuffer;
unsigned int i;
(void) timeInfo; /* Prevent unused variable warnings. */
(void) statusFlags;
(void) userData;
if( inputBuffer == NULL )
{
for( i=0; i<framesPerBuffer; i++ )
{
*out++ = *in++; /* left - clean */
*out++ = *in++; /* right - clean */
}
}
return paContinue;
}
EDIT: I have created a ticket for this which has data on an alternative to this way of doing things.
I have updated the code in an attempt to use MY_CXT's callback as gcxt was not storing across threads. However this segfaults at ENTER.
#include "EXTERN.h"
#include "perl.h"
#include "XSUB.h"
#ifndef aTHX_
#define aTHX_
#endif
#ifdef USE_THREADS
#define HAVE_TLS_CONTEXT
#endif
/* For windows */
#ifndef SDL_PERL_DEFINES_H
#define SDL_PERL_DEFINES_H
#ifdef HAVE_TLS_CONTEXT
PerlInterpreter *parent_perl = NULL;
extern PerlInterpreter *parent_perl;
#define GET_TLS_CONTEXT parent_perl = PERL_GET_CONTEXT;
#define ENTER_TLS_CONTEXT \
PerlInterpreter *current_perl = PERL_GET_CONTEXT; \
PERL_SET_CONTEXT(parent_perl); { \
PerlInterpreter *my_perl = parent_perl;
#define LEAVE_TLS_CONTEXT \
} PERL_SET_CONTEXT(current_perl);
#else
#define GET_TLS_CONTEXT /* TLS context not enabled */
#define ENTER_TLS_CONTEXT /* TLS context not enabled */
#define LEAVE_TLS_CONTEXT /* TLS context not enabled */
#endif
#endif
#include <SDL.h>
#define MY_CXT_KEY "SDL::Time::_guts" XS_VERSION
typedef struct {
void* data;
SV* callback;
Uint32 retval;
} my_cxt_t;
static my_cxt_t gcxt;
START_MY_CXT
static Uint32 add_timer_cb ( Uint32 interval, void* param )
{
ENTER_TLS_CONTEXT
dMY_CXT;
dSP;
int back;
ENTER; //SEGFAULTS RIGHT HERE!
SAVETMPS;
PUSHMARK(SP);
XPUSHs(sv_2mortal(newSViv(interval)));
PUTBACK;
if (0 != (back = call_sv(MY_CXT.callback,G_SCALAR))) {
SPAGAIN;
if (back != 1 ) Perl_croak (aTHX_ "Timer Callback failed!");
MY_CXT.retval = POPi;
} else {
Perl_croak(aTHX_ "Timer Callback failed!");
}
FREETMPS;
LEAVE;
LEAVE_TLS_CONTEXT
dMY_CXT;
return MY_CXT.retval;
}
MODULE = SDL::Time PACKAGE = SDL::Time PREFIX = time_
BOOT:
{
MY_CXT_INIT;
}
SDL_TimerID
time_add_timer ( interval, cmd )
Uint32 interval
void *cmd
PREINIT:
dMY_CXT;
CODE:
MY_CXT.callback=cmd;
gcxt = MY_CXT;
RETVAL = SDL_AddTimer(interval,add_timer_cb,(void *)cmd);
OUTPUT:
RETVAL
void
CLONE(...)
CODE:
MY_CXT_CLONE;
This segfaults as soon as I go into ENTER for the callback.
use SDL;
use SDL::Time;
SDL::init(SDL_INIT_TIMER);
my $time = 0;
SDL::Timer::add_timer(100, sub { $time++; return $_[0]} );
sleep(10);
print "Never Prints";
Output is
$
it should be
$ Never Prints
Quick comments:
Do not use Perl structs (SV, AV, HV, ...) outside of the context of a Perl interpreter object. I.e. do not use it as C-level static data. It will blow up in a threading context. Trust me, I've been there.
Check out the "Safely Storing Static Data in XS" section in the perlxs manpage.
Some of that stuff you're doing looks rather non-public from the point of view of the perlapi. I'm not quite certain, though.
$time needs to be a shared variable - otherwise perl works with separate copies of the variable.
