2008-12-13 18:28:00 +00:00
|
|
|
/*
|
|
|
|
* This file is part of MPlayer.
|
|
|
|
*
|
|
|
|
* MPlayer is free software; you can redistribute it and/or modify
|
|
|
|
* it under the terms of the GNU General Public License as published by
|
|
|
|
* the Free Software Foundation; either version 2 of the License, or
|
|
|
|
* (at your option) any later version.
|
|
|
|
*
|
|
|
|
* MPlayer is distributed in the hope that it will be useful,
|
|
|
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
|
|
* GNU General Public License for more details.
|
|
|
|
*
|
|
|
|
* You should have received a copy of the GNU General Public License along
|
|
|
|
* with MPlayer; if not, write to the Free Software Foundation, Inc.,
|
|
|
|
* 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
|
|
|
|
*/
|
|
|
|
|
2005-10-25 21:23:45 +00:00
|
|
|
#include "config.h"
|
2002-01-30 12:46:03 +00:00
|
|
|
|
|
|
|
#include <stdlib.h>
|
|
|
|
#include <string.h>
|
|
|
|
#include <stdio.h>
|
2009-03-14 21:52:45 +00:00
|
|
|
#include <stdbool.h>
|
2002-01-30 12:46:03 +00:00
|
|
|
#include <unistd.h>
|
|
|
|
#include <errno.h>
|
|
|
|
#include <sys/types.h>
|
2002-10-22 14:32:17 +00:00
|
|
|
#include <sys/stat.h>
|
2002-02-04 14:48:32 +00:00
|
|
|
#include <sys/time.h>
|
2002-01-30 12:46:03 +00:00
|
|
|
#include <fcntl.h>
|
2013-11-28 18:28:38 +00:00
|
|
|
#include <pthread.h>
|
2011-07-17 01:47:50 +00:00
|
|
|
#include <assert.h>
|
2002-01-30 12:46:03 +00:00
|
|
|
|
2013-02-16 21:51:10 +00:00
|
|
|
#include <libavutil/avstring.h>
|
|
|
|
#include <libavutil/common.h>
|
|
|
|
|
2012-02-03 07:05:11 +00:00
|
|
|
#include "osdep/io.h"
|
2014-09-10 01:16:43 +00:00
|
|
|
#include "osdep/semaphore.h"
|
2014-09-09 21:44:38 +00:00
|
|
|
#include "misc/rendezvous.h"
|
2012-02-03 07:05:11 +00:00
|
|
|
|
2002-01-30 12:46:03 +00:00
|
|
|
#include "input.h"
|
2011-04-25 06:43:59 +00:00
|
|
|
#include "keycodes.h"
|
2013-12-26 16:10:35 +00:00
|
|
|
#include "cmd_list.h"
|
|
|
|
#include "cmd_parse.h"
|
2014-01-31 18:50:25 +00:00
|
|
|
#include "osdep/threads.h"
|
2005-10-25 21:23:45 +00:00
|
|
|
#include "osdep/timer.h"
|
2013-12-17 01:39:45 +00:00
|
|
|
#include "common/msg.h"
|
|
|
|
#include "common/global.h"
|
2014-09-27 14:25:29 +00:00
|
|
|
#include "options/m_config.h"
|
2013-12-17 01:02:25 +00:00
|
|
|
#include "options/m_option.h"
|
|
|
|
#include "options/path.h"
|
2008-04-30 04:15:52 +00:00
|
|
|
#include "talloc.h"
|
2013-12-17 01:02:25 +00:00
|
|
|
#include "options/options.h"
|
2014-08-29 10:09:04 +00:00
|
|
|
#include "misc/bstr.h"
|
2012-08-01 19:50:24 +00:00
|
|
|
#include "stream/stream.h"
|
2013-12-17 01:39:45 +00:00
|
|
|
#include "common/common.h"
|
2002-01-30 12:46:03 +00:00
|
|
|
|
2013-07-16 11:28:28 +00:00
|
|
|
#if HAVE_COCOA
|
2013-06-02 15:39:05 +00:00
|
|
|
#include "osdep/macosx_events.h"
|
|
|
|
#endif
|
|
|
|
|
2013-09-05 11:02:25 +00:00
|
|
|
#define input_lock(ictx) pthread_mutex_lock(&ictx->mutex)
|
|
|
|
#define input_unlock(ictx) pthread_mutex_unlock(&ictx->mutex)
|
2013-08-13 12:40:17 +00:00
|
|
|
|
2013-08-26 17:15:35 +00:00
|
|
|
#define MP_MAX_KEY_DOWN 4
|
|
|
|
|
2011-07-16 15:17:48 +00:00
|
|
|
struct cmd_bind {
|
2013-06-22 22:07:45 +00:00
|
|
|
int keys[MP_MAX_KEY_DOWN];
|
|
|
|
int num_keys;
|
2011-04-25 08:38:46 +00:00
|
|
|
char *cmd;
|
2012-10-13 19:09:42 +00:00
|
|
|
char *location; // filename/line number of definition
|
2013-06-22 21:26:44 +00:00
|
|
|
bool is_builtin;
|
2012-10-13 19:09:42 +00:00
|
|
|
struct cmd_bind_section *owner;
|
2011-07-16 15:17:48 +00:00
|
|
|
};
|
2008-04-30 06:53:12 +00:00
|
|
|
|
2013-12-26 16:10:35 +00:00
|
|
|
struct cmd_bind_section {
|
|
|
|
struct cmd_bind *binds;
|
|
|
|
int num_binds;
|
|
|
|
char *section;
|
|
|
|
struct mp_rect mouse_area; // set at runtime, if at all
|
|
|
|
bool mouse_area_set; // mouse_area is valid and should be tested
|
|
|
|
struct cmd_bind_section *next;
|
2010-12-19 10:12:20 +00:00
|
|
|
};
|
|
|
|
|
2014-09-10 01:16:43 +00:00
|
|
|
#define MP_MAX_SOURCES 10
|
2002-11-14 23:41:44 +00:00
|
|
|
|
2014-05-01 23:27:17 +00:00
|
|
|
#define MAX_ACTIVE_SECTIONS 50
|
input: handle mouse movement differently
Before this commit, mouse movement events emitted a special command
("set_mouse_pos"), which was specially handled in command.c. This was
once special-cased to the dvdnav and menu code, and did nothing after
libmenu and dvdnav were removed.
Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"),
which then can be bound to an arbitrary command. The mouse position is
now managed in input.c. A command which actually needs the mouse
position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos()
to query it. The former returns raw window-space coordinates, while the
latter returns coordinates transformed to OSD- space. (Both are the same
for most VOs, except vo_xv and vo_x11, which can't render OSD in
window-space. These require extra code for mapping mouse position.)
As of this commit, there is still nothing that uses mouse movement, so
MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the
mouse (much like MOUSE_BTN0).
Extend the concept of input sections. Allow multiple sections to be
active at once, and organize them as stack. Bindings from the top of
the stack are preferred to lower ones.
Each section has a mouse input section associated, inside which mouse
events are associated with the bindings. If the mouse pointer is
outside of a section's mouse area, mouse events will be dispatched to
an input section lower on the stack of active sections. This is intended
for scripting, which is to be added later. Two scripts could occupy
different areas of the screen without conflicting with each other. (If
it turns out that this mechanism is useless, we'll just remove it
again.)
2013-04-26 00:13:30 +00:00
|
|
|
|
|
|
|
struct active_section {
|
|
|
|
char *name;
|
|
|
|
int flags;
|
|
|
|
};
|
|
|
|
|
2011-07-17 01:47:50 +00:00
|
|
|
struct cmd_queue {
|
|
|
|
struct mp_cmd *first;
|
|
|
|
};
|
|
|
|
|
2008-04-30 04:15:52 +00:00
|
|
|
struct input_ctx {
|
2013-08-26 17:15:35 +00:00
|
|
|
pthread_mutex_t mutex;
|
2014-09-10 01:16:43 +00:00
|
|
|
sem_t wakeup;
|
2013-09-10 06:29:45 +00:00
|
|
|
struct mp_log *log;
|
2013-12-21 18:33:45 +00:00
|
|
|
struct mpv_global *global;
|
2014-07-27 19:53:29 +00:00
|
|
|
struct input_opts *opts;
|
2013-08-26 17:15:35 +00:00
|
|
|
|
2013-12-01 05:23:39 +00:00
|
|
|
bool using_alt_gr;
|
2013-07-27 19:26:00 +00:00
|
|
|
bool using_ar;
|
|
|
|
bool using_cocoa_media_keys;
|
2014-05-20 00:40:00 +00:00
|
|
|
bool win_drag;
|
2013-07-27 19:26:00 +00:00
|
|
|
|
2008-04-30 04:15:52 +00:00
|
|
|
// Autorepeat stuff
|
|
|
|
short ar_state;
|
2013-05-25 16:31:06 +00:00
|
|
|
int64_t last_ar;
|
Add initial Lua scripting support
This is preliminary. There are still tons of issues, and any aspect
of scripting may change in the future. I decided to merge this
(preliminary) work now because it makes it easier to develop it, not
because it's done. lua.rst is clear enough about it (plus some
sarcasm).
This requires linking to Lua. Lua has no official pkg-config file, but
there are distribution specific .pc files, all with different names.
Adding a non-pkg-config based configure test was considered, but we'd
rather not.
One major complication is that libquvi links against Lua too, and if
the Lua version is different from mpv's, you will get a crash as soon
as libquvi uses Lua. (libquvi by design always runs when a file is
opened.) I would consider this the problem of distros and whoever
builds mpv, but to make things easier for users, we add a terrible
runtime test to the configure script, which probes whether libquvi
will crash. This is disabled when cross-compiling, but in that case
we hope the user knows what he is doing.
2013-09-25 22:41:14 +00:00
|
|
|
|
2014-04-18 14:27:02 +00:00
|
|
|
// history of key downs - the newest is in position 0
|
|
|
|
int key_history[MP_MAX_KEY_DOWN];
|
2014-04-18 15:45:41 +00:00
|
|
|
// key code of the last key that triggered MP_KEY_STATE_DOWN
|
|
|
|
int last_key_down;
|
2014-04-18 15:16:33 +00:00
|
|
|
int64_t last_key_down_time;
|
input: handle mouse movement differently
Before this commit, mouse movement events emitted a special command
("set_mouse_pos"), which was specially handled in command.c. This was
once special-cased to the dvdnav and menu code, and did nothing after
libmenu and dvdnav were removed.
Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"),
which then can be bound to an arbitrary command. The mouse position is
now managed in input.c. A command which actually needs the mouse
position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos()
to query it. The former returns raw window-space coordinates, while the
latter returns coordinates transformed to OSD- space. (Both are the same
for most VOs, except vo_xv and vo_x11, which can't render OSD in
window-space. These require extra code for mapping mouse position.)
As of this commit, there is still nothing that uses mouse movement, so
MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the
mouse (much like MOUSE_BTN0).
Extend the concept of input sections. Allow multiple sections to be
active at once, and organize them as stack. Bindings from the top of
the stack are preferred to lower ones.
Each section has a mouse input section associated, inside which mouse
events are associated with the bindings. If the mouse pointer is
outside of a section's mouse area, mouse events will be dispatched to
an input section lower on the stack of active sections. This is intended
for scripting, which is to be added later. Two scripts could occupy
different areas of the screen without conflicting with each other. (If
it turns out that this mechanism is useless, we'll just remove it
again.)
2013-04-26 00:13:30 +00:00
|
|
|
struct mp_cmd *current_down_cmd;
|
|
|
|
|
2013-07-02 12:00:24 +00:00
|
|
|
int last_doubleclick_key_down;
|
|
|
|
double last_doubleclick_time;
|
|
|
|
|
input: handle mouse movement differently
Before this commit, mouse movement events emitted a special command
("set_mouse_pos"), which was specially handled in command.c. This was
once special-cased to the dvdnav and menu code, and did nothing after
libmenu and dvdnav were removed.
Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"),
which then can be bound to an arbitrary command. The mouse position is
now managed in input.c. A command which actually needs the mouse
position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos()
to query it. The former returns raw window-space coordinates, while the
latter returns coordinates transformed to OSD- space. (Both are the same
for most VOs, except vo_xv and vo_x11, which can't render OSD in
window-space. These require extra code for mapping mouse position.)
As of this commit, there is still nothing that uses mouse movement, so
MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the
mouse (much like MOUSE_BTN0).
Extend the concept of input sections. Allow multiple sections to be
active at once, and organize them as stack. Bindings from the top of
the stack are preferred to lower ones.
Each section has a mouse input section associated, inside which mouse
events are associated with the bindings. If the mouse pointer is
outside of a section's mouse area, mouse events will be dispatched to
an input section lower on the stack of active sections. This is intended
for scripting, which is to be added later. Two scripts could occupy
different areas of the screen without conflicting with each other. (If
it turns out that this mechanism is useless, we'll just remove it
again.)
2013-04-26 00:13:30 +00:00
|
|
|
// Mouse position on the consumer side (as command.c sees it)
|
|
|
|
int mouse_x, mouse_y;
|
2013-06-19 16:19:45 +00:00
|
|
|
char *mouse_section; // last section to receive mouse event
|
input: handle mouse movement differently
Before this commit, mouse movement events emitted a special command
("set_mouse_pos"), which was specially handled in command.c. This was
once special-cased to the dvdnav and menu code, and did nothing after
libmenu and dvdnav were removed.
Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"),
which then can be bound to an arbitrary command. The mouse position is
now managed in input.c. A command which actually needs the mouse
position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos()
to query it. The former returns raw window-space coordinates, while the
latter returns coordinates transformed to OSD- space. (Both are the same
for most VOs, except vo_xv and vo_x11, which can't render OSD in
window-space. These require extra code for mapping mouse position.)
As of this commit, there is still nothing that uses mouse movement, so
MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the
mouse (much like MOUSE_BTN0).
Extend the concept of input sections. Allow multiple sections to be
active at once, and organize them as stack. Bindings from the top of
the stack are preferred to lower ones.
Each section has a mouse input section associated, inside which mouse
events are associated with the bindings. If the mouse pointer is
outside of a section's mouse area, mouse events will be dispatched to
an input section lower on the stack of active sections. This is intended
for scripting, which is to be added later. Two scripts could occupy
different areas of the screen without conflicting with each other. (If
it turns out that this mechanism is useless, we'll just remove it
again.)
2013-04-26 00:13:30 +00:00
|
|
|
|
|
|
|
// Mouse position on the producer side (as the VO sees it)
|
|
|
|
// Unlike mouse_x/y, this can be used to resolve mouse click bindings.
|
|
|
|
int mouse_vo_x, mouse_vo_y;
|
2008-04-30 07:21:50 +00:00
|
|
|
|
2014-07-27 19:33:11 +00:00
|
|
|
bool mouse_mangle, mouse_src_mangle;
|
|
|
|
struct mp_rect mouse_src, mouse_dst;
|
|
|
|
|
2008-04-30 05:56:45 +00:00
|
|
|
// List of command binding sections
|
2011-07-16 15:17:48 +00:00
|
|
|
struct cmd_bind_section *cmd_bind_sections;
|
input: handle mouse movement differently
Before this commit, mouse movement events emitted a special command
("set_mouse_pos"), which was specially handled in command.c. This was
once special-cased to the dvdnav and menu code, and did nothing after
libmenu and dvdnav were removed.
Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"),
which then can be bound to an arbitrary command. The mouse position is
now managed in input.c. A command which actually needs the mouse
position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos()
to query it. The former returns raw window-space coordinates, while the
latter returns coordinates transformed to OSD- space. (Both are the same
for most VOs, except vo_xv and vo_x11, which can't render OSD in
window-space. These require extra code for mapping mouse position.)
As of this commit, there is still nothing that uses mouse movement, so
MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the
mouse (much like MOUSE_BTN0).
Extend the concept of input sections. Allow multiple sections to be
active at once, and organize them as stack. Bindings from the top of
the stack are preferred to lower ones.
Each section has a mouse input section associated, inside which mouse
events are associated with the bindings. If the mouse pointer is
outside of a section's mouse area, mouse events will be dispatched to
an input section lower on the stack of active sections. This is intended
for scripting, which is to be added later. Two scripts could occupy
different areas of the screen without conflicting with each other. (If
it turns out that this mechanism is useless, we'll just remove it
again.)
2013-04-26 00:13:30 +00:00
|
|
|
|
|
|
|
// List currently active command sections
|
|
|
|
struct active_section active_sections[MAX_ACTIVE_SECTIONS];
|
|
|
|
int num_active_sections;
|
2008-04-30 08:06:55 +00:00
|
|
|
|
2013-05-25 16:31:06 +00:00
|
|
|
unsigned int mouse_event_counter;
|
2013-05-16 21:17:46 +00:00
|
|
|
|
2014-09-10 01:16:43 +00:00
|
|
|
struct mp_input_src *sources[MP_MAX_SOURCES];
|
input: redo how --input-file is handled
Abandon the "old" infrastructure for --input-file (mp_input_add_fd(),
select() loop, non-blocking reads). Replace it with something that
starts a reader thread, using blocking input.
This is for the sake of Windows. Windows is a truly insane operating
system, and there's not even a way to read a pipe in a non-blocking
way, or to wait for new input in an interruptible way (like with
poll()). And unfortunately, some want to use pipe to send input to
mpv. There are probably (slightly) better IPC mechanisms available
on Windows, but for the sake of platform uniformity, make this work
again for now.
On Vista+, CancelIoEx() could probably be used. But there's no way on
XP. Also, that function doesn't work on wine, making development
harder. We could forcibly terminate the thread, which might work, but
is unsafe. So what we do is starting a thread, and if we don't want
the pipe input anymore, we just abandon the thread. The thread might
remain blocked forever, but if we exit the process, the kernel will
forcibly kill it. On Unix, just use poll() to handle this.
Unfortunately the code is pretty crappy, but it's ok, because it's late
and I wanted to stop working on this an hour ago.
Tested on wine; might not work on a real Windows.
2014-08-24 21:50:43 +00:00
|
|
|
int num_sources;
|
2008-04-30 10:14:03 +00:00
|
|
|
|
2013-09-01 23:09:52 +00:00
|
|
|
struct cmd_queue cmd_queue;
|
stream: redo playback abort handling
This mechanism originates from MPlayer's way of dealing with blocking
network, but it's still useful. On opening and closing, mpv waits for
network synchronously, and also some obscure commands and use-cases can
lead to such blocking. In these situations, the stream is asynchronously
forced to stop by "interrupting" it.
The old design interrupting I/O was a bit broken: polling with a
callback, instead of actively interrupting it. Change the direction of
this. There is no callback anymore, and the player calls
mp_cancel_trigger() to force the stream to return.
libavformat (via stream_lavf.c) has the old broken design, and fixing it
would require fixing libavformat, which won't happen so quickly. So we
have to keep that part. But everything above the stream layer is
prepared for a better design, and more sophisticated methods than
mp_cancel_test() could be easily introduced.
There's still one problem: commands are still run in the central
playback loop, which we assume can block on I/O in the worst case.
That's not a problem yet, because we simply mark some commands as being
able to stop playback of the current file ("quit" etc.), so input.c
could abort playback as soon as such a command is queued. But there are
also commands abort playback only conditionally, and the logic for that
is in the playback core and thus "unreachable". For example,
"playlist_next" aborts playback only if there's a next file. We don't
want it to always abort playback.
As a quite ugly hack, abort playback only if at least 2 abort commands
are queued - this pretty much happens only if the core is frozen and
doesn't react to input.
2014-09-13 12:23:08 +00:00
|
|
|
|
|
|
|
struct mp_cancel *cancel;
|
2008-04-30 04:15:52 +00:00
|
|
|
};
|
|
|
|
|
2013-12-26 16:41:28 +00:00
|
|
|
static int parse_config(struct input_ctx *ictx, bool builtin, bstr data,
|
|
|
|
const char *location, const char *restrict_section);
|
input: redo how --input-file is handled
Abandon the "old" infrastructure for --input-file (mp_input_add_fd(),
select() loop, non-blocking reads). Replace it with something that
starts a reader thread, using blocking input.
This is for the sake of Windows. Windows is a truly insane operating
system, and there's not even a way to read a pipe in a non-blocking
way, or to wait for new input in an interruptible way (like with
poll()). And unfortunately, some want to use pipe to send input to
mpv. There are probably (slightly) better IPC mechanisms available
on Windows, but for the sake of platform uniformity, make this work
again for now.
On Vista+, CancelIoEx() could probably be used. But there's no way on
XP. Also, that function doesn't work on wine, making development
harder. We could forcibly terminate the thread, which might work, but
is unsafe. So what we do is starting a thread, and if we don't want
the pipe input anymore, we just abandon the thread. The thread might
remain blocked forever, but if we exit the process, the kernel will
forcibly kill it. On Unix, just use poll() to handle this.
