Community
Participate
Working Groups
Currently the POC for finding all references shows the results as a tree of files with the references listed in order as children. It would be better / easy to understand if we grouped them as: 1. a flat list sorted by confidence 2. a tree, with the root elements being where the matches were found, for example 'In Comments', 'In Code', etc 3. a tree, with the root elements being a descriptor of the confidence of the matches, like "Perfect Match", "Possible Match", etc
As per the latest discussion, I like 3) (SSQ's idea). I still have a feeling deep down that a flat list might be interesting/needed but certainly, a descriptor at the root means I can ignore whole categories (leave the collapsed). The downside is that I need to expand/collapse categories to get at the "real data". Using it for a while will help. Let's implement 3) for now.
Pushed support for showing alternate result models. The UI is not hooked up yet to switch them yet. http://git.eclipse.org/c/orion/org.eclipse.orion.client.git/commit/?id=b71dee3068ab02abe78307799ada1d60ca421f56
From bug 479177 comment 1, a possible new layout: Exact Matches Possible Matches Unrelated Matches This grouping will apply to what was the flat list layout. For the other grouping (by location), we might use the same grouping above, but inside the location grouping. That way we can avoid cluttering the UI with piles of percentages. For example we would could have: Function Calls Exact Matches Possible Matches Unrelated Matches JSDoc Comments Exact Matches Possible Matches Unrelated Matches etc. And possible new formatting of matches (also from bug 479177): Each match would also look a bit different like: (filename: line#) matchText or matchText (filename: line#)
Pushed support for the new groups mentioned in bug 476906 and some additional support for sync plugins in Tern and improved type loading support if the file client has not answered Tern yet. I'm going to close this bug as fixed and further work will be in bug 476906