Oswald Buddenhagen [Tue, 8 Oct 2013 15:43:32 +0000 (17:43 +0200)]
correctly resolve INCLUDEPATH relative to current project
Task-number: QTBUG-24587
Change-Id: I1384403e5c5e3313649fa4abb98e7578cae4677b
Reviewed-by: hjk <hjk121@nokiamail.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Oswald Buddenhagen [Tue, 8 Oct 2013 15:43:20 +0000 (17:43 +0200)]
INCLUDEPATH does *not* act as VPATH for HEADERS
Change-Id: Ibbcc5d40d475d88694b527ef9fab6bc75075d97c
Reviewed-by: hjk <hjk121@nokiamail.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Oswald Buddenhagen [Tue, 8 Oct 2013 15:42:20 +0000 (17:42 +0200)]
don't look for sources in FORMS3
amends
17af244afa7
Change-Id: Ie718df1bed4e1c78390c1a6e76a99f13f5e4f136
Reviewed-by: hjk <hjk121@nokiamail.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Oswald Buddenhagen [Tue, 8 Oct 2013 13:06:29 +0000 (15:06 +0200)]
don't unnecessarily claim that gui is needed
Change-Id: Ie5d2c0d74f3b47e7f5cf0a3d34b84b382f01f7bd
Reviewed-by: Sergio Ahumada <sergio.ahumada@digia.com>
Thiago Macieira [Mon, 7 Oct 2013 21:51:54 +0000 (14:51 -0700)]
Keep the copied files as writable
Mac's strip tool doesn't work if the file is read-only. Some of the
libraries we copy might be read-only (like those installed by
HomeBrew) and that would cause stripping to fail.
Change-Id: Ie8caa853a26794ec97ec84354d9f7bb142cce993
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@digia.com>
Eskil Abrahamsen Blomfeldt [Mon, 7 Oct 2013 08:20:47 +0000 (10:20 +0200)]
androiddeployqt: Don't detect directories as executable files
When looking for an executable on the PATH we would mistakenly
report directories with the same filename (like "ant") as
the correct executable.
Task-number: QTBUG-33907
Change-Id: I7122e6f8da7a3027666e9dfc3040a17c2876e981
Reviewed-by: Paul Olav Tvete <paul.tvete@digia.com>
Friedemann Kleint [Mon, 7 Oct 2013 12:12:57 +0000 (14:12 +0200)]
Qt Designer: Temporarily disable loading of QDeclarativeView plugin.
Avoid symbol clashes between QtQuick1 and QtQml.
Task-number: QTCREATORBUG-10273
Change-Id: I84d5b378b9e246d67d8114f701ba5d012cbf5cff
Reviewed-by: Eike Ziller <eike.ziller@digia.com>
Morten Johan Sørvig [Thu, 15 Nov 2012 06:08:44 +0000 (07:08 +0100)]
Macdeployqt: Deploy QML imports.
Use qmlimportscanner from QtDeclarative to find used
imports.
New option:
-qmldir=<path> : Deploy imports used by .qml files
in the given path. (defaults to ".")
Change-Id: I3efa47557a1b5aae0ad838024ba5fe1a98be3b16
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@digia.com>
Reviewed-by: Gabriel de Dietrich <gabriel.dedietrich@digia.com>
Paul Olav Tvete [Tue, 1 Oct 2013 13:59:59 +0000 (15:59 +0200)]
androiddeployqt: make it easier to diagnose problems
Enumerate the enum so we don't have to count the values
each time we need to make sense of the return code.
Change-Id: Ibe168798fb167659688519b2cd833a1894a9deec
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@digia.com>
Paul Olav Tvete [Tue, 1 Oct 2013 13:53:40 +0000 (15:53 +0200)]
Make deployment on Android work again when device is specified
When fixing the case of executable path names with spaces, we
accidentally put an option inside the quotation marks.
Change-Id: Ia3bcf8415d1a936a85fe1b0900d47ffd87daf667
Reviewed-by: Daniel Teske <daniel.teske@digia.com>
Reviewed-by: Christian Stromme <christian.stromme@digia.com>
Oswald Buddenhagen [Mon, 30 Sep 2013 16:11:43 +0000 (18:11 +0200)]
add missing file
amends
35139d45
Change-Id: I53a65e9410a7c6bcd3d28c3b4b416d275679436f
Reviewed-by: hjk <hjk121@nokiamail.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Friedemann Kleint [Mon, 26 Aug 2013 14:09:06 +0000 (16:09 +0200)]
Long live windeployqt!
Add windeployqt for deploying Windows/Windows Runtime applications.
Change-Id: Ia6ca4af13a93fdc75ef6fe4f794cbe228533e85f
Reviewed-by: Oliver Wolff <oliver.wolff@digia.com>
Reviewed-by: Maurice Kalinowski <maurice.kalinowski@digia.com>
Eskil Abrahamsen Blomfeldt [Thu, 26 Sep 2013 07:06:15 +0000 (09:06 +0200)]
androiddeployqt: Fix overriding auto-detected deployment
The platform plugin is set as part of the whole autodetection
step. When this step is not run because deployment is explicitly
set in the .pro file, we should still detect the platform plugin.
The way we do this is by checking which platform plugin is
included in the deployment settings and using this.