Unfortunately the code is pretty crappy, but it's ok, because it's late
and I wanted to stop working on this an hour ago.
Tested on wine; might not work on a real Windows.
2014-08-24 21:50:43 +00:00
|
|
|
static void close_input_sources(struct input_ctx *ictx);
|
2013-12-26 16:41:28 +00:00
|
|
|
|
2014-06-10 23:54:03 +00:00
|
|
|
#define OPT_BASE_STRUCT struct input_opts
|
|
|
|
struct input_opts {
|
|
|
|
char *config_file;
|
|
|
|
int doubleclick_time;
|
2014-09-27 14:25:29 +00:00
|
|
|
// Maximum number of queued commands from keypresses (limit to avoid
|
|
|
|
// repeated slow commands piling up)
|
2014-06-10 23:54:03 +00:00
|
|
|
int key_fifo_size;
|
2014-09-27 14:25:29 +00:00
|
|
|
// Autorepeat config (be aware of mp_input_set_repeat_info())
|
2014-06-10 23:54:03 +00:00
|
|
|
int ar_delay;
|
|
|
|
int ar_rate;
|
|
|
|
char *js_dev;
|
|
|
|
int use_joystick;
|
|
|
|
int use_lirc;
|
|
|
|
char *lirc_configfile;
|
|
|
|
int use_lircc;
|
|
|
|
int use_alt_gr;
|
|
|
|
int use_appleremote;
|
|
|
|
int use_media_keys;
|
2014-10-09 19:23:54 +00:00
|
|
|
int use_app_events;
|
2014-06-10 23:54:03 +00:00
|
|
|
int default_bindings;
|
2014-07-27 19:53:29 +00:00
|
|
|
int enable_mouse_movements;
|
2014-10-09 16:28:37 +00:00
|
|
|
int vo_key_input;
|
2014-06-10 23:54:03 +00:00
|
|
|
int test;
|
2002-02-11 11:42:08 +00:00
|
|
|
};
|
|
|
|
|
2014-06-10 23:54:03 +00:00
|
|
|
const struct m_sub_options input_config = {
|
|
|
|
.opts = (const m_option_t[]) {
|
|
|
|
OPT_STRING("conf", config_file, CONF_GLOBAL),
|
|
|
|
OPT_INT("ar-delay", ar_delay, CONF_GLOBAL),
|
|
|
|
OPT_INT("ar-rate", ar_rate, CONF_GLOBAL),
|
|
|
|
OPT_PRINT("keylist", mp_print_key_list),
|
|
|
|
OPT_PRINT("cmdlist", mp_print_cmd_list),
|
|
|
|
OPT_STRING("js-dev", js_dev, CONF_GLOBAL),
|
|
|
|
OPT_FLAG("default-bindings", default_bindings, CONF_GLOBAL),
|
|
|
|
OPT_FLAG("test", test, CONF_GLOBAL),
|
|
|
|
OPT_INTRANGE("doubleclick-time", doubleclick_time, 0, 0, 1000),
|
|
|
|
OPT_FLAG("joystick", use_joystick, CONF_GLOBAL),
|
|
|
|
OPT_FLAG("lirc", use_lirc, CONF_GLOBAL),
|
|
|
|
OPT_FLAG("right-alt-gr", use_alt_gr, CONF_GLOBAL),
|
|
|
|
OPT_INTRANGE("key-fifo-size", key_fifo_size, CONF_GLOBAL, 2, 65000),
|
2014-07-27 19:53:29 +00:00
|
|
|
OPT_FLAG("cursor", enable_mouse_movements, CONF_GLOBAL),
|
2014-10-09 16:28:37 +00:00
|
|
|
OPT_FLAG("vo-keyboard", vo_key_input, CONF_GLOBAL),
|
|
|
|
OPT_FLAG("x11-keyboard", vo_key_input, CONF_GLOBAL), // old alias
|
2014-09-09 22:51:36 +00:00
|
|
|
#if HAVE_LIRC
|
2014-06-10 23:54:03 +00:00
|
|
|
OPT_STRING("lirc-conf", lirc_configfile, CONF_GLOBAL),
|
2014-09-09 22:51:36 +00:00
|
|
|
#endif
|
|
|
|
#if HAVE_COCOA
|
2014-06-10 23:54:03 +00:00
|
|
|
OPT_FLAG("appleremote", use_appleremote, CONF_GLOBAL),
|
|
|
|
OPT_FLAG("media-keys", use_media_keys, CONF_GLOBAL),
|
2014-10-09 19:23:54 +00:00
|
|
|
OPT_FLAG("app-events", use_app_events, CONF_GLOBAL),
|
2014-09-09 22:51:36 +00:00
|
|
|
#endif
|
2014-06-10 23:54:03 +00:00
|
|
|
{0}
|
|
|
|
},
|
|
|
|
.size = sizeof(struct input_opts),
|
|
|
|
.defaults = &(const struct input_opts){
|
|
|
|
.key_fifo_size = 7,
|
|
|
|
.doubleclick_time = 300,
|
|
|
|
.ar_delay = 200,
|
|
|
|
.ar_rate = 40,
|
|
|
|
.use_lirc = 1,
|
|
|
|
.use_alt_gr = 1,
|
2014-07-27 19:53:29 +00:00
|
|
|
.enable_mouse_movements = 1,
|
2013-07-16 11:28:28 +00:00
|
|
|
#if HAVE_COCOA
|
2014-06-10 23:54:03 +00:00
|
|
|
.use_appleremote = 1,
|
|
|
|
.use_media_keys = 1,
|
2014-10-09 19:23:54 +00:00
|
|
|
.use_app_events = 1,
|
2013-06-03 21:02:16 +00:00
|
|
|
#endif
|
2014-06-10 23:54:03 +00:00
|
|
|
.default_bindings = 1,
|
2014-10-09 16:28:37 +00:00
|
|
|
.vo_key_input = 1,
|
2014-06-10 23:54:03 +00:00
|
|
|
},
|
2002-02-11 11:42:08 +00:00
|
|
|
};
|
2002-02-08 18:10:56 +00:00
|
|
|
|
2012-08-01 20:52:28 +00:00
|
|
|
static const char builtin_input_conf[] =
|
2013-12-17 00:23:09 +00:00
|
|
|
#include "input/input_conf.h"
|
2012-08-01 20:52:28 +00:00
|
|
|
;
|
|
|
|
|
input: handle mouse movement differently
Before this commit, mouse movement events emitted a special command
("set_mouse_pos"), which was specially handled in command.c. This was
once special-cased to the dvdnav and menu code, and did nothing after
libmenu and dvdnav were removed.
Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"),
which then can be bound to an arbitrary command. The mouse position is
now managed in input.c. A command which actually needs the mouse
position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos()
to query it. The former returns raw window-space coordinates, while the
latter returns coordinates transformed to OSD- space. (Both are the same
for most VOs, except vo_xv and vo_x11, which can't render OSD in
window-space. These require extra code for mapping mouse position.)
As of this commit, there is still nothing that uses mouse movement, so
MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the
mouse (much like MOUSE_BTN0).
Extend the concept of input sections. Allow multiple sections to be
active at once, and organize them as stack. Bindings from the top of
the stack are preferred to lower ones.
Each section has a mouse input section associated, inside which mouse
events are associated with the bindings. If the mouse pointer is
outside of a section's mouse area, mouse events will be dispatched to
an input section lower on the stack of active sections. This is intended
for scripting, which is to be added later. Two scripts could occupy
different areas of the screen without conflicting with each other. (If
it turns out that this mechanism is useless, we'll just remove it
again.)
2013-04-26 00:13:30 +00:00
|
|
|
static bool test_rect(struct mp_rect *rc, int x, int y)
|
|
|
|
{
|
|
|
|
return x >= rc->x0 && y >= rc->y0 && x < rc->x1 && y < rc->y1;
|
|
|
|
}
|
|
|
|
|
2012-01-18 01:35:40 +00:00
|
|
|
static int queue_count_cmds(struct cmd_queue *queue)
|
2011-07-17 01:47:50 +00:00
|
|
|
{
|
2012-01-18 01:35:40 +00:00
|
|
|
int res = 0;
|
|
|
|
for (struct mp_cmd *cmd = queue->first; cmd; cmd = cmd->queue_next)
|
|
|
|
res++;
|
|
|
|
return res;
|
|
|
|
}
|
|
|
|
|
2011-10-24 00:28:56 +00:00
|
|
|
static void queue_remove(struct cmd_queue *queue, struct mp_cmd *cmd)
|
|
|
|
{
|
|
|
|
struct mp_cmd **p_prev = &queue->first;
|
|
|
|
while (*p_prev != cmd) {
|
|
|
|
p_prev = &(*p_prev)->queue_next;
|
|
|
|
}
|
|
|
|
// if this fails, cmd was not in the queue
|
|
|
|
assert(*p_prev == cmd);
|
|
|
|
*p_prev = cmd->queue_next;
|
|
|
|
}
|
|
|
|
|
2014-09-12 23:08:35 +00:00
|
|
|
static struct mp_cmd *queue_remove_head(struct cmd_queue *queue)
|
|
|
|
{
|
|
|
|
struct mp_cmd *ret = queue->first;
|
|
|
|
if (ret)
|
|
|
|
queue_remove(queue, ret);
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2013-09-02 00:29:20 +00:00
|
|
|
static void queue_add_tail(struct cmd_queue *queue, struct mp_cmd *cmd)
|
|
|
|
{
|
|
|
|
struct mp_cmd **p_prev = &queue->first;
|
|
|
|
while (*p_prev)
|
|
|
|
p_prev = &(*p_prev)->queue_next;
|
|
|
|
*p_prev = cmd;
|
|
|
|
cmd->queue_next = NULL;
|
2013-08-13 12:40:17 +00:00
|
|
|
}
|
|
|
|
|
2013-09-02 00:40:05 +00:00
|
|
|
static struct mp_cmd *queue_peek_tail(struct cmd_queue *queue)
|
|
|
|
{
|
|
|
|
struct mp_cmd *cur = queue->first;
|
|
|
|
while (cur && cur->queue_next)
|
|
|
|
cur = cur->queue_next;
|
|
|
|
return cur;
|
|
|
|
}
|
|
|
|
|
2013-09-27 13:30:59 +00:00
|
|
|
static void append_bind_info(struct input_ctx *ictx, char **pmsg,
|
|
|
|
struct cmd_bind *bind)
|
2012-10-13 19:10:20 +00:00
|
|
|
{
|
|
|
|
char *msg = *pmsg;
|
2013-09-27 13:30:59 +00:00
|
|
|
struct mp_cmd *cmd = mp_input_parse_cmd(ictx, bstr0(bind->cmd),
|
|
|
|
bind->location);
|
2012-10-13 19:10:20 +00:00
|
|
|
bstr stripped = cmd ? cmd->original : bstr0(bind->cmd);
|
|
|
|
msg = talloc_asprintf_append(msg, " '%.*s'", BSTR_P(stripped));
|
|
|
|
if (!cmd)
|
|
|
|
msg = talloc_asprintf_append(msg, " (invalid)");
|
|
|
|
if (strcmp(bind->owner->section, "default") != 0)
|
|
|
|
msg = talloc_asprintf_append(msg, " in section {%s}",
|
|
|
|
bind->owner->section);
|
input: handle mouse movement differently
Before this commit, mouse movement events emitted a special command
("set_mouse_pos"), which was specially handled in command.c. This was
once special-cased to the dvdnav and menu code, and did nothing after
libmenu and dvdnav were removed.
Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"),
which then can be bound to an arbitrary command. The mouse position is
now managed in input.c. A command which actually needs the mouse
position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos()
to query it. The former returns raw window-space coordinates, while the
latter returns coordinates transformed to OSD- space. (Both are the same
for most VOs, except vo_xv and vo_x11, which can't render OSD in
window-space. These require extra code for mapping mouse position.)
As of this commit, there is still nothing that uses mouse movement, so
MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the
mouse (much like MOUSE_BTN0).
Extend the concept of input sections. Allow multiple sections to be
active at once, and organize them as stack. Bindings from the top of
the stack are preferred to lower ones.
Each section has a mouse input section associated, inside which mouse
events are associated with the bindings. If the mouse pointer is
outside of a section's mouse area, mouse events will be dispatched to
an input section lower on the stack of active sections. This is intended
for scripting, which is to be added later. Two scripts could occupy
different areas of the screen without conflicting with each other. (If
it turns out that this mechanism is useless, we'll just remove it
again.)
2013-04-26 00:13:30 +00:00
|
|
|
msg = talloc_asprintf_append(msg, " in %s", bind->location);
|
2013-06-22 21:26:44 +00:00
|
|
|
if (bind->is_builtin)
|
input: handle mouse movement differently
Before this commit, mouse movement events emitted a special command
("set_mouse_pos"), which was specially handled in command.c. This was
once special-cased to the dvdnav and menu code, and did nothing after
libmenu and dvdnav were removed.
Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"),
which then can be bound to an arbitrary command. The mouse position is
now managed in input.c. A command which actually needs the mouse
position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos()
to query it. The former returns raw window-space coordinates, while the
latter returns coordinates transformed to OSD- space. (Both are the same
for most VOs, except vo_xv and vo_x11, which can't render OSD in
window-space. These require extra code for mapping mouse position.)
As of this commit, there is still nothing that uses mouse movement, so
MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the
mouse (much like MOUSE_BTN0).
Extend the concept of input sections. Allow multiple sections to be
active at once, and organize them as stack. Bindings from the top of
the stack are preferred to lower ones.
Each section has a mouse input section associated, inside which mouse
events are associated with the bindings. If the mouse pointer is
outside of a section's mouse area, mouse events will be dispatched to
an input section lower on the stack of active sections. This is intended
for scripting, which is to be added later. Two scripts could occupy
different areas of the screen without conflicting with each other. (If
it turns out that this mechanism is useless, we'll just remove it
again.)
2013-04-26 00:13:30 +00:00
|
|
|
msg = talloc_asprintf_append(msg, " (default)");
|
|
|
|
talloc_free(cmd);
|
2012-10-13 19:10:20 +00:00
|
|
|
*pmsg = msg;
|
|
|
|
}
|
|
|
|
|
2014-04-18 14:27:02 +00:00
|
|
|
static mp_cmd_t *handle_test(struct input_ctx *ictx, int code)
|
2012-10-13 19:10:20 +00:00
|
|
|
{
|
2014-04-18 14:37:27 +00:00
|
|
|
if (code == MP_KEY_CLOSE_WIN) {
|
|
|
|
MP_WARN(ictx,
|
|
|
|
"CLOSE_WIN was received. This pseudo key can be remapped too,\n"
|
|
|
|
"but --input-test will always quit when receiving it.\n");
|
|
|
|
const char *args[] = {"quit", NULL};
|
2014-10-10 20:58:28 +00:00
|
|
|
mp_cmd_t *res = mp_input_parse_cmd_strv(ictx->log, args);
|
2014-04-18 14:37:27 +00:00
|
|
|
return res;
|
|
|
|
}
|
|
|
|
|
2014-04-18 14:27:02 +00:00
|
|
|
char *key_buf = mp_input_get_key_combo_name(&code, 1);
|
2014-02-19 14:39:59 +00:00
|
|
|
char *msg = talloc_asprintf(NULL, "Key %s is bound to:\n", key_buf);
|
2012-10-13 19:10:20 +00:00
|
|
|
talloc_free(key_buf);
|
|
|
|
|
|
|
|
int count = 0;
|
|
|
|
for (struct cmd_bind_section *bs = ictx->cmd_bind_sections;
|
|
|
|
bs; bs = bs->next)
|
|
|
|
{
|
2013-06-22 22:40:35 +00:00
|
|
|
for (int i = 0; i < bs->num_binds; i++) {
|
2014-04-18 14:27:02 +00:00
|
|
|
if (bs->binds[i].num_keys && bs->binds[i].keys[0] == code) {
|
2012-10-13 19:10:20 +00:00
|
|
|
count++;
|
2014-08-24 17:17:33 +00:00
|
|
|
if (count > 1)
|
|
|
|
msg = talloc_asprintf_append(msg, "\n");
|
2012-10-13 19:10:20 +00:00
|
|
|
msg = talloc_asprintf_append(msg, "%d. ", count);
|
2013-09-27 13:30:59 +00:00
|
|
|
append_bind_info(ictx, &msg, &bs->binds[i]);
|
2012-10-13 19:10:20 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
if (!count)
|
|
|
|
msg = talloc_asprintf_append(msg, "(nothing)");
|
|
|
|
|
2014-08-24 17:17:33 +00:00
|
|
|
MP_INFO(ictx, "%s\n", msg);
|
2014-02-19 14:39:59 +00:00
|
|
|
const char *args[] = {"show_text", msg, NULL};
|
2014-10-10 20:58:28 +00:00
|
|
|
mp_cmd_t *res = mp_input_parse_cmd_strv(ictx->log, args);
|
2014-02-19 14:39:59 +00:00
|
|
|
talloc_free(msg);
|
2012-10-13 19:10:20 +00:00
|
|
|
return res;
|
|
|
|
}
|
|
|
|
|
2011-07-16 15:17:48 +00:00
|
|
|
static struct cmd_bind_section *get_bind_section(struct input_ctx *ictx,
|
2013-06-22 21:26:44 +00:00
|
|
|
bstr section)
|
2008-04-30 05:56:45 +00:00
|
|
|
{
|
2011-07-16 15:17:48 +00:00
|
|
|
struct cmd_bind_section *bind_section = ictx->cmd_bind_sections;
|
2011-04-25 08:38:46 +00:00
|
|
|
|
2012-09-22 13:17:15 +00:00
|
|
|
if (section.len == 0)
|
|
|
|
section = bstr0("default");
|
2011-04-25 08:38:46 +00:00
|
|
|
while (bind_section) {
|
2013-06-22 21:26:44 +00:00
|
|
|
if (bstrcmp0(section, bind_section->section) == 0)
|
2011-04-25 08:38:46 +00:00
|
|
|
return bind_section;
|
|
|
|
if (bind_section->next == NULL)
|
|
|
|
break;
|
|
|
|
bind_section = bind_section->next;
|
|
|
|
}
|
|
|
|
if (bind_section) {
|
|
|
|
bind_section->next = talloc_ptrtype(ictx, bind_section->next);
|
|
|
|
bind_section = bind_section->next;
|
|
|
|
} else {
|
|
|
|
ictx->cmd_bind_sections = talloc_ptrtype(ictx, ictx->cmd_bind_sections);
|
|
|
|
bind_section = ictx->cmd_bind_sections;
|
|
|
|
}
|
input: handle mouse movement differently
Before this commit, mouse movement events emitted a special command
("set_mouse_pos"), which was specially handled in command.c. This was
once special-cased to the dvdnav and menu code, and did nothing after
libmenu and dvdnav were removed.
Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"),
which then can be bound to an arbitrary command. The mouse position is
now managed in input.c. A command which actually needs the mouse
position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos()
to query it. The former returns raw window-space coordinates, while the
latter returns coordinates transformed to OSD- space. (Both are the same
for most VOs, except vo_xv and vo_x11, which can't render OSD in
window-space. These require extra code for mapping mouse position.)
As of this commit, there is still nothing that uses mouse movement, so
MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the
mouse (much like MOUSE_BTN0).
Extend the concept of input sections. Allow multiple sections to be
active at once, and organize them as stack. Bindings from the top of
the stack are preferred to lower ones.
Each section has a mouse input section associated, inside which mouse
events are associated with the bindings. If the mouse pointer is
outside of a section's mouse area, mouse events will be dispatched to
an input section lower on the stack of active sections. This is intended
for scripting, which is to be added later. Two scripts could occupy
different areas of the screen without conflicting with each other. (If
it turns out that this mechanism is useless, we'll just remove it
again.)
2013-04-26 00:13:30 +00:00
|
|
|
*bind_section = (struct cmd_bind_section) {
|
|
|
|
.section = bstrdup0(bind_section, section),
|
2014-11-23 08:10:51 +00:00
|
|
|
.mouse_area = {INT_MIN, INT_MIN, INT_MAX, INT_MAX},
|
|
|
|
.mouse_area_set = true,
|
input: handle mouse movement differently
Before this commit, mouse movement events emitted a special command
("set_mouse_pos"), which was specially handled in command.c. This was
once special-cased to the dvdnav and menu code, and did nothing after
libmenu and dvdnav were removed.
Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"),
which then can be bound to an arbitrary command. The mouse position is
now managed in input.c. A command which actually needs the mouse
position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos()
to query it. The former returns raw window-space coordinates, while the
latter returns coordinates transformed to OSD- space. (Both are the same
for most VOs, except vo_xv and vo_x11, which can't render OSD in
window-space. These require extra code for mapping mouse position.)
As of this commit, there is still nothing that uses mouse movement, so
MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the
mouse (much like MOUSE_BTN0).
