org.eclipse.emf.compare.git - Model comparison integrated tool and framework

path: root/plugins/org.eclipse.emf.compare/.settings

Age	Commit message (Collapse)	Author	Files	Lines
2014-01-29	Change BREEs to Java 1.6 and activate API error for missing descriptions	Laurent Goubet	2	-471/+471
	Change-Id: Iead5dbf879f0cae1702a125aa791b73f02bb387b
2013-03-12	checkstyle and findbugs	Mikaël Barbero	2	-0/+135
	re-activate checkstyle on some projects add findbugs configuration to some projects
2013-01-14	set API Tooling 'compatibility' and 'version' level to warning	Mikaël Barbero	1	-0/+97

2012-08-21	3874611 : Non ID Based Matching	cbrun	1	-1/+1
	https://bugs.eclipse.org/bugs/show_bug.cgi?id=387461 Introduces another implementation of EObject matcher which : - browse the scope, indexing the model elements. - browse the indexes, finding matches along the way The indexes are able to return the closest element of an other side. This is implemented by computing the number of differences we have between two elements. This is not free regarding CPU cycles but we are sure to get the optimal result. The new matcher is plugged in the current implementation by introducing the ability for the ID based matcher to delegate to another matcher when an ID is not found. I named it "ProximityEObjectMatcher" as its based on a distance function used to compare two EObjects. The current implementation of the distance function we are using is roughly an "EditionDistance" : it counts the number of changes you have to make to transform one object to the other. How do we do that ? By calling EMF Compare's diffing during the match by the way ! (ok, it may sound twisted). It allows us to fine tune a lot of thing like ignoring some EStructuralFeatures for the distance, or having different weight for different features. No UI is provided to change this so far, its "framework only" but the default behavior will use IDs when available and fall back to the Proximity matching on the remaining objects. Performance-wise the ProximityMatcher is about 10 times slowest than the ID based one on a decent UML model which have very few differences. Have a look on the corresponding test case if you want. By the way all this implementation is fairly provisional. ps : while coding I introduced a few builder interfaces to configure some instances.
2012-06-18	Disabling compiler warnings :	lgoubet	1	-2/+2
	- Method can be static - Unnecessary else
2012-06-07	First draft of the merge implementation. We now handle merging of the ↵	lgoubet	1	-1/+1
	ReferenceChanges : ADD, DELETE and MOVE.
2012-06-01	Acting the change to CRLF since all projects' settings set this as the file ↵	lgoubet	2	-4/+4
	delimiter. Adding the necessary logic to determine the insertion index when merging reference and attribute values.
2012-04-20	Regenerating compare metamodel with containment reference for equivalences, ↵	lgoubet	1	-1/+1
	reference to the three eResources from the MatchResource, and removed useless Comparison.addDiff method. Restoring the logical model tests plugins. Restoring the logical model build. Make sure the logical model tests are launched at build time.
2012-04-19	Configuring all projects to use "UTF-8" as their default encoding.	lgoubet	2	-0/+4
	Configuring all projects to use CRLF as their default line ending. Modifying compare.ecore to add missing references and remove useless addDiff method. Code has yet to be re-generated. Migrating the logical model implementation to the new metamodel. Tentative implementation of ResourceCompareInput.
2012-03-13	Introduce the new versions of the metamodel and tests for the EMF Compare core.	lgoubet	2	-0/+434

2012-03-12	Move all features and plugins that will be deprecated by the incoming ↵	lgoubet	4	-546/+0
	restructuration into a separate folder.
2011-10-10	Compilation errors	lgoubet	1	-333/+337

2011-03-08	revert back the compilation issue with the deprecated DefaultScope()	Cedric Brun	1	-2/+14
	to keep Eclipse 3.6 compatibility
2009-04-08	Activating API tooling on all core projects of EMF Compare	lgoubet	1	-0/+150

2009-04-02	fixing ambiguity between formatter and checkstyle	lgoubet	1	-1/+1

2009-03-13	Formatting all code	lgoubet	2	-1/+313
	Sharing format configuration Fixing bundle versions and names of o.e.epatch.*
2008-12-10	adding checkstyle checks for generic regexes	lgoubet	1	-3/+7

2007-12-07	Code cleaned up for Eclipse 3.4 - EMF 2.4 compatibility	lgoubet	1	-1/+6

2007-11-07	Fully refactored to enable internationalization	lgoubet	1	-2/+2
	Fully checked for standalone usage
2007-10-04	[199060] All properties are now shown by the properties tab	lgoubet	1	-0/+56
	[205144] Content viewers are now synchronized for their horizontal scrolling Various performance improvements and code optimization FIX : Potential ClassCastException in metrics computing FIX : Potential NullPointerException with the properties tab FIX : Error preventing content metric to be taken into account during matching
2007-07-13	ADD: Three way comparison	jmusset	1	-0/+3

Breadcrumbs