Adding deduplication in merge command #2280
Conversation
updating tests
| Results = new List<Result>() | ||
| }; | ||
| emptyRun.Results = new List<Result>(); | ||
| splitRun = _ruleIdToRunsMap[key] = emptyRun; |
There was a problem hiding this comment.
emptyRun [](start = 71, length = 8)
Instead of creating a new run, I changed to emptyRun object to preserve artifacts and other fields #ByDesign
| Results = new List<Result>() | ||
| }; | ||
| emptyRun.Results = new List<Result>(); | ||
| splitRun = _ruleIdToRunsMap[key] = emptyRun; |
There was a problem hiding this comment.
emptyRun [](start = 71, length = 8)
This might be the right start but I think we're not comprehensively handling all that we should. What are we doing about rule indexes and artifact location indexes for example? When we add a result from one run into another, we should be fixing up those indices, expanding the target run with new rule/artifact data if it comes in, etc. Our baselining/result matching code does this and has some utilities for it. This is all to say: we probably have more work to do in this command (which previously simply moved entire run objects around, a very simple operation). Let's open an issue noting the open work and keep moving. :)
Fixes #2279