Extend the concept of input sections. Allow multiple sections to be
active at once, and organize them as stack. Bindings from the top of
the stack are preferred to lower ones.
Each section has a mouse input section associated, inside which mouse
events are associated with the bindings. If the mouse pointer is
outside of a section's mouse area, mouse events will be dispatched to
an input section lower on the stack of active sections. This is intended
for scripting, which is to be added later. Two scripts could occupy
different areas of the screen without conflicting with each other. (If
it turns out that this mechanism is useless, we'll just remove it
again.)
2013-04-26 00:13:30 +00:00
|
|
|
};
|
2011-04-25 08:38:46 +00:00
|
|
|
return bind_section;
|
2007-06-07 18:06:53 +00:00
|
|
|
}
|
|
|
|
|
2014-04-18 14:27:02 +00:00
|
|
|
static void key_buf_add(int *buf, int code)
|
|
|
|
{
|
|
|
|
for (int n = MP_MAX_KEY_DOWN - 1; n > 0; n--)
|
|
|
|
buf[n] = buf[n - 1];
|
|
|
|
buf[0] = code;
|
|
|
|
}
|
|
|
|
|
2013-06-22 21:26:44 +00:00
|
|
|
static struct cmd_bind *find_bind_for_key_section(struct input_ctx *ictx,
|
2014-04-18 14:27:02 +00:00
|
|
|
char *section, int code)
|
2013-06-22 21:26:44 +00:00
|
|
|
{
|
|
|
|
struct cmd_bind_section *bs = get_bind_section(ictx, bstr0(section));
|
2013-06-22 22:40:35 +00:00
|
|
|
|
2014-04-18 14:27:02 +00:00
|
|
|
if (!bs->num_binds)
|
2013-06-22 22:40:35 +00:00
|
|
|
return NULL;
|
|
|
|
|
2014-04-18 14:27:02 +00:00
|
|
|
int keys[MP_MAX_KEY_DOWN];
|
|
|
|
memcpy(keys, ictx->key_history, sizeof(keys));
|
|
|
|
key_buf_add(keys, code);
|
|
|
|
|
2014-04-19 12:31:54 +00:00
|
|
|
struct cmd_bind *best = NULL;
|
|
|
|
|
2013-06-22 22:40:35 +00:00
|
|
|
// Prefer user-defined keys over builtin bindings
|
|
|
|
for (int builtin = 0; builtin < 2; builtin++) {
|
2014-09-27 14:25:29 +00:00
|
|
|
if (builtin && !ictx->opts->default_bindings)
|
2013-08-30 00:24:14 +00:00
|
|
|
break;
|
2014-04-19 12:31:54 +00:00
|
|
|
if (best)
|
|
|
|
break;
|
2013-06-22 22:40:35 +00:00
|
|
|
for (int n = 0; n < bs->num_binds; n++) {
|
2014-04-18 14:27:02 +00:00
|
|
|
if (bs->binds[n].is_builtin == (bool)builtin) {
|
|
|
|
struct cmd_bind *b = &bs->binds[n];
|
|
|
|
// we have: keys=[key2 key1 keyX ...]
|
|
|
|
// and: b->keys=[key1 key2] (and may be just a prefix)
|
|
|
|
for (int i = 0; i < b->num_keys; i++) {
|
|
|
|
if (b->keys[i] != keys[b->num_keys - 1 - i])
|
|
|
|
goto skip;
|
|
|
|
}
|
2014-04-19 12:31:54 +00:00
|
|
|
if (!best || b->num_keys >= best->num_keys)
|
|
|
|
best = b;
|
2014-04-18 14:27:02 +00:00
|
|
|
skip: ;
|
|
|
|
}
|
2013-06-22 22:40:35 +00:00
|
|
|
}
|
|
|
|
}
|
2014-04-19 12:31:54 +00:00
|
|
|
return best;
|
2013-06-22 21:26:44 +00:00
|
|
|
}
|
|
|
|
|
2013-03-30 19:07:15 +00:00
|
|
|
static struct cmd_bind *find_any_bind_for_key(struct input_ctx *ictx,
|
2014-04-18 14:27:02 +00:00
|
|
|
char *force_section, int code)
|
2008-04-30 05:56:45 +00:00
|
|
|
{
|
2013-06-22 21:26:44 +00:00
|
|
|
if (force_section)
|
2014-04-18 14:27:02 +00:00
|
|
|
return find_bind_for_key_section(ictx, force_section, code);
|
2013-06-19 16:19:45 +00:00
|
|
|
|
2014-04-18 14:27:02 +00:00
|
|
|
bool use_mouse = MP_KEY_DEPENDS_ON_MOUSE_POS(code);
|
2013-08-30 00:24:14 +00:00
|
|
|
|
|
|
|
// First look whether a mouse section is capturing all mouse input
|
|
|
|
// exclusively (regardless of the active section stack order).
|
2014-04-26 18:27:52 +00:00
|
|
|
if (use_mouse && MP_KEY_IS_MOUSE_BTN_SINGLE(ictx->last_key_down)) {
|
2013-08-30 00:24:14 +00:00
|
|
|
struct cmd_bind *bind =
|
2014-04-18 14:27:02 +00:00
|
|
|
find_bind_for_key_section(ictx, ictx->mouse_section, code);
|
2013-08-30 00:24:14 +00:00
|
|
|
if (bind)
|
|
|
|
return bind;
|
|
|
|
}
|
|
|
|
|
2014-02-17 01:35:00 +00:00
|
|
|
struct cmd_bind *best_bind = NULL;
|
input: handle mouse movement differently
Before this commit, mouse movement events emitted a special command
("set_mouse_pos"), which was specially handled in command.c. This was
once special-cased to the dvdnav and menu code, and did nothing after
libmenu and dvdnav were removed.
Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"),
which then can be bound to an arbitrary command. The mouse position is
now managed in input.c. A command which actually needs the mouse
position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos()
to query it. The former returns raw window-space coordinates, while the
latter returns coordinates transformed to OSD- space. (Both are the same
for most VOs, except vo_xv and vo_x11, which can't render OSD in
window-space. These require extra code for mapping mouse position.)
As of this commit, there is still nothing that uses mouse movement, so
MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the
mouse (much like MOUSE_BTN0).
Extend the concept of input sections. Allow multiple sections to be
active at once, and organize them as stack. Bindings from the top of
the stack are preferred to lower ones.
Each section has a mouse input section associated, inside which mouse
events are associated with the bindings. If the mouse pointer is
outside of a section's mouse area, mouse events will be dispatched to
an input section lower on the stack of active sections. This is intended
for scripting, which is to be added later. Two scripts could occupy
different areas of the screen without conflicting with each other. (If
it turns out that this mechanism is useless, we'll just remove it
again.)
2013-04-26 00:13:30 +00:00
|
|
|
for (int i = ictx->num_active_sections - 1; i >= 0; i--) {
|
2013-06-22 21:26:44 +00:00
|
|
|
struct active_section *s = &ictx->active_sections[i];
|
2014-04-18 14:27:02 +00:00
|
|
|
struct cmd_bind *bind = find_bind_for_key_section(ictx, s->name, code);
|
2013-06-22 21:26:44 +00:00
|
|
|
if (bind) {
|
|
|
|
struct cmd_bind_section *bs = bind->owner;
|
input: don't deliver mouse events if mouse area is not set
This caused the OSC to be always visible at startup on X11:
- EnterNotify event send a mouse event to input.c
- OSC has not completely initialized yet, and no mouse area is set
- mouse event is dispatched to "showhide" OSC section
- OSC becomes visible, regardless of mouse position
Fix this by treating the mouse area as empty if it's not set, instead of
infinite as it was before this commit. This means an input section must
set a mouse area to receive mouse events at all. We also have to change
the default section to receive mouse events with the new behavior.
Also, if MOUSE_MOVE is unmapped (or mapped to something that doesn't
parse), and produces no command, the mouse position wouldn't be updated
(because the mouse position is bound to input commands), so we have to
generate a dummy command in this case.
(This matters only for the OSC, On Screen Controller, which isn't merged
yet, so these changes shouldn't have much effect right now.)
2013-09-05 21:58:51 +00:00
|
|
|
if (!use_mouse || (bs->mouse_area_set && test_rect(&bs->mouse_area,
|
|
|
|
ictx->mouse_vo_x,
|
|
|
|
ictx->mouse_vo_y)))
|
2014-02-17 01:35:00 +00:00
|
|
|
{
|
|
|
|
if (!best_bind || (best_bind->is_builtin && !bind->is_builtin))
|
|
|
|
best_bind = bind;
|
|
|
|
}
|
input: handle mouse movement differently
Before this commit, mouse movement events emitted a special command
("set_mouse_pos"), which was specially handled in command.c. This was
once special-cased to the dvdnav and menu code, and did nothing after
libmenu and dvdnav were removed.
Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"),
which then can be bound to an arbitrary command. The mouse position is
now managed in input.c. A command which actually needs the mouse
position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos()
to query it. The former returns raw window-space coordinates, while the
latter returns coordinates transformed to OSD- space. (Both are the same
for most VOs, except vo_xv and vo_x11, which can't render OSD in
window-space. These require extra code for mapping mouse position.)
As of this commit, there is still nothing that uses mouse movement, so
MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the
mouse (much like MOUSE_BTN0).
Extend the concept of input sections. Allow multiple sections to be
active at once, and organize them as stack. Bindings from the top of
the stack are preferred to lower ones.
Each section has a mouse input section associated, inside which mouse
events are associated with the bindings. If the mouse pointer is
outside of a section's mouse area, mouse events will be dispatched to
an input section lower on the stack of active sections. This is intended
for scripting, which is to be added later. Two scripts could occupy
different areas of the screen without conflicting with each other. (If
it turns out that this mechanism is useless, we'll just remove it
again.)
2013-04-26 00:13:30 +00:00
|
|
|
}
|
2013-06-22 21:26:44 +00:00
|
|
|
if (s->flags & MP_INPUT_EXCLUSIVE)
|
input: handle mouse movement differently
Before this commit, mouse movement events emitted a special command
("set_mouse_pos"), which was specially handled in command.c. This was
once special-cased to the dvdnav and menu code, and did nothing after
libmenu and dvdnav were removed.
Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"),
which then can be bound to an arbitrary command. The mouse position is
now managed in input.c. A command which actually needs the mouse
position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos()
to query it. The former returns raw window-space coordinates, while the
latter returns coordinates transformed to OSD- space. (Both are the same
for most VOs, except vo_xv and vo_x11, which can't render OSD in
window-space. These require extra code for mapping mouse position.)
As of this commit, there is still nothing that uses mouse movement, so
MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the
mouse (much like MOUSE_BTN0).
Extend the concept of input sections. Allow multiple sections to be
active at once, and organize them as stack. Bindings from the top of
the stack are preferred to lower ones.
Each section has a mouse input section associated, inside which mouse
events are associated with the bindings. If the mouse pointer is
outside of a section's mouse area, mouse events will be dispatched to
an input section lower on the stack of active sections. This is intended
for scripting, which is to be added later. Two scripts could occupy
different areas of the screen without conflicting with each other. (If
it turns out that this mechanism is useless, we'll just remove it
again.)
2013-04-26 00:13:30 +00:00
|
|
|
break;
|
2014-12-04 19:59:01 +00:00
|
|
|
if (best_bind && (s->flags & MP_INPUT_ON_TOP))
|
|
|
|
break;
|
2012-08-24 11:29:28 +00:00
|
|
|
}
|
2013-06-19 16:19:45 +00:00
|
|
|
|
2014-02-17 01:35:00 +00:00
|
|
|
return best_bind;
|
2013-03-30 19:07:15 +00:00
|
|
|
}
|
|
|
|
|
2013-06-19 16:19:45 +00:00
|
|
|
static mp_cmd_t *get_cmd_from_keys(struct input_ctx *ictx, char *force_section,
|
2014-04-18 14:27:02 +00:00
|
|
|
int code)
|
2013-03-30 19:07:15 +00:00
|
|
|
{
|
2014-09-27 14:25:29 +00:00
|
|
|
if (ictx->opts->test)
|
2014-04-18 14:27:02 +00:00
|
|
|
return handle_test(ictx, code);
|
2011-04-25 08:38:46 +00:00
|
|
|
|
2014-04-18 14:27:02 +00:00
|
|
|
struct cmd_bind *cmd = find_any_bind_for_key(ictx, force_section, code);
|
2011-04-25 08:38:46 +00:00
|
|
|
if (cmd == NULL) {
|
2015-01-12 11:53:49 +00:00
|
|
|
if (code == MP_KEY_CLOSE_WIN)
|
|
|
|
return mp_input_parse_cmd_strv(ictx->log, (const char*[]){"quit", 0});
|
2013-09-05 22:28:31 +00:00
|
|
|
int msgl = MSGL_WARN;
|
2014-04-18 14:27:02 +00:00
|
|
|
if (code == MP_KEY_MOUSE_MOVE || code == MP_KEY_MOUSE_LEAVE)
|
2013-12-21 20:41:18 +00:00
|
|
|
msgl = MSGL_DEBUG;
|
2014-04-18 14:27:02 +00:00
|
|
|
char *key_buf = mp_input_get_key_combo_name(&code, 1);
|
2013-09-27 13:30:59 +00:00
|
|
|
MP_MSG(ictx, msgl, "No bind found for key '%s'.\n", key_buf);
|
2011-04-25 08:38:46 +00:00
|
|
|
talloc_free(key_buf);
|
|
|
|
return NULL;
|
2002-02-11 11:42:08 +00:00
|
|
|
}
|
2013-09-27 13:30:59 +00:00
|
|
|
mp_cmd_t *ret = mp_input_parse_cmd(ictx, bstr0(cmd->cmd), cmd->location);
|
2013-06-19 16:19:45 +00:00
|
|
|
if (ret) {
|
|
|
|
ret->input_section = cmd->owner->section;
|
2013-12-21 20:49:13 +00:00
|
|
|
if (mp_msg_test(ictx->log, MSGL_DEBUG)) {
|
2014-04-18 14:27:02 +00:00
|
|
|
char *keyname = mp_input_get_key_combo_name(&code, 1);
|
2013-09-10 06:29:45 +00:00
|
|
|
MP_DBG(ictx, "key '%s' -> '%s' in '%s'\n",
|
2013-08-30 00:30:17 +00:00
|
|
|
keyname, cmd->cmd, ret->input_section);
|
|
|
|
talloc_free(keyname);
|
|
|
|
}
|
2014-11-23 14:08:49 +00:00
|
|
|
ret->is_mouse_button = code & MP_KEY_EMIT_ON_UP;
|
2013-06-19 16:19:45 +00:00
|
|
|
} else {
|
2014-04-18 14:27:02 +00:00
|
|
|
char *key_buf = mp_input_get_key_combo_name(&code, 1);
|
2013-09-27 13:30:59 +00:00
|
|
|
MP_ERR(ictx, "Invalid command for bound key '%s': '%s'\n",
|
|
|
|
key_buf, cmd->cmd);
|
2011-04-25 08:38:46 +00:00
|
|
|
talloc_free(key_buf);
|
2002-02-24 21:22:12 +00:00
|
|
|
}
|
2011-04-25 08:38:46 +00:00
|
|
|
return ret;
|
2002-02-11 11:42:08 +00:00
|
|
|
}
|
|
|
|
|
2013-08-30 00:24:14 +00:00
|
|
|
static void update_mouse_section(struct input_ctx *ictx)
|
|
|
|
{
|
2013-11-01 12:00:15 +00:00
|
|
|
struct cmd_bind *bind =
|
2014-04-18 14:27:02 +00:00
|
|
|
find_any_bind_for_key(ictx, NULL, MP_KEY_MOUSE_MOVE);
|
2013-08-30 00:24:14 +00:00
|
|
|
|
2013-11-01 12:00:15 +00:00
|
|
|
char *new_section = bind ? bind->owner->section : "default";
|
2013-08-30 00:24:14 +00:00
|
|
|
|
|
|
|
char *old = ictx->mouse_section;
|
|
|
|
ictx->mouse_section = new_section;
|
|
|
|
|
|
|
|
if (strcmp(old, ictx->mouse_section) != 0) {
|
2013-09-10 06:29:45 +00:00
|
|
|
MP_DBG(ictx, "input: switch section %s -> %s\n",
|
2013-08-30 00:30:17 +00:00
|
|
|
old, ictx->mouse_section);
|
2014-09-12 23:08:35 +00:00
|
|
|
mp_input_queue_cmd(ictx, get_cmd_from_keys(ictx, old, MP_KEY_MOUSE_LEAVE));
|
2013-08-30 00:24:14 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2013-12-01 16:59:20 +00:00
|
|
|
// Called when the currently held-down key is released. This (usually) sends
|
2014-05-26 19:49:14 +00:00
|
|
|
// the a key-up version of the command associated with the keys that were held
|
2013-12-01 16:59:20 +00:00
|
|
|
// down.
|
|
|
|
// If the drop_current parameter is set to true, then don't send the key-up
|
|
|
|
// command. Unless we've already sent a key-down event, in which case the
|
|
|
|
// input receiver (the player) must get a key-up event, or it would get stuck
|
|
|
|
// thinking a key is still held down.
|
|
|
|
static void release_down_cmd(struct input_ctx *ictx, bool drop_current)
|
input: handle mouse movement differently
Before this commit, mouse movement events emitted a special command
("set_mouse_pos"), which was specially handled in command.c. This was
once special-cased to the dvdnav and menu code, and did nothing after
libmenu and dvdnav were removed.
Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"),
which then can be bound to an arbitrary command. The mouse position is
now managed in input.c. A command which actually needs the mouse
position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos()
to query it. The former returns raw window-space coordinates, while the
latter returns coordinates transformed to OSD- space. (Both are the same
for most VOs, except vo_xv and vo_x11, which can't render OSD in
window-space. These require extra code for mapping mouse position.)
As of this commit, there is still nothing that uses mouse movement, so
MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the
mouse (much like MOUSE_BTN0).
Extend the concept of input sections. Allow multiple sections to be
active at once, and organize them as stack. Bindings from the top of
the stack are preferred to lower ones.
Each section has a mouse input section associated, inside which mouse
events are associated with the bindings. If the mouse pointer is
outside of a section's mouse area, mouse events will be dispatched to
an input section lower on the stack of active sections. This is intended
for scripting, which is to be added later. Two scripts could occupy
different areas of the screen without conflicting with each other. (If
it turns out that this mechanism is useless, we'll just remove it
again.)
2013-04-26 00:13:30 +00:00
|
|
|
{
|
2014-11-24 15:48:34 +00:00
|
|
|
if (ictx->current_down_cmd && ictx->current_down_cmd->emit_on_up &&
|
|
|
|
(!drop_current || ictx->current_down_cmd->def->on_updown))
|
2013-12-01 16:59:20 +00:00
|
|
|
{
|
input: restore ability to combine mouse buttons
Key bindings are decided on the "down" event, so if the prefix is not
unique, the first/shortest will be used (e.g. when both "a" and "a-b"
are mapped, "a" will always be chosen).
This also breaks combining multiple mouse buttons. But it seems users
expect it to work, and it's indeed a bit strange that it shouldn't work,
as mouse bindings are emitted on the key "up" event, not "down" (if the
shorter binding didn't emit a command yet, why shouldn't it be
combinable).
Deal with this by clearing the key history when a command is actually
emitted, instead of when a command is decided. This means if both
MOUSE_BTN0 and MOUSE_BTN0-MOUSE_BTN1 are mapped, the sequence of holding
down BTN0 and then BTN1 will redecide the current command. On the other
hand, if BTN0 is released before BTN1 is pressed, the command is
emitted, and the key history is deleted. So the BTN1 press will not
trigger BTN0-BTN1.
For normal keys, nothing should change, because commands are emitted on
the "down" event already, so the key history is always cleared.
Might fix #902.
CC: @mpv-player/stable (if this fix is successful)
2014-07-07 16:18:41 +00:00
|
|
|
memset(ictx->key_history, 0, sizeof(ictx->key_history));
|
2014-11-23 14:08:49 +00:00
|
|
|
ictx->current_down_cmd->is_up = true;
|
2014-09-12 23:08:35 +00:00
|
|
|
mp_input_queue_cmd(ictx, ictx->current_down_cmd);
|
input: handle mouse movement differently
Before this commit, mouse movement events emitted a special command
("set_mouse_pos"), which was specially handled in command.c. This was
once special-cased to the dvdnav and menu code, and did nothing after
libmenu and dvdnav were removed.
Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"),
which then can be bound to an arbitrary command. The mouse position is
now managed in input.c. A command which actually needs the mouse
position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos()
to query it. The former returns raw window-space coordinates, while the
latter returns coordinates transformed to OSD- space. (Both are the same
for most VOs, except vo_xv and vo_x11, which can't render OSD in
window-space. These require extra code for mapping mouse position.)
As of this commit, there is still nothing that uses mouse movement, so
MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the
mouse (much like MOUSE_BTN0).
