* TaskBuildIndex starts seperate indexer processes and communicates via shared memory
* indexer processes get restarted if they fail
* indexer processes are killed when the app closes or crashes
* added utility class SharedMemory to allocate and access shared memory, providing basic data structures
* added SharedMemoryGarbageCollector for keeping track of running instances and cleaning up shared memory in case of a crash
* show errors for source files that crashed during indexing
* added basic exception handling to task scheduling
* added option to turn off processes and use threads in preferences, but still use shared memory
fortune cookie message = Good news from someone dear is coming soon.
* use canonical filepath cache in cxx indexer
* made constructors of FilePath explicit
* use FilePath at more places instead of string
* forward declares FilePath wherever possible
* implemented handling of aggregations in bookmarks
* split index Storage and bookmark storage
* each sqlite database has its own version now
* added bookmark tests
* fixed creating bookmark tables when project can be loaded but bookmark db does not exist
* fixed font size for bookmarkview tooltips
* using bookmark cache now
* using default category name when none specified by the user
* raising bookmark windows when they should be shown
* clearing a bookmark or category in the editor will now update the star icon accordingly
* non-persistent bookmarks use ids of nodes and edges now (instead of names)
* git ignore .srctrlbm files
* mark all files with errors as incomplete
* fix incomplete label in codeview tooltip not updated properly
* added setting for canceling indexing for files with fatal errors
* avoid AST traversial for CXX translation units with fatal errors
* save all files within those translation units as incomplete
* show number of complete files in finished indexing dialog and overview stats
* display incomplete files hatched in graph view
* display incomplete file titles hatched in code view
* mention file incompleteness in tooltips
* removed derived Indexer classes and set ParserType via template parameter instead
* moved getSourceFileFromCDB from IndexerCxxCdb to IndexerCommandCxxCdb
bug id = 358
* project can not have multiple source groups with different types
* language of a project is now inferred from source groups
* setting global NameHierarchy delimiter of first sourceGroup in project
* added project settings migrations for source groups
* Add Settings Migration classes (MoveKey, DeleteKey, Lambda)
* Update sample projects with migrations
* added method to config to retrieve sublevel key names
* implemented recursive removing of keys in settings
* moved project files out of top level
* removed solution parsing code as it is not used anymore
fortune cookie message = The secret of getting ahead is getting started.
* added option to create a Coati project from a Maven POM file
* show list of indexed file paths in maven project setup
* use FileManager to detect indexed files in project setup (except cdb)
* added Maven path to preferences window and added path detectors.
* fixed: QtDialogView::hideStatusDialog now also hides loader in status bar.
* refactored ProjectType
fortune cookie message = If you continually give, you will continually have.
* replaced parser arguments by project specific IndexerCommands that know everything the indexer needs to know to do its job
* added different language and project specific indexers that know how to handle a certain kind of indexer command
* refactored FileManager and FileRegister to have less overhead
* removed no-rtti restriction for clang files in lib cxx
* uncommented deprecated interprocess code.
fortune cookie message = Happiness will bring you good luck.
* renamed DefinitionType to DefinitionKind
* passing DefinitionKind from Java to C++ instead of just a single bool
* omitting to create a symbol for undefined types
fortune cookie message = Happy news is on the way to you.
* recording of usages of FunctionProtoTypes (e.g. "int(void)" can be used as a template argument)
* implemented solving names of dependent template specialization types (these are not as bad as it sounds ;) )
* implemented correct solving or decayed types
* fixed: qualified type usages are recorded in the correct context
* changed cxx indexer tests not to crash anymore when something is not found
* added classes that represent different name kinds
* rewrote NameResolvers to return the new name classes instead of NameElement and NameHierarchy
* removed DataType classes since their functionality moved to the CxxTypeName class.
* removed old AstVisitor
* added new implementation called CxxAstVisitor
* added CxxVerboseAstVisitor that may wrap the CxxAstVisitor to log the AST while visiting
* added the same structure for the Java visitors
* removed all the logs that were fired when recording stuff
* Added UI setting for verbose indexer logging.