Change-Id: I189972a34f03b9059db84bc41855ac662071ca11
Reviewed-by: Paul Olav Tvete <paul.tvete@digia.com>
Eskil Abrahamsen Blomfeldt [Thu, 26 Sep 2013 06:31:44 +0000 (08:31 +0200)]
androideployqt: Allow executable paths with spaces
When the path for an executable contains a space (like when running
jarsigner from "Program Files" on Windows, we need to quote it before
passing it to the shell. We just quote all executables to make it simple.
Change-Id: I1a387c2d846ed0f0ee6aca36b7414b0da827dde3
Reviewed-by: Paul Olav Tvete <paul.tvete@digia.com>
ABBAPOH [Mon, 23 Sep 2013 16:58:40 +0000 (20:58 +0400)]
Add QKeySequenceEdit to Qt Designer
Change-Id: I0efe3f58539fef59154d1c07e6c044bb45fe53cf
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@digia.com>
Reviewed-by: Marc Mutz <marc.mutz@kdab.com>
Friedemann Kleint [Tue, 24 Sep 2013 14:19:40 +0000 (16:19 +0200)]
Remove QtAlgorithms usage from Qt Assistant.
QtAlgorithms is getting deprecated,
see http://www.mail-archive.com/development@qt-project.org/msg01603.html
Change-Id: I6f02d91289c3f455d8643b145fb2c6477863056c
Reviewed-by: Giuseppe D'Angelo <giuseppe.dangelo@kdab.com>
Reviewed-by: Karsten Heimrich <karsten.heimrich@digia.com>
Friedemann Kleint [Tue, 24 Sep 2013 08:17:34 +0000 (10:17 +0200)]
Remove QtAlgorithms usage from Qt Designer.
QtAlgorithms is getting deprecated,
see http://www.mail-archive.com/development@qt-project.org/msg01603.html
Change-Id: I2c7057557295a9bbc83b3d7c625b04400d9e6c2a
Reviewed-by: Jarek Kobus <jaroslaw.kobus@digia.com>
Reviewed-by: Giuseppe D'Angelo <giuseppe.dangelo@kdab.com>
Friedemann Kleint [Wed, 25 Sep 2013 09:45:35 +0000 (11:45 +0200)]
Remove QtAlgorithms usage from Qt Linguist and its tools.
QtAlgorithms is getting deprecated,
see http://www.mail-archive.com/development@qt-project.org/msg01603.html
Change-Id: I3d74510693a5144b536cf515077d870e380d47df
Reviewed-by: Giuseppe D'Angelo <giuseppe.dangelo@kdab.com>
Paul Olav Tvete [Tue, 24 Sep 2013 08:30:23 +0000 (10:30 +0200)]
Remove usage of deprecated function qSort
Change-Id: Ic99f7627430546a44a5290f913a35b2ba6d664b5
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@digia.com>
Paul Olav Tvete [Tue, 24 Sep 2013 08:32:32 +0000 (10:32 +0200)]
Work around wrong return code from pclose
When we do not read the process output, pclose returns an error code even
if the process completes successfully. The simple fix is to always
read the output, even if we don't need it.
Change-Id: Idaec75acbdea6a911aede04506cbd403120b3d92
Reviewed-by: Eskil Abrahamsen Blomfeldt <eskil.abrahamsen-blomfeldt@digia.com>
Reviewed-by: Daniel Teske <daniel.teske@digia.com>
Eskil Abrahamsen Blomfeldt [Fri, 30 Aug 2013 06:23:03 +0000 (08:23 +0200)]
Add tool to deploy Qt applications on Android
This tool automates building an Android project with
either bundling deployment or Ministro. It can be used
from the command line and will also be used from Creator
to create packages.
Task-number: QTBUG-32856
Change-Id: I89716127d192618054317d1412f3bb224a4a81a4
Reviewed-by: Daniel Teske <daniel.teske@digia.com>
Sergio Ahumada [Wed, 18 Sep 2013 19:50:01 +0000 (21:50 +0200)]
Merge remote-tracking branch 'origin/stable' into dev
Change-Id: I8f0a0ffd5e83828d3d0910ece4ec0485c85f2d14
Mitch Curtis [Tue, 10 Sep 2013 08:16:43 +0000 (10:16 +0200)]
Fix no such signal error on assistant startup with no webkit.
QObject::connect: No such signal HelpViewer::printRequested() in
qttools/src/assistant/assistant/centralwidget.cpp:626
Change-Id: If7dee3b8d38f659ca037b9d4ed73e01779f07c67
Reviewed-by: Karsten Heimrich <karsten.heimrich@digia.com>
Sergio Ahumada [Fri, 13 Sep 2013 14:43:11 +0000 (16:43 +0200)]
Merge remote-tracking branch 'origin/stable' into dev
Change-Id: Ia3604754fdb32b0e358a635b0bb16d4fa312bf2d
Sune Vuorela [Fri, 5 Jul 2013 12:07:18 +0000 (14:07 +0200)]
Command line client to standard paths
This tool is usable within scripts and for debugging to query
paths and executables and whatever QStandardPaths will report.
Also a few additional outputs.
See also: kde4-config tool.
