...so that SemanticInfoUpdater does not depend anymore on the
EditorDocumentParser. Accessing the snapshot was a blocking operation
that delayed the semantic info update longer than actually needed.
Change-Id: I348d22ef83ab310d4319b2e8b9678fe90ee24d6a
Reviewed-by: Erik Verbruggen <erik.verbruggen@digia.com>
The CppRefactoringChanges::fileNoEditor takes a Document from the
snapshot it has. Although this snapshot is a safe copy, it might have
been gotten right after it was updated by indexing. Such a document will
still have its AST. If this AST is used by any refactoring action
without retaining it, the pointers will be dangling after a short while
(specifically: after the locator has extracted all the information).
The fileNoEditor method is called by the declDefLinkFinder to search
the target document. The snapshot is obtained before, and contains the
document for the semantic info. However, the target document will not
come from the semantic info, but from the indexer.
Change-Id: I212ff41dde6910e94e80552b2c3e5911fe9496ae
Task-number: QTCREATORBUG-11262
Reviewed-by: Nikolai Kosjar <nikolai.kosjar@digia.com>
Replace all* remaining deprecated Qt 4 functions with
their Qt 5 counterparts. This means we no longer need to
define the QT_DISABLE_DEPRECATED_BEFORE macro.
This patch is relatively small because most source-compatible
changes of this kind have been done before.
* The one exception is the QmlDesigner, which uses QWeakPointer
in a deprecated way all over the place.
Change-Id: Id4b839c6685f3b5bdf2b89137f95231758ec53c7
Reviewed-by: Eike Ziller <eike.ziller@digia.com>
Reviewed-by: hjk <hjk121@nokiamail.com>
This will hopefully help a bit in resolving the crashes that happen in
QCoreApplication::notifyInternal.
Change-Id: Ib3aa8c1a1f50778bd89938d7c529d8399ccf91ea
Task-number: QTCREATORBUG-11262
Reviewed-by: Nikolai Kosjar <nikolai.kosjar@digia.com>
And change the code-model inspector dialog to use them in the
ProjectParts tab to sync the master/details view, and in the
CppPreProcessorDialog to save/reload the selected part.
Change-Id: I7cc8ac1c21d31122f8650985d0b85177ebabdfb8
Reviewed-by: Nikolai Kosjar <nikolai.kosjar@digia.com>
The issue only appeared in the test due to the different
editor construction there. Now use the same factory access.
Change-Id: I3a8534fbe683bb88f04ad68850cecdfe32b11433
Reviewed-by: Christian Stenger <christian.stenger@digia.com>
Remove usage of the Q_EXPORT_PLUGIN macros, which do not exist in Qt 5.
Change-Id: I678c3cf10b9c5d5c1b9f252b0ecd1c97dc810a47
Reviewed-by: hjk <hjk121@nokiamail.com>
Reviewed-by: Christian Stenger <christian.stenger@digia.com>
This mainly takes CppEditorSupport apart.
* Parsing is now invoked by CPPEditorDocument itself by listening to
QTextDocument::contentsChanged().
* Upon construction and destruction CPPEditorDocument creates and
deletes an EditorDocumentHandle for (un)registration in the model
manager. This handle provides everything to generate the working copy
and to access the editor document processor.
* A CPPEditorDocument owns a BaseEditorDocumentProcessor instance that
controls parsing, semantic info recalculation and the semantic
highlighting for the document. This is more or less what is left from
CppEditorSupport and can be considered as the backend of a
CPPEditorDocument. CPPEditorDocument itself is quite small.
* BuiltinEditorDocumentProcessor and ClangEditorDocumentProcessor
derive from BaseEditorDocumentProcessor and implement the gaps.
* Since the semantic info calculation was bound to the widget, it
also calculated the local uses, which depend on the cursor
position. This calculation got moved into the extracted class
UseSeletionsUpdater in the cppeditor plugin, which is run once the
cursor position changes or the semantic info document is updated.
* Some more logic got extracted:
- SemanticInfoUpdater (logic was in CppEditorSupport)
- SemanticHighlighter (logic was in CppEditorSupport)
* The *Parser and *Processor classes can be easily accessed by the
static function get().
