Added auto-detection of Git installations on Windows
Change-Id: I42702126ffb550dd8b289a45ee552d5095a1a615
Signed-off-by: Dirk Fauth <dirk.fauth@googlemail.com>
Added external executables to the terminal combobox on the Open Terminal
Dialog (Ctrl+Alt+Shift+T)
Change-Id: Ia9d9a3959fd95b666c4a1ad4474b519103297920
Signed-off-by: Dirk Fauth <dirk.fauth@googlemail.com>
Moved the code and all related configurations regarding showin from
o.e.tm.terminal.connector.local to o.e.tm.terminal.view.ui
Change-Id: If548ae729173c6127f7cf918aa2f783cd6cb3961
Signed-off-by: Dirk Fauth <dirk.fauth@googlemail.com>
Introduced an ISourceProvider to check if external executables are
configured. Renamed the Terminal context menu "Show In" to "Show in
Local Terminal". Extended the visibleWhen check so that the "Show in
Local Terminal" menu is only visible in the context menu if an external
executable is configured. Otherwise it is possible to open the Terminal
via the inclusion of the command in the default "Show In" menu.
Change-Id: I618b2352a8973d6591cf9e673dc4eb01aeec072e
Signed-off-by: Dirk Fauth <dirk.fauth@googlemail.com>
When TM Terminal sees this:
ESC [ 5 n
It replies indicating that the terminal is OK:
ESC [ 0 n
This escape sequence is documented here:
8.3.35 DSR - DEVICE STATUS REPORT. Page 40.
http://www.ecma-international.org/publications/files/ECMA-ST/Ecma-048.pdf
Motivation: I'd like to switch winpty from using DSR(6) to DSR(5) as its
ESC flushing command. When winpty sees a bare ESC, it can either
synthesize a VK_ESCAPE keypress or wait to see if the ESC was part of some
other escape sequence. It tries to deal with this by generating a DSR(6),
which should flush out the remaining bytes of the sequence, if there are
any. DSR(6) is suboptimal, though, because the reply collides with the
sequence some terminals use for F3 with a modifier:
ESC [ nn ; nn R
DSR(5) generates a DSR(0) reply, which does not have the same problem.
DSR(5) is supported by all the terminal emulators I tested, including
xterm, rxvt, rxvt-unicode, gnome-terminal, konsole, putty, mintty,
OS X's Terminal.app, and IntelliJ/jediterm.
Signed-off-by: Ryan Prichard <ryan.prichard@gmail.com>
Mark Terminal Interfaces as @noextend @noimplement where implementation
doesn't make sense (since it's just constants or an abstract base class
exists for extending). This helps giving room for safely extending
Terminal interfaces in a backward-compatible way in the future, by
leveraging Eclipse API Tooling.
Change-Id: Ia5a99f9379ddd73dd58afe4dec4540ae24633b09
Signed-off-by: Martin Oberhuber <martin.oberhuber@windriver.com>
This fixes API related issues and warnings by adding appropriate markup
where it had been forgotten. This is formally breaking binary
compatibility in some cases, but we are OK releasing as 4.1 since it
just codifies the original intent of the API.
1. org.eclipse.tm.terminal.control/MANIFEST.MF:
- Replace unintuitive "x-friends" with more obvious "x-internal".
- Update the terminal.test/.classpath instead for giving access.
- Add @noimplement keywords on interfaces where appropriate.
2. org.eclipse.tm.terminal.view.ui:
- Add missing @since tags, and mark the view.showin package as
"x-internal", since it was never meant to be API.
- Add @noimplement keywords on interfaces where appropriate.
3. Generally, remove unnecessary @suppress("restriction") warnings
and fix one "Type Safety (ArrayList)" warning.
Change-Id: I75f8b0fb3a5ff0383bff3bc0a4a8d9fd07b71ac9
Signed-off-by: Martin Oberhuber <martin.oberhuber@windriver.com>
As discussed in bugzilla, this extension allows terminal connectors to
implement special features on mouse events (such as clicking on hyperlinks).
Feature-request: https://bugs.eclipse.org/bugs/show_bug.cgi?id=475267
Signed-off-by: Davy Landman <davy.landman@cwi.nl>
Enable API Analysis on all Terminal Connectors, and mark
o.e.tm.terminal.control/provisional/api as API in MANIFEST.MF
such that full API Analysis is available on our de-facto API.
Since the conversion of provisional.api to full API triggers errors,
this needs a patched 4.0 baseline to properly compare against. Use
http://archive.eclipse.org/tm/downloads/tm_4.0_api_baseline.zip
for analysis in 4.1 -- after 4.1 a normal baseline will do.
Change-Id: Ibbc2addd59a629d45f676a97339bf27bba07014e
Signed-off-by: Martin Oberhuber <martin.oberhuber@windriver.com>
The "Pin&Clone" concept is percieved complicated by the user. It is
replaced by a more simpler concept where the user has an action to open
a new "Terminals" view on demand. A new console is always opened in the
last active "Terminals" view.
opening the view
Use the secondary id the caller passed in if it's not null or '*',
otherwise use the auto generated secondary id.
Change-Id: I25a884bcb6841cb5fc7e1e6f191800b99e42873a
Signed-off-by: Q.S. Wang <wangqs_eclipse@yahoo.com>
4.Remove the super freference, there's no reason to call that way.
5. Clean up the getDefaultTerminalTitle method in
AbstractLauncherDelegate.
Change-Id: I5211dd26fa6691c666bb457a1ef10e19a190441d
Signed-off-by: Q.S. Wang <wangqs_eclipse@yahoo.com>
