Commit graph

2399 commits

Author SHA1 Message Date
Tor Lillqvist
26ceda2f15 More specific message 2016-04-13 11:49:00 +03:00
Miklos Vajna
4fac6c0126 DocumentBroker: can pass these by reference
Change-Id: I327e515e39671239b8699cfadcf61e088af68b13
2016-04-13 10:05:10 +02:00
Tor Lillqvist
8f3ec463b8 Change some cppunit assertion messages to match what is actually being asserted
If you can't come up with a meaningful message to use in a
CPPUNIT_ASSERT_MESSAGE(), just use CPPUNIT_ASSERT() instead. These
messages aren't intended for end-users but for developers, so it is
pointless to make them high-level and dumbed-down.
2016-04-13 09:58:18 +03:00
Pranav Kant
8f9d812806 loleaflet: We do not use underscores in identifier names
Change-Id: Ic466bb3067353c9d271ea264a323502a03ed2c2b
2016-04-13 11:47:22 +05:30
Pranav Kant
df0315f317 loleaflet: Change message on status bar when document is saved
... and empty the status bar when document is modified and not
yet saved on server.

Change-Id: I976405783f744b875b9f6ee7e700006bddf5bc6f
2016-04-13 11:45:56 +05:30
Pranav Kant
9ea4624839 loolwsd: Handle socket exceptions while trying to send frame
In case of abnormal termination of session from client-side,
we might still have data being processed in the kit process, for
example, during auto-save. Trying to send such data over an
expired socket towards the client would throw exceptions which
need to be handled, otherwise the auto-save process would not be
successful. Further, the unhandled exception can bring the document
broker in an unstable state with dockey still present in the
object.

Also do not send 'editlock: 0' to a websocket session which is
going to be removed because the session might have been
forcefully terminated from the client-side, in which case sending
data would go nowhere.

Change-Id: I10eb9c818bc81d4db26d5a19dc8bd44f6fbdf32c
2016-04-13 09:56:32 +05:30
Henry Castro
6701dc62ce loolwsd: response bad request 2016-04-12 17:45:25 -04:00
Michael Meeks
25a092d7de Add memory accounting metric to UnitPrefork.
Also add hooks to allow custom LOK <-> WSD message for this test.
2016-04-12 21:58:27 +01:00
Henry Castro
87639db612 loolwsd: test: remove temporary files 2016-04-12 14:34:25 -04:00
Henry Castro
c7abd16399 loolwsd: replace CPPUNIT_ASSERT_MESSAGE -> CPPUNIT_FAIL 2016-04-12 13:56:32 -04:00
Michael Meeks
6a990dfb61 Security bits and test cleanup.
Enforce user being 'lool' for setcap binaries loolmount and loolforkit.

Add warnings if configured without --enable-debug.

Developers should pass --enable-debug to configure.
2016-04-12 17:13:56 +01:00
Pranav Kant
ce64cab64d Update README about Admin changes
... since we have stopped using pipes for admin console.

Change-Id: If2c1814be453cf8c83da16dc78a287c856d9f278
2016-04-12 21:04:54 +05:30
Tor Lillqvist
ec9bc15de8 Make the exit status of run_test.sh reflect its result
Make it match the exit status of the actual ./test program.
2016-04-12 18:29:02 +03:00
Tor Lillqvist
241d421fc7 Kill the "lool_admin_notify.fifo" thing as it is not used
The loolwsd process created it and opened it for reading, but nothing
opened it for writing.

There is still documentation for it in README, that needs to be either
rewritten to match reality or removed.
2016-04-12 18:07:29 +03:00
Tor Lillqvist
745950ceb2 Improve error logging for the fifo creation 2016-04-12 17:57:58 +03:00
Tor Lillqvist
5c6e209567 Log creation and opening of fifos in more detail 2016-04-12 17:49:20 +03:00
Tor Lillqvist
6342c33dda ClientPortNumber is not static any longer 2016-04-12 16:04:59 +03:00
Henry Castro
53aea60ab1 loolwsd: fix -Werror,-Wshadow 2016-04-12 08:32:22 -04:00
Tor Lillqvist
9086f378d8 Pedantry 2016-04-12 14:48:26 +03:00
Tor Lillqvist
8dcd7d22e4 Remove a few more more leftovers from the "broker" terminology 2016-04-12 14:24:47 +03:00
Tor Lillqvist
479d461348 Add a debug option --careerspan to make loolwsd retire after that many seconds
Comes in handly in some testing situations where you don't want to
send a signal to get loolwsd to finish. Option is present only in an
--enable-debug build.
2016-04-12 13:57:45 +03:00
Pranav Kant
9efed29c1d loolwsd: Save using session holding the edit-lock
If the first session used to save does not hold the edit-lock,
all the .uno:Save commands will get converted to dummy messages
and never reach the kit process.