My preferred way of handling this is storing the data in the PL_modglobal hash. It's automatically tied to the current interpreter.
We have found a solution to this using Perl interpreter threads and threads::shared. Please see these
Time.xs
Also here is an example of a script using this code.
TestTimer.pl
I am looking for a command line parser for Qt4.
I did a small google search, and found this: http://www.froglogic.com/pg?id=PublicationsFreeware&category=getopt however it lacks support for "--enable-foo" and "--disable-foo" switches. Besides that, it looks like a real winner.
EDIT:
It seems Frologic removed this. So the best options I see are using Boost (which is not API nor ABI stable) or forking the support for kdelibs. Yay...
QCoreApplication's constructors require (int &argc, char **argv) (and QApplication inherits from QCoreApplication). As the documentation states, it is highly recommended that
Since QApplication also deals with common command line arguments, it is usually a good idea to create it before any interpretation or modification of argv is done in the application itself.
And if you're letting Qt get the first pass at handling arguments anyways, it would also be a good idea to use QStringList QCoreApplication::arguments() instead of walking through argv; QApplication may remove some of the arguments that it has taken for its own use.
This doesn't lend itself to being very compatible with other argument-parsing libraries...
However, kdelibs does come with a nice argument parser, KCmdLineArgs. It is LGPL and can be used without KApplication if you really want (call KCmdLineArgs::init).
KCmdLineOptions options;
options.add("enable-foo", ki18n("enables foo"));
options.add("nodisable-foo", ki18n("disables foo"));
// double negatives are confusing, but this makes disable-foo enabled by default
KCmdLineArgs::addCmdLineOptions(options);
KApplication app;
KCmdLineArgs *args = KCmdLineArgs::parsedArgs();
if (args->isSet("enable-foo") && !args->isSet("disable-foo"))
cout << "foo enabled" << endl;
else
cout << "foo disabled" << endl;
Untested (who ever tests what they post on S.O.?).
Since Qt 5.2 you can finally find a solution in QtCore itself: I contributed QCommandLineParser there.
This is more or less the same answer as ephemient, but with a simple regexp to help parse the args. (This way could be useful if you only need a handful of args)
Run with this:
./QArgTest --pid=45 --enable-foo
And the code:
int main(int argc, char *argv[]) {
QApplication app(argc, argv, false);
qDebug() << "QApp arg test app";
QStringList args = app.arguments();
int pid = 0;
QRegExp rxArgPid("--pid=([0-9]{1,})");
QRegExp rxArgFooEna("--enable-foo");
QRegExp rxArgFooDis("--disable-foo");
for (int i = 1; i < args.size(); ++i) {
if (rxArgPid.indexIn(args.at(i)) != -1 ) {
pid = rxArgPid.cap(1).toInt();
qDebug() << i << ":" << args.at(i) << rxArgPid.cap(1) << pid;
}
else if (rxArgFooEna.indexIn(args.at(i)) != -1 ) {
qDebug() << i << ":" << args.at(i) << "Enable Foo";
}
else if (rxArgFooDis.indexIn(args.at(i)) != -1 ) {
qDebug() << i << ":" << args.at(i) << "Disable Foo";
}
else {
qDebug() << "Uknown arg:" << args.at(i);
}
}
return 0;
}
There is also QxtCommandOptions from http://www.libqxt.org/
That package does support --disable-foo and --enable-foo via opts.addSwitch("disable-foo", &foo_disabled); and opts.addSwitch("enable-foo", &foo_enabled);. You need handle checking both, and dealing with someone specifying both (oops).
What I don't understand is how this has anything to do with QT4...
Look at this: http://code.google.com/p/qtargparser/
A really simple method is to scan "key=value" args,
put them in a table say zz.map: QString -> QVariant,
and get their values with zz.map.value( key, default ).
An example:
#include "ztest.h"
Ztest zz;
int main( int argc, char* argv[] )
{
zz.eqargs( ++ argv ); // scan test=2 x=str ... to zz.map
QString xx = zz.map.value( "xx", "" );
if( Zint( Size, 10 )) // a #def -> zz.map.value( "Size", 10 )
...
ztest.h is < 1 page, below; same for Python ~ 10 lines.
(Everybody has his/her favorite options parser;
this one's about the simplest.
