* Notification: Drop the fencing from #10083; it never actually helped, and had subtle side-effects we could do without.
* VirtualKeyBoard: Flash on close, otherwise, some of the fast refresh glitches may be burned into the working buffer until a flash. Making sure we flash ourselves prevent it from sticking around on the page ;).
* util: Move `writeToSysfs` to base (i.e., `ffi/util`), as we need it there (and it actually makes more sense there anyway ;p).
* Bump base for https://github.com/koreader/koreader-base/pull/1645, which is where the actual workaround (hopefully) lives.
Re #8414, #9806, #10558
Notification is a toast, so it doesn't stop event popagation.
If we don't disable thoses handlers inside Notification, we get spurious
duplicate handlers being fired ;).
Fix#10461
* Rewrite the loop mechanism to use scheduled tasks instead of a single blocking-ish subprocess.
* Change the actual logic to be more pleasing to the eye, especially on newer devices, as those *may* natively ramp on set; and fix a bad interaction with that behavior that could lead to no ramp at all on ramp down.
* Simplify Generic's Suspend handling to deal with the refresh ordering in a saner manner. The screensaver might be visible a tad longer than before this change before the frontlight actually ramps off.
Re: https://github.com/koreader/koreader/issues/9806#issuecomment-1416827447
Depends on https://github.com/koreader/koreader-base/pull/1576
Includes assorted cosmetics tweaks related to duplicate `setDirty` calls when instantiating widgets that already have an `onShow` handler doing it. (I left widgets doing it in `update` instead of `init` alone, on the assumption that callers *may* be relying on that behavior when updating widgets at runtime. This might actually never matter, and it certainly didn't for ScreenSaverWidget, which is why I removed it from there ;p).
Get rid of the doc & seqtext fields, as they are not actually used (nor
are they particularly useful, the event handler's name should be pretty
self-explanatory).
Also, tweak the key_events documentation to highlight the quirks of the
API, especially as far as array nesting is involved...
Random drive-by cleanup of the declarations of key_events & ges_events
to re-use the existing instance object (now that we know they're sane
;p) for tables with a single member (less GC pressure).
The ultimate goal is for toast widgets (i.e., Notification when flagged as such) to:
* Not stop event propagation
* Close themselves when the event was emitted by user input.
Instead of doing event filtering in UIManager, we simply overload the onGesture & onKey* handlers in Notification to do just that, and just make sure UIManager will *send* those events to toasts, but without affecting the usual semantics of top widget selection and event propagation (which is as simple as just calling handleEvent on them unchecked ;p).
Thanks to @poire-z for the brainstorming in https://github.com/koreader/koreader/issues/9594 ;).
This also happens to fix a bug in which we might have looped on the top widget twice, because of an array vs. hash mishap ;).
Basically:
* Use `extend` for class definitions
* Use `new` for object instantiations
That includes some minor code cleanups along the way:
* Updated `Widget`'s docs to make the semantics clearer.
* Removed `should_restrict_JIT` (it's been dead code since https://github.com/koreader/android-luajit-launcher/pull/283)
* Minor refactoring of LuaSettings/LuaData/LuaDefaults/DocSettings to behave (mostly, they are instantiated via `open` instead of `new`) like everything else and handle inheritance properly (i.e., DocSettings is now a proper LuaSettings subclass).
* Default to `WidgetContainer` instead of `InputContainer` for stuff that doesn't actually setup key/gesture events.
* Ditto for explicit `*Listener` only classes, make sure they're based on `EventListener` instead of something uselessly fancier.
* Unless absolutely necessary, do not store references in class objects, ever; only values. Instead, always store references in instances, to avoid both sneaky inheritance issues, and sneaky GC pinning of stale references.
* ReaderUI: Fix one such issue with its `active_widgets` array, with critical implications, as it essentially pinned *all* of ReaderUI's modules, including their reference to the `Document` instance (i.e., that was a big-ass leak).
* Terminal: Make sure the shell is killed on plugin teardown.
* InputText: Fix Home/End/Del physical keys to behave sensibly.
* InputContainer/WidgetContainer: If necessary, compute self.dimen at paintTo time (previously, only InputContainers did, which might have had something to do with random widgets unconcerned about input using it as a baseclass instead of WidgetContainer...).
* OverlapGroup: Compute self.dimen at *init* time, because for some reason it needs to do that, but do it directly in OverlapGroup instead of going through a weird WidgetContainer method that it was the sole user of.
* ReaderCropping: Under no circumstances should a Document instance member (here, self.bbox) risk being `nil`ed!