Change-Id: I1c10257df80da537eaa85354810dcc051211c602
Reviewed-by: David Faure <david.faure@kdab.com>
Jerome Pasion [Thu, 12 Sep 2013 12:50:10 +0000 (14:50 +0200)]
Doc: Fixing "clcik" to "click"
Change-Id: Ia6c59c286e08ccef85c2e7c73ab07073544f5ac3
Reviewed-by: Leena Miettinen <riitta-leena.miettinen@digia.com>
Sergio Ahumada [Fri, 6 Sep 2013 12:33:37 +0000 (14:33 +0200)]
Merge remote-tracking branch 'origin/stable' into dev
Change-Id: Ie422cd4d1e634ac7bbd37e98b6cdffa2e75ab58e
Oswald Buddenhagen [Mon, 2 Sep 2013 14:43:49 +0000 (16:43 +0200)]
print project errors after all. sometimes.
we evaluate prf files precisely, so we need to see their errors, as they
may give a hint why they terminated prematurely.
however, we prefix the errors with "WARNING:" to make them less scary.
Change-Id: Ia14953a5a9dc31af56ad6c338017dd5b85bb4494
Reviewed-by: hjk <hjk121@nokiamail.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Oswald Buddenhagen [Mon, 2 Sep 2013 14:46:10 +0000 (16:46 +0200)]
remove dead code for printing parser/evaluator warnings
we print only errors anyway.
amends
3d960a16dc9
Change-Id: Iac1c69e1ae94229e3702d7e67163f9a60d04decc
Reviewed-by: hjk <hjk121@nokiamail.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Oswald Buddenhagen [Mon, 2 Sep 2013 14:52:47 +0000 (16:52 +0200)]
make CONFIG feature evaluation failure non-fatal in cumulative mode
while we evaluate the features themselves in precise mode (which is the
reason why they can error out), we do not want them to terminate
cumulative project evaluation.
Change-Id: I70f3e1bcb2ca04a70c74ff484749ca92c1cf6372
Reviewed-by: hjk <hjk121@nokiamail.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Oswald Buddenhagen [Mon, 2 Sep 2013 14:59:34 +0000 (16:59 +0200)]
purge phonon config
Change-Id: If6d2d99b04345c29ff3c51616ff913771f3b8c11
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Friedemann Kleint [Fri, 12 Apr 2013 13:24:55 +0000 (15:24 +0200)]
Use QLineEdit's clear button feature in in Qt Designer.
Remove old filter widget.
Change-Id: Id65eee844d85df2d3c5a87f5acc26ba154d202c9
Reviewed-by: David Faure (KDE) <faure@kde.org>
Reviewed-by: Kevin Ottens <kevin.ottens@kdab.com>
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@digia.com>
Frederik Gladhorn [Tue, 27 Aug 2013 20:46:13 +0000 (22:46 +0200)]
Merge remote-tracking branch 'origin/stable' into dev
Change-Id: I5db3f62185e7e99ce2c701debd00a44f40a9f5e5
Takumi Asaki [Thu, 1 Aug 2013 08:09:40 +0000 (17:09 +0900)]
Linguist: Fix layout of message editor for length variants
Task-number: QTBUG-14348
Change-Id: I6475cd87df4bd1fe1729d437fa01f5f46d362395
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Oswald Buddenhagen [Fri, 23 Aug 2013 12:37:10 +0000 (14:37 +0200)]
update the form preview upon switching the current translation
Task-number: QTBUG-7964
Change-Id: I6555f7f1d51329923bfa24b44c048e47c4606e7f
Reviewed-by: hjk <hjk121@nokiamail.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Oswald Buddenhagen [Mon, 26 Aug 2013 16:55:02 +0000 (18:55 +0200)]
don't disable error view just because of lacking location info
amends
94584dcdd
Change-Id: I5fe662a54e140a25894975dbefc74b221b16d3d8
Reviewed-by: hjk <hjk121@nokiamail.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Oswald Buddenhagen [Wed, 21 Aug 2013 08:51:57 +0000 (10:51 +0200)]
actually recognize -drop-tags argument
it's what is documented.
-drop-tag is still supported, both for backwards compat and because it's
kinda intuitive.
Task-number: QTBUG-33092
Change-Id: I8bf1a7873b8f2c39c98e8676433d2fe7fabf62c4
Reviewed-by: hjk <hjk121@nokiamail.com>
Oswald Buddenhagen [Mon, 19 Aug 2013 15:47:40 +0000 (17:47 +0200)]
search in plural forms after all
our search is not fuzzy, so for some languages it makes a lot of sense
to search in the plurals, as they may use different grammatical forms.
Task-number: QTBUG-32218
Change-Id: I9ba2e370f9bcc564e707b2339ab0b1e73935c03d
Reviewed-by: hjk <hjk121@nokiamail.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Oswald Buddenhagen [Mon, 19 Aug 2013 15:41:50 +0000 (17:41 +0200)]
search translator comments of every translation
Change-Id: I45ce7c0d22826ec9d383da65a8018356569ebd3a
Reviewed-by: hjk <hjk121@nokiamail.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Oswald Buddenhagen [Mon, 19 Aug 2013 15:41:27 +0000 (17:41 +0200)]
de-duplicate code paths
Change-Id: I368b0adb7cc96f40465d176a01196697eb09db4c
Reviewed-by: hjk <hjk121@nokiamail.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Oswald Buddenhagen [Mon, 19 Aug 2013 11:04:48 +0000 (13:04 +0200)]
fix search occasionally skipping messages
the bizarre persistence of m_foundWhere would make us skip a hit in a
directly subsequent message if the match was in a field that is searched
before the field the last hit was in.
Task-number: QTBUG-32196
Change-Id: I5007d435cfc3a96bb2b61875a5fc40ac6a23c548
Reviewed-by: hjk <hjk121@nokiamail.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Oswald Buddenhagen [Wed, 24 Jul 2013 17:49:38 +0000 (19:49 +0200)]
change reporting of missing files
use the new parser flags to report all i/o errors directly.
as a notable side effect, the "WARNING" prefix is gone (even though
it is still treated like that, which is mildly insane to start with).
