98a7baf16e
Merge pull request #2945 from stapelberg/i3-config
...
simplify tests which use launch_with_config precisely once
2017-09-14 22:08:36 +02:00
24462f98ae
tests: unflake tests by not starting i3bar
2017-09-14 21:46:20 +02:00
363417e010
tests: remove the (broken) exit_gracefully check
...
I previously tried to fix the check, but could only come up with a fix which
required removing our module pre-loading, which makes the tests considerably
more expensive. Instead, let’s just remove the check.
2017-09-14 21:46:20 +02:00
02786155dc
tests: use i3_config arg instead of precisely one launch_with_config
...
This way, tests are shorter, and i3test’s invocation of launch_with_config
parallelizes work better, using dont_block => 1.
2017-09-14 21:46:20 +02:00
53e7f18710
i3test::XTEST: don’t “use i3test” to avoid clobbering state
...
Before this commit, the $i3_autostart variable was accidentally overridden.
2017-09-14 21:45:48 +02:00
3a31a76b2a
Reorder tests to not use the same number ( #2947 )
...
Distinct numbers make re-running individual tests easier by helping with
tab-completion.
Completeness verified using:
% for i in $(seq 0 600)
do
files=$(ls testcases/t/$(printf "%03d" $i)-*.t 2>&- | wc -l)
[ "$files" != "0" ] && [ "$files" != "1" ] && echo "clash: $i"
done
2017-09-14 17:49:02 +02:00
a5aed10d09
tests: unflake t/257-keypress-group1-fallback.t ( #2946 )
...
fixes #2944
2017-09-14 17:48:37 +02:00
d0b8f93d4f
travis: install xrandr in Docker container for t/533-randr15.t ( #2949 )
2017-09-14 17:48:22 +02:00
75d118203b
Merge pull request #2942 from stapelberg/append
...
Improve append_layout’s invalid JSON handling
2017-09-13 18:38:47 +02:00
100fad741f
Add .editorconfig (see editorconfig.org) ( #2943 )
...
This file uses the same settings which we currently have in our vim modeline,
but can be picked up by many different editors without having to add and
maintain editor-specific modelines in all of our source files.
2017-09-13 18:38:27 +02:00
c45c002bad
validate JSON before loading
...
This commit also introduces slurp() which reads a file in its entirety. Using
this function instead of doing IO in the functions in load_layout.c again and
again makes the code cleaner (fixing at least two memory leaks) and avoids
re-reading the same file 3 times.
related to #2755
2017-09-13 17:41:05 +02:00
72c972a36c
AnyEvent-I3: use Carp for easier debugging (includes stacktraces)
2017-09-13 16:39:44 +02:00
b23e1f644f
Bugfix: free incomplete containers when JSON parsing fails
...
related to #2755
2017-09-13 16:39:13 +02:00
17627a5861
Merge pull request #2935 from orestisf1993/issue-2934
...
Make sure that parse_config input terminates with '\0'
2017-09-13 15:07:21 +02:00
7b0f4abf4f
scalloc parse_config input to make sure it terminates with '\0'
...
Otherwise strchr() can crash for files that don't end with '\n' because
it won't find a null char to terminate at.
Fixes #2934
2017-09-13 15:31:53 +03:00
fd7319683d
Don't insert newline at end of config with launch_with_config
2017-09-13 15:30:37 +03:00
44e4ad52f6
Merge pull request #2920 from CyberShadow/monitor-output-names
...
Consider RandR 1.5's monitors' output names in addition to monitor names
2017-09-13 10:58:38 +02:00
3b75be1334
docs/userguide: Document that i3 can accept RandR output names
2017-09-13 08:23:27 +00:00
9be4199e13
533-randr15.t: Add test for bar output name canonicalization
2017-09-13 08:23:27 +00:00
fbeded23a9
533-randr15.t: Stop hard-coding the output name
...
Refactor away all mentions of DP3.
2017-09-13 08:23:27 +00:00
8a0e5f89f9
533-randr15.t: Add a fake output connected to the fake monitor
...
Add an output ID to the simulated RRGetMonitors reply, then add a
simulated RRGetOutputInfo reply describing the added output.
2017-09-13 08:23:27 +00:00
c6e3fbcf19
testcases/lib: Add inject_randr15_outputinfo argument
...
Allow tests to specify a file name for inject_randr15's
--getoutputinfo_reply command-line parameter.
2017-09-13 08:23:26 +00:00
de1b578f75
inject_randr1.5: Intercept X11 error responses in addition to replies
...
Allow clients to send garbage to the server, then intercept the
server's error response and substitute it with the supplied simulated
reply data.
2017-09-13 08:23:26 +00:00
f88f4fb5d4
inject_randr1.5: Add RRGetOutputInfo reply injection
...
Add a --getoutputinfo_reply switch to indicate a filename containing
the RRGetOutputInfo reply data to inject.
2017-09-13 08:23:26 +00:00
dfc5ad9850
inject_randr1.5: Refactor reading and storing reply buffer to a struct
...
Allows easier introduction of additional reply buffers in upcoming
changes.
2017-09-13 08:23:26 +00:00
6de9cdd96f
ipc: Canonicalize output names in bar configuration
...