* Kobo: Minor code cleanups.
- ButtonTable, ButtonDialog, ButtonDialogTitle
- ConfirmBox, MultiConfirmBox, SkimToWidget
- KeyboardLayoutDialog (and initially move the dialog
down to show the title in landscape mode)
- InputText's Clipboard dialog
Also: Notification: truncate long text
The hyphenation of a word can be changed from its default
by long pressing for 3 seconds and selecting 'Hyphenate'.
These overrides are stored in a per-language file, i.e:
koreader/settings/user-German.hyph.
That seems like a rather terrible idea to beign with, and that that may actually have fatal consequences.
Re #7738
Co-authored-by: Frans de Jonge <fransdejonge@gmail.com>
Notification: adds some functions so it can be used as
a notification manager.
Have various bits of code emitting events that may generate
notifications advertize themselves as the source for following
notifications.
Add a menu to allow selecting some subsets of sources
to show or hide.
* Input: Don't create a new TimeVal object for input frame timestamps, just promote our existing table by assigning it the `TimeVal` metatable.
* TimeVal: Export (const) `zero` & `huge` TimeVal objects, because they're common enough in our codebase. (NOTE: not actually const, that's a Lua 5.4 feature ;p).
* GestureDetector: Explain the behavior of the `last_tevs` & `first_tevs` tables, and why one needs a new object and not the other.
* Speaking of, simplify the copy method for `first_tevs`, because it doesn't need to create a new TimeVal object, we can just reference the original, it's unique and re-assigned for each frame.
68a5fcdb moved them a bit to near the top than before,
and it can overlap with CRE rendering progress bar.
Also makes it nicer when multiple notications are stacked.
* ReaderDictionary: Port delay computations to TimeVal
* ReaderHighlight: Port delay computations to TimeVal
* ReaderView: Port delay computations to TimeVal
* Android: Reset gesture detection state on APP_CMD_TERM_WINDOW.
This prevents potentially being stuck in bogus gesture states when switching apps.
* GestureDetector:
* Port delay computations to TimeVal
* Fixed delay computations to handle time warps (large and negative deltas).
* Simplified timed callback handling to invalidate timers much earlier, preventing accumulating useless timers that no longer have any chance of ever detecting a gesture.
* Fixed state clearing to handle the actual effective slots, instead of hard-coding slot 0 & slot 1.
* Simplified timed callback handling in general, and added support for a timerfd backend for better performance and accuracy.
* The improved timed callback handling allows us to detect and honor (as much as possible) the three possible clock sources usable by Linux evdev events.
The only case where synthetic timestamps are used (and that only to handle timed callbacks) is limited to non-timerfd platforms where input events use
a clock source that is *NOT* MONOTONIC.
AFAICT, that's pretty much... PocketBook, and that's it?
* Input:
* Use the <linux/input.h> FFI module instead of re-declaring every constant
* Fixed (verbose) debug logging of input events to actually translate said constants properly.
* Completely reset gesture detection state on suspend. This should prevent bogus gesture detection on resume.
* Refactored the waitEvent loop to make it easier to comprehend (hopefully) and much more efficient.
Of specific note, it no longer does a crazy select spam every 100µs, instead computing and relying on sane timeouts,
as afforded by switching the UI event/input loop to the MONOTONIC time base, and the refactored timed callbacks in GestureDetector.
* reMarkable: Stopped enforcing synthetic timestamps on input events, as it should no longer be necessary.
* TimeVal:
* Refactored and simplified, especially as far as metamethods are concerned (based on <bsd/sys/time.h>).
* Added a host of new methods to query the various POSIX clock sources, and made :now default to MONOTONIC.
* Removed the debug guard in __sub, as time going backwards can be a perfectly normal occurrence.
* New methods:
* Clock sources: :realtime, :monotonic, :monotonic_coarse, :realtime_coarse, :boottime
* Utility: :tonumber, :tousecs, :tomsecs, :fromnumber, :isPositive, :isZero
* UIManager:
* Ported event loop & scheduling to TimeVal, and switched to the MONOTONIC time base.
This ensures reliable and consistent scheduling, as time is ensured never to go backwards.
* Added a :getTime() method, that returns a cached TimeVal:now(), updated at the top of every UI frame.
It's used throughout the codebase to cadge a syscall in circumstances where we are guaranteed that a syscall would return a mostly identical value,
because very few time has passed.