Change-Id: I084375d5e7a3314ae763795f7c318804a9fb84b6
Reviewed-by: Joerg Bornemann <joerg.bornemann@digia.com>
(cherry picked from qtbase/
b4c0d9166e7a8f5dfa5ca052002a5ab318787aa2)
Oswald Buddenhagen [Mon, 19 Aug 2013 16:54:03 +0000 (18:54 +0200)]
collect messages from INSTALL and DEPLOYMENT file lists
Task-number: QTBUG-14056
Change-Id: I1e014885f92b8587462d926e552834915fc1be60
Reviewed-by: hjk <hjk121@nokiamail.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Oswald Buddenhagen [Mon, 19 Aug 2013 16:54:38 +0000 (18:54 +0200)]
fix arguments of main() in test data
not that it would matter, but ...
Change-Id: I798eb118e623cf22c7288cda7b5d806eba8a6065
Reviewed-by: hjk <hjk121@nokiamail.com>
Oswald Buddenhagen [Thu, 8 Aug 2013 16:33:23 +0000 (18:33 +0200)]
add -pro-debug option: trace .pro evaluation (like qmake -d)
Change-Id: I2f6a3d43fa5e65fb6bfa44c17d833d9b8a30e06a
Reviewed-by: hjk <hjk121@nokiamail.com>
Oswald Buddenhagen [Mon, 26 Aug 2013 13:36:34 +0000 (15:36 +0200)]
fix setting the output directory for the qmake evaluator
not so wise to use the project file instead of its parent directory ...
amends
f51f7b7b95
Change-Id: I35b18f81a86d097b82b8184669b0231810c66f73
Reviewed-by: hjk <hjk121@nokiamail.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Oswald Buddenhagen [Wed, 3 Jul 2013 11:54:17 +0000 (13:54 +0200)]
make split_value_list() even less sane again
contrary to what one may expect, it's actually *not* supposed to remove
the meta-characters it interprets.
luckily, this function is not used much any more ...
Change-Id: I2b60f9b173140da78db2b07b596cc2e5f6e6d555
Reviewed-by: Joerg Bornemann <joerg.bornemann@digia.com>
(cherry picked from qtbase/
783181cfc11e995ad678237ebc0025a2a023c88c)
Oswald Buddenhagen [Mon, 12 Aug 2013 19:12:36 +0000 (21:12 +0200)]
avoid deadlock in cache() during spec/cache loading
sync up; doesn't do anything in lupdate.
if we end up calling cache() from within the initialization of the base
context, we cannot wait for for the completion of that initialization
before we proceed, obviously.
Change-Id: If30c6f3665fe423e767373a8821c406b2f5e0eca
Reviewed-by: Joerg Bornemann <joerg.bornemann@digia.com>
Reviewed-by: Daniel Teske <daniel.teske@digia.com>
(cherry picked from qtcreator/
b64b4431c20afd9e39c1463e736f998ef450688f)
Oswald Buddenhagen [Tue, 23 Jul 2013 15:57:44 +0000 (17:57 +0200)]
lock baseEnv in cache()
sync up; doesn't do anything in lupdate.
as we modify the environment, it must be properly locked.
this implies that initFrom() also needs to be called with a lock.
Change-Id: I48bae9af9adaa0518e5a9db0ba08ff057ae14f9f
Reviewed-by: Joerg Bornemann <joerg.bornemann@digia.com>
(cherry picked from qtcreator/
d022a2d19cecb00397c2a215fc4e3bf64b1e627b)
Oswald Buddenhagen [Tue, 23 Jul 2013 15:27:34 +0000 (17:27 +0200)]
always check for baseEnv init failure
sync up; doesn't do anything in lupdate.
creator's file watcher can trigger many parallel, entirely
non-hierarchical project reloads. if there is enough of them to exceed
the thread pool size, some will be serialized already by qtconcurrent,
not by our wait condition. these should notice a faulty spec, too.
Change-Id: I8ce40cb90fbc28045127881d57ec94e125df79af
Reviewed-by: Joerg Bornemann <joerg.bornemann@digia.com>
(cherry picked from qtcreator/
759d0a69dd3973b4785b6f9412f46666a05cdf85)
Chris Stylianou [Fri, 21 Jun 2013 14:02:39 +0000 (15:02 +0100)]
Fix incorrect uname usage.
The POSIX function uname() can return any non-negative value for success.
- Example: Solaris 10 returns 1 on success.
See: http://pubs.opengroup.org/onlinepubs/
009695399/functions/uname.html
Change-Id: Ic3d51f94a31d8b562b28a2d09df3d31b04c77dc8
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Reviewed-by: Daniel Teske <daniel.teske@digia.com>
(cherry picked from qtbase/
e97c37a37ac27f9ab901e974f0a882e81ffd571d)
Oswald Buddenhagen [Tue, 11 Jun 2013 17:06:18 +0000 (19:06 +0200)]
find qmakespecs in source directory of shadow built qt
qt 5.2 does not copy the qmakespecs to the build directory any more.
a fallback to the /get property variant ensures that other build types
continue to work. consequently, this has no effect whatsoever on
installed qt versions.