* CppHighlightingSupport is gone since it turned out to be useless.
* The editor dependency in CompletionAssistProviders is gone since we
actually only need the file path now.
Change-Id: I49d3a7bd138c5ed9620123e34480772535156508
Reviewed-by: Orgad Shaneh <orgads@gmail.com>
Reviewed-by: Erik Verbruggen <erik.verbruggen@digia.com>
Less typing and less cycles than join(QString) where appropriate
Change-Id: I6ebc0e17e4d7fd9845864dd95b7de4ba4dad6906
Reviewed-by: Eike Ziller <eike.ziller@digia.com>
...in CheckSymbols (highlighting backend). CheckSymbols might run
concurrently several times with the same CPlusPlus::Document and modify
the diagnostic messages in an unsave manner. See stack straces below.
While the generated diagnostic messages ("Only virtual functions can be
marked 'final'", "Too few arguments", ...) are useful, they weren't
propagated to the editor widget since several minor versions.
=================================================================
==23724==ERROR: AddressSanitizer: attempting double-free on 0x60c0072fcd00 in thread T528 (Thread (pooled)):
#0 0x4787c1 in __interceptor_free (/home/nik/dev/creator/creator-ut_clang-qt5/bin/qtcreator+0x4787c1)
#1 0x7fa15e5da4ac in QTypedArrayData<unsigned short>::deallocate(QArrayData*) /home/nik/usr/qt-5.3.1/include/QtCore/qarraydata.h:234
#2 0x7fa15e5d87db in QString::~QString() /home/nik/usr/qt-5.3.1/include/QtCore/qstring.h:995
#3 0x7fa15e5f4f6a in CPlusPlus::Document::DiagnosticMessage::~DiagnosticMessage() /home/nik/dev/creator/creator-ut/src/libs/cplusplus/CppDocument.h:140
#4 0x7fa15e5f4de2 in QList<CPlusPlus::Document::DiagnosticMessage>::node_destruct(QList<CPlusPlus::Document::DiagnosticMessage>::Node*, QList<CPlusPlus::Document::DiagnosticMessage>::Node*) /home/nik/usr/qt-5.3.1/include/QtCore/qlist.h:432
#5 0x7fa15e5f4ae9 in QList<CPlusPlus::Document::DiagnosticMessage>::dealloc(QListData::Data*) /home/nik/usr/qt-5.3.1/include/QtCore/qlist.h:784
#6 0x7fa15e5f47ed in QList<CPlusPlus::Document::DiagnosticMessage>::~QList() /home/nik/usr/qt-5.3.1/include/QtCore/qlist.h:760
#7 0x7fa15e7533c2 in QList<CPlusPlus::Document::DiagnosticMessage>::clear() /home/nik/usr/qt-5.3.1/include/QtCore/qlist.h:793
#8 0x7fa15e6f2ed2 in CPlusPlus::Document::clearDiagnosticMessages() /home/nik/dev/creator/creator-ut/src/libs/cplusplus/CppDocument.h:205
#9 0x7fa15e6c5f5b in CppTools::CheckSymbols::run() /home/nik/dev/creator/creator-ut/src/plugins/cpptools/cppchecksymbols.cpp:337
#10 0x7fa15e6c83c2 in non-virtual thunk to CppTools::CheckSymbols::run() /home/nik/dev/creator/creator-ut/src/plugins/cpptools/cppchecksymbols.cpp:348
#11 0x7fa17f0ccab1 (/home/nik/usr/qt-5.3.1/lib/libQt5Core.so.5+0x98ab1)
#12 0x7fa17f0cfa5e (/home/nik/usr/qt-5.3.1/lib/libQt5Core.so.5+0x9ba5e)
#13 0x7fa17eaeb181 in start_thread (/lib/x86_64-linux-gnu/libpthread.so.0+0x8181)
#14 0x7fa17dbec38c (/lib/x86_64-linux-gnu/libc.so.6+0xfb38c)
0x60c0072fcd00 is located 0 bytes inside of 128-byte region [0x60c0072fcd00,0x60c0072fcd80)
freed by thread T539 (Thread (pooled)) here:
#0 0x4787c1 in __interceptor_free (/home/nik/dev/creator/creator-ut_clang-qt5/bin/qtcreator+0x4787c1)
