Summary: To me, it seems that these methods should never be overwritten in subclasses.
Test Plan: `arc unit`
Reviewers: epriestley
Reviewed By: epriestley
Subscribers: aran, epriestley, Korvin
Differential Revision: https://secure.phabricator.com/D7958
Summary:
Currently, the Phutil XHPAST Linter and vanilla XHPAST Linter reuse the same parse tree, but do this by having explicit knowledge of one another.
Instead, let them synchronize by writing to a glorified array of globals on the Engine. They no longer require knowledge of one another, so this can work under `.arclint`.
(This could probably be a little cleaner by putting more logic in the shared base class, but Facebook has some kind of