Extend the concept of input sections. Allow multiple sections to be
active at once, and organize them as stack. Bindings from the top of
the stack are preferred to lower ones.
Each section has a mouse input section associated, inside which mouse
events are associated with the bindings. If the mouse pointer is
outside of a section's mouse area, mouse events will be dispatched to
an input section lower on the stack of active sections. This is intended
for scripting, which is to be added later. Two scripts could occupy
different areas of the screen without conflicting with each other. (If
it turns out that this mechanism is useless, we'll just remove it
again.)
2013-04-26 00:13:30 +00:00
|
|
|
} else {
|
|
|
|
talloc_free(ictx->current_down_cmd);
|
|
|
|
}
|
|
|
|
ictx->current_down_cmd = NULL;
|
2014-04-18 15:45:41 +00:00
|
|
|
ictx->last_key_down = 0;
|
2014-04-18 15:16:33 +00:00
|
|
|
ictx->last_key_down_time = 0;
|
input: handle mouse movement differently
Before this commit, mouse movement events emitted a special command
("set_mouse_pos"), which was specially handled in command.c. This was
once special-cased to the dvdnav and menu code, and did nothing after
libmenu and dvdnav were removed.
Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"),
which then can be bound to an arbitrary command. The mouse position is
now managed in input.c. A command which actually needs the mouse
position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos()
to query it. The former returns raw window-space coordinates, while the
latter returns coordinates transformed to OSD- space. (Both are the same
for most VOs, except vo_xv and vo_x11, which can't render OSD in
window-space. These require extra code for mapping mouse position.)
As of this commit, there is still nothing that uses mouse movement, so
MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the
mouse (much like MOUSE_BTN0).
Extend the concept of input sections. Allow multiple sections to be
active at once, and organize them as stack. Bindings from the top of
the stack are preferred to lower ones.
Each section has a mouse input section associated, inside which mouse
events are associated with the bindings. If the mouse pointer is
outside of a section's mouse area, mouse events will be dispatched to
an input section lower on the stack of active sections. This is intended
for scripting, which is to be added later. Two scripts could occupy
different areas of the screen without conflicting with each other. (If
it turns out that this mechanism is useless, we'll just remove it
again.)
2013-04-26 00:13:30 +00:00
|
|
|
ictx->ar_state = -1;
|
2014-04-18 15:45:41 +00:00
|
|
|
update_mouse_section(ictx);
|
2013-07-01 21:54:59 +00:00
|
|
|
}
|
2002-11-14 23:41:44 +00:00
|
|
|
|
2013-09-02 00:26:41 +00:00
|
|
|
// We don't want the append to the command queue indefinitely, because that
|
|
|
|
// could lead to situations where recovery would take too long. On the other
|
|
|
|
// hand, don't drop commands that will abort playback.
|
|
|
|
static bool should_drop_cmd(struct input_ctx *ictx, struct mp_cmd *cmd)
|
|
|
|
{
|
|
|
|
struct cmd_queue *queue = &ictx->cmd_queue;
|
2014-09-27 14:25:29 +00:00
|
|
|
return queue_count_cmds(queue) >= ictx->opts->key_fifo_size &&
|
stream: redo playback abort handling
This mechanism originates from MPlayer's way of dealing with blocking
network, but it's still useful. On opening and closing, mpv waits for
network synchronously, and also some obscure commands and use-cases can
lead to such blocking. In these situations, the stream is asynchronously
forced to stop by "interrupting" it.
The old design interrupting I/O was a bit broken: polling with a
callback, instead of actively interrupting it. Change the direction of
this. There is no callback anymore, and the player calls
mp_cancel_trigger() to force the stream to return.
libavformat (via stream_lavf.c) has the old broken design, and fixing it
would require fixing libavformat, which won't happen so quickly. So we
have to keep that part. But everything above the stream layer is
prepared for a better design, and more sophisticated methods than
mp_cancel_test() could be easily introduced.
There's still one problem: commands are still run in the central
playback loop, which we assume can block on I/O in the worst case.
That's not a problem yet, because we simply mark some commands as being
able to stop playback of the current file ("quit" etc.), so input.c
could abort playback as soon as such a command is queued. But there are
also commands abort playback only conditionally, and the logic for that
is in the playback core and thus "unreachable". For example,
"playlist_next" aborts playback only if there's a next file. We don't
want it to always abort playback.
As a quite ugly hack, abort playback only if at least 2 abort commands
are queued - this pretty much happens only if the core is frozen and
doesn't react to input.
2014-09-13 12:23:08 +00:00
|
|
|
!mp_input_is_abort_cmd(cmd);
|
2013-09-02 00:26:41 +00:00
|
|
|
}
|
|
|
|
|
2014-04-19 12:11:55 +00:00
|
|
|
static struct mp_cmd *resolve_key(struct input_ctx *ictx, int code)
|
|
|
|
{
|
|
|
|
update_mouse_section(ictx);
|
|
|
|
struct mp_cmd *cmd = get_cmd_from_keys(ictx, NULL, code);
|
2014-04-19 12:21:57 +00:00
|
|
|
key_buf_add(ictx->key_history, code);
|
input: restore ability to combine mouse buttons
Key bindings are decided on the "down" event, so if the prefix is not
unique, the first/shortest will be used (e.g. when both "a" and "a-b"
are mapped, "a" will always be chosen).
This also breaks combining multiple mouse buttons. But it seems users
expect it to work, and it's indeed a bit strange that it shouldn't work,
as mouse bindings are emitted on the key "up" event, not "down" (if the
shorter binding didn't emit a command yet, why shouldn't it be
combinable).
Deal with this by clearing the key history when a command is actually
emitted, instead of when a command is decided. This means if both
MOUSE_BTN0 and MOUSE_BTN0-MOUSE_BTN1 are mapped, the sequence of holding
down BTN0 and then BTN1 will redecide the current command. On the other
hand, if BTN0 is released before BTN1 is pressed, the command is
emitted, and the key history is deleted. So the BTN1 press will not
trigger BTN0-BTN1.
For normal keys, nothing should change, because commands are emitted on
the "down" event already, so the key history is always cleared.
Might fix #902.
CC: @mpv-player/stable (if this fix is successful)
2014-07-07 16:18:41 +00:00
|
|
|
if (cmd && cmd->id != MP_CMD_IGNORE && !should_drop_cmd(ictx, cmd))
|
|
|
|
return cmd;
|
|
|
|
talloc_free(cmd);
|
2014-04-19 12:21:57 +00:00
|
|
|
return NULL;
|
2014-04-19 12:11:55 +00:00
|
|
|
}
|
|
|
|
|
2013-09-02 00:26:41 +00:00
|
|
|
static void interpret_key(struct input_ctx *ictx, int code, double scale)
|
2013-07-01 21:54:59 +00:00
|
|
|
{
|
2014-04-18 15:45:41 +00:00
|
|
|
int state = code & (MP_KEY_STATE_DOWN | MP_KEY_STATE_UP);
|
|
|
|
code = code & ~(unsigned)state;
|
|
|
|
|
2013-12-21 20:49:13 +00:00
|
|
|
if (mp_msg_test(ictx->log, MSGL_DEBUG)) {
|
2014-04-18 15:45:41 +00:00
|
|
|
char *key = mp_input_get_key_name(code);
|
2013-09-10 06:29:45 +00:00
|
|
|
MP_DBG(ictx, "key code=%#x '%s'%s%s\n",
|
2014-04-18 15:45:41 +00:00
|
|
|
code, key, (state & MP_KEY_STATE_DOWN) ? " down" : "",
|
|
|
|
(state & MP_KEY_STATE_UP) ? " up" : "");
|
2013-08-30 00:30:17 +00:00
|
|
|
talloc_free(key);
|
|
|
|
}
|
|
|
|
|
2014-09-12 23:08:35 +00:00
|
|
|
if (MP_KEY_DEPENDS_ON_MOUSE_POS(code & ~MP_KEY_MODIFIER_MASK)) {
|
2013-09-02 00:26:41 +00:00
|
|
|
ictx->mouse_event_counter++;
|
2014-09-12 23:08:35 +00:00
|
|
|
mp_input_wakeup(ictx);
|
|
|
|
}
|
2013-09-02 00:26:41 +00:00
|
|
|
|
|
|
|
struct mp_cmd *cmd = NULL;
|
2013-07-01 21:54:59 +00:00
|
|
|
|
2014-04-18 15:45:41 +00:00
|
|
|
if (state == MP_KEY_STATE_DOWN) {
|
|
|
|
// Protect against VOs which send STATE_DOWN with autorepeat
|
|
|
|
if (ictx->last_key_down == code)
|
2013-09-02 00:26:41 +00:00
|
|
|
return;
|
2013-07-01 21:54:59 +00:00
|
|
|
// Cancel current down-event (there can be only one)
|
2013-12-01 16:59:20 +00:00
|
|
|
release_down_cmd(ictx, true);
|
2014-04-19 12:11:55 +00:00
|
|
|
cmd = resolve_key(ictx, code);
|
2014-11-23 14:08:49 +00:00
|
|
|
if (cmd) {
|
|
|
|
cmd->is_up_down = true;
|
2014-11-24 15:48:34 +00:00
|
|
|
cmd->emit_on_up = (code & MP_KEY_EMIT_ON_UP) || cmd->def->on_updown;
|
|
|
|
ictx->current_down_cmd = mp_cmd_clone(cmd);
|
2014-11-23 14:08:49 +00:00
|
|
|
}
|
2014-04-18 15:45:41 +00:00
|
|
|
ictx->last_key_down = code;
|
|
|
|
ictx->last_key_down_time = mp_time_us();
|
|
|
|
ictx->ar_state = 0;
|
2014-09-12 23:02:09 +00:00
|
|
|
mp_input_wakeup(ictx); // possibly start timer for autorepeat
|
2014-04-18 15:45:41 +00:00
|
|
|
} else if (state == MP_KEY_STATE_UP) {
|
|
|
|
// Most VOs send RELEASE_ALL anyway
|
2014-04-18 16:06:19 +00:00
|
|
|
release_down_cmd(ictx, false);
|
2013-07-01 21:54:59 +00:00
|
|
|
} else {
|
|
|
|
// Press of key with no separate down/up events
|
2014-04-18 15:45:41 +00:00
|
|
|
if (ictx->last_key_down == code) {
|
2013-07-01 21:54:59 +00:00
|
|
|
// Mixing press events and up/down with the same key is not allowed
|
2013-09-27 13:30:59 +00:00
|
|
|
MP_WARN(ictx, "Mixing key presses and up/down.\n");
|
2013-07-01 21:54:59 +00:00
|
|
|
}
|
2014-04-19 12:11:55 +00:00
|
|
|
cmd = resolve_key(ictx, code);
|
2013-09-02 00:26:41 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
if (!cmd)
|
|
|
|
return;
|
|
|
|
|
2013-12-01 16:34:43 +00:00
|
|
|
// Don't emit a command on key-down if the key is designed to emit commands
|
|
|
|
// on key-up (like mouse buttons). Also, if the command specifically should
|
|
|
|
// be sent both on key down and key up, still emit the command.
|
2014-11-24 15:48:34 +00:00
|
|
|
if (cmd->emit_on_up && !cmd->def->on_updown) {
|
2013-09-02 00:26:41 +00:00
|
|
|
talloc_free(cmd);
|
|
|
|
return;
|
2011-04-25 11:12:04 +00:00
|
|
|
}
|
2013-09-02 00:26:41 +00:00
|
|
|
|
input: restore ability to combine mouse buttons
Key bindings are decided on the "down" event, so if the prefix is not
unique, the first/shortest will be used (e.g. when both "a" and "a-b"
are mapped, "a" will always be chosen).
This also breaks combining multiple mouse buttons. But it seems users
expect it to work, and it's indeed a bit strange that it shouldn't work,
as mouse bindings are emitted on the key "up" event, not "down" (if the
shorter binding didn't emit a command yet, why shouldn't it be
combinable).
Deal with this by clearing the key history when a command is actually
emitted, instead of when a command is decided. This means if both
MOUSE_BTN0 and MOUSE_BTN0-MOUSE_BTN1 are mapped, the sequence of holding
down BTN0 and then BTN1 will redecide the current command. On the other
hand, if BTN0 is released before BTN1 is pressed, the command is
emitted, and the key history is deleted. So the BTN1 press will not
trigger BTN0-BTN1.
For normal keys, nothing should change, because commands are emitted on
the "down" event already, so the key history is always cleared.
Might fix #902.
CC: @mpv-player/stable (if this fix is successful)
2014-07-07 16:18:41 +00:00
|
|
|
memset(ictx->key_history, 0, sizeof(ictx->key_history));
|
|
|
|
|
2013-09-02 00:26:41 +00:00
|
|
|
cmd->scale = scale;
|
2014-09-12 23:08:35 +00:00
|
|
|
mp_input_queue_cmd(ictx, cmd);
|
2007-08-25 04:28:11 +00:00
|
|
|
}
|
2002-02-11 11:42:08 +00:00
|
|
|
|
2013-09-07 23:44:47 +00:00
|
|
|
static void mp_input_feed_key(struct input_ctx *ictx, int code, double scale)
|
2011-07-17 01:47:50 +00:00
|
|
|
{
|
2014-09-27 14:25:29 +00:00
|
|
|
struct input_opts *opts = ictx->opts;
|
|
|
|
|
2014-08-26 18:39:28 +00:00
|
|
|
code = mp_normalize_keycode(code);
|
2014-04-19 12:11:25 +00:00
|
|
|
int unmod = code & ~MP_KEY_MODIFIER_MASK;
|
2011-07-17 01:47:50 +00:00
|
|
|
if (code == MP_INPUT_RELEASE_ALL) {
|
2013-09-10 06:29:45 +00:00
|
|
|
MP_DBG(ictx, "release all\n");
|
2014-04-18 16:06:19 +00:00
|
|
|
release_down_cmd(ictx, false);
|
2011-07-17 01:47:50 +00:00
|
|
|
return;
|
|
|
|
}
|
2014-09-27 14:25:29 +00:00
|
|
|
if (!opts->enable_mouse_movements && MP_KEY_IS_MOUSE(unmod))
|
2014-07-27 20:00:55 +00:00
|
|
|
return;
|
2014-04-19 12:11:25 +00:00
|
|
|
if (unmod == MP_KEY_MOUSE_LEAVE) {
|
2013-09-27 13:33:47 +00:00
|
|
|
update_mouse_section(ictx);
|
2014-09-12 23:08:35 +00:00
|
|
|
mp_input_queue_cmd(ictx, get_cmd_from_keys(ictx, NULL, code));
|
2013-09-27 13:33:47 +00:00
|
|
|
return;
|
|
|
|
}
|
2013-07-02 12:00:24 +00:00
|
|
|
double now = mp_time_sec();
|
|
|
|
// ignore system-doubleclick if we generate these events ourselves
|
2014-09-27 14:25:29 +00:00
|
|
|
if (opts->doubleclick_time && MP_KEY_IS_MOUSE_BTN_DBL(unmod))
|
2013-07-02 12:00:24 +00:00
|
|
|
return;
|
2013-09-07 23:44:47 +00:00
|
|
|
interpret_key(ictx, code, scale);
|
2013-07-02 12:00:24 +00:00
|
|
|
if (code & MP_KEY_STATE_DOWN) {
|
|
|
|
code &= ~MP_KEY_STATE_DOWN;
|
2014-09-27 14:25:29 +00:00
|
|
|
if (ictx->last_doubleclick_key_down == code &&
|
|
|
|
now - ictx->last_doubleclick_time < opts->doubleclick_time / 1000.0)
|
2013-07-02 12:00:24 +00:00
|
|
|
{
|
|
|
|
if (code >= MP_MOUSE_BTN0 && code <= MP_MOUSE_BTN2)
|
2013-09-07 23:44:47 +00:00
|
|
|
interpret_key(ictx, code - MP_MOUSE_BTN0 + MP_MOUSE_BTN0_DBL, 1);
|
2013-07-02 12:00:24 +00:00
|
|
|
}
|
|
|
|
ictx->last_doubleclick_key_down = code;
|
|
|
|
ictx->last_doubleclick_time = now;
|
|
|
|
}
|
2013-09-07 23:44:47 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
void mp_input_put_key(struct input_ctx *ictx, int code)
|
|
|
|
{
|
|
|
|
input_lock(ictx);
|
|
|
|
mp_input_feed_key(ictx, code, 1);
|
2013-08-26 17:15:35 +00:00
|
|
|
input_unlock(ictx);
|
2013-07-02 12:00:24 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
void mp_input_put_key_utf8(struct input_ctx *ictx, int mods, struct bstr t)
|
|
|
|
{
|
|
|
|
while (t.len) {
|
|
|
|
int code = bstr_decode_utf8(t, &t);
|
|
|
|
if (code < 0)
|
|
|
|
break;
|
|
|
|
mp_input_put_key(ictx, code | mods);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2013-07-25 16:08:57 +00:00
|
|
|
void mp_input_put_axis(struct input_ctx *ictx, int direction, double value)
|
|
|
|
{
|
2013-11-25 18:31:13 +00:00
|
|
|
if (value == 0.0)
|
|
|
|
return;
|
2013-08-26 17:15:35 +00:00
|
|
|
input_lock(ictx);
|
2013-09-07 23:44:47 +00:00
|
|
|
mp_input_feed_key(ictx, direction, value);
|
2013-08-26 17:15:35 +00:00
|
|
|
input_unlock(ictx);
|
2013-07-25 16:08:57 +00:00
|
|
|
}
|
|
|
|
|
2014-07-27 19:33:11 +00:00
|
|
|
void mp_input_set_mouse_transform(struct input_ctx *ictx, struct mp_rect *dst,
|
|
|
|
struct mp_rect *src)
|
|
|
|
{
|
|
|
|
input_lock(ictx);
|
|
|
|
ictx->mouse_mangle = dst || src;
|
|
|
|
if (ictx->mouse_mangle) {
|
|
|
|
ictx->mouse_dst = *dst;
|
|
|
|
ictx->mouse_src_mangle = !!src;
|
|
|
|
if (ictx->mouse_src_mangle)
|
|
|
|
ictx->mouse_src = *src;
|
|
|
|
}
|
|
|
|
input_unlock(ictx);
|
|
|
|
}
|
|
|
|
|
2014-07-27 19:53:29 +00:00
|
|
|
bool mp_input_mouse_enabled(struct input_ctx *ictx)
|
|
|
|
{
|
|
|
|
input_lock(ictx);
|
|
|
|
bool r = ictx->opts->enable_mouse_movements;
|
|
|
|
input_unlock(ictx);
|
|
|
|
return r;
|
|
|
|
}
|
|
|
|
|
2014-10-09 16:28:37 +00:00
|
|
|
bool mp_input_vo_keyboard_enabled(struct input_ctx *ictx)
|
client API, X11: change default keyboard input handling again
Commit 64b7811c tried to do the "right thing" with respect to whether
keyboard input should be enabled or not. It turns out that X11 does
something stupid by design. All modern toolkits work around this native
X11 behavior, but embedding breaks these workarounds.
The only way to handle this correctly is the XEmbed protocol. It needs
to be supported by the toolkit, and probably also some mpv support. But
Qt has inconsistent support for it. In Qt 4, a X11 specific embedding
widget was needed. Qt 5.0 doesn't support it at all. Qt 5.1 apparently
supports it via QWindow, but if it really does, I couldn't get it to
work.
So add a hack instead. The new --input-x11-keyboard option controls
whether mpv should enable keyboard input on the X11 window or not. In
the command line player, it's enabled by default, but in libmpv it's
disabled.
This hack has the same problem as all previous embedding had: move the
mouse outside of the window, and you don't get keyboard input anymore.
Likewise, mpv will steal all keyboard input from the parent application
as long as the mouse is inside of the mpv window.
Also see issue #1090.
2014-09-28 18:11:00 +00:00
|
|
|
{
|
|
|
|
input_lock(ictx);
|
2014-10-09 16:28:37 +00:00
|
|
|
bool r = ictx->opts->vo_key_input;
|
client API, X11: change default keyboard input handling again
Commit 64b7811c tried to do the "right thing" with respect to whether
keyboard input should be enabled or not. It turns out that X11 does
something stupid by design. All modern toolkits work around this native
X11 behavior, but embedding breaks these workarounds.
The only way to handle this correctly is the XEmbed protocol. It needs
to be supported by the toolkit, and probably also some mpv support. But
Qt has inconsistent support for it. In Qt 4, a X11 specific embedding
widget was needed. Qt 5.0 doesn't support it at all. Qt 5.1 apparently
supports it via QWindow, but if it really does, I couldn't get it to
work.
So add a hack instead. The new --input-x11-keyboard option controls
whether mpv should enable keyboard input on the X11 window or not. In
the command line player, it's enabled by default, but in libmpv it's
disabled.