Worth repeating: however you specify options, echo them to output files --
"every scientist I know has trouble keeping track
of what parameters they used last time they ran a script".)
To make QPoints etc work one of course needs a QString -> QPoint parser.
Anyone know offhand why this doesn't work (in Qt 4.4.3) ?
QPoint pt(0,0);
QDataStream s( "QPoint(1,2)" );
s >> pt;
qDebug() << "pt:" << pt; // QPoint(1364225897,1853106225) ??
Added 25nov --
// ztest.h: scan args x=2 s=str ... to a key -> string table
// usage:
// Ztest ztest;
// int main( int argc, char* argv[] )
// {
// QApplication app( argc, argv );
// ztest.eqargs( ++ argv ); // scan leading args name=value ...
// int x = Zint( x, 10 ); // arg x= or default 10
// qreal ff = Zreal( ff, 3.14 );
// QString s = Zstr( s, "default" );
// care: int misspelled = Zint( misspellled ) -- you lose
//version: 2009-06-09 jun denis
#ifndef ztest_h
#define ztest_h
#include <QHash>
#include <QString>
#include <QVariant>
#include <QRegExp>
//------------------------------------------------------------------------------
class Ztest {
public:
QHash< QString, QVariant > map;
int test; // arg test=num, if( ztest.test )
Ztest() : test( 0 ) {}
QVariant val( const QString& key, const QVariant& default_ = 0 )
{
return map.value( key, default_ );
}
void setval( const QString& key, const QVariant& val )
{
map[key] = val;
if( key == "test" || key == "Test" )
test = val.toInt();
}
//------------------------------------------------------------------------------
// ztest.eqargs( ++ argv ) scans test=2 x=3 ... -> ztest table
void eqargs( char** argv )
{
char** argv0 = argv;
char *arg;
QRegExp re( "(\\w+)=(.*)" ); // name= anything, but not ./file=name
for( ; (arg = *argv) && re.exactMatch( arg ); argv ++ ){
setval( re.cap(1), re.cap(2) );
}
// change argv[0..] -> args after all name=values
while(( *argv0++ = *argv++) != 0 ) {}
}
};
extern Ztest ztest;
// macros: int x = Zint( x, 10 ): x= arg or default 10
#define Zstr( key, default ) ztest.val( #key, default ).toString()
#define Zint( key, default ) ztest.val( #key, default ).toInt()
#define Zreal( key, default ) ztest.val( #key, default ).toDouble()
#endif
It's 2013 and still no "1st party" arg parser. Anyways..if anyone finds themselves facing the same problem and would like to avoid the learning curves that come with cmd parser libs, here is a "quick & dirty" fix for you:-
QString QArgByKey(QString key, QChar sep = QChar('\0') ) //prototype usually in separate header
QString QArgByKey(QString key, QChar sep )
{
bool sepd=sep!=QChar('\0');
int pos=sepd?qApp->arguments().indexOf(QRegExp('^'+key+sep+"\\S*")):qApp->arguments().indexOf(QRegExp(key));
return pos==-1?QString::null:
(sepd?qApp->arguments().at(pos).split(sep).at(1):(++pos<qApp->arguments().size()?qApp->arguments().at(pos):QString::null));
}
Example:-
user#box:~$ ./myApp firstKey=Value1 --secondKey Value2 thirdKey=val3.1,val3.2,val3.3 --enable-foo
Usage:
QString param1 = QArgByKey("firstkey",'='); // Returns `Value1` from first pair
QString param2 = QArgByKey("--secondkey"); // Returns `Value2` from second pair
QString param3-1 = QArgByKey("thirdkey",'=').split(',').at(0); // Returns `val3.1`
bool fooEnabled = qApp->arguments().contains("--enable-foo"); //To check for `--enable-foo`
Params can be passed in any order
Edit: Updates to this snippet will be found here
Does it have to be Qt4 specific? If not, GNU Getopt is really nice, although licensing may be a problem if you are not doing open source software.
Also for some fancy options parsing you can try gperf.
IBM has a nice tutorial on it.
Another option I ran across while looking to do this, too:
http://code.google.com/p/qgetopts/
I haven't used it though.