The only code left that does live syscalls does it because it's actually necessary for accuracy,
and the only code left that does that in a REALTIME time base is code that *actually* deals with calendar time (e.g., Statistics).
* DictQuickLookup: Port delay computations to TimeVal
* FootNoteWidget: Port delay computations to TimeVal
* HTMLBoxWidget: Port delay computations to TimeVal
* Notification: Port delay computations to TimeVal
* TextBoxWidget: Port delay computations to TimeVal
* AutoSuspend: Port to TimeVal
* AutoTurn:
* Fix it so that settings are actually honored.
* Port to TimeVal
* BackgroundRunner: Port to TimeVal
* Calibre: Port benchmarking code to TimeVal
* BookInfoManager: Removed unnecessary yield in the metadata extraction subprocess now that subprocesses get scheduled properly.
* All in all, these changes reduced the CPU cost of a single tap by a factor of ten (!), and got rid of an insane amount of weird poll/wakeup cycles that must have been hell on CPU schedulers and batteries..
Make Notification have toast=true, and UIManager deal
specifically with such widget:
a "toast" widget gets closed by any event, and let
the event be handled by a lower widget.
This should allow us to not wait or tap to get rid
of a notification, and just go on with what we're
doing.
Also make them have a default timeout of 2s,
used with all existing ones.
* Fix HW dithered partial refreshes sometimes behaving as if the refreshed
content had moved a few pixels to the side...
Probably a kernel issue with the alignment fixup in the EPDC?
* Get rid of the legacy coordinates fixup
It shouldn't be necessary anymore.
And I'd rather fix the root cause, anyway.
* Bump base
(https://github.com/koreader/koreader-base/pull/1116)
* Missed a few DIVs in #6224
* Switch all initial highlights to "fast" update
i.e., everything that does an invert
Plus a few other things that refresh small UI elements onTap
Re #3130
* Tweak refreshtype for a number of widgets:
* Fix iconbutton dimen
* Make touchmenu flash on close & initial menu popup. Full-screen on close.
* Use flashing updates when opening/closing dictionary popup. Full-screen on close.
* Switch FileManager to partial.
It's mostly text, and we want flash promotion there.
* Make configdialog & menu flash on exit
* Make FLWidget flash on close
* virtualkeyboard: flash on layout change & popup.
* Potentially not that great workaround to ensure we actually see the
highlights in the FM's chevrons
* Flash when closing BookStatus Widget
* Optimize away a quirk of the dual "fast" update in touchmenu
* Promote updates to flashing slightly more agressively.
* Document what each refreshtype actually does.
With a few guidelines on their optimal usecases.
* Switch remaining scheduleIn(0.0) to nextTick()
* Tighter scheduling timers
Shaving a hundred ms off UI callbacks...
* Cache FFI C Library namespace
* Ask MuPDF to convert pixmaps to BGR on Kobo
Fix#3949
* Mention koxtoolchain in the README
re #3972
* Kindle: Handle *all* fonts via EXT_FONT_DIR instead of bind mounts insanity
* Make black flashes in UI elements user-configurable
(All or nothing).
* Jot down some random KOA2 sysfs path
Changed from 1 second to 2.
I also fixed the problem with blocking the UI when displaying the message. Now notification box is closed after taping anywhere (like InfoMessage).
More closely matches native behavior on REAGL devices.
Closing those widgets should still trigger a partial refresh though,
because we usually get back to the reader, and text, so we want REAGL
;).
This is a larger clean-up of the refresh situation.
The general shift is that refreshes are now mainly triggered by
the (top-level) widgets when they get shown or closed via UIManager.
All refreshes for the widgets when they are in use were handled by
themselves before. This adds the case of showing/closing.
It is the desired result of not having UIManager:show()/:close()
do (full screen) refreshes on its own.
This is a major overhaul of the hardware abstraction layer.
A few notes:
General platform distinction happens in
frontend/device.lua
which will delegate everything else to
frontend/device/<platform_name>/device.lua
which should extend
frontend/device/generic/device.lua
Screen handling is implemented in
frontend/device/screen.lua
which includes the *functionality* to support device specifics.
Actually setting up the device specific functionality, however,
is done in the device specific setup code in the relevant
device.lua file.
The same goes for input handling.
colors were a mixture of 4bpp integers (0=white, 15=black) and
fractional blackness levels (0=white, 1.0=black) before. This is
now unified to use the color specification of the Blitbuffer API.
This should finish the work to make all globals to local variables.
That allows LuaJIT to properly compile things by interning the
references to the relevant parts (rather than looking up globals
all the time which stops a trace).