#1 0x7fa15e5da4ac in QTypedArrayData<unsigned short>::deallocate(QArrayData*) /home/nik/usr/qt-5.3.1/include/QtCore/qarraydata.h:234
#2 0x7fa15e5d87db in QString::~QString() /home/nik/usr/qt-5.3.1/include/QtCore/qstring.h:995
#3 0x7fa15e5f4f6a in CPlusPlus::Document::DiagnosticMessage::~DiagnosticMessage() /home/nik/dev/creator/creator-ut/src/libs/cplusplus/CppDocument.h:140
#4 0x7fa15e5f4de2 in QList<CPlusPlus::Document::DiagnosticMessage>::node_destruct(QList<CPlusPlus::Document::DiagnosticMessage>::Node*, QList<CPlusPlus::Document::DiagnosticMessage>::Node*) /home/nik/usr/qt-5.3.1/include/QtCore/qlist.h:432
#5 0x7fa15e5f4ae9 in QList<CPlusPlus::Document::DiagnosticMessage>::dealloc(QListData::Data*) /home/nik/usr/qt-5.3.1/include/QtCore/qlist.h:784
#6 0x7fa15e5f47ed in QList<CPlusPlus::Document::DiagnosticMessage>::~QList() /home/nik/usr/qt-5.3.1/include/QtCore/qlist.h:760
#7 0x7fa15e7533c2 in QList<CPlusPlus::Document::DiagnosticMessage>::clear() /home/nik/usr/qt-5.3.1/include/QtCore/qlist.h:793
#8 0x7fa15e6f2ed2 in CPlusPlus::Document::clearDiagnosticMessages() /home/nik/dev/creator/creator-ut/src/libs/cplusplus/CppDocument.h:205
#9 0x7fa15e6c5f5b in CppTools::CheckSymbols::run() /home/nik/dev/creator/creator-ut/src/plugins/cpptools/cppchecksymbols.cpp:337
#10 0x7fa15e6c83c2 in non-virtual thunk to CppTools::CheckSymbols::run() /home/nik/dev/creator/creator-ut/src/plugins/cpptools/cppchecksymbols.cpp:348
Change-Id: Ifab2842ea43aeb26099835966b02d8afc4b85df4
Reviewed-by: Erik Verbruggen <erik.verbruggen@digia.com>
This is in preparation for changes to the ProjectParts, where one part
can only hold files for 1 language.
Change-Id: I5e9a1a803ecc3ecbb36added0d0094df63922549
Reviewed-by: Nikolai Kosjar <nikolai.kosjar@digia.com>
A<int[]> now prefer second specialization for
template<typename T> class A
template<typename T> class A<[]>
Change-Id: I32e874f78b2f5b363d088fbab6a8897e42e44035
Reviewed-by: Nikolai Kosjar <nikolai.kosjar@digia.com>
They do not really depend on CppEditorWidget and will be re-used in a
follow-up change.
CppEditorWidget::identifierUnderCursor()
CppEditorWidget::findCanonicalMacro()
Change-Id: I553bb68694fb4b5f1baa331debf33ae234e3dedf
Reviewed-by: Erik Verbruggen <erik.verbruggen@digia.com>
This is the first step in actually making "closeEditor(s)" close editors
instead of documents.
Change-Id: I02761e0cef950b8fc093f65a90df04a9a7550681
Reviewed-by: Daniel Teske <daniel.teske@digia.com>
Reviewed-by: hjk <hjk121@nokiamail.com>
Adding QTC_FIND_ERRORS_INDEXING=1 to the run environment will Creator
force to do some "extended indexing" in order to find bugs:
1) The project files are parsed as if they would be opened in an
editor. That is, parsing happens as precisely as possible, based on
the ProjectPart information.
2) Symbols are looked up by invoking CheckSymbols, the backend of the
semantic highlighter.
3) A "Task List File" (*.tasks) will be written with all diagnostic
messages.
This special indexing mode is meant for testing purposes. E.g. it can be
run nightly on some bigger projects to find regressions.