This hack has the same problem as all previous embedding had: move the
mouse outside of the window, and you don't get keyboard input anymore.
Likewise, mpv will steal all keyboard input from the parent application
as long as the mouse is inside of the mpv window.
Also see issue #1090.
2014-09-28 18:11:00 +00:00
|
|
|
input_unlock(ictx);
|
|
|
|
return r;
|
|
|
|
}
|
|
|
|
|
input: handle mouse movement differently
Before this commit, mouse movement events emitted a special command
("set_mouse_pos"), which was specially handled in command.c. This was
once special-cased to the dvdnav and menu code, and did nothing after
libmenu and dvdnav were removed.
Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"),
which then can be bound to an arbitrary command. The mouse position is
now managed in input.c. A command which actually needs the mouse
position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos()
to query it. The former returns raw window-space coordinates, while the
latter returns coordinates transformed to OSD- space. (Both are the same
for most VOs, except vo_xv and vo_x11, which can't render OSD in
window-space. These require extra code for mapping mouse position.)
As of this commit, there is still nothing that uses mouse movement, so
MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the
mouse (much like MOUSE_BTN0).
Extend the concept of input sections. Allow multiple sections to be
active at once, and organize them as stack. Bindings from the top of
the stack are preferred to lower ones.
Each section has a mouse input section associated, inside which mouse
events are associated with the bindings. If the mouse pointer is
outside of a section's mouse area, mouse events will be dispatched to
an input section lower on the stack of active sections. This is intended
for scripting, which is to be added later. Two scripts could occupy
different areas of the screen without conflicting with each other. (If
it turns out that this mechanism is useless, we'll just remove it
again.)
2013-04-26 00:13:30 +00:00
|
|
|
void mp_input_set_mouse_pos(struct input_ctx *ictx, int x, int y)
|
|
|
|
{
|
2013-08-26 17:15:35 +00:00
|
|
|
input_lock(ictx);
|
2013-09-10 06:29:45 +00:00
|
|
|
MP_DBG(ictx, "mouse move %d/%d\n", x, y);
|
2013-08-30 00:30:17 +00:00
|
|
|
|
2014-07-27 19:53:29 +00:00
|
|
|
if (!ictx->opts->enable_mouse_movements) {
|
|
|
|
input_unlock(ictx);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
2014-07-27 19:33:11 +00:00
|
|
|
if (ictx->mouse_mangle) {
|
|
|
|
struct mp_rect *src = &ictx->mouse_src;
|
|
|
|
struct mp_rect *dst = &ictx->mouse_dst;
|
|
|
|
x = MPCLAMP(x, dst->x0, dst->x1) - dst->x0;
|
|
|
|
y = MPCLAMP(y, dst->y0, dst->y1) - dst->y0;
|
|
|
|
if (ictx->mouse_src_mangle) {
|
|
|
|
x = x * 1.0 / (dst->x1 - dst->x0) * (src->x1 - src->x0) + src->x0;
|
|
|
|
y = y * 1.0 / (dst->y1 - dst->y0) * (src->y1 - src->y0) + src->y0;
|
|
|
|
}
|
|
|
|
MP_DBG(ictx, "-> %d/%d\n", x, y);
|
|
|
|
}
|
|
|
|
|
input: handle mouse movement differently
Before this commit, mouse movement events emitted a special command
("set_mouse_pos"), which was specially handled in command.c. This was
once special-cased to the dvdnav and menu code, and did nothing after
libmenu and dvdnav were removed.
Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"),
which then can be bound to an arbitrary command. The mouse position is
now managed in input.c. A command which actually needs the mouse
position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos()
to query it. The former returns raw window-space coordinates, while the
latter returns coordinates transformed to OSD- space. (Both are the same
for most VOs, except vo_xv and vo_x11, which can't render OSD in
window-space. These require extra code for mapping mouse position.)
As of this commit, there is still nothing that uses mouse movement, so
MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the
mouse (much like MOUSE_BTN0).
Extend the concept of input sections. Allow multiple sections to be
active at once, and organize them as stack. Bindings from the top of
the stack are preferred to lower ones.
Each section has a mouse input section associated, inside which mouse
events are associated with the bindings. If the mouse pointer is
outside of a section's mouse area, mouse events will be dispatched to
an input section lower on the stack of active sections. This is intended
for scripting, which is to be added later. Two scripts could occupy
different areas of the screen without conflicting with each other. (If
it turns out that this mechanism is useless, we'll just remove it
again.)
2013-04-26 00:13:30 +00:00
|
|
|
ictx->mouse_event_counter++;
|
|
|
|
ictx->mouse_vo_x = x;
|
|
|
|
ictx->mouse_vo_y = y;
|
2013-06-19 16:19:45 +00:00
|
|
|
|
2013-08-30 00:24:14 +00:00
|
|
|
update_mouse_section(ictx);
|
2014-04-18 14:27:02 +00:00
|
|
|
struct mp_cmd *cmd = get_cmd_from_keys(ictx, NULL, MP_KEY_MOUSE_MOVE);
|
input: don't deliver mouse events if mouse area is not set
This caused the OSC to be always visible at startup on X11:
- EnterNotify event send a mouse event to input.c
- OSC has not completely initialized yet, and no mouse area is set
- mouse event is dispatched to "showhide" OSC section
- OSC becomes visible, regardless of mouse position
Fix this by treating the mouse area as empty if it's not set, instead of
infinite as it was before this commit. This means an input section must
set a mouse area to receive mouse events at all. We also have to change
the default section to receive mouse events with the new behavior.
Also, if MOUSE_MOVE is unmapped (or mapped to something that doesn't
parse), and produces no command, the mouse position wouldn't be updated
(because the mouse position is bound to input commands), so we have to
generate a dummy command in this case.
(This matters only for the OSC, On Screen Controller, which isn't merged
yet, so these changes shouldn't have much effect right now.)
2013-09-05 21:58:51 +00:00
|
|
|
if (!cmd)
|
2013-09-27 13:30:59 +00:00
|
|
|
cmd = mp_input_parse_cmd(ictx, bstr0("ignore"), "<internal>");
|
2013-06-19 16:19:45 +00:00
|
|
|
|
2013-08-26 17:15:35 +00:00
|
|
|
if (cmd) {
|
|
|
|
cmd->mouse_move = true;
|
|
|
|
cmd->mouse_x = x;
|
|
|
|
cmd->mouse_y = y;
|
2013-09-02 00:26:41 +00:00
|
|
|
if (should_drop_cmd(ictx, cmd)) {
|
|
|
|
talloc_free(cmd);
|
|
|
|
} else {
|
2013-09-02 00:40:05 +00:00
|
|
|
// Coalesce with previous mouse move events (i.e. replace it)
|
|
|
|
struct mp_cmd *tail = queue_peek_tail(&ictx->cmd_queue);
|
|
|
|
if (tail && tail->mouse_move) {
|
|
|
|
queue_remove(&ictx->cmd_queue, tail);
|
|
|
|
talloc_free(tail);
|
|
|
|
}
|
2014-09-12 23:08:35 +00:00
|
|
|
mp_input_queue_cmd(ictx, cmd);
|
2013-09-02 00:26:41 +00:00
|
|
|
}
|
2013-08-26 17:15:35 +00:00
|
|
|
}
|
|
|
|
input_unlock(ictx);
|
2011-07-17 01:47:50 +00:00
|
|
|
}
|
|
|
|
|
2013-12-26 16:41:28 +00:00
|
|
|
unsigned int mp_input_get_mouse_event_counter(struct input_ctx *ictx)
|
2011-07-17 01:47:50 +00:00
|
|
|
{
|
2013-12-26 16:41:28 +00:00
|
|
|
// Make the frontend always display the mouse cursor (as long as it's not
|
|
|
|
// forced invisible) if mouse input is desired.
|
|
|
|
input_lock(ictx);
|
|
|
|
if (mp_input_test_mouse_active(ictx, ictx->mouse_x, ictx->mouse_y))
|
|
|
|
ictx->mouse_event_counter++;
|
|
|
|
int ret = ictx->mouse_event_counter;
|
|
|
|
input_unlock(ictx);
|
|
|
|
return ret;
|
2011-07-17 01:47:50 +00:00
|
|
|
}
|
2002-01-30 12:46:03 +00:00
|
|
|
|
2014-09-10 01:16:43 +00:00
|
|
|
// adjust min time to wait until next repeat event
|
|
|
|
static void adjust_max_wait_time(struct input_ctx *ictx, double *time)
|
2013-07-07 14:53:35 +00:00
|
|
|
{
|
2014-09-27 14:25:29 +00:00
|
|
|
struct input_opts *opts = ictx->opts;
|
|
|
|
if (ictx->last_key_down && opts->ar_rate > 0 && ictx->ar_state >= 0) {
|
|
|
|
*time = FFMIN(*time, 1.0 / opts->ar_rate);
|
|
|
|
*time = FFMIN(*time, opts->ar_delay / 1000.0);
|
2002-01-30 12:46:03 +00:00
|
|
|
}
|
2011-07-17 01:47:50 +00:00
|
|
|
}
|
2007-08-25 04:28:11 +00:00
|
|
|
|
2014-09-13 14:47:30 +00:00
|
|
|
static bool test_abort_cmd(struct input_ctx *ictx, struct mp_cmd *new)
|
|
|
|
{
|
|
|
|
if (!mp_input_is_maybe_abort_cmd(new))
|
|
|
|
return false;
|
|
|
|
if (mp_input_is_abort_cmd(new))
|
|
|
|
return true;
|
|
|
|
// Abort only if there are going to be at least 2 commands in the queue.
|
|
|
|
for (struct mp_cmd *cmd = ictx->cmd_queue.first; cmd; cmd = cmd->queue_next) {
|
|
|
|
if (mp_input_is_maybe_abort_cmd(cmd))
|
|
|
|
return true;
|
|
|
|
}
|
|
|
|
return false;
|
|
|
|
}
|
|
|
|
|
2011-04-25 08:38:46 +00:00
|
|
|
int mp_input_queue_cmd(struct input_ctx *ictx, mp_cmd_t *cmd)
|
2008-04-30 10:00:59 +00:00
|
|
|
{
|
2013-08-26 17:15:35 +00:00
|
|
|
input_lock(ictx);
|
2014-09-12 23:08:35 +00:00
|
|
|
if (cmd) {
|
2014-09-13 14:47:30 +00:00
|
|
|
if (ictx->cancel && test_abort_cmd(ictx, cmd))
|
stream: redo playback abort handling
This mechanism originates from MPlayer's way of dealing with blocking
network, but it's still useful. On opening and closing, mpv waits for
network synchronously, and also some obscure commands and use-cases can
lead to such blocking. In these situations, the stream is asynchronously
forced to stop by "interrupting" it.
The old design interrupting I/O was a bit broken: polling with a
callback, instead of actively interrupting it. Change the direction of
this. There is no callback anymore, and the player calls
mp_cancel_trigger() to force the stream to return.
libavformat (via stream_lavf.c) has the old broken design, and fixing it
would require fixing libavformat, which won't happen so quickly. So we
have to keep that part. But everything above the stream layer is
prepared for a better design, and more sophisticated methods than
mp_cancel_test() could be easily introduced.
There's still one problem: commands are still run in the central
playback loop, which we assume can block on I/O in the worst case.
That's not a problem yet, because we simply mark some commands as being
able to stop playback of the current file ("quit" etc.), so input.c
could abort playback as soon as such a command is queued. But there are
also commands abort playback only conditionally, and the logic for that
is in the playback core and thus "unreachable". For example,
"playlist_next" aborts playback only if there's a next file. We don't
want it to always abort playback.
As a quite ugly hack, abort playback only if at least 2 abort commands
are queued - this pretty much happens only if the core is frozen and
doesn't react to input.
2014-09-13 12:23:08 +00:00
|
|
|
mp_cancel_trigger(ictx->cancel);
|
2013-09-02 00:29:20 +00:00
|
|
|
queue_add_tail(&ictx->cmd_queue, cmd);
|
2014-09-12 23:08:35 +00:00
|
|
|
mp_input_wakeup(ictx);
|
|
|
|
}
|
2013-08-26 17:15:35 +00:00
|
|
|
input_unlock(ictx);
|
2011-04-25 08:38:46 +00:00
|
|
|
return 1;
|
2002-02-23 21:13:35 +00:00
|
|
|
}
|
|
|
|
|
2013-09-07 23:44:47 +00:00
|
|
|
static mp_cmd_t *check_autorepeat(struct input_ctx *ictx)
|
|
|
|
{
|
2014-09-27 14:25:29 +00:00
|
|
|
struct input_opts *opts = ictx->opts;
|
|
|
|
|
2013-09-07 23:44:47 +00:00
|
|
|
// No input : autorepeat ?
|
2014-09-27 14:25:29 +00:00
|
|
|
if (opts->ar_rate <= 0 || !ictx->current_down_cmd || !ictx->last_key_down ||
|
2014-09-12 23:08:35 +00:00
|
|
|
(ictx->last_key_down & MP_NO_REPEAT_KEY) ||
|
|
|
|
!mp_input_is_repeatable_cmd(ictx->current_down_cmd))
|
2014-04-18 16:13:58 +00:00
|
|
|
ictx->ar_state = -1; // disable
|
2014-11-24 15:48:34 +00:00
|
|
|
|
2014-04-18 16:13:58 +00:00
|
|
|
if (ictx->ar_state >= 0) {
|
2013-09-07 23:44:47 +00:00
|
|
|
int64_t t = mp_time_us();
|
|
|
|
if (ictx->last_ar + 2000000 < t)
|
|
|
|
ictx->last_ar = t;
|
|
|
|
// First time : wait delay
|
|
|
|
if (ictx->ar_state == 0
|
2014-09-27 14:25:29 +00:00
|
|
|
&& (t - ictx->last_key_down_time) >= opts->ar_delay * 1000)
|
2013-09-07 23:44:47 +00:00
|
|
|
{
|
|
|
|
ictx->ar_state = 1;
|
2014-09-27 14:25:29 +00:00
|
|
|
ictx->last_ar = ictx->last_key_down_time + opts->ar_delay * 1000;
|
2013-09-07 23:44:47 +00:00
|
|
|
// Then send rate / sec event
|
|
|
|
} else if (ictx->ar_state == 1
|
2014-09-27 14:25:29 +00:00
|
|
|
&& (t - ictx->last_ar) >= 1000000 / opts->ar_rate) {
|
|
|
|
ictx->last_ar += 1000000 / opts->ar_rate;
|
2014-11-24 15:48:34 +00:00
|
|
|
} else {
|
|
|
|
return NULL;
|
2013-09-07 23:44:47 +00:00
|
|
|
}
|
2014-11-24 15:48:34 +00:00
|
|
|
struct mp_cmd *ret = mp_cmd_clone(ictx->current_down_cmd);
|
|
|
|
ret->repeated = true;
|
|
|
|
return ret;
|
2013-09-07 23:44:47 +00:00
|
|
|
}
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
2014-09-07 18:44:54 +00:00
|
|
|
void mp_input_wait(struct input_ctx *ictx, double seconds)
|
|
|
|
{
|
|
|
|
input_lock(ictx);
|
2014-09-10 01:16:43 +00:00
|
|
|
adjust_max_wait_time(ictx, &seconds);
|
|
|
|
input_unlock(ictx);
|
|
|
|
while (sem_trywait(&ictx->wakeup) == 0)
|
|
|
|
seconds = -1;
|
2014-09-08 22:25:54 +00:00
|
|
|
if (seconds > 0) {
|
|
|
|
MP_STATS(ictx, "start sleep");
|
2014-09-10 01:16:43 +00:00
|
|
|
struct timespec ts =
|
|
|
|
mp_time_us_to_timespec(mp_add_timeout(mp_time_us(), seconds));
|
|
|
|
sem_timedwait(&ictx->wakeup, &ts);
|
2014-09-08 22:25:54 +00:00
|
|
|
MP_STATS(ictx, "end sleep");
|
|
|
|
}
|
2014-09-10 01:16:43 +00:00
|
|
|
}
|
|
|
|
|
2014-09-20 02:19:41 +00:00
|
|
|
void mp_input_wakeup_nolock(struct input_ctx *ictx)
|
2014-09-10 01:16:43 +00:00
|
|
|
{
|
2014-09-20 02:19:41 +00:00
|
|
|
// Some audio APIs discourage use of locking in their audio callback,
|
|
|
|
// and these audio callbacks happen to call mp_input_wakeup_nolock()
|
|
|
|
// when new data is needed. This is why we use semaphores here.
|
2014-09-10 01:16:43 +00:00
|
|
|
sem_post(&ictx->wakeup);
|
|
|
|
}
|
|
|
|
|
2014-09-20 02:19:41 +00:00
|
|
|
void mp_input_wakeup(struct input_ctx *ictx)
|
2014-09-10 01:16:43 +00:00
|
|
|
{
|
2014-09-20 02:19:41 +00:00
|
|
|
mp_input_wakeup_nolock(ictx);
|
2014-09-07 18:44:54 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
mp_cmd_t *mp_input_read_cmd(struct input_ctx *ictx)
|
2008-04-30 04:15:52 +00:00
|
|
|
{
|
2013-08-26 17:15:35 +00:00
|
|
|
input_lock(ictx);
|
2014-09-12 23:08:35 +00:00
|
|
|
struct mp_cmd *ret = queue_remove_head(&ictx->cmd_queue);
|
2014-11-24 15:48:34 +00:00
|
|
|
if (!ret)
|
2014-09-12 23:08:35 +00:00
|
|
|
ret = check_autorepeat(ictx);
|
|
|
|
if (ret && ret->mouse_move) {
|
|
|
|
ictx->mouse_x = ret->mouse_x;
|
|
|
|
ictx->mouse_y = ret->mouse_y;
|
input: handle mouse movement differently
Before this commit, mouse movement events emitted a special command
("set_mouse_pos"), which was specially handled in command.c. This was
once special-cased to the dvdnav and menu code, and did nothing after
libmenu and dvdnav were removed.
Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"),
which then can be bound to an arbitrary command. The mouse position is
now managed in input.c. A command which actually needs the mouse
position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos()
to query it. The former returns raw window-space coordinates, while the
latter returns coordinates transformed to OSD- space. (Both are the same
for most VOs, except vo_xv and vo_x11, which can't render OSD in
window-space. These require extra code for mapping mouse position.)
As of this commit, there is still nothing that uses mouse movement, so
MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the
mouse (much like MOUSE_BTN0).
Extend the concept of input sections. Allow multiple sections to be
active at once, and organize them as stack. Bindings from the top of
the stack are preferred to lower ones.
Each section has a mouse input section associated, inside which mouse
events are associated with the bindings. If the mouse pointer is
outside of a section's mouse area, mouse events will be dispatched to
an input section lower on the stack of active sections. This is intended
for scripting, which is to be added later. Two scripts could occupy
different areas of the screen without conflicting with each other. (If
it turns out that this mechanism is useless, we'll just remove it
again.)
2013-04-26 00:13:30 +00:00
|
|
|
}
|
2013-08-26 17:15:35 +00:00
|
|
|
input_unlock(ictx);
|
2011-04-25 08:38:46 +00:00
|
|
|
return ret;
|
2002-01-30 12:46:03 +00:00
|
|
|
}
|
|
|
|
|
input: handle mouse movement differently
Before this commit, mouse movement events emitted a special command
("set_mouse_pos"), which was specially handled in command.c. This was
once special-cased to the dvdnav and menu code, and did nothing after
libmenu and dvdnav were removed.
Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"),
which then can be bound to an arbitrary command. The mouse position is
now managed in input.c. A command which actually needs the mouse
position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos()
to query it. The former returns raw window-space coordinates, while the
latter returns coordinates transformed to OSD- space. (Both are the same
for most VOs, except vo_xv and vo_x11, which can't render OSD in
window-space. These require extra code for mapping mouse position.)
As of this commit, there is still nothing that uses mouse movement, so
MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the
mouse (much like MOUSE_BTN0).
Extend the concept of input sections. Allow multiple sections to be
active at once, and organize them as stack. Bindings from the top of
the stack are preferred to lower ones.
Each section has a mouse input section associated, inside which mouse
events are associated with the bindings. If the mouse pointer is
outside of a section's mouse area, mouse events will be dispatched to
an input section lower on the stack of active sections. This is intended
for scripting, which is to be added later. Two scripts could occupy
different areas of the screen without conflicting with each other. (If
it turns out that this mechanism is useless, we'll just remove it
again.)
2013-04-26 00:13:30 +00:00
|
|
|
void mp_input_get_mouse_pos(struct input_ctx *ictx, int *x, int *y)
|
|
|
|
{
|
2013-08-26 17:15:35 +00:00
|
|
|
input_lock(ictx);
|
input: handle mouse movement differently
Before this commit, mouse movement events emitted a special command
("set_mouse_pos"), which was specially handled in command.c. This was
once special-cased to the dvdnav and menu code, and did nothing after
libmenu and dvdnav were removed.
Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"),
which then can be bound to an arbitrary command. The mouse position is
now managed in input.c. A command which actually needs the mouse
position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos()
to query it. The former returns raw window-space coordinates, while the
latter returns coordinates transformed to OSD- space. (Both are the same
for most VOs, except vo_xv and vo_x11, which can't render OSD in
window-space. These require extra code for mapping mouse position.)
As of this commit, there is still nothing that uses mouse movement, so
MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the
mouse (much like MOUSE_BTN0).
Extend the concept of input sections. Allow multiple sections to be
active at once, and organize them as stack. Bindings from the top of
the stack are preferred to lower ones.
Each section has a mouse input section associated, inside which mouse
events are associated with the bindings. If the mouse pointer is
outside of a section's mouse area, mouse events will be dispatched to
an input section lower on the stack of active sections. This is intended
for scripting, which is to be added later. Two scripts could occupy
different areas of the screen without conflicting with each other. (If
it turns out that this mechanism is useless, we'll just remove it
again.)
2013-04-26 00:13:30 +00:00
|
|
|
*x = ictx->mouse_x;
|
|
|
|
*y = ictx->mouse_y;
|
2013-08-26 17:15:35 +00:00
|
|
|
input_unlock(ictx);
|
input: handle mouse movement differently
Before this commit, mouse movement events emitted a special command
("set_mouse_pos"), which was specially handled in command.c. This was
once special-cased to the dvdnav and menu code, and did nothing after
libmenu and dvdnav were removed.
Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"),
which then can be bound to an arbitrary command. The mouse position is
now managed in input.c. A command which actually needs the mouse
position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos()
to query it. The former returns raw window-space coordinates, while the
latter returns coordinates transformed to OSD- space. (Both are the same
for most VOs, except vo_xv and vo_x11, which can't render OSD in
window-space. These require extra code for mapping mouse position.)
As of this commit, there is still nothing that uses mouse movement, so
MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the
mouse (much like MOUSE_BTN0).
Extend the concept of input sections. Allow multiple sections to be
active at once, and organize them as stack. Bindings from the top of
the stack are preferred to lower ones.
Each section has a mouse input section associated, inside which mouse
events are associated with the bindings. If the mouse pointer is
outside of a section's mouse area, mouse events will be dispatched to
an input section lower on the stack of active sections. This is intended
for scripting, which is to be added later. Two scripts could occupy
different areas of the screen without conflicting with each other. (If
it turns out that this mechanism is useless, we'll just remove it
again.)
2013-04-26 00:13:30 +00:00
|
|
|
}
|
|
|
|
|
2013-12-26 16:41:28 +00:00
|
|
|
// If name is NULL, return "default".
|
|
|
|
// Return a statically allocated name of the section (i.e. return value never
|
|
|
|
// gets deallocated).
|
|
|
|
static char *normalize_section(struct input_ctx *ictx, char *name)
|
|
|
|
{
|
|
|
|
return get_bind_section(ictx, bstr0(name))->section;
|
|
|
|
}
|
|
|
|
|
|
|
|
void mp_input_disable_section(struct input_ctx *ictx, char *name)
|
|
|
|
{
|
|
|
|
input_lock(ictx);
|
|
|
|
name = normalize_section(ictx, name);
|
|
|
|
|
|
|
|
// Remove old section, or make sure it's on top if re-enabled
|
|
|
|
for (int i = ictx->num_active_sections - 1; i >= 0; i--) {
|
|
|
|
struct active_section *as = &ictx->active_sections[i];
|
|
|
|
if (strcmp(as->name, name) == 0) {
|
|
|
|
MP_TARRAY_REMOVE_AT(ictx->active_sections,
|
|
|
|
ictx->num_active_sections, i);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
input_unlock(ictx);
|
|
|
|
}
|
|
|
|
|
|
|
|
void mp_input_enable_section(struct input_ctx *ictx, char *name, int flags)
|
|
|
|
{
|
|
|
|
input_lock(ictx);
|
|
|
|
name = normalize_section(ictx, name);
|
|
|
|
|
|
|
|
mp_input_disable_section(ictx, name);
|
|
|
|
|
2014-08-24 17:17:33 +00:00
|
|
|
MP_DBG(ictx, "enable section '%s'\n", name);
|
2013-12-26 16:41:28 +00:00
|
|
|
|
|
|
|
if (ictx->num_active_sections < MAX_ACTIVE_SECTIONS) {
|
2014-02-19 14:40:04 +00:00
|
|
|
int top = ictx->num_active_sections;
|
|
|
|
if (!(flags & MP_INPUT_ON_TOP)) {
|
|
|
|
// insert before the first top entry
|
|
|
|
for (top = 0; top < ictx->num_active_sections; top++) {
|
|
|
|
if (ictx->active_sections[top].flags & MP_INPUT_ON_TOP)
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
for (int n = ictx->num_active_sections; n > top; n--)
|
|
|
|
ictx->active_sections[n] = ictx->active_sections[n - 1];
|
|
|
|
}
|
|
|
|
ictx->active_sections[top] = (struct active_section){name, flags};
|
|
|
|
ictx->num_active_sections++;
|
2013-12-26 16:41:28 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
MP_DBG(ictx, "active section stack:\n");
|
|
|
|
for (int n = 0; n < ictx->num_active_sections; n++) {
|
|
|
|
MP_DBG(ictx, " %s %d\n", ictx->active_sections[n].name,
|
|
|
|
ictx->active_sections[n].flags);
|
|
|
|
}
|
|
|
|
|
|
|
|
input_unlock(ictx);
|
|
|
|
}
|
|
|
|
|
|
|
|
void mp_input_disable_all_sections(struct input_ctx *ictx)
|
|
|
|
{
|
|
|
|
input_lock(ictx);
|
|
|
|
ictx->num_active_sections = 0;
|
|
|
|
input_unlock(ictx);
|
|
|
|
}
|
|
|
|
|
|
|
|
void mp_input_set_section_mouse_area(struct input_ctx *ictx, char *name,
|
|
|
|
int x0, int y0, int x1, int y1)
|
|
|
|
{
|
|
|
|
input_lock(ictx);
|
|
|
|
struct cmd_bind_section *s = get_bind_section(ictx, bstr0(name));
|
|
|
|
s->mouse_area = (struct mp_rect){x0, y0, x1, y1};
|
|
|
|
s->mouse_area_set = x0 != x1 && y0 != y1;
|
|
|
|
input_unlock(ictx);
|
|
|
|
}
|
|
|
|
|
|
|
|
static bool test_mouse(struct input_ctx *ictx, int x, int y, int rej_flags)
|
|
|
|
{
|
|
|
|
input_lock(ictx);
|
|
|
|
bool res = false;
|
|
|
|
for (int i = 0; i < ictx->num_active_sections; i++) {
|
|
|
|
struct active_section *as = &ictx->active_sections[i];
|
|
|
|
if (as->flags & rej_flags)
|
|
|
|
continue;
|
|
|
|
struct cmd_bind_section *s = get_bind_section(ictx, bstr0(as->name));
|
|
|
|
if (s->mouse_area_set && test_rect(&s->mouse_area, x, y)) {
|
|
|
|
res = true;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
input_unlock(ictx);
|
|
|
|
return res;
|
|
|
|
}
|
|
|
|
|
|
|
|
bool mp_input_test_mouse_active(struct input_ctx *ictx, int x, int y)
|
|
|
|
{
|
|
|
|
return test_mouse(ictx, x, y, MP_INPUT_ALLOW_HIDE_CURSOR);
|
|
|
|
}
|
|
|
|
|
|
|
|
bool mp_input_test_dragging(struct input_ctx *ictx, int x, int y)
|
|
|
|
{
|
2014-05-20 00:40:00 +00:00
|
|
|
input_lock(ictx);
|
|
|
|
bool r = !ictx->win_drag || test_mouse(ictx, x, y, MP_INPUT_ALLOW_VO_DRAGGING);
|
|
|
|
input_unlock(ictx);
|
|
|
|
return r;
|
2013-12-26 16:41:28 +00:00
|
|
|
}
|
|
|
|
|
2013-06-22 22:40:35 +00:00
|
|
|
static void bind_dealloc(struct cmd_bind *bind)
|
|
|
|
{
|
|
|
|
talloc_free(bind->cmd);
|
|
|
|
talloc_free(bind->location);
|
|
|
|
}
|
|
|
|
|
2013-12-26 16:41:28 +00:00
|
|
|
// builtin: if true, remove all builtin binds, else remove all user binds
|
|
|
|
static void remove_binds(struct cmd_bind_section *bs, bool builtin)
|
|
|
|
{
|
|
|
|
for (int n = bs->num_binds - 1; n >= 0; n--) {
|
|
|
|
if (bs->binds[n].is_builtin == builtin) {
|
|
|
|
bind_dealloc(&bs->binds[n]);
|
|
|
|
assert(bs->num_binds >= 1);
|
|
|
|
bs->binds[n] = bs->binds[bs->num_binds - 1];
|
|
|
|
bs->num_binds--;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
void mp_input_define_section(struct input_ctx *ictx, char *name, char *location,
|
|
|
|
char *contents, bool builtin)
|
|
|
|
{
|
|
|
|
if (!name || !name[0])
|
|
|
|
return; // parse_config() changes semantics with restrict_section==empty
|
|
|
|
input_lock(ictx);
|
2014-12-03 14:44:41 +00:00
|
|
|
// Delete:
|
|
|
|
struct cmd_bind_section *bs = get_bind_section(ictx, bstr0(name));
|
|
|
|
remove_binds(bs, builtin);
|
2013-12-26 16:41:28 +00:00
|
|
|
if (contents) {
|
2014-12-03 14:44:41 +00:00
|
|
|
// Redefine:
|
2013-12-26 16:41:28 +00:00
|
|
|
parse_config(ictx, builtin, bstr0(contents), location, name);
|
|
|
|
} else {
|
|
|
|
// Disable:
|
|
|
|
mp_input_disable_section(ictx, name);
|
|
|
|
}
|
|
|
|
input_unlock(ictx);
|
|
|
|
}
|
|
|
|
|
2014-04-18 14:48:13 +00:00
|
|
|
static bool bind_matches_key(struct cmd_bind *bind, int num_keys, const int *keys)
|
|
|
|
{
|
|
|
|
if (bind->num_keys != num_keys)
|
|
|
|
return false;
|
|
|
|
for (int i = 0; i < num_keys; i++) {
|
|
|
|
if (bind->keys[i] != keys[i])
|
|
|
|
return false;
|
|
|
|
}
|
|
|
|
return true;
|
|
|
|
}
|
|
|
|
|
2012-09-22 13:17:15 +00:00
|
|
|
static void bind_keys(struct input_ctx *ictx, bool builtin, bstr section,
|
2013-06-22 22:07:45 +00:00
|
|
|
const int *keys, int num_keys, bstr command,
|
2012-10-13 19:09:42 +00:00
|
|
|
const char *loc)
|
2008-04-30 05:56:45 +00:00
|
|
|
{
|
2013-06-22 22:40:35 +00:00
|
|
|
struct cmd_bind_section *bs = get_bind_section(ictx, section);
|
2011-07-16 15:17:48 +00:00
|
|
|
struct cmd_bind *bind = NULL;
|
2011-04-25 08:38:46 +00:00
|
|
|
|
2013-06-22 22:07:45 +00:00
|
|
|
assert(num_keys <= MP_MAX_KEY_DOWN);
|
|
|
|
|
2013-06-22 22:40:35 +00:00
|
|
|
for (int n = 0; n < bs->num_binds; n++) {
|
|
|
|
struct cmd_bind *b = &bs->binds[n];
|
|
|
|
if (bind_matches_key(b, num_keys, keys) && b->is_builtin == builtin) {
|
|
|
|
bind = b;
|
|
|
|
break;
|
2011-04-25 08:38:46 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
if (!bind) {
|
2013-06-22 22:40:35 +00:00
|
|
|
struct cmd_bind empty = {{0}};
|
|
|
|
MP_TARRAY_APPEND(bs, bs->binds, bs->num_binds, empty);
|
|
|
|
bind = &bs->binds[bs->num_binds - 1];
|
2002-02-24 16:31:27 +00:00
|
|
|
}
|
2013-06-22 22:40:35 +00:00
|
|
|
|
|
|
|
bind_dealloc(bind);
|
|
|
|
|
|
|
|
*bind = (struct cmd_bind) {
|
|
|
|
.cmd = bstrdup0(bs->binds, command),
|
|
|
|
.location = talloc_strdup(bs->binds, loc),
|
|
|
|
.owner = bs,
|
|
|
|
.is_builtin = builtin,
|
|
|
|
.num_keys = num_keys,
|
|
|
|
};
|
2013-06-22 22:07:45 +00:00
|
|
|
memcpy(bind->keys, keys, num_keys * sizeof(bind->keys[0]));
|
2014-02-17 01:35:16 +00:00
|
|
|
if (mp_msg_test(ictx->log, MSGL_DEBUG)) {
|
|
|
|
char *s = mp_input_get_key_combo_name(keys, num_keys);
|
|
|
|
MP_DBG(ictx, "add: section='%s' key='%s'%s cmd='%s' location='%s'\n",
|
|
|
|
bind->owner->section, s, bind->is_builtin ? " builtin" : "",
|
|
|
|
bind->cmd, bind->location);
|
|
|
|
talloc_free(s);
|
|
|
|
}
|
2002-02-24 16:31:27 +00:00
|
|
|
}
|
|
|
|
|
input: handle mouse movement differently
Before this commit, mouse movement events emitted a special command
("set_mouse_pos"), which was specially handled in command.c. This was
once special-cased to the dvdnav and menu code, and did nothing after
libmenu and dvdnav were removed.
Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"),
which then can be bound to an arbitrary command. The mouse position is
now managed in input.c. A command which actually needs the mouse
position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos()
to query it. The former returns raw window-space coordinates, while the
latter returns coordinates transformed to OSD- space. (Both are the same
for most VOs, except vo_xv and vo_x11, which can't render OSD in
window-space. These require extra code for mapping mouse position.)
As of this commit, there is still nothing that uses mouse movement, so
MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the
mouse (much like MOUSE_BTN0).
Extend the concept of input sections. Allow multiple sections to be
active at once, and organize them as stack. Bindings from the top of
the stack are preferred to lower ones.
Each section has a mouse input section associated, inside which mouse
events are associated with the bindings. If the mouse pointer is
outside of a section's mouse area, mouse events will be dispatched to
an input section lower on the stack of active sections. This is intended
for scripting, which is to be added later. Two scripts could occupy
different areas of the screen without conflicting with each other. (If
it turns out that this mechanism is useless, we'll just remove it
again.)
2013-04-26 00:13:30 +00:00
|
|
|
// restrict_section: every entry is forced to this section name
|
|
|
|
// if NULL, load normally and allow any sections
|
2012-10-13 19:09:42 +00:00
|
|
|
static int parse_config(struct input_ctx *ictx, bool builtin, bstr data,
|
input: handle mouse movement differently
Before this commit, mouse movement events emitted a special command
("set_mouse_pos"), which was specially handled in command.c. This was
once special-cased to the dvdnav and menu code, and did nothing after
libmenu and dvdnav were removed.
Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"),
which then can be bound to an arbitrary command. The mouse position is
now managed in input.c. A command which actually needs the mouse
position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos()
to query it. The former returns raw window-space coordinates, while the
latter returns coordinates transformed to OSD- space. (Both are the same
for most VOs, except vo_xv and vo_x11, which can't render OSD in
window-space. These require extra code for mapping mouse position.)
As of this commit, there is still nothing that uses mouse movement, so
MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the
mouse (much like MOUSE_BTN0).
Extend the concept of input sections. Allow multiple sections to be
active at once, and organize them as stack. Bindings from the top of
the stack are preferred to lower ones.
Each section has a mouse input section associated, inside which mouse
events are associated with the bindings. If the mouse pointer is
outside of a section's mouse area, mouse events will be dispatched to
an input section lower on the stack of active sections. This is intended
for scripting, which is to be added later. Two scripts could occupy
different areas of the screen without conflicting with each other. (If
it turns out that this mechanism is useless, we'll just remove it
again.)
2013-04-26 00:13:30 +00:00
|
|
|
const char *location, const char *restrict_section)
|
2008-04-30 05:56:45 +00:00
|
|
|
{
|
2013-06-22 22:07:45 +00:00
|
|
|
int n_binds = 0;
|
2012-10-13 19:09:42 +00:00
|
|
|
int line_no = 0;
|
|
|
|
char *cur_loc = NULL;
|
2011-07-20 01:41:51 +00:00
|
|
|
|
2012-08-01 19:50:24 +00:00
|
|
|
while (data.len) {
|
2012-10-13 19:09:42 +00:00
|
|
|
line_no++;
|
|
|
|
if (cur_loc)
|
|
|
|
talloc_free(cur_loc);
|
|
|
|
cur_loc = talloc_asprintf(NULL, "%s:%d", location, line_no);
|
|
|
|
|
2012-08-01 19:50:24 +00:00
|
|
|
bstr line = bstr_strip_linebreaks(bstr_getline(data, &data));
|
|
|
|
line = bstr_lstrip(line);
|
|
|
|
if (line.len == 0 || bstr_startswith0(line, "#"))
|
2011-04-25 08:38:46 +00:00
|
|
|
continue;
|
2014-02-25 21:09:20 +00:00
|
|
|
if (bstr_eatstart0(&line, "default-bindings ")) {
|
|
|
|
bstr orig = line;
|
|
|
|
bstr_split_tok(line, "#", &line, &(bstr){0});
|
|
|
|
line = bstr_strip(line);
|
|
|
|
if (bstr_equals0(line, "start")) {
|
|
|
|
builtin = true;
|
|
|
|
} else {
|
|
|
|
MP_ERR(ictx, "Broken line: %.*s at %s\n", BSTR_P(orig), cur_loc);
|
|
|
|
}
|
|
|
|
continue;
|
|
|
|
}
|
2011-07-20 01:41:51 +00:00
|
|
|
struct bstr command;
|
|
|
|
// Find the key name starting a line
|
2012-08-01 21:41:21 +00:00
|
|
|
struct bstr keyname = bstr_split(line, WHITESPACE, &command);
|
2011-07-20 01:41:51 +00:00
|
|
|
command = bstr_strip(command);
|
|
|
|
if (command.len == 0) {
|
2013-09-27 13:30:59 +00:00
|
|
|
MP_ERR(ictx, "Unfinished key binding: %.*s at %s\n", BSTR_P(line),
|
|
|
|
cur_loc);
|
2011-07-20 01:41:51 +00:00
|
|
|
continue;
|
|
|
|
}
|
|
|
|
char *name = bstrdup0(NULL, keyname);
|
2013-06-22 22:07:45 +00:00
|
|
|
int keys[MP_MAX_KEY_DOWN];
|
|
|
|
int num_keys = 0;
|
2013-12-26 16:10:35 +00:00
|
|
|
if (!mp_input_get_keys_from_string(name, MP_MAX_KEY_DOWN, &num_keys, keys))
|
|
|
|
{
|
2011-07-20 01:41:51 +00:00
|
|
|
talloc_free(name);
|
2013-09-27 13:30:59 +00:00
|
|
|
MP_ERR(ictx, "Unknown key '%.*s' at %s\n", BSTR_P(keyname), cur_loc);
|
2011-04-25 08:38:46 +00:00
|
|
|
continue;
|
2010-12-19 10:12:20 +00:00
|
|
|
}
|
2011-07-20 01:41:51 +00:00
|
|
|
talloc_free(name);
|
2012-09-22 13:17:15 +00:00
|
|
|
|
input: handle mouse movement differently
Before this commit, mouse movement events emitted a special command
("set_mouse_pos"), which was specially handled in command.c. This was
once special-cased to the dvdnav and menu code, and did nothing after
libmenu and dvdnav were removed.
Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"),
which then can be bound to an arbitrary command. The mouse position is
now managed in input.c. A command which actually needs the mouse
position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos()
to query it. The former returns raw window-space coordinates, while the
latter returns coordinates transformed to OSD- space. (Both are the same
for most VOs, except vo_xv and vo_x11, which can't render OSD in
window-space. These require extra code for mapping mouse position.)
As of this commit, there is still nothing that uses mouse movement, so
MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the
mouse (much like MOUSE_BTN0).
Extend the concept of input sections. Allow multiple sections to be
active at once, and organize them as stack. Bindings from the top of
the stack are preferred to lower ones.
Each section has a mouse input section associated, inside which mouse
events are associated with the bindings. If the mouse pointer is
outside of a section's mouse area, mouse events will be dispatched to
an input section lower on the stack of active sections. This is intended
for scripting, which is to be added later. Two scripts could occupy
different areas of the screen without conflicting with each other. (If
it turns out that this mechanism is useless, we'll just remove it
again.)
