summaryrefslogtreecommitdiffstats
path: root/qmake/project.cpp
Commit message (Collapse)AuthorAgeFilesLines
* fix path separator matchingOswald Buddenhagen2010-12-101-2/+2
| | | | use QDir::separator() instead of Option::dir_sep in localOS context.
* fix misleading uppercasing deprecation warningOswald Buddenhagen2010-12-031-3/+2
| | | | | | | | don't complain about uppercasing in the function's name if the real problem is that it's not defined at all. this is also slighly more efficient, as we try to lowercase only as a fallback now. Reviewed-by: joerg
* Fixed namespace issues related to epocroot.cppMiikka Heikkinen2010-11-191-2/+2
| | | | | Task-number: QTBUG-15393 Reviewed-by: axis
* Make qmake to pass all UTF-8 characters unchanged through parser.Miikka Heikkinen2010-11-081-1/+17
| | | | | | | | | | | | | | | | | | | | | | | QMakeProject::parse() calls QString::simplified(), which assumes any byte 0xA0 in multibyte UTF-8 characters is a space character (0x00A0 is unicode character NBSP [non-breaking space]) and replaces those bytes with regular space, corrupting the UTF-8 string. Fixed by temporarily changing all 0xA0 bytes in parser input to another non-space character 0x01A0. This is safe replacement as qmake doesn't accept unicode .pro files, so there should never be actual NBSP or the replacement 0x01A0 characters in a valid .pro file. Note that there are a couple of more uses of QString::simplified() in qmake, but those do not deal with strings that need to be UTF-8 compatible as far as I can tell, so no need to touch them. Cherry picked to 4.7 branch from master branch as part of QTBUG-15068, original commit: 15a7626480b64d85992bed819fe6052e0c5c8fa9 Task-number: QTBUG-15068 Task-number: QTBUG-14357 Reviewed-by: Oswald Buddenhagen
* Fixed many spelling errors.Rohan McGovern2010-10-251-2/+2
|
* don't remove the path from the name of included filesOswald Buddenhagen2010-06-231-1/+0
| | | | | | | | it is not done for the top-level file, either. this is minimally behavior-incompatible, but anyone relying on the old behavior (and thus not using $$basename()) should be shot anyway. :) Reviewed-by: joerg
* add some commentsOswald Buddenhagen2010-06-231-2/+2
|
* remove support for QMAKE_POST_INCLUDE_FILESOswald Buddenhagen2010-06-231-10/+0
| | | | | | | google has no mention of it except for a single orphaned instance in our own source code => trash. Reviewed-by: joerg
* fix prompt() stdin safety checkOswald Buddenhagen2010-06-231-1/+1
| | | | Reviewed-by: joerg
* complain about unescaped backslashesOswald Buddenhagen2010-05-261-0/+6
|
* clarify wording of warning messageOswald Buddenhagen2010-05-201-1/+1
| | | | Reviewed-by: joerg
* close scope while inside assignment only if the last char is a closing braceOswald Buddenhagen2010-05-201-1/+1
| | | | | | the code assumed it anyway and would make a mess if it was wrong Reviewed-by: joerg
* unify QMAKE_QMAKE path separator fixingOswald Buddenhagen2010-05-201-7/+6
| | | | | | | | | | | the value of the variable in Option is only ever accessed via the project variable, so there is no point in early fixing. as it happens, this fixes mingw+sh generating makefiles with the wrong separator, as the fixing is delayed to a point where QMAKE_DIR_SEP was read back into Option. Reviewed-by: joerg
* bye bye QMakeProjectEnvOswald Buddenhagen2010-04-301-41/+5
| | | | | | | | | | | | | | qmake variables would have been exported to the command run by $$system() and - optionally - to the command run by system(). however, this was a unix-only feature and made the kernel barf at the huge environment on older linuxes. as we don't like platform-specific hacks which are unreliable, in particular when a workaround exists (the commands execute shell code after all, so one can inject arbitrary env variables), just blow it away - it was undocumented, after all. Reviewed-by: joerg
* short-cut evaluation inside if() testsOswald Buddenhagen2010-04-301-6/+3
| | | | | | | | this is consistent with the top-level scope evaluation. if() is undocumented and the old behavior would be pretty unexpected for anyone, so i feel free to break compatibility. Reviewed-by: joerg
* eliminate special splitting of INCLUDEPATH and DEPENDPATHOswald Buddenhagen2010-04-301-5/+3
| | | | | | | | | | | it wouldn't work as expected anyway, as the splitting rule (using the semicolon in addition to whitespace) applied only to the string literal from the pro file, but not any expanded [environment] variables, etc. (i.e., where it might make any sense). so just drop it, as it would considerably complicate later optimizations. it wasn't documented anyway ... Reviewed-by: joerg