(based on qtbase/
214b55d5d4c8fdf975b3c36f69bac4b07d8d39d0)
Task-number: QTCREATORBUG-9572
Change-Id: I9edac11f8997fcb0594d0a67419d4733dd4ed86b
Reviewed-by: Mitch Curtis <mitch.curtis@digia.com>
Reviewed-by: Daniel Teske <daniel.teske@digia.com>
(cherry picked from qtcreator/
958672e4607acf654b0fe5f35a385e95a3042f23)
Oswald Buddenhagen [Tue, 11 Jun 2013 14:08:30 +0000 (16:08 +0200)]
introduce /src qmake property variants
this is for shadow builds during build time, where the respective files
are expected in the source dir.
Change-Id: I18dcfbdef99e1562a51dacac333642cae8105ebd
(cherry picked from qtbase/
4531c83376c1a3546305a9fa5203e0a885e8860d)
Reviewed-by: Daniel Teske <daniel.teske@digia.com>
Oswald Buddenhagen [Wed, 5 Jun 2013 13:44:46 +0000 (15:44 +0200)]
remove pointless string conversion
Change-Id: I1640083007ed0c428dfb8ab5ce8b68a7d7e2a749
(cherry picked from qtbase/
d1b1b31c9237e0bd11c61fc54ad45551b7d862da)
Reviewed-by: Daniel Teske <daniel.teske@digia.com>
Oswald Buddenhagen [Mon, 3 Jun 2013 19:59:27 +0000 (21:59 +0200)]
make .qmake.{super,cache,conf} find "anchored" features
these files function as "anchors" for mkspecs/ and features/ directories
used by projects which load these files. ironically, these files didn't
see these feature files themselves.
Change-Id: I590855eb4a9d2c72b9abfcaa431d2f85a719c6e2
(cherry picked from qtbase/
7456dfb4552c4e4f900af9fa8923828a23737280)
Reviewed-by: Daniel Teske <daniel.teske@digia.com>
Oswald Buddenhagen [Mon, 3 Jun 2013 19:56:51 +0000 (21:56 +0200)]
find features/ under mkspecs/ of source & build root
we were finding features/ directly in the source & build root, and we
were finding features/ under mkspecs/ from $QMAKEPATH and other mkspecs
locations, but we omitted the "transitive hull". this was
counterintuitive.
Change-Id: I9823e6606467c98f264c81385250da92311f51ca
(cherry picked from qtbase/
0bf65180b8710e4516496d0bc53a265ff0dd6a01)
Reviewed-by: Daniel Teske <daniel.teske@digia.com>
Reviewed-by: Joerg Bornemann <joerg.bornemann@digia.com>
Oswald Buddenhagen [Fri, 14 Jun 2013 07:50:46 +0000 (09:50 +0200)]
make Makefiles not depend on .qmake.super
sync with qmake. this is a no-op in lupdate.
Change-Id: Id5158d7e272fdee4f4a041fb7c828295a0a86684
Reviewed-by: Joerg Bornemann <joerg.bornemann@digia.com>
(cherry picked from qtbase/
d135a311a34e693d761a3607663b0879be9bbf65)
Oswald Buddenhagen [Mon, 3 Jun 2013 17:22:23 +0000 (19:22 +0200)]
update m_featureRoots less aggressively
it's not necessary to immediately re-calculate the roots after assigning
QMAKE_PLATFORM - it's sufficient to do it on-demand, so merely
invalidate them. this avoids that we re-calculate them multiple times
without ever using them in between while processing specs with
distributed platform assignments.
Change-Id: If508594764811b96a577fc81c5ded34ab0453148
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@digia.com>
(cherry picked from qtcreator/
28df27d924bb407791a76de8159c9ffa6efde283)
Reviewed-by: Joerg Bornemann <joerg.bornemann@digia.com>
Oswald Buddenhagen [Mon, 3 Jun 2013 17:08:10 +0000 (19:08 +0200)]
cache results of feature search
looking up the same files in the same locations over and over again
is a rather significant waste. in particular, looking up the CONFIG
flags that don't correspond with features has a measurable impact on qt
creator's project loading time.
Task-number: QTCREATORBUG-9154
Change-Id: Ibae3d8b7797e706a6416a7d45c77734ab1281b51
Reviewed-by: Daniel Teske <daniel.teske@digia.com>
(cherry picked from qtcreator/
fa27cd79e05aed4ebd16d5648480cc7d48fefd43)
Oswald Buddenhagen [Fri, 31 May 2013 18:04:31 +0000 (20:04 +0200)]
minor optimization: don't concatenate strings needlessly
Change-Id: Iddec1a818ff9f3ad8b12491100883e433e4b8246
Reviewed-by: Daniel Teske <daniel.teske@digia.com>
(cherry picked from qtcreator/
12652c20711fd29dcba62b8d5ba71c077d8bd06c)
Oswald Buddenhagen [Wed, 29 May 2013 18:18:51 +0000 (20:18 +0200)]
implement simple VFS to support caching during project parsing
sync up with qt creator.
this counts as a bugfix, because it makes write_file() and cache() do
something remotely useful.
Change-Id: I833253f81c3159056fab2ff888f293b36cc2ef56
Reviewed-by: Daniel Teske <daniel.teske@digia.com>
Reviewed-by: Joerg Bornemann <joerg.bornemann@digia.com>
(cherry picked from qtbase/
b215176da34661969015e4950815fe8297885163)
(cherry picked from qtcreator/
66802ef8bf7989dc025e34bf91d93576189c483c)
Oswald Buddenhagen [Wed, 29 May 2013 15:12:11 +0000 (17:12 +0200)]
don't unnecessarily use QFileInfo::filePath()
just directly use the filename we constructed it from.