2013-04-26 00:13:30 +00:00
|
|
|
bstr section = bstr0(restrict_section);
|
|
|
|
if (!section.len) {
|
|
|
|
if (bstr_startswith0(command, "{")) {
|
|
|
|
int p = bstrchr(command, '}');
|
|
|
|
if (p != -1) {
|
|
|
|
section = bstr_strip(bstr_splice(command, 1, p));
|
|
|
|
command = bstr_lstrip(bstr_cut(command, p + 1));
|
|
|
|
}
|
2012-09-22 13:17:15 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2013-06-22 22:07:45 +00:00
|
|
|
bind_keys(ictx, builtin, section, keys, num_keys, command, cur_loc);
|
2011-07-20 01:41:51 +00:00
|
|
|
n_binds++;
|
2012-09-22 04:33:04 +00:00
|
|
|
|
|
|
|
// Print warnings if invalid commands are encountered.
|
2013-09-27 13:30:59 +00:00
|
|
|
talloc_free(mp_input_parse_cmd(ictx, command, cur_loc));
|
2002-01-30 12:46:03 +00:00
|
|
|
}
|
2012-08-01 19:50:24 +00:00
|
|
|
|
2012-10-13 19:09:42 +00:00
|
|
|
talloc_free(cur_loc);
|
|
|
|
|
2012-08-01 19:50:24 +00:00
|
|
|
return n_binds;
|
|
|
|
}
|
|
|
|
|
2013-02-08 22:52:06 +00:00
|
|
|
static int parse_config_file(struct input_ctx *ictx, char *file, bool warn)
|
2012-08-01 19:50:24 +00:00
|
|
|
{
|
2013-12-14 18:50:00 +00:00
|
|
|
int r = 0;
|
|
|
|
void *tmp = talloc_new(NULL);
|
|
|
|
stream_t *s = NULL;
|
|
|
|
|
2013-12-21 19:45:19 +00:00
|
|
|
file = mp_get_user_path(tmp, ictx->global, file);
|
2014-10-29 21:54:03 +00:00
|
|
|
|
2013-12-21 19:36:45 +00:00
|
|
|
s = stream_open(file, ictx->global);
|
2012-08-01 19:50:24 +00:00
|
|
|
if (!s) {
|
2013-09-10 06:29:45 +00:00
|
|
|
MP_ERR(ictx, "Can't open input config file %s.\n", file);
|
2013-12-14 18:50:00 +00:00
|
|
|
goto done;
|
2012-08-01 19:50:24 +00:00
|
|
|
}
|
2014-07-12 19:25:32 +00:00
|
|
|
stream_skip_bom(s);
|
2013-12-14 18:50:00 +00:00
|
|
|
bstr data = stream_read_complete(s, tmp, 1000000);
|
2013-12-28 14:06:43 +00:00
|
|
|
if (data.start) {
|
|
|
|
MP_VERBOSE(ictx, "Parsing input config file %s\n", file);
|
|
|
|
int num = parse_config(ictx, false, data, file, NULL);
|
|
|
|
MP_VERBOSE(ictx, "Input config file %s parsed: %d binds\n", file, num);
|
|
|
|
r = 1;
|
|
|
|
} else {
|
|
|
|
MP_ERR(ictx, "Error reading input config file %s\n", file);
|
|
|
|
}
|
2013-12-14 18:50:00 +00:00
|
|
|
|
|
|
|
done:
|
|
|
|
free_stream(s);
|
|
|
|
talloc_free(tmp);
|
|
|
|
return r;
|
2002-01-30 12:46:03 +00:00
|
|
|
}
|
|
|
|
|
2013-09-10 06:29:45 +00:00
|
|
|
struct input_ctx *mp_input_init(struct mpv_global *global)
|
2008-04-30 04:15:52 +00:00
|
|
|
{
|
2013-07-27 19:26:00 +00:00
|
|
|
|
2008-04-30 04:15:52 +00:00
|
|
|
struct input_ctx *ictx = talloc_ptrtype(NULL, ictx);
|
|
|
|
*ictx = (struct input_ctx){
|
2013-12-21 18:33:45 +00:00
|
|
|
.global = global,
|
2014-09-27 13:47:49 +00:00
|
|
|
.opts = talloc_zero(ictx, struct input_opts), // replaced later
|
2008-04-30 04:15:52 +00:00
|
|
|
.ar_state = -1,
|
2014-09-27 13:47:49 +00:00
|
|
|
.log = mp_log_new(ictx, global->log, "input"),
|
2013-08-30 00:24:14 +00:00
|
|
|
.mouse_section = "default",
|
2008-04-30 04:15:52 +00:00
|
|
|
};
|
2013-08-26 17:15:35 +00:00
|
|
|
|
2014-09-10 01:16:43 +00:00
|
|
|
if (sem_init(&ictx->wakeup, 0, 0)) {
|
|
|
|
MP_FATAL(ictx, "mpv doesn't work on systems without POSIX semaphores.\n");
|
|
|
|
abort();
|
|
|
|
}
|
|
|
|
|
2014-01-31 18:50:25 +00:00
|
|
|
mpthread_mutex_init_recursive(&ictx->mutex);
|
2013-08-26 17:15:35 +00:00
|
|
|
|
input: don't deliver mouse events if mouse area is not set
This caused the OSC to be always visible at startup on X11:
- EnterNotify event send a mouse event to input.c
- OSC has not completely initialized yet, and no mouse area is set
- mouse event is dispatched to "showhide" OSC section
- OSC becomes visible, regardless of mouse position
Fix this by treating the mouse area as empty if it's not set, instead of
infinite as it was before this commit. This means an input section must
set a mouse area to receive mouse events at all. We also have to change
the default section to receive mouse events with the new behavior.
Also, if MOUSE_MOVE is unmapped (or mapped to something that doesn't
parse), and produces no command, the mouse position wouldn't be updated
(because the mouse position is bound to input commands), so we have to
generate a dummy command in this case.
(This matters only for the OSC, On Screen Controller, which isn't merged
yet, so these changes shouldn't have much effect right now.)
2013-09-05 21:58:51 +00:00
|
|
|
// Setup default section, so that it does nothing.
|
|
|
|
mp_input_enable_section(ictx, NULL, MP_INPUT_ALLOW_VO_DRAGGING |
|
|
|
|
MP_INPUT_ALLOW_HIDE_CURSOR);
|
2012-08-01 20:45:35 +00:00
|
|
|
|
2014-09-27 13:47:49 +00:00
|
|
|
return ictx;
|
|
|
|
}
|
|
|
|
|
|
|
|
void mp_input_load(struct input_ctx *ictx)
|
|
|
|
{
|
2014-09-27 14:25:29 +00:00
|
|
|
struct input_opts *input_conf =
|
|
|
|
m_sub_options_copy(ictx, &input_config, ictx->global->opts->input_opts);
|
2014-09-27 13:47:49 +00:00
|
|
|
|
2014-09-27 14:25:29 +00:00
|
|
|
talloc_free(ictx->opts);
|
2014-09-27 13:47:49 +00:00
|
|
|
ictx->opts = input_conf;
|
|
|
|
|
2013-09-01 15:06:11 +00:00
|
|
|
// "Uncomment" the default key bindings in etc/input.conf and add them.
|
|
|
|
// All lines that do not start with '# ' are parsed.
|
|
|
|
bstr builtin = bstr0(builtin_input_conf);
|
|
|
|
while (builtin.len) {
|
|
|
|
bstr line = bstr_getline(builtin, &builtin);
|
|
|
|
bstr_eatstart0(&line, "#");
|
|
|
|
if (!bstr_startswith0(line, " "))
|
|
|
|
parse_config(ictx, true, line, "<builtin>", NULL);
|
|
|
|
}
|
2008-04-30 04:15:52 +00:00
|
|
|
|
2013-02-08 22:52:06 +00:00
|
|
|
bool config_ok = false;
|
|
|
|
if (input_conf->config_file)
|
|
|
|
config_ok = parse_config_file(ictx, input_conf->config_file, true);
|
2014-09-27 13:47:49 +00:00
|
|
|
if (!config_ok && ictx->global->opts->load_config) {
|
2011-04-25 08:38:46 +00:00
|
|
|
// Try global conf dir
|
2014-10-29 21:54:03 +00:00
|
|
|
void *tmp = talloc_new(NULL);
|
|
|
|
char **files = mp_find_all_config_files(tmp, ictx->global, "input.conf");
|
|
|
|
for (int n = 0; files && files[n]; n++)
|
|
|
|
config_ok = config_ok | parse_config_file(ictx, files[n], false);
|
|
|
|
talloc_free(tmp);
|
2013-02-08 22:52:06 +00:00
|
|
|
}
|
|
|
|
if (!config_ok) {
|
2013-12-21 18:33:45 +00:00
|
|
|
MP_VERBOSE(ictx, "Falling back on default (hardcoded) input config\n");
|
2004-08-03 12:21:14 +00:00
|
|
|
}
|
2002-01-30 12:46:03 +00:00
|
|
|
|
2013-07-16 11:28:28 +00:00
|
|
|
#if HAVE_JOYSTICK
|
2013-12-21 18:33:45 +00:00
|
|
|
if (input_conf->use_joystick)
|
2014-09-09 22:28:53 +00:00
|
|
|
mp_input_joystick_add(ictx, input_conf->js_dev);
|
2002-01-30 12:46:03 +00:00
|
|
|
#endif
|
|
|
|
|
2013-07-16 11:28:28 +00:00
|
|
|
#if HAVE_LIRC
|
2013-12-21 18:33:45 +00:00
|
|
|
if (input_conf->use_lirc)
|
2014-09-09 22:26:45 +00:00
|
|
|
mp_input_lirc_add(ictx, input_conf->lirc_configfile);
|
2002-01-31 09:39:11 +00:00
|
|
|
#endif
|
|
|
|
|
2013-12-01 05:23:39 +00:00
|
|
|
if (input_conf->use_alt_gr) {
|
|
|
|
ictx->using_alt_gr = true;
|
|
|
|
}
|
|
|
|
|
2013-07-16 11:28:28 +00:00
|
|
|
#if HAVE_COCOA
|
2014-10-09 19:23:54 +00:00
|
|
|
if (input_conf->use_app_events) {
|
|
|
|
cocoa_start_event_monitor();
|
|
|
|
}
|
|
|
|
|
2014-04-24 15:27:27 +00:00
|
|
|
if (input_conf->use_appleremote) {
|
2013-06-04 20:12:23 +00:00
|
|
|
cocoa_init_apple_remote();
|
2013-07-27 19:26:00 +00:00
|
|
|
ictx->using_ar = true;
|
2013-06-04 20:12:23 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
if (input_conf->use_media_keys) {
|
|
|
|
cocoa_init_media_keys();
|
2013-07-27 19:26:00 +00:00
|
|
|
ictx->using_cocoa_media_keys = true;
|
2013-06-02 22:52:40 +00:00
|
|
|
}
|
|
|
|
#endif
|
|
|
|
|
2014-09-27 13:47:49 +00:00
|
|
|
ictx->win_drag = ictx->global->opts->allow_win_drag;
|
2014-05-20 00:40:00 +00:00
|
|
|
|
2014-10-15 15:48:47 +00:00
|
|
|
#if defined(__MINGW32__)
|
|
|
|
if (ictx->global->opts->input_file && *ictx->global->opts->input_file)
|
|
|
|
#if HAVE_WAIO
|
|
|
|
mp_input_pipe_add(ictx, ictx->global->opts->input_file);
|
2014-09-14 14:21:04 +00:00
|
|
|
#else
|
|
|
|
MP_ERR(ictx, "Pipes not available.\n");
|
|
|
|
#endif
|
2014-10-15 15:48:47 +00:00
|
|
|
#endif
|
2002-01-31 09:39:11 +00:00
|
|
|
}
|
|
|
|
|
2013-02-03 16:43:47 +00:00
|
|
|
static void clear_queue(struct cmd_queue *queue)
|
|
|
|
{
|
|
|
|
while (queue->first) {
|
|
|
|
struct mp_cmd *item = queue->first;
|
|
|
|
queue_remove(queue, item);
|
|
|
|
talloc_free(item);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2013-07-27 19:26:00 +00:00
|
|
|
void mp_input_uninit(struct input_ctx *ictx)
|
2008-04-30 04:15:52 +00:00
|
|
|
{
|
2008-04-30 06:14:06 +00:00
|
|
|
if (!ictx)
|
|
|
|
return;
|
|
|
|
|
2013-07-16 11:28:28 +00:00
|
|
|
#if HAVE_COCOA
|
2013-07-27 19:26:00 +00:00
|
|
|
if (ictx->using_ar) {
|
2013-06-04 20:12:23 +00:00
|
|
|
cocoa_uninit_apple_remote();
|
|
|
|
}
|
|
|
|
|
2013-07-27 19:26:00 +00:00
|
|
|
if (ictx->using_cocoa_media_keys) {
|
2013-06-04 20:12:23 +00:00
|
|
|
cocoa_uninit_media_keys();
|
2013-06-02 22:52:40 +00:00
|
|
|
}
|
|
|
|
#endif
|
|
|
|
|
input: redo how --input-file is handled
Abandon the "old" infrastructure for --input-file (mp_input_add_fd(),
select() loop, non-blocking reads). Replace it with something that
starts a reader thread, using blocking input.
This is for the sake of Windows. Windows is a truly insane operating
system, and there's not even a way to read a pipe in a non-blocking
way, or to wait for new input in an interruptible way (like with
poll()). And unfortunately, some want to use pipe to send input to
mpv. There are probably (slightly) better IPC mechanisms available
on Windows, but for the sake of platform uniformity, make this work
again for now.
On Vista+, CancelIoEx() could probably be used. But there's no way on
XP. Also, that function doesn't work on wine, making development
harder. We could forcibly terminate the thread, which might work, but
is unsafe. So what we do is starting a thread, and if we don't want
the pipe input anymore, we just abandon the thread. The thread might
remain blocked forever, but if we exit the process, the kernel will
forcibly kill it. On Unix, just use poll() to handle this.
Unfortunately the code is pretty crappy, but it's ok, because it's late
and I wanted to stop working on this an hour ago.
Tested on wine; might not work on a real Windows.
2014-08-24 21:50:43 +00:00
|
|
|
close_input_sources(ictx);
|
2013-09-01 23:09:52 +00:00
|
|
|
clear_queue(&ictx->cmd_queue);
|
input: handle mouse movement differently
Before this commit, mouse movement events emitted a special command
("set_mouse_pos"), which was specially handled in command.c. This was
once special-cased to the dvdnav and menu code, and did nothing after
libmenu and dvdnav were removed.
Change it so that mouse movement triggers a pseudo-key ("MOUSE_MOVE"),
which then can be bound to an arbitrary command. The mouse position is
now managed in input.c. A command which actually needs the mouse
position can use either mp_input_get_mouse_pos() or mp_get_osd_mouse_pos()
to query it. The former returns raw window-space coordinates, while the
latter returns coordinates transformed to OSD- space. (Both are the same
for most VOs, except vo_xv and vo_x11, which can't render OSD in
window-space. These require extra code for mapping mouse position.)
As of this commit, there is still nothing that uses mouse movement, so
MOUSE_MOVE is mapped to "ignore" to silence warnings when moving the
mouse (much like MOUSE_BTN0).
Extend the concept of input sections. Allow multiple sections to be
active at once, and organize them as stack. Bindings from the top of
the stack are preferred to lower ones.
Each section has a mouse input section associated, inside which mouse
events are associated with the bindings. If the mouse pointer is
outside of a section's mouse area, mouse events will be dispatched to
an input section lower on the stack of active sections. This is intended
for scripting, which is to be added later. Two scripts could occupy
different areas of the screen without conflicting with each other. (If
it turns out that this mechanism is useless, we'll just remove it
again.)
2013-04-26 00:13:30 +00:00
|
|
|
talloc_free(ictx->current_down_cmd);
|
2013-11-28 18:28:38 +00:00
|
|
|
pthread_mutex_destroy(&ictx->mutex);
|
2014-09-10 01:16:43 +00:00
|
|
|
sem_destroy(&ictx->wakeup);
|
2011-04-25 08:38:46 +00:00
|
|
|
talloc_free(ictx);
|
2002-01-30 12:46:03 +00:00
|
|
|
}
|
|
|
|
|
stream: redo playback abort handling
This mechanism originates from MPlayer's way of dealing with blocking
network, but it's still useful. On opening and closing, mpv waits for
network synchronously, and also some obscure commands and use-cases can
lead to such blocking. In these situations, the stream is asynchronously
forced to stop by "interrupting" it.
The old design interrupting I/O was a bit broken: polling with a
callback, instead of actively interrupting it. Change the direction of
this. There is no callback anymore, and the player calls
mp_cancel_trigger() to force the stream to return.
libavformat (via stream_lavf.c) has the old broken design, and fixing it
would require fixing libavformat, which won't happen so quickly. So we
have to keep that part. But everything above the stream layer is
prepared for a better design, and more sophisticated methods than
mp_cancel_test() could be easily introduced.
There's still one problem: commands are still run in the central
playback loop, which we assume can block on I/O in the worst case.
That's not a problem yet, because we simply mark some commands as being
able to stop playback of the current file ("quit" etc.), so input.c
could abort playback as soon as such a command is queued. But there are
also commands abort playback only conditionally, and the logic for that
is in the playback core and thus "unreachable". For example,
"playlist_next" aborts playback only if there's a next file. We don't
want it to always abort playback.
As a quite ugly hack, abort playback only if at least 2 abort commands
are queued - this pretty much happens only if the core is frozen and
doesn't react to input.
2014-09-13 12:23:08 +00:00
|
|
|
void mp_input_set_cancel(struct input_ctx *ictx, struct mp_cancel *cancel)
|
2008-04-30 04:15:52 +00:00
|
|
|
{
|
2013-08-26 17:15:35 +00:00
|
|
|
input_lock(ictx);
|
stream: redo playback abort handling
This mechanism originates from MPlayer's way of dealing with blocking
network, but it's still useful. On opening and closing, mpv waits for
network synchronously, and also some obscure commands and use-cases can
lead to such blocking. In these situations, the stream is asynchronously
forced to stop by "interrupting" it.
The old design interrupting I/O was a bit broken: polling with a
callback, instead of actively interrupting it. Change the direction of
this. There is no callback anymore, and the player calls
mp_cancel_trigger() to force the stream to return.
libavformat (via stream_lavf.c) has the old broken design, and fixing it
would require fixing libavformat, which won't happen so quickly. So we
have to keep that part. But everything above the stream layer is
prepared for a better design, and more sophisticated methods than
mp_cancel_test() could be easily introduced.
There's still one problem: commands are still run in the central
playback loop, which we assume can block on I/O in the worst case.
That's not a problem yet, because we simply mark some commands as being
able to stop playback of the current file ("quit" etc.), so input.c
could abort playback as soon as such a command is queued. But there are
also commands abort playback only conditionally, and the logic for that
is in the playback core and thus "unreachable". For example,
"playlist_next" aborts playback only if there's a next file. We don't
want it to always abort playback.
As a quite ugly hack, abort playback only if at least 2 abort commands
are queued - this pretty much happens only if the core is frozen and
doesn't react to input.
2014-09-13 12:23:08 +00:00
|
|
|
ictx->cancel = cancel;
|
2013-08-26 17:15:35 +00:00
|
|
|
input_unlock(ictx);
|
2002-10-23 14:46:20 +00:00
|
|
|
}
|
2013-05-16 21:17:46 +00:00
|
|
|
|
2013-12-01 05:23:39 +00:00
|
|
|
bool mp_input_use_alt_gr(struct input_ctx *ictx)
|
|
|
|
{
|
|
|
|
return ictx->using_alt_gr;
|
|
|
|
}
|
2013-12-26 16:10:35 +00:00
|
|
|
|
|
|
|
struct mp_cmd *mp_input_parse_cmd(struct input_ctx *ictx, bstr str,
|
|
|
|
const char *location)
|
|
|
|
{
|
|
|
|
return mp_input_parse_cmd_(ictx->log, str, location);
|
|
|
|
}
|
2014-01-04 15:59:22 +00:00
|
|
|
|
2014-10-10 20:58:28 +00:00
|
|
|
void mp_input_run_cmd(struct input_ctx *ictx, const char **cmd)
|
2014-01-04 15:59:22 +00:00
|
|
|
{
|
2014-10-10 20:58:28 +00:00
|
|
|
mp_input_queue_cmd(ictx, mp_input_parse_cmd_strv(ictx->log, cmd));
|
2014-01-04 15:59:22 +00:00
|
|
|
}
|
input: redo how --input-file is handled
Abandon the "old" infrastructure for --input-file (mp_input_add_fd(),
select() loop, non-blocking reads). Replace it with something that
starts a reader thread, using blocking input.