* do not env-expand cache file pathOswald Buddenhagen2010-04-301-1/+1
| | | | | | | it makes totally no sense - if one wanted to expand env variables, one would let the shell do it. Reviewed-by: joerg
* warn about usage of deprecated variablesOswald Buddenhagen2010-04-301-0/+5
| | | | Reviewed-by: joerg
* warn about using non-lowercased replace $$function()sOswald Buddenhagen2010-04-301-1/+5
| | | | | | planning to kill off that (mis-)feature at some point. Reviewed-by: joerg
* make QMakeProject::isEmpty() consider legacy mappingsOswald Buddenhagen2010-04-301-0/+6
| | | | Reviewed-by: joerg
* fix $$size() not using function-scoped variablesOswald Buddenhagen2010-04-301-2/+1
| | | | | Reviewed-by: Janne Anttila Reviewed-by: joerg
* Merge branch '4.7' of scm.dev.nokia.troll.no:qt/oslo-staging-1 into ↵Qt Continuous Integration System2010-04-181-4/+1
|\ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | 4.7-integration * '4.7' of scm.dev.nokia.troll.no:qt/oslo-staging-1: (71 commits) Autotest: increase wait time to 3s on Windows to bypass 2s-granularity limitation Autotest: update to the new values for the locale do not expand variables in read()'s file name remove ability to use break() a block outside any loop don't add -unix to the qmake command line Increase the timeout for the QNAM getter test to 30 seconds Remove unstable hosts from the list qdoc: Output TOC for more class ref pages. Other fixes to the remote network stresstester Fix SSL connection problem. Make sure we don't try URLs that aren't HTTP or HTTPS Fix menu bar visibility. Add SSL remote host tests Split the remote and the local tests in two, in preparation for SSL tests Add tests for remote hosts Change the way we calculate the average transfer rates. Finish renaming Move these tests to tests/manual. Add a QNetworkAccessManager stresstest Add a non-blocking native function too ...
| * do not expand variables in read()'s file nameOswald Buddenhagen2010-04-161-2/+1
| | | | | | | | | | | | | | it is positively backwards to apply any expansions at such a low level - they have already been applied where necessary. Reviewed-by: mariusSO
| * remove ability to use break() a block outside any loopOswald Buddenhagen2010-04-161-2/+0
| | | | | | | | | | | | | | it is a completely bizarre feature which is nowhere documented and no justification for it is provided anywhere. Reviewed-by: mariusSO
* | Fix option(recursive)Miikka Heikkinen2010-04-151-1/+1
|/ | | | | | | | | QMakeProject::init initializes the whole project, while QMakeProject::reset initializes the parser for a single file. "recursive" needs to apply to the whole project. Task-number: QTBUG-9847 Reviewed-by: Oswald Buddenhagen
* fix cetest build properlyJoerg Bornemann2010-04-121-2/+0
| | | | | | | | | | | | Commit f5b19c173109c53bf3d8167573f7276cf39262d2 broke the build for cetest. Reverting my initial naive attempt to fix this bd5d323373dbaf9d827126b77895da253128c1e5. We're introducing a new define for building qmake without generators. QT_QMAKE_PARSER_ONLY is used for cetest and the qmake COM wrapper of the Visual Studio Add-in. Reviewed-by: ossi
* Added variable QMAKE_TARGET.arch to qmake to be used with MSVC++Zeno Albisser2010-04-061-0/+15
| | | | | | | | The variable QMAKE_TARGET.arch shall be used to determine cross compiling for x64 on a x32 system. Reviewed-by: Marius Storm-Olsen Task-number: QTBUG-9160
* cetest build fixJoerg Bornemann2010-03-251-0/+2
| | | | | | | | Introducing a new define for building qmake without generators. QT_BUILD_QMAKE_NO_GENERATORS is used for cetest and the qmake COM wrapper of the Visual Studio Add-in. Reviewed-by: mauricek
* make the fallback value of QMAKE_QMAKE absoluteOswald Buddenhagen2010-03-031-1/+2
| | | | | | | | inspired by the pbx generator. currently this has no effect, as all generators build their own fallbacks anyway. Reviewed-by: mariusSO
* make the value of QMAKE_QMAKE somewhat less magicOswald Buddenhagen2010-03-021-1/+0
| | | | | | | | | the generators change the value of QMAKE_QMAKE, so it is unwise to "redirect" it to a hidden builtin which is reset each time. in particular, this fixes qmake generating makefiles without an absolute path to qmake itself - the initial quoting of the filename will make the variable "real", so contains() will start working for it.