Convert the output names specified in the "output" and "tray_output"
fields in bar blocks in i3's configuration to the referred output's
primary name. This allows specifying names other than the primary
output's name in the given fields without changing the IPC protocol.
2017-09-13 08:23:26 +00:00
ae8f3c2679
randr: Look up alternative output names when searching outputs
...
Update get_output_by_name to look at all additional names added by the
change in the previous commit, not just the primary one.
2017-09-13 08:23:26 +00:00
8d51780d13
randr: Register monitors' output names as additional i3 output names
...
In addition to the name of the monitor itself (which is still used as
the i3 output's primary name), register RandR output names associated
with the RandR monitor as alternative i3 output names.
2017-09-13 08:23:26 +00:00
d01a59b922
Store output names as a linked list
...
Currently, only one name is ever added, and only the first name is
ever accessed; actually using the capability to store and access
multiple names comes in the following commits.
2017-09-13 08:23:24 +00:00
30b1ab38b3
Introduce output_primary_name function
...
Currently simply returns output->name, but this will make it easier to
change how output names are stored in the following commits.
Also replace reading output->name with invocations of
output_primary_name. Code which writes output->name is unchanged. Done
using a mostly mechanical replacement of output->name to
output_primary_name(output).
2017-09-13 08:22:57 +00:00
48fbb3e9ef
testsuite: install Module::Install so that AnyEvent-I3/Makefile.PL works ( #2940 )
...
As per https://perlmaven.com/cant-locate-inc-module-install-in-inc , the inc/
directory should not be under version control.
fixes #2914
2017-09-12 22:16:36 +02:00
bcaec96414
Fix userguide bug ( #2932 )
...
Fixes #2931
2017-09-11 21:31:29 +02:00
5e5444613f
Merge pull request #2928 from CyberShadow/pull-20170911-130032
...
docs/hacking-howto: Promote "Using git / sending patches" section
2017-09-11 20:49:47 +02:00
7116bbaa12
docs/hacking-howto: Update section topology
...
- Promote the "How to build?" sub-section to a top-level
section ("Building i3")
- Convert the "Introduction" sub-section as the intro to the remaining
contents of the "Using git / sending patches" section
- Keep "Which branch to use?" as a level-3 sub-section, thus making it
a sub-section of what used to be the "Introduction" sub-section.
2017-09-11 13:16:44 +00:00
e799bda2da
docs/hacking-howto: Promote "How to build?" sub-section
...
Move the "How to build?" sub-section to the top of its parent section.
2017-09-11 13:16:34 +00:00
5c693ec2ae
docs/hacking-howto: Promote "Using git / sending patches" section
...
Move the contents of the "Using git / sending patches" section to the
top of the document.
2017-09-11 13:04:58 +00:00
0f2bce3916
Merge pull request #2926 from stapelberg/unflake-tests
...
Unflake tests
2017-09-10 22:08:05 +02:00
1c94d189c3
tests: run 533-randr15.t at the very end
...
The test runs `xrandr setmonitor`, which will otherwise affect any test
scheduled after 533-randr15.t, causing flakyness in t/217-NET_CURRENT_DESKTOP.t
for example.
2017-09-10 21:03:14 +02:00
06e8f75d52
tests: unflake t/263-edge-borders.t
2017-09-10 21:03:14 +02:00
bed5c9e03a
tests: re-seed random number generator in workers
2017-09-10 21:03:13 +02:00
eba177342f
Revert "Use OVER operator for drawing text ( #2908 )" ( #2925 )
...
This reverts commit 16160462a3
2017-09-10 17:03:09 +02:00
1b18f89422
Merge pull request #2923 from stapelberg/tests-split
...
Testsuite: avoid starting new i3 instances where possible
2017-09-10 13:41:46 +02:00
828c759489
docs/testsuite: Correct Xephyr package name on Arch Linux ( #2913 )
...
The package is called `xorg-server-xephyr`, not `xorg-xserver-xephyr`.
2017-09-10 13:02:10 +02:00
f7565b5f32
Kill windows between tests
2017-09-10 11:31:10 +02:00
94c76d9e30
i3test: add kill_all_windows convenience function
2017-09-10 11:30:56 +02:00
6caf4e1cb5
testcases/Makefile.PL: tell MakeMaker this is a pure-Perl distribution ( #2922 )
...
fixes #2914
2017-09-10 11:25:43 +02:00
65e8c02f2f
Add README.md ( #2910 )
2017-09-10 07:43:53 +02:00
899ffd872f
529-net-wm-desktop: avoid timeout, avoid restarts, split
...
This shaves off almost half a second of the wall-clock time (from 5.9s to 5.6s).
2017-09-09 18:39:56 +02:00
1a2ce3edaf
travis: remove clang-analyze ( #2917 )
...
• The output currently contains a large number of false-positives and — AFAICT —
no actual issues.
• Upstream shows little interest in addressing the long-standing issues with the
TAILQ macros, so the false-positive situation probably won’t change soon:
https://bugs.llvm.org/show_bug.cgi?id=18222
Currently, we largely spend travis CPU cycles on this, for no additional value.
2017-09-09 18:30:17 +02:00
ffd0ebd85c
165-for_window: merge config and re-use i3 instance, split remainder
...
This reduces total test wall-clock time by 1.5s (from 7.5s down to 5.9s).
2017-09-09 18:21:59 +02:00
Michael Stapelberg