Change-Id: I36cbee778f8c2c5866dcf276cf156fdf9ed8388e
2016-04-12 16:12:23 +05:30
Tor Lillqvist
9c8e4988fa Bin the old --test mode of LOOLWSD
It hardly was relevant any more. The 'connect' test program can be
used for effectively the same kind of testing anyway.
2016-04-12 13:34:33 +03:00
Michael Meeks
cd2f70521b Cleanup configure.ac status messages. 2016-04-12 10:00:44 +01:00
Jan Holesovsky
280fafe03c Add a trivial autogen.sh. 2016-04-12 10:17:20 +02:00
Jan Holesovsky
3d17f3cc4d Add a missing autoheader call. 2016-04-12 10:09:10 +02:00
Tor Lillqvist
d7f107c532 Add m4, and re-order for clarity
Only have under the "Autofoo" comment stuff actually related to
autofoo or configury in general.
2016-04-12 10:49:57 +03:00
Jan Holesovsky
cc3ee1033b Make the autotools steps reliable. 2016-04-12 09:39:53 +02:00
Miklos Vajna
73985418c7 ${systemplate} -> @SYSTEMPLATE_PATH@ to fix make check
Change-Id: I5cb78fa3820c2e48fe071ba2b030b824a6780e10
2016-04-12 08:37:24 +02:00
Tor Lillqvist
255f9304e9 Fix build without --with-lo-path 2016-04-12 09:29:57 +03:00
Miklos Vajna
773dad57ba AdminModel: can pass these by reference
Change-Id: I1c9ff03c06a8de00162cb2987b7218209f514850
2016-04-12 08:09:42 +02:00
Michael Meeks
ea1e2493bf Re-factor tests, move jails & systemplate up a level into loolwsd.
Also generate the test scripts with config.status containing absolute
paths to make them easier to run, and enable the older unit tests.
2016-04-11 22:01:49 +01:00
Tor Lillqvist
0f3fa0e9d4 Don't log the payload inside a CLOSE frame
It is a status code, a 2-byte unsigned integer in network byte order,
potentially followed by a textual reason. But we never include any
specific status codes in the CLOSE frames anyway. (With a Poco-based
WebSocket peer it is always WS_NORMAL_CLOSE, 1000, 0x03 0xE8.)
2016-04-11 17:21:57 +03:00
Tor Lillqvist
ccf73d159d Mention libtoolize, and that building from tarball is simpler
(Is it?)
2016-04-11 17:04:19 +03:00
Tor Lillqvist
9929842a69 It is loolforkit that has capabilities now, not loolwsd 2016-04-11 17:00:11 +03:00
Tor Lillqvist
eb5fcdba93 Consistenly put a colon after the "editlock" in the message sent to the client 2016-04-11 13:31:03 +03:00
Pranav Kant
ef2638298a loleaflet: listen to composition events
Treat 'compositionend' event as 'keypress' event with composited
data.

Change-Id: Ia1d35e3ce2c583259dc6b6179741050ed173fe64
2016-04-11 15:38:19 +05:30
Tor Lillqvist
a1f4ded793 YAGNI: Drop parameter with default value that is never overridden
No non-default value is never passed. Just use the default value as such in the function.
2016-04-11 11:30:54 +03:00
Tor Lillqvist
6729bb7726 PipeReader::readLine() is used only internally by processOnce() 2016-04-11 11:23:33 +03:00
Tor Lillqvist
077a4436df Bin unused function PipeReader::process() 2016-04-11 11:21:18 +03:00
Tor Lillqvist
2623208c8f Rename MAINTENANCE_INTERVAL to something more descriptive 2016-04-11 11:17:27 +03:00
Tor Lillqvist
a598ea869c Use less misleading function and variable names
Also, get rid of one default parameter that was passed a non-default
value only 'internally'.
2016-04-11 10:14:26 +03:00
Miklos Vajna
ee1e036aa9 Util: strcpy -> strncpy
Change-Id: I03f50f4e5e7f404364fdf90197750bf7cd762c9c
2016-04-11 09:12:59 +02:00
Tor Lillqvist
1c9a6400b6 Make the testBadLoad() unit test pass again
The server sends an "editlock: 1" message even if no document has been
loaded. Ignore that while waiting for the "error:" message.
2016-04-11 10:00:45 +03:00
Tor Lillqvist
bde585a879 Make the time log messages make more sense semantically 2016-04-11 09:25:32 +03:00
Tor Lillqvist
9d4aeb86d1 I doubt we need nanosecond precision here in the logging
(As in "Most recent inactivity was 8096.000000 ms ago".)
2016-04-11 09:15:12 +03:00
Tor Lillqvist
dfd1782300 Probably we can drop the "30-second check" log output 2016-04-11 09:09:18 +03:00
Ashod Nakashian
abca88c7cf loolwsd: correction to prefork unit-test
Since WSD now takes into account the fact that
ForKit always spawns one child, the unit-test
shouldn't expect +1 preforked children.

Change-Id: I5cbe9d817a0d2ffdf9fb0953ef85450f7b8b224f
Reviewed-on: https://gerrit.libreoffice.org/23980
Reviewed-by: Ashod Nakashian <ashnakash@gmail.com>
Tested-by: Ashod Nakashian <ashnakash@gmail.com>
2016-04-11 03:12:26 +00:00
Ashod Nakashian
63068a5dda loolwsd: command timeout is now a const
Change-Id: I9115f465fe760f4c388294bfb5e14cbb441c1174
Reviewed-on: https://gerrit.libreoffice.org/23979
Reviewed-by: Ashod Nakashian <ashnakash@gmail.com>
Tested-by: Ashod Nakashian <ashnakash@gmail.com>
2016-04-11 03:12:06 +00:00
Ashod Nakashian
280669c253 loolwsd: save on disconnection
Last client disconnection now correctly issues a save
and waits for the confirmation before tearing down
the sockets, queues and threads.

Change-Id: I28c28d79a17d359e9aa1fe67b983ca9fb592b847
Reviewed-on: https://gerrit.libreoffice.org/23978
Reviewed-by: Ashod Nakashian <ashnakash@gmail.com>
Tested-by: Ashod Nakashian <ashnakash@gmail.com>
2016-04-11 03:11:38 +00:00