Rather than pinning the time at the center of the audio display, pin the
time at the mouse's position since that's the area the user probably
cares about. If the mouse is not over the audio display (such as if the
user is adjusting the zoom via the sliders rather than the mouse wheel),
just use the center as it used to.
Use boost::filesystem::path for all paths, and std::string for all other
strings, converting to/from wxString as close to the actual uses of wx
as possible.
Where possible, replace the uses of non-UI wxWidgets functionality with
the additions to the standard library in C++11, or the equivalents in
boost.
Move the path token management logic to libaegisub (and rewrite it in
the process).
Add a basic thread pool based on asio and std::thread to libaegisub.
This touches nearly every file in the project and a nontrivial amount of
code had to be rewritten entirely, so there's probably a lot of broken
stuff.
Nearly all of the files including main.h are doing so only for OPT_GET
and friends, which are rather unrelated to the main things that main.h
declares.
When auto scrolling is enabled, scroll the audio display after a click
near an edge to mimic the effect of scrolling to show the new audio
selection (as 2.1.9 did), but hopefully without all the weird
undesirable cases that had.
wx doesn't call the keyboard hook when a window has capture, which
results in hotkeys not working while a mouse button is held down. This
usually doesn't matter much, but in the audio display one might want to
play audio while moving a marker. In general checking for hotkeys in
keydown doesn't work, but in the specific cast of marker dragging
keyboard focus is guaranteed to be on the audio display so it works out.
Originally committed to SVN as r6931.
Char hook events propagate by default, removing the need for the event
filter to make key down events propagate, which was causing some funny
issues.
On Windows, the char hook handler runs before menu accelerators, fixing
a bug where Default context hotkeys would override more specific ones
when they appeared on a menu. Unfortunately, this is not the case on
GTK, so the dumb accelerator-disabling hack is still required.
Originally committed to SVN as r6724.
Clicking on the audio display (to change line timing) now never scrolls
the display, rather than jumping around if the line is too long to fit
on the display.
Dragged markers are now always kept visible in the display, even if auto
scroll is off.
Originally committed to SVN as r6541.
In practice, painting audio from the cache isn't all that expensive, so
the limited refreshing was just introducing places for bugs without
improving performance all that much.
Closes#1432.
Originally committed to SVN as r6439.
Keeping track of audio markers and labels to be shown in the audio
display is not particularly related to AudioController's primary
responsiblity of controlling the playback of audio.
Originally committed to SVN as r6427.
The sample rate of the currently open audio is not something that things
which do not interact with the raw audio data should have to care about,
or even know about.
Originally committed to SVN as r6426.
Thomas Goyne