Change-Id: Ia428a2cb4b192fea6bde62dfbf35361bcfc4b21e
Reviewed-by: Daniel Teske <daniel.teske@digia.com>
(cherry picked from qtcreator/
571234786a3ff7e8e3a9220f12d22a9f74f7a53c)
Oswald Buddenhagen [Wed, 3 Jul 2013 14:54:44 +0000 (16:54 +0200)]
don't crash if builddir is suffix of sourcedir (or vice-versa)
Change-Id: I97a6e2ebd51350cbf39c86efa5c26a376c49da95
Reviewed-by: Joerg Bornemann <joerg.bornemann@digia.com>
(cherry picked from qtbase/
b30273ec4a4ce65794fd402c6e3eb21de7bae1b9)
Friedemann Kleint [Fri, 23 Aug 2013 13:41:54 +0000 (15:41 +0200)]
Fix QUiLoader not applying empty string properties.
Introduced by
6b02fc50ec6a5236f628b08ebeca50e62b22f362 .
Task-number: QTBUG-33130
Change-Id: Ida5ac8c3bf043cb04650680a2fc3777eff3b3f3f
Reviewed-by: Jarek Kobus <jaroslaw.kobus@digia.com>
Oswald Buddenhagen [Mon, 19 Aug 2013 10:17:09 +0000 (12:17 +0200)]
prune unused obsolete tests
the code has been absorbed into the proper autotest a long time ago.
Task-number: QTBUG-26510
Change-Id: I199b26331645cea51e91cf9f8be15c3d4628aa7b
Reviewed-by: hjk <hjk121@nokiamail.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Oswald Buddenhagen [Fri, 16 Aug 2013 19:33:22 +0000 (21:33 +0200)]
fix handling of country-less languages like esperanto
it's impossible to instantiate QLocale objects like that, so the
language setting would be simply clobbered. instead, we now convert
between numerical and string codes directly.
unfortunately QLocale has no public api for that. but we are including
private QTranslator headers already anyway, so whatever.
Task-number: QTBUG-14592
Change-Id: I95189c1898aa1fb5520ecf7057521597ed9331f1
Reviewed-by: hjk <hjk121@nokiamail.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Oswald Buddenhagen [Thu, 15 Aug 2013 14:51:46 +0000 (16:51 +0200)]
make TR_EXCLUDE work for #include statements as well
Change-Id: I299043c63c6e8b2f24eeee7d3f03a514b96bca5f
Reviewed-by: hjk <hjk121@nokiamail.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Oswald Buddenhagen [Thu, 8 Aug 2013 16:01:22 +0000 (18:01 +0200)]
don't complain about missing include()s during cumulative evaluation
it produces simply too many false positives.
Change-Id: I3dfa109866450babe5c16f84665ad22024d99e42
Reviewed-by: Joerg Bornemann <joerg.bornemann@digia.com>
Dmitry Shachnev [Fri, 9 Aug 2013 17:08:01 +0000 (21:08 +0400)]
qtconfig: Update for QFont::removeSubstitutions() renaming in Qt 5
Change-Id: I8c6dce1fad8ce4b4dceb25ccb95b7607ef94286f
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@digia.com>
Simon Hausmann [Tue, 13 Aug 2013 08:02:19 +0000 (10:02 +0200)]
Remove unnecessary dependency to qtjsbackend from sync.profile
Change-Id: I48e7799ba86b95e76270b97bc2eecf2c9d36049f
Reviewed-by: Sergio Ahumada <sergio.ahumada@digia.com>
Oswald Buddenhagen [Fri, 9 Aug 2013 10:17:11 +0000 (12:17 +0200)]
execute some loops even in cumulative mode
we execute foreach loops now. this is (mostly) safe nowadays, because
a previous change added precautions against exponential value list
growth, so it's unlikely that two nested loops would keep the cpu busy
for a day as before.
we continue to exclude forever loops and loops with excessive integer
counts.
Task-number: QTBUG-8550
Change-Id: Iaa116086986cc7fd5023834753f791dd205102e5
Reviewed-by: Daniel Teske <daniel.teske@digia.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Oswald Buddenhagen [Wed, 7 Aug 2013 13:54:29 +0000 (15:54 +0200)]
don't print warnings from parser & evaluator
they are pretty much noise in lupdate context.
we still print error messages: the ones from the parser are always
genuine, while the ones from the evaluator are at least occasionally
helpful.
Task-number: QTBUG-28159
Change-Id: I6cdb11b51c0a3f569ec430f98a0068f933db903d
Reviewed-by: Joerg Bornemann <joerg.bornemann@digia.com>
Oswald Buddenhagen [Wed, 7 Aug 2013 13:54:12 +0000 (15:54 +0200)]
don't print project messages/warnings/errors
they are meaningless during the cumulative parse we are doing
Task-number: QTBUG-28159
Change-Id: I501e7e9b0066fc3f36543147d71632b0d20cca08
Reviewed-by: Joerg Bornemann <joerg.bornemann@digia.com>
Frederik Gladhorn [Mon, 12 Aug 2013 10:48:24 +0000 (12:48 +0200)]
Merge remote-tracking branch 'origin/stable' into dev
Conflicts:
.qmake.conf
src/linguist/lupdate/cpp.cpp
Change-Id: Id27ac2ea8438e60a940d4ae3719ad393d7ad88ea
Sergio Ahumada [Fri, 9 Aug 2013 20:39:27 +0000 (22:39 +0200)]
Merge remote-tracking branch 'origin/release' into stable
Change-Id: Ida20e2ebeb6316e2c6916570ce2e166951af9ee2
Topi Reinio [Thu, 8 Aug 2013 13:21:55 +0000 (15:21 +0200)]
Doc: Fix qthelp URL used in Remotecontrol example
Task-number: QTBUG-32877
Change-Id: Iecd50007e5790f0c938febe4e98e36275fe35a80
Reviewed-by: Jerome Pasion <jerome.pasion@digia.com>
Oswald Buddenhagen [Mon, 5 Aug 2013 14:58:39 +0000 (16:58 +0200)]
introduce new message type "vanished"
this state is equivalent to "previously finished, but now obsolete", or
in gettext terms "not fuzzy and obsolete".