* give symbian an own platform modeOswald Buddenhagen2010-02-261-85/+14
| | | | | | ... instead of using the gross isForSymbian() magic Reviewed-by: mariusSO
* decouple host platform mode from target platform modeOswald Buddenhagen2010-02-261-0/+45
| | | | | | | | | | | derive the host mode from the generator - this doesn't work *too* well if the mode is different from the real host platform, so it's only for testing. get the target platform mode from the qmakespec, falling back to the host platform mode. Reviewed-by: mariusSO
* make QMAKE_QMAKE and QMAKE_EXT_OBJ magic builtinsOswald Buddenhagen2010-02-261-9/+13
| | | | | | | this will allow setting their actual value depending on a delayed determination of the platform. Reviewed-by: mariusSO
* use QDir::separator() instead of Option::dir_sep where appropriateOswald Buddenhagen2010-02-261-1/+1
| | | | | | | | | in these cases, the separator depends on the *real* host platform, not on the one that will host the build. there are many more cases like that, but that's for (much) later ... Reviewed-by: mariusSO
* sanitize evaluation of OS scopesOswald Buddenhagen2010-02-261-13/+14
| | | | | | | | | this is marginally behavior-incompatible in that adding the name of an OS scope to CONFIG will not make it true any longer. the cleaned up semantics (besides having merit by themselves) will enable optimizations. Reviewed-by: mariusSO
* instead of hard-coding recursion for symbian, add it to the specsOswald Buddenhagen2010-02-261-4/+0
| | | | Reviewed-by: mariusSO
* add possibility to request project recursion from within a pro fileOswald Buddenhagen2010-02-261-2/+19
| | | | | | | | | | | this is the same as specifying -r on the command line, but it can be set in pro and spec files for cases where no recursion would plain not work. the implementation is via a new option() instruction which at some point will be used to set other "modes of operation" as well. for now, only the "recursive" option is recognized. Reviewed-by: mariusSO
* Factored epocRoot implementation out of qmakeGareth Stockwell2010-02-231-2/+3
| | | | | | | | | | | | This function is now implemented in its own source file, rather than being embedded within the Symbian qmake generator. The motivation for this is to allow code to be shared between qmake and configure - the latter needs to determine the epoc root path in order to perform feature detection on Symbian SDKs. Reviewed-by: Miikka Heikkinen (cherry picked from commit 6ebcf2c24b43fdc1d6da50e9d7ec9dd63dd507d7)
* Revert "Factored epocRoot implementation out of qmake"Thiago Macieira2010-02-211-3/+2
| | | | | | | | | This reverts commit 13cb80be958c40077245cbc4b36448a661e30c64. Conflicts: qmake/Makefile.unix qmake/generators/symbian/symmake.cpp
* Merge remote branch 'origin/4.6' into qt-master-from-4.6Thiago Macieira2010-02-201-2/+3
|\ | | | | | | | | | | Conflicts: qmake/Makefile.unix qmake/generators/symbian/symmake.cpp
| * Factored epocRoot implementation out of qmakeGareth Stockwell2010-02-191-2/+3
| | | | | | | | | | | | | | | | | | | | | | This function is now implemented in its own source file, rather than being embedded within the Symbian qmake generator. The motivation for this is to allow code to be shared between qmake and configure - the latter needs to determine the epoc root path in order to perform feature detection on Symbian SDKs. Reviewed-by: Miikka Heikkinen
| * Revert "Factored epocRoot implementation out of qmake"axis2010-02-191-3/+2
| | | | | | | | | | | | This reverts commit 13cb80be958c40077245cbc4b36448a661e30c64. It breaks non-Symbian platforms.
| * Factored epocRoot implementation out of qmakeGareth Stockwell2010-02-181-2/+3
| | | | | | | | | | | | | | | | | | | | | | This function is now implemented in its own source file, rather than being embedded within the Symbian qmake generator. The motivation for this is to allow code to be shared between qmake and configure - the latter needs to determine the epoc root path in order to perform feature detection on Symbian SDKs. Reviewed-by: Miikka Heikkinen
* | remove remainder of mac9 modeOswald Buddenhagen2010-02-121-10/+1
| | | | | | | | Reviewed-by: mariusSO
* | remove the most blatant tmake compat pathsOswald Buddenhagen2010-02-121-8/+1
|/ | | | Reviewed-By: mariusSO
* Fixed QFileInfo::absolutePath() warning when running "qmake -project"Andreas Kling2010-02-031-1/+1
| | | | | Task-number: QTBUG-7176 Reviewed-by: Benjamin Poulain <benjamin.poulain@nokia.com>
* Add command to qmake to do symbian-uid generationThomas Zander2010-01-201-1/+13
| | | | | | | | The UID3 as required by symbian has to be auto-generated when the user doesn't supply one. To allow this to be done in a mkspec we need a function to do the hashing. This method adds that. Reviewed-By: Marius Storm-Olsen
* Update copyright year to 2010Jason McDonald2010-01-061-1/+1
| | | | Reviewed-by: Trust Me
* Refactored SymbianSubdirsMetaMakefileGenerator out of qmake.Miikka Heikkinen2009-10-161-23/+24
| | | | | | | | | There was no need to have SymbianSubdirsMetaMakefileGenerator in cross-platform metamakefile.cpp, so moved the Symbian specific functionality to symmake.cpp as suggested by qmake reviewers. Task-number: QT-822 Reviewed-by: Janne Anttila
* Update license headers again.Jason McDonald2009-09-091-4/+4
| | | | Reviewed-by: Trust Me
generated by cgit v0.12 at 2024-11-23 12:48:23 (GMT)