Age | Commit message (Collapse) | Author |
|
|
|
|
|
|
|
|
|
Thanks to Oliver Garcia Albertos <oliverarafo@gmail.com>
|
|
|
|
Alt-sends-esc is no longer the default.
The opposite of alt-sends-esc is to pass through Alt un-altered, which
allows phone keyboards to use alt to provide extra characters.
There's no longer any way to have Alt set the high bit.
|
|
|
|
Conflicts:
libraries/emulatorview/src/jackpal/androidterm/emulatorview/EmulatorView.java
|
|
Re-routing Alt and Meta seems to be required on this device, but not
on other devices (such as Nexus 7 using USB OTG.)
We should refine this test as we find out which devices need this
special handling.
|
|
Re-routing Alt and Meta seems to be required on this device, but not
on other devices (such as Nexus 7 using USB OTG.)
We should refine this test as we find out which devices need this
special handling.
|
|
This patch handles the alt sends ESC in IME mode by routing the key event
through the onKeyDown() and onKeyUp() methods if Alt pressed and
the alt-sends-ESC mode is enabled.
|
|
Added new Alt sends strings to 'hu' localization.
|
|
|
|
|
|
(Found by running fuzzer.go and trimming down the repro case.)
adb push issue145repro-2.txt /data/local/a
Run Android terminal emulator.
$ cat /data/local/a
|
|
Added new strings.
|
|
+ Move to the Keyboard section of preferences.
+ On by default. (and now we actually honor the default.)
+ Rework preference text to make it clearer.
+ Rename string resource id names to make their role clearer.
+ Add support for setting the high bit instead of sending ESC.
- I think there are more people who want this than the
original composing logic.
+ Treat the Meta key the same way as the Alt key.
- I considered having this be a separate preference, but
there's a flaw: Some keyboards such as the popular
Hacker's keyboard have a key labeled "Alt" that actually
reports as "Meta" in software.
So let's treat both Alt and Meta the same for now.
+ Make this code work on Android SDK 11+ physical keyboards
Android SDK 11+ physical keyboards are "chording" as opposed to
"toggling". Therefore we have to handle the !allowToggle case.
Also, the Meta flags can come in as a combination of
3 bits, so to mask them off we need to mask with META_ALT_MASK
rather than just META_ALT_ON.
|
|
|
|
|
|
This will allow the user to choose whether she wants to keep the old
behavior or send a esc character before the string.
|
|
This change attempts full code coverage. However the Fn key
functionality cannot be reached by my emulator version. All the tests
with Function key fail.
|
|
|
|
I think for most users of Android Terminal emulator the default initial
command (which adds /data/local/bin to the path) was not useful,
because the directory /data/local/bin does not exist by default.
The /data/local directory is special because its permissions are set
so that it can be written to by the Android shell even on non-rooted
phones.
Therefore, it is possible for users of non-rooted phones to create the
/data/local/bin directory, and then put executables into that dir,
and then add it to their path, and then execute them as shell commands.
But, in reality, most people who would go to the trouble of installing
command-line applications tend to root their phone and set their
phone up more like a normal Linux system.
So there were very few people (possibly none) for which the default
initial command string was useful.
|
|
|
|
Miscellaneous bug fixes
|
|
Contains a handful of characters from the Supplementary Multilingual
Plane (some musical and mathematical symbols, since I don't know any of
the other scripts in the SMP).
|
|
We currently discover a non-BMP character at the start or end of a
text run only after we process both halves of the surrogate pair
representing the character individually, resulting in off-by-one errors
in calculating the portion of the char array containing the run to be
drawn (we erroneously omit any high surrogate at the start of a run, and
erroneously include any high surrogate immediately following the end of
a run).
Fix this by processing both halves of the surrogate pair immediately
when we see the high surrogate.
|
|
Since an East Asian wide character takes two screen columns, if we're
inserting one into a spot previously occupied by a normal-width
character, we need to overwrite the contents of the column to the right
of the position into which the character is inserted. We do this
shifting the contents of the rest of the line left in the array, then
updating our idea of the storage used -- or at least that's what's
supposed to happen.
There are at least two problems with the current implementation:
(1) findStartOfColumn() is used in a few places to get the spot in the
array where a column starts, but this relies on the offset array, which
is potentially out of date at this point. Fix this by adding the shift
(which will be used to update the offset array later) to the value
whenever we use findStartOfColumn().
(2) When truncating the line by reducing the storage used count, we
currently effectively reduce the count by the size of the last column in
the line, which is not necessarily the same as the size of the
overwritten column. We should instead adjust by the size of the
just-overwritten column (which is readily available to us at this
point).
Fixes #145, and potentially a class of other difficult-to-reproduce bugs
involving East Asian wide character support.
|
|
This occurs, for instance, when we try to retrieve text from the
transcript without caring about color information. Fixes crash when
selecting/copying text introduced by 54d2530... ("Make blank lines track
the current foreground/background color.").
|
|
|
|
This fixes emulation errors related to scrolling with a non-default
background color.
|
|
Should fix Issue 149.
|
|
Keep track of ESC ( x and ESC ) x sequences.
|
|
This reverts commit fd719e10d6b542ebf0be28cb58bbda484c5834b7, reversing
changes made to 113d77e5fa711cfcdc50953b70a7df6803a467d8.
This change was reverted because it caused a regression in the way soft
keyboard modifier keys worked. (e.g. using the volume keys as control
keys.) I think the regression happened because the change introduced
multiple copies of TermKeyListener. There should be only one instance
of TermKeyListener for each terminal window.
I would still like the feature (of having Alt send Esc), but without the
regression to the soft keyboard.
|
|
|
|
(Which is an exception thrown while drawing UTF8 text)
|
|
I'm guessing this broke when the keyboard handling code was re-written to
support sending Alt-as-esc.
|
|
Allows the Hacker Keyboard IME's Alt (which seems in testing to
actually send Meta rather than Alt when I tested it on a Nexus 7
running Android 4.1) to work with the AltKey-sends-escape feature.
|
|
Conflicts:
libraries/emulatorview/src/jackpal/androidterm/emulatorview/EmulatorView.java
This was caused by a rebase interactive after a push.
|
|
Save/restore current text effect, origin mode and wraparound mode.
|
|
|
|
|
|
|
|
Fixes vttest test of cursor movement.
|
|
This is useful for running vttest.
|
|
Fixes issue 156
|
|
(By popular demand. It's more traditional than White text on blue.)
|
|
Add a comment describing what escape sequences we're trying to match.
This patch is noisy because the previous version of this file didn't use
project-standard whitespace. The whitespace has been fixed.
Thanks to Yi Zhang <yiz.cheapwine@gmail.com> for reporting the bug and
providing a patch that fixed it.
|
|
Rename mCharcodes to mCharSequence (matches the public getter's name.)
Rename the public getter from getCharSequence to extractCharSequence, to
signify that each time you call it the current character sequence is
reset.
Bug repro:
Switch to Alt-key-does-not-send-ESC mode.
Press and release the letter X
Hold down the alt key.
Expected behavior: nothing happens.
Actual behavior: The letter x repeats as long as you hold down the Alt key.
|