this bumps the ts format version to 2.1.
Task-number: QTBUG-12948
Change-Id: I6c89bbfb9b16053b4e1219ec7611cb7a34189d95
Reviewed-by: hjk <hjk121@nokiamail.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Oswald Buddenhagen [Fri, 2 Aug 2013 17:02:55 +0000 (19:02 +0200)]
complain about missing sources in id-based messages
i considered making this optional on the basis that the source strings
may be assigned externally, but in that scenario lupdate wouldn't be
used to start with.
Task-number: QTBUG-11843
Change-Id: I3f09db902462b8e94f7a134f6fe950562b300307
Reviewed-by: hjk <hjk121@nokiamail.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Oswald Buddenhagen [Tue, 6 Aug 2013 13:11:37 +0000 (15:11 +0200)]
fix strings from #included sources not being collected
Task-number: QTBUG-27974
Change-Id: I30edae2db5449605984a118458b1ccce7ad67eb1
Reviewed-by: hjk <hjk121@nokiamail.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Oswald Buddenhagen [Tue, 25 Jun 2013 15:44:35 +0000 (17:44 +0200)]
make TR_EXCLUDE work for SUBDIRS
this is needed for excluding subprojects we have no direct control over,
e.g., git submodules.
note that unlike TR_EXCLUDE for files, this matches SUBDIRS entries, not
actual project file names. wildcards also work, though.
Change-Id: I1d0a0410a28af4a60f045ee5bbac34540c14f0bd
Reviewed-by: Friedemann Kleint <Friedemann.Kleint@digia.com>
Reviewed-by: hjk <hjk121@nokiamail.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Oswald Buddenhagen [Wed, 7 Aug 2013 11:47:13 +0000 (13:47 +0200)]
use correct AND operator
amends
066b08fc62 and
e1f3732197
Change-Id: I248c1b3b85b9857552bfba34fdb5299024bb55eb
Reviewed-by: hjk <hjk121@nokiamail.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Oswald Buddenhagen [Wed, 7 Aug 2013 11:26:15 +0000 (13:26 +0200)]
purge pointless QT_DISABLE_DEPRECATED_BEFORE=0
neither the tool test code nor the lupdate test data need any of it.
Change-Id: I7604596f17308f7d49850f683b8ab84a0bc88131
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Oswald Buddenhagen [Wed, 7 Aug 2013 11:23:51 +0000 (13:23 +0200)]
build without QT_DISABLE_DEPRECATED_BEFORE=0
Change-Id: I76f23c2bb29bba1bcdb9f49cecc7aba36716161f
Reviewed-by: hjk <hjk121@nokiamail.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Stephen Kelly [Tue, 23 Jul 2013 07:22:17 +0000 (09:22 +0200)]
CMake: Use path manipulation features from qtbase.
The qtbase commit
4ce32f66 (Use new path manipulation functions
from qmake., 2013-01-18) used new qmake functions for path manipulation,
and commits
7ac58d1f (Make cmake packages installed to /usr
non-relocatable., 2013-02-11) and
5aa34962 (Handle usr-move without
forcing absolute paths, 2013-04-17) added handling of symlinks which
would otherwise break relative path calculation.
Port those patches to the qttools cmake files.
Task-number: QTBUG-32570
Change-Id: I7d22609c3a03be4b9197dd030d7e9bd6c2e04731
Reviewed-by: Stephen Kelly <stephen.kelly@kdab.com>
Oswald Buddenhagen [Fri, 2 Aug 2013 14:06:01 +0000 (16:06 +0200)]
remove dead files
Change-Id: I3735795bc51ad42ff7ad242221ef1dd174d882f1
Reviewed-by: hjk <hjk121@nokiamail.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Oswald Buddenhagen [Thu, 1 Aug 2013 15:48:58 +0000 (17:48 +0200)]
don't ignore changes to meta data when updating translation units
Task-number: QTBUG-18890
Change-Id: Iceb466105d5fae56da0e2d26cb245fa610fed865
Reviewed-by: hjk <hjk121@nokiamail.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Oswald Buddenhagen [Thu, 1 Aug 2013 15:30:07 +0000 (17:30 +0200)]
reshuffle the code for clarity
less spaghetti, less nesting
Change-Id: Ib60214e23e20c9592f4eba362ea7a55fe92d8805
Reviewed-by: hjk <hjk121@nokiamail.com>
Oswald Buddenhagen [Thu, 1 Aug 2013 14:44:59 +0000 (16:44 +0200)]
de-duplicate extra comments when adding a message multiple times
Change-Id: I1ed80166be3cb53f865c083b0a693148b6000731
Reviewed-by: hjk <hjk121@nokiamail.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Oswald Buddenhagen [Thu, 1 Aug 2013 12:41:08 +0000 (14:41 +0200)]
update all data of existing messages in extend()
when the message identifier (msgid or context/source/disambiguation)
appears in multiple places, the remaining meta data may be inconsistent
between the occurrences. we need to handle partial data on one hand, and
contradicting data on the other.