Note that Qt Creator will quit exactly after the first time some source
files are indexed. E.g. that will happen if you open a file manually or
if you open a new unconfigured project. Therefore it's required to
configure projects as needed before invoking in this indexing mode.
Change-Id: If25b83e67d24df9e28e107cb062f21cbf3b4c643
Reviewed-by: Erik Verbruggen <erik.verbruggen@digia.com>
In case the list of source files to be parsed is empty, the
CppSourceProcessor is never deleted.
This was never a real problem because of the guard in
CppModelManager::updateSourceFiles().
Change-Id: Icfd6962d11f2b2bf2ac28825f2fa0af0838a09c8
Reviewed-by: Erik Verbruggen <erik.verbruggen@digia.com>
There was document(), textDocument() and baseTextDocument().
Two should be enough...
Change-Id: Id9e41c8d857c5cb3269a9fce5ab594d34448c982
Reviewed-by: Tobias Hunger <tobias.hunger@digia.com>
There are conceptually only two: one that operates a new document,
and one that shares one. Being explicit makes moving data over to
the Editor hierarchy easier. Convenience can be re-added there,
later.
Change-Id: I9b34ff26628c99ffff01201dcf99332d5e7253e9
Reviewed-by: Christian Stenger <christian.stenger@digia.com>
Reviewed-by: hjk <hjk121@nokiamail.com>
The parent of instantiation of nested class of template class
should be the instantiation of enclosing template class.
To prevent the infinite loop for case with local typedef of enclosing
template we should not change a parent of typedefed instatiation of
enclosing template. Example:
template <typename T>
struct Enclosing
{
typedef Enclosing<T> EnclosingT;// first case
struct Nested
{
typedef Enclosing<T> EnclosingT;// second case
};
};
Task-number: QTCREATORBUG-11752
Task-number: QTCREATORBUG-11999
Change-Id: Iadd7b5ef73ee0c4881f59c9dabfe03339f55827b
Reviewed-by: Nikolai Kosjar <nikolai.kosjar@digia.com>
Not just for selected documents. This also removes the need for
the separate setupAsPlainTextDocument function.
Change-Id: I51a04eebbad37a7598068ce6e157a53c1519464f
Reviewed-by: Christian Stenger <christian.stenger@digia.com>
Reviewed-by: hjk <hjk121@nokiamail.com>
Same patter. Plain is fully merged now, so remove the files, too.
Change-Id: Id8c0ba5689ad9980a0db3580cb9833344fd911f3
Reviewed-by: Tobias Hunger <tobias.hunger@digia.com>
Reviewed-by: Christian Stenger <christian.stenger@digia.com>
So far the pre-compiled headers were processed (thus defines from those
headers were visible), but the actual includes for the documents were
not added, which is necessary for lookup/completion.
Note that this will be only done if pre-compiled headers are not ignored
(Options > C++ > Code Model > [] Ignore pre-compiled headers).
Change-Id: I54a8e6b00597af164d958e3e9f2a1075ea187788
Reviewed-by: Erik Verbruggen <erik.verbruggen@digia.com>
Check only other projects if the current file is not part of any
project.
Task-number: QTCREATORBUG-12328
Change-Id: I7db65b26313a26773bbbf17e966abd668ea9a1a5
Reviewed-by: Erik Verbruggen <erik.verbruggen@digia.com>
@classes, @protocols, @properties and -methods (and +methods) will now
show up in the locator.
Change-Id: I33f904c456c321fc80b8aba7ff52d5f1207c58be
Reviewed-by: Nikolai Kosjar <nikolai.kosjar@digia.com>
Turns out we don't even need two-phase initialization,
by transparently postponing registration until we get
a non-empty file name, either at constuction, or at
file name change times.
Change-Id: I3e87e47c820066e6707e946fc474ab9c1993e61f
Reviewed-by: Eike Ziller <eike.ziller@digia.com>
Looks like a compiler bug, but it's a supported environment.
Change-Id: I73d67a74e1adbac6747cbc96e3fbd96eeee9556c
Reviewed-by: Nikolai Kosjar <nikolai.kosjar@digia.com>