This is for the sake of Windows. Windows is a truly insane operating
system, and there's not even a way to read a pipe in a non-blocking
way, or to wait for new input in an interruptible way (like with
poll()). And unfortunately, some want to use pipe to send input to
mpv. There are probably (slightly) better IPC mechanisms available
on Windows, but for the sake of platform uniformity, make this work
again for now.
On Vista+, CancelIoEx() could probably be used. But there's no way on
XP. Also, that function doesn't work on wine, making development
harder. We could forcibly terminate the thread, which might work, but
is unsafe. So what we do is starting a thread, and if we don't want
the pipe input anymore, we just abandon the thread. The thread might
remain blocked forever, but if we exit the process, the kernel will
forcibly kill it. On Unix, just use poll() to handle this.
Unfortunately the code is pretty crappy, but it's ok, because it's late
and I wanted to stop working on this an hour ago.
Tested on wine; might not work on a real Windows.
2014-08-24 21:50:43 +00:00
|
|
|
|
2014-09-09 20:11:45 +00:00
|
|
|
struct mp_input_src_internal {
|
2014-09-09 21:44:38 +00:00
|
|
|
pthread_t thread;
|
|
|
|
bool thread_running;
|
|
|
|
int wakeup[2];
|
|
|
|
bool init_done;
|
|
|
|
|
2014-09-09 20:11:45 +00:00
|
|
|
char *cmd_buffer;
|
|
|
|
size_t cmd_buffer_size;
|
|
|
|
bool drop;
|
|
|
|
};
|
|
|
|
|
input: redo how --input-file is handled
Abandon the "old" infrastructure for --input-file (mp_input_add_fd(),
select() loop, non-blocking reads). Replace it with something that
starts a reader thread, using blocking input.
This is for the sake of Windows. Windows is a truly insane operating
system, and there's not even a way to read a pipe in a non-blocking
way, or to wait for new input in an interruptible way (like with
poll()). And unfortunately, some want to use pipe to send input to
mpv. There are probably (slightly) better IPC mechanisms available
on Windows, but for the sake of platform uniformity, make this work
again for now.
On Vista+, CancelIoEx() could probably be used. But there's no way on
XP. Also, that function doesn't work on wine, making development
harder. We could forcibly terminate the thread, which might work, but
is unsafe. So what we do is starting a thread, and if we don't want
the pipe input anymore, we just abandon the thread. The thread might
remain blocked forever, but if we exit the process, the kernel will
forcibly kill it. On Unix, just use poll() to handle this.
Unfortunately the code is pretty crappy, but it's ok, because it's late
and I wanted to stop working on this an hour ago.
Tested on wine; might not work on a real Windows.
2014-08-24 21:50:43 +00:00
|
|
|
struct mp_input_src *mp_input_add_src(struct input_ctx *ictx)
|
|
|
|
{
|
|
|
|
input_lock(ictx);
|
2014-09-10 01:16:43 +00:00
|
|
|
if (ictx->num_sources == MP_MAX_SOURCES) {
|
input: redo how --input-file is handled
Abandon the "old" infrastructure for --input-file (mp_input_add_fd(),
select() loop, non-blocking reads). Replace it with something that
starts a reader thread, using blocking input.
This is for the sake of Windows. Windows is a truly insane operating
system, and there's not even a way to read a pipe in a non-blocking
way, or to wait for new input in an interruptible way (like with
poll()). And unfortunately, some want to use pipe to send input to
mpv. There are probably (slightly) better IPC mechanisms available
on Windows, but for the sake of platform uniformity, make this work
again for now.
On Vista+, CancelIoEx() could probably be used. But there's no way on
XP. Also, that function doesn't work on wine, making development
harder. We could forcibly terminate the thread, which might work, but
is unsafe. So what we do is starting a thread, and if we don't want
the pipe input anymore, we just abandon the thread. The thread might
remain blocked forever, but if we exit the process, the kernel will
forcibly kill it. On Unix, just use poll() to handle this.
Unfortunately the code is pretty crappy, but it's ok, because it's late
and I wanted to stop working on this an hour ago.
Tested on wine; might not work on a real Windows.
2014-08-24 21:50:43 +00:00
|
|
|
input_unlock(ictx);
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
char name[80];
|
|
|
|
snprintf(name, sizeof(name), "#%d", ictx->num_sources + 1);
|
|
|
|
struct mp_input_src *src = talloc_ptrtype(NULL, src);
|
|
|
|
*src = (struct mp_input_src){
|
|
|
|
.global = ictx->global,
|
|
|
|
.log = mp_log_new(src, ictx->log, name),
|
|
|
|
.input_ctx = ictx,
|
2014-09-09 21:44:38 +00:00
|
|
|
.in = talloc(src, struct mp_input_src_internal),
|
|
|
|
};
|
|
|
|
*src->in = (struct mp_input_src_internal){
|
|
|
|
.wakeup = {-1, -1},
|
input: redo how --input-file is handled
Abandon the "old" infrastructure for --input-file (mp_input_add_fd(),
select() loop, non-blocking reads). Replace it with something that
starts a reader thread, using blocking input.
This is for the sake of Windows. Windows is a truly insane operating
system, and there's not even a way to read a pipe in a non-blocking
way, or to wait for new input in an interruptible way (like with
poll()). And unfortunately, some want to use pipe to send input to
mpv. There are probably (slightly) better IPC mechanisms available
on Windows, but for the sake of platform uniformity, make this work
again for now.
On Vista+, CancelIoEx() could probably be used. But there's no way on
XP. Also, that function doesn't work on wine, making development
harder. We could forcibly terminate the thread, which might work, but
is unsafe. So what we do is starting a thread, and if we don't want
the pipe input anymore, we just abandon the thread. The thread might
remain blocked forever, but if we exit the process, the kernel will
forcibly kill it. On Unix, just use poll() to handle this.
Unfortunately the code is pretty crappy, but it's ok, because it's late
and I wanted to stop working on this an hour ago.
Tested on wine; might not work on a real Windows.
2014-08-24 21:50:43 +00:00
|
|
|
};
|
|
|
|
|
|
|
|
ictx->sources[ictx->num_sources++] = src;
|
|
|
|
|
|
|
|
input_unlock(ictx);
|
|
|
|
return src;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void close_input_sources(struct input_ctx *ictx)
|
|
|
|
{
|
|
|
|
// To avoid lock-order issues, we first remove each source from the context,
|
|
|
|
// and then destroy it.
|
|
|
|
while (1) {
|
|
|
|
input_lock(ictx);
|
|
|
|
struct mp_input_src *src = ictx->num_sources ? ictx->sources[0] : NULL;
|
|
|
|
input_unlock(ictx);
|
|
|
|
if (!src)
|
|
|
|
break;
|
|
|
|
mp_input_src_kill(src);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
void mp_input_src_kill(struct mp_input_src *src)
|
|
|
|
{
|
|
|
|
if (!src)
|
|
|
|
return;
|
|
|
|
struct input_ctx *ictx = src->input_ctx;
|
|
|
|
input_lock(ictx);
|
|
|
|
for (int n = 0; n < ictx->num_sources; n++) {
|
|
|
|
if (ictx->sources[n] == src) {
|
|
|
|
MP_TARRAY_REMOVE_AT(ictx->sources, ictx->num_sources, n);
|
|
|
|
input_unlock(ictx);
|
2014-09-09 21:44:38 +00:00
|
|
|
write(src->in->wakeup[1], &(char){0}, 1);
|
2014-09-14 14:21:04 +00:00
|
|
|
if (src->cancel)
|
|
|
|
src->cancel(src);
|
2014-09-09 21:44:38 +00:00
|
|
|
if (src->in->thread_running)
|
|
|
|
pthread_join(src->in->thread, NULL);
|
2014-09-14 14:21:04 +00:00
|
|
|
if (src->uninit)
|
|
|
|
src->uninit(src);
|
input: redo how --input-file is handled
Abandon the "old" infrastructure for --input-file (mp_input_add_fd(),
select() loop, non-blocking reads). Replace it with something that
starts a reader thread, using blocking input.
This is for the sake of Windows. Windows is a truly insane operating
system, and there's not even a way to read a pipe in a non-blocking
way, or to wait for new input in an interruptible way (like with
poll()). And unfortunately, some want to use pipe to send input to
mpv. There are probably (slightly) better IPC mechanisms available
on Windows, but for the sake of platform uniformity, make this work
again for now.
On Vista+, CancelIoEx() could probably be used. But there's no way on
XP. Also, that function doesn't work on wine, making development
harder. We could forcibly terminate the thread, which might work, but
is unsafe. So what we do is starting a thread, and if we don't want
the pipe input anymore, we just abandon the thread. The thread might
remain blocked forever, but if we exit the process, the kernel will
forcibly kill it. On Unix, just use poll() to handle this.
Unfortunately the code is pretty crappy, but it's ok, because it's late
and I wanted to stop working on this an hour ago.
Tested on wine; might not work on a real Windows.
2014-08-24 21:50:43 +00:00
|
|
|
talloc_free(src);
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
abort();
|
|
|
|
}
|
|
|
|
|
2014-09-09 21:44:38 +00:00
|
|
|
void mp_input_src_init_done(struct mp_input_src *src)
|
|
|
|
{
|
|
|
|
assert(!src->in->init_done);
|
|
|
|
assert(src->in->thread_running);
|
|
|
|
assert(pthread_equal(src->in->thread, pthread_self()));
|
|
|
|
src->in->init_done = true;
|
|
|
|
mp_rendezvous(&src->in->init_done, 0);
|
|
|
|
}
|
|
|
|
|
|
|
|
static void *input_src_thread(void *ptr)
|
|
|
|
{
|
|
|
|
void **args = ptr;
|
|
|
|
struct mp_input_src *src = args[0];
|
|
|
|
void (*loop_fn)(struct mp_input_src *src, void *ctx) = args[1];
|
|
|
|
void *ctx = args[2];
|
|
|
|
|
2014-10-19 21:32:34 +00:00
|
|
|
mpthread_set_name("input source");
|
|
|
|
|
2014-09-09 21:44:38 +00:00
|
|
|
src->in->thread_running = true;
|
|
|
|
|
|
|
|
loop_fn(src, ctx);
|
|
|
|
|
|
|
|
if (!src->in->init_done)
|
|
|
|
mp_rendezvous(&src->in->init_done, -1);
|
|
|
|
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
int mp_input_add_thread_src(struct input_ctx *ictx, void *ctx,
|
|
|
|
void (*loop_fn)(struct mp_input_src *src, void *ctx))
|
|
|
|
{
|
|
|
|
struct mp_input_src *src = mp_input_add_src(ictx);
|
|
|
|
if (!src)
|
|
|
|
return -1;
|
|
|
|
|
|
|
|
#ifndef __MINGW32__
|
|
|
|
// Always create for convenience.
|
|
|
|
if (mp_make_wakeup_pipe(src->in->wakeup) < 0) {
|
|
|
|
mp_input_src_kill(src);
|
|
|
|
return -1;
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
|
|
|
|
void *args[] = {src, loop_fn, ctx};
|
|
|
|
if (pthread_create(&src->in->thread, NULL, input_src_thread, args)) {
|
|
|
|
mp_input_src_kill(src);
|
|
|
|
return -1;
|
|
|
|
}
|
|
|
|
if (mp_rendezvous(&src->in->init_done, 0) < 0) {
|
|
|
|
mp_input_src_kill(src);
|
|
|
|
return -1;
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
int mp_input_src_get_wakeup_fd(struct mp_input_src *src)
|
|
|
|
{
|
|
|
|
return src->in->wakeup[0];
|
|
|
|
}
|
|
|
|
|
input: redo how --input-file is handled
Abandon the "old" infrastructure for --input-file (mp_input_add_fd(),
select() loop, non-blocking reads). Replace it with something that
starts a reader thread, using blocking input.
This is for the sake of Windows. Windows is a truly insane operating
system, and there's not even a way to read a pipe in a non-blocking
way, or to wait for new input in an interruptible way (like with
poll()). And unfortunately, some want to use pipe to send input to
mpv. There are probably (slightly) better IPC mechanisms available
on Windows, but for the sake of platform uniformity, make this work
again for now.
On Vista+, CancelIoEx() could probably be used. But there's no way on
XP. Also, that function doesn't work on wine, making development
harder. We could forcibly terminate the thread, which might work, but
is unsafe. So what we do is starting a thread, and if we don't want
the pipe input anymore, we just abandon the thread. The thread might
remain blocked forever, but if we exit the process, the kernel will
forcibly kill it. On Unix, just use poll() to handle this.
Unfortunately the code is pretty crappy, but it's ok, because it's late
and I wanted to stop working on this an hour ago.
Tested on wine; might not work on a real Windows.
2014-08-24 21:50:43 +00:00
|
|
|
#define CMD_BUFFER (4 * 4096)
|
|
|
|
|
|
|
|
void mp_input_src_feed_cmd_text(struct mp_input_src *src, char *buf, size_t len)
|
|
|
|
{
|
2014-09-09 20:11:45 +00:00
|
|
|
struct mp_input_src_internal *in = src->in;
|
|
|
|
if (!in->cmd_buffer)
|
|
|
|
in->cmd_buffer = talloc_size(in, CMD_BUFFER);
|
input: redo how --input-file is handled
Abandon the "old" infrastructure for --input-file (mp_input_add_fd(),
select() loop, non-blocking reads). Replace it with something that
starts a reader thread, using blocking input.
This is for the sake of Windows. Windows is a truly insane operating
system, and there's not even a way to read a pipe in a non-blocking
way, or to wait for new input in an interruptible way (like with
poll()). And unfortunately, some want to use pipe to send input to
mpv. There are probably (slightly) better IPC mechanisms available
on Windows, but for the sake of platform uniformity, make this work
again for now.
On Vista+, CancelIoEx() could probably be used. But there's no way on
XP. Also, that function doesn't work on wine, making development
harder. We could forcibly terminate the thread, which might work, but
is unsafe. So what we do is starting a thread, and if we don't want
the pipe input anymore, we just abandon the thread. The thread might
remain blocked forever, but if we exit the process, the kernel will
forcibly kill it. On Unix, just use poll() to handle this.
Unfortunately the code is pretty crappy, but it's ok, because it's late
and I wanted to stop working on this an hour ago.
Tested on wine; might not work on a real Windows.
2014-08-24 21:50:43 +00:00
|
|
|
while (len) {
|
|
|
|
char *next = memchr(buf, '\n', len);
|
|
|
|
bool term = !!next;
|
|
|
|
next = next ? next + 1 : buf + len;
|
|
|
|
size_t copy = next - buf;
|
2014-09-09 20:11:45 +00:00
|
|
|
bool overflow = copy > CMD_BUFFER - in->cmd_buffer_size;
|
|
|
|
if (overflow || in->drop) {
|
|
|
|
in->cmd_buffer_size = 0;
|
|
|
|
in->drop = overflow || !term;
|
input: redo how --input-file is handled
Abandon the "old" infrastructure for --input-file (mp_input_add_fd(),
select() loop, non-blocking reads). Replace it with something that
starts a reader thread, using blocking input.
This is for the sake of Windows. Windows is a truly insane operating
system, and there's not even a way to read a pipe in a non-blocking
way, or to wait for new input in an interruptible way (like with
poll()). And unfortunately, some want to use pipe to send input to
mpv. There are probably (slightly) better IPC mechanisms available
on Windows, but for the sake of platform uniformity, make this work
again for now.
On Vista+, CancelIoEx() could probably be used. But there's no way on
XP. Also, that function doesn't work on wine, making development
harder. We could forcibly terminate the thread, which might work, but
is unsafe. So what we do is starting a thread, and if we don't want
the pipe input anymore, we just abandon the thread. The thread might
remain blocked forever, but if we exit the process, the kernel will
forcibly kill it. On Unix, just use poll() to handle this.
Unfortunately the code is pretty crappy, but it's ok, because it's late
and I wanted to stop working on this an hour ago.
Tested on wine; might not work on a real Windows.
2014-08-24 21:50:43 +00:00
|
|
|
MP_WARN(src, "Dropping overlong line.\n");
|
|
|
|
} else {
|
2014-09-09 20:11:45 +00:00
|
|
|
memcpy(in->cmd_buffer + in->cmd_buffer_size, buf, copy);
|
|
|
|
in->cmd_buffer_size += copy;
|
input: redo how --input-file is handled
Abandon the "old" infrastructure for --input-file (mp_input_add_fd(),
select() loop, non-blocking reads). Replace it with something that
starts a reader thread, using blocking input.
This is for the sake of Windows. Windows is a truly insane operating
system, and there's not even a way to read a pipe in a non-blocking
way, or to wait for new input in an interruptible way (like with
poll()). And unfortunately, some want to use pipe to send input to
mpv. There are probably (slightly) better IPC mechanisms available
on Windows, but for the sake of platform uniformity, make this work
again for now.
On Vista+, CancelIoEx() could probably be used. But there's no way on
XP. Also, that function doesn't work on wine, making development
harder. We could forcibly terminate the thread, which might work, but
is unsafe. So what we do is starting a thread, and if we don't want
the pipe input anymore, we just abandon the thread. The thread might
remain blocked forever, but if we exit the process, the kernel will
forcibly kill it. On Unix, just use poll() to handle this.
Unfortunately the code is pretty crappy, but it's ok, because it's late
and I wanted to stop working on this an hour ago.
Tested on wine; might not work on a real Windows.
2014-08-24 21:50:43 +00:00
|
|
|
buf += copy;
|
|
|
|
len -= copy;
|
|
|
|
if (term) {
|
2014-09-09 20:11:45 +00:00
|
|
|
bstr s = {in->cmd_buffer, in->cmd_buffer_size};
|
input: redo how --input-file is handled
Abandon the "old" infrastructure for --input-file (mp_input_add_fd(),
select() loop, non-blocking reads). Replace it with something that
starts a reader thread, using blocking input.
This is for the sake of Windows. Windows is a truly insane operating
system, and there's not even a way to read a pipe in a non-blocking
way, or to wait for new input in an interruptible way (like with
poll()). And unfortunately, some want to use pipe to send input to
mpv. There are probably (slightly) better IPC mechanisms available
on Windows, but for the sake of platform uniformity, make this work
again for now.
On Vista+, CancelIoEx() could probably be used. But there's no way on
XP. Also, that function doesn't work on wine, making development
harder. We could forcibly terminate the thread, which might work, but
is unsafe. So what we do is starting a thread, and if we don't want
the pipe input anymore, we just abandon the thread. The thread might
remain blocked forever, but if we exit the process, the kernel will
forcibly kill it. On Unix, just use poll() to handle this.
Unfortunately the code is pretty crappy, but it's ok, because it's late
and I wanted to stop working on this an hour ago.
Tested on wine; might not work on a real Windows.
2014-08-24 21:50:43 +00:00
|
|
|
s = bstr_strip(s);
|
|
|
|
struct mp_cmd *cmd= mp_input_parse_cmd_(src->log, s, "<>");
|
|
|
|
if (cmd)
|
|
|
|
mp_input_queue_cmd(src->input_ctx, cmd);
|
2014-09-09 20:11:45 +00:00
|
|
|
in->cmd_buffer_size = 0;
|
input: redo how --input-file is handled
Abandon the "old" infrastructure for --input-file (mp_input_add_fd(),
select() loop, non-blocking reads). Replace it with something that
starts a reader thread, using blocking input.
This is for the sake of Windows. Windows is a truly insane operating
system, and there's not even a way to read a pipe in a non-blocking
way, or to wait for new input in an interruptible way (like with
poll()). And unfortunately, some want to use pipe to send input to
mpv. There are probably (slightly) better IPC mechanisms available
on Windows, but for the sake of platform uniformity, make this work
again for now.
On Vista+, CancelIoEx() could probably be used. But there's no way on
XP. Also, that function doesn't work on wine, making development
harder. We could forcibly terminate the thread, which might work, but
is unsafe. So what we do is starting a thread, and if we don't want
the pipe input anymore, we just abandon the thread. The thread might
remain blocked forever, but if we exit the process, the kernel will
forcibly kill it. On Unix, just use poll() to handle this.
Unfortunately the code is pretty crappy, but it's ok, because it's late
and I wanted to stop working on this an hour ago.
Tested on wine; might not work on a real Windows.
2014-08-24 21:50:43 +00:00
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
2014-09-19 15:36:37 +00:00
|
|
|
|
|
|
|
void mp_input_set_repeat_info(struct input_ctx *ictx, int rate, int delay)
|
|
|
|
{
|
2014-09-19 15:48:06 +00:00
|
|
|
input_lock(ictx);
|
2014-09-27 14:25:29 +00:00
|
|
|
ictx->opts->ar_rate = rate;
|
|
|
|
ictx->opts->ar_delay = delay;
|
2014-09-19 15:48:06 +00:00
|
|
|
input_unlock(ictx);
|
2014-09-19 15:36:37 +00:00
|
|
|
}
|
2014-09-19 15:48:06 +00:00
|
|
|
|