Task-number: QTBUG-29131
Task-number: QTBUG-32725
Change-Id: I1ed08e42e266b0b5f014400c137abbcec9ffb93d
Reviewed-by: hjk <hjk121@nokiamail.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Oswald Buddenhagen [Thu, 1 Aug 2013 10:05:47 +0000 (12:05 +0200)]
avoid false positives when looking for magic comments
expect meta data only at the beginning of statements.
this required syntax fixes in some unrelated autotests.
Task-number: QTBUG-11866
Change-Id: Ic0c0c150ac8648df6953a510b8eb0c199c8d3ae2
Reviewed-by: hjk <hjk121@nokiamail.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Oswald Buddenhagen [Thu, 1 Aug 2013 09:04:48 +0000 (11:04 +0200)]
make comments not confuse the c++ parser
comments can appear pretty much everywhere. consequently, handling them
as proper tokens is rather tedious (and was of course not done
consistently, leading to numerous bugs).
so take them out of the token stream and handle magic comments "in the
background".
this necessitates that we use match() a bit more sparingly, as we must
not call getToken() after a "tr()-terminating" token too early, as it
might erroneously collect subsequent magic comments too early otherwise.
Task-number: QTBUG-21876
Change-Id: I3d2168f019dfc06b0778142bcd5eb619b61e6b2d
Reviewed-by: hjk <hjk121@nokiamail.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Oswald Buddenhagen [Wed, 31 Jul 2013 14:22:10 +0000 (16:22 +0200)]
cancel complex constructs at end of #define
it's possible that we encounter the end of a #define statement while in
the middle of a complex construct (which is part of a macro, obviously).
we should discard those, as otherwise Weird Things may happen.
Change-Id: Ie9608ad0215e105ab5e5bac110e4118029fa6c15
Reviewed-by: hjk <hjk121@nokiamail.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Oswald Buddenhagen [Wed, 31 Jul 2013 14:21:51 +0000 (16:21 +0200)]
micro-cleanup: exit directly upon unexpected eof
there is not much point in going on.
Change-Id: Ic278941c4b8bbce4ccd93ab92628ab98c2ab3ae2
Reviewed-by: hjk <hjk121@nokiamail.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Oswald Buddenhagen [Wed, 31 Jul 2013 12:45:57 +0000 (14:45 +0200)]
remove pointless additions
yyWord is always empty at this point.
Change-Id: I1bf897c61e25d52978a77a0d6121078ef264b503
Reviewed-by: hjk <hjk121@nokiamail.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Oswald Buddenhagen [Wed, 31 Jul 2013 08:43:32 +0000 (10:43 +0200)]
don't claim that single slashes are comments
Change-Id: Iae5c66ebebe580595a16caa765eeee9f5e97f2c4
Reviewed-by: hjk <hjk121@nokiamail.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Jerome Pasion [Thu, 1 Aug 2013 10:38:41 +0000 (12:38 +0200)]
Doc: Set the \module pages' \qtvariable.
-Tells QDoc the module's QT variable value
Change-Id: I91cbaa6eb75ea181dc1b049d92a8acd97cbf1344
Reviewed-by: Martin Smith <martin.smith@digia.com>
Oswald Buddenhagen [Fri, 26 Jul 2013 17:46:15 +0000 (19:46 +0200)]
fix handling of c'tors with namespaced parent classes
this is achieved by setting yyTokColonSeen under much narrower
conditions (when we are pretty sure that we are seeing the start of a
c'tor's initializer list), which allows us to unset it under much
narrower conditions (when we are pretty sure we are seeing the start of
the c'tor's body), so we now actually *do* skip over qualifications
which would confuse us.
Change-Id: Ic04bad57a08cc66d41094775570d25e7288e6324
Reviewed-by: hjk <hjk121@nokiamail.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Oswald Buddenhagen [Fri, 26 Jul 2013 17:37:41 +0000 (19:37 +0200)]
clear prospectiveContext after "consuming" it
otherwise it would trigger useless assignments later on.
Change-Id: I8af23747c8f1abcadc3fd86810093303db65117e
Reviewed-by: hjk <hjk121@nokiamail.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Oswald Buddenhagen [Fri, 26 Jul 2013 15:54:36 +0000 (17:54 +0200)]
make QT_TR_NOOP work in static initializers, take 2
simply ignoring equal signs turns out to be not so smart: there are
static initializers without braces, which consequently do not look like
function definitions with a weird additional char.
Task-number: QTBUG-9276
Change-Id: I7d200aee7b81f17b42866eca4830be388c8edb64
Reviewed-by: hjk <hjk121@nokiamail.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>
Oswald Buddenhagen [Tue, 30 Jul 2013 10:05:31 +0000 (12:05 +0200)]
ignore bracketed expressions only outside function scope
static initializers obviously cannot appear in scopes, so there is no
reason to suppress parsing tr() statements used for indexing inside
functions.
Task-number: QTBUG-29998
Change-Id: I3a0330e1382cd5a74106328ae6722a02af282f7f
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@digia.com>