* implemented AST name obfuscation for Java and Cxx
* unified test visitors for Java and Cxx
* implemented recording using directive decl
* removed most of the old onXxxParsed methods
* Added TaskParse as base for TaskParseCxx and TaskParseJava
* Fixed increment and decrement of indexer_count on blackboard not atomic
* Fixed TaskInjectStorage finished before TaskParse started
* Fixed project not cleared when refreshing on now empty project
* Fixed TaskInjectStorage overheating when nothing to do
* Define exit state for TaskRepeatWhileSuccess
* Added TaskFinishParsing to end of parsing task sequence
* Also measure time spend clearing files in total index time
* reimplemented interrupting TaskParseCxx by adding a listener for the MessageInterruptTask and returning a failure status code on update. This cancels the parent sequence task which results in the indexed items not getting inserted into the persistent storage
* removed the capability for interrupting from TaskScheduler
* changed task system to be closer to the standard behavior tree implementation
* changed task system to accommodate the 3 status return types: Running, Success and Failure
* made TaskGroupSequential fail once a member task fails
* made TaskGroupParallel fail once a member task fails
* split TaskParse... into one task for indexing and one task for injecting
* added TaskRunner that handles updating and resetting the managed task
* fixed numbers that are shown as parsed file count in indexing ui
* fixed deadlock that originated from interaction between TaskScheduler and MessageQueue (one thread wanted to destroy a message listener on a task while the other one wanted to send as message as a task)
* implemented a dynamic way (that does not involve the linker) to load the Java runtime library
* implemented JavaPathDetectorLinux
* Filepicker for Java runtime library now asks to pick JVM library instead of folder
* extended error handling for JavaEnvironmentFactory
* implemented a more generic approach to detecting filepaths
* reimplemented standard header detection using this approach
* added auto detection of JVM dynamic library on windows
* fixed java tests that were broken when updating the java_indexer project before this commit.
* Added DialogView with QtDialogView implementation to handle information dialogs
* Implemented QtIndexingDialog class that covers all different dialogs for indexing
* Directly communicate with DialogView from Prpject and all Tasks involved in indexing
* Block QtMainWindow UI while indexing
* implemented java will not be used if it is not found on system
* add jni libraries weak linked for Mac
* use dlopen for lazy library loading
* added java path to preferences
* expanded appsettings to include java specific stuff
* added java indexer as separate project that can be build via its build script
* added dependencies for java indexer to the setup folder
* added license files for these dependencies.
* added sample project for java. this is just to test java on other machines and can be removed again in the future.
* the java parser test suite is currently uncommented because java is not running on every machine.
* added some more node types
* removed TokenComponentAccess::AccessType and replaced all its usages by using AccessKind
* moved access components from edges to nodes
* using recordReference of ParserClient for java
* removed recording access specifier of inheritance edges.
* added styles for new node types
* modified the TaskParseCxx to be able to run multiple times in parallel
* made FileRegister threadsafe and changed a lot of its mechanisms
* added TaskGroupParallel that runs all children in parallel
* added TaskParseWrapper that acts as a decorator to execute some code before and after parsing.
* implemented task setup in project with 4 parsing threads
* removed SimpleTask as it was only used as interface for the LambdaTask
* renamed Storage to PersistentStorage
* added Storage as baseclass that implements data injection
* cleaned ParserClient interface
* added default constructors for storage types
* cleaned SqlteStorage by implementing the private getAll methods as template specializations
* added CXX flags to CMakeLists (this is needed for sqlite with visual studio)
* removed TestStorage.h/.cpp since these files were not used anymore.
* fixed deleted project could still be selected in the recent project menu
* fixed project file not loaded when declining for reload
* fixed main in tutorial description not clickable
* fixed clang warnings
* fixed commands not showing up in autocompletion
* fixed classes were expanded when activating aggregation
* fixed crash when messagelisteners tried to unregister after messagequeue was destructed
* added SymbolType that can be used by ASTVisitor and ParserClient
* reordered some of the ParserClient's methods.
* removed return type of ParserClient methods.
* added recording of references to the ASTVisitor.
* refactored reference recording in ASTVisitor.
* merged all the onXxxUsageParsed methods of the ParserClient.
malte_langkabel