Skip to main content

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
[mdt-papyrus.dev] Stable line breaks and formatting for generated code of papyrus

Hello,

Yesterday I discussed with Remi the problem described in the bugzilla #435997 [1], and he suggested to move the discussion to this mailing list.

We are actually looking at the re-generation of some Papyrus diagrams with xtend-based templates.
We have identified a few things which make regeneration process difficult, especially for non-committer who wants to contribute via Gerrit: 

a) the generated diagram files have different line breaks (win linux)
b) the GMF-T artifacts (*.gmfgen files) have different line breaks within the single file
c) some of gmfgen files committed with a concise formatting (e.g, for activity diagram [2]), while some have more useful formatting (e.g for class diagram [3]). Here obviously [3] is MUCH better than [2] for future changes that will come to Gerrit
d) custom xtend templates and their corresponding xtend-generated java files have different line breaks
e) different formatting options for generated code of different diagrams

This asks for one time fix, by the commit with huge diff consisting of only spaces, line breaks and other formatting changes.

Unfortunately by the nature of this change it is difficult to use gerrit for verification of the proposed changes. E.g, I prepared just the change for the item c) for ActivityDiagram gmfgen at Gerrit [4] (now abandoned, it also contain unrelated changes for other classes). The relevant part of the change consists of 6K+ modified lines and now someone would need to check them one by one to verify that there are indeed no semantic changes hidden somewhere inside.

I believe it would be more simple if someone with actual committer rights made and commit this changes, because the person would know HOW she produced the changes.

I also believe that it is important to make this commit before branching for Luna SR stream, because otherwise it would be twice as difficult to backport that changes for Luna SR into the master for Mars.

Regards,
Michael

[1] https://bugs.eclipse.org/bugs/show_bug.cgi?id=435997
[2] http://git.eclipse.org/c/papyrus/org.eclipse.papyrus.git/tree/plugins/uml/diagram/org.eclipse.papyrus.uml.diagram.activity/model/activityDiagram.gmfgen
[3] http://git.eclipse.org/c/papyrus/org.eclipse.papyrus.git/tree/plugins/uml/diagram/org.eclipse.papyrus.uml.diagram.clazz/model/classdiagram.gmfgen
[4] https://git.eclipse.org/r/#/c/28954/1/plugins/uml/diagram/org.eclipse.papyrus.uml.diagram.activity/model/activityDiagram.gmfgen
--

Michael "Borlander" Golubev
Eclipse Committer (GMF, UML2Tools)
at Montages Think Tank, 
Prague, Czech Republic
1165/1 Dvorecka, 14700, Prague-4 Podoli

tek: +420 602 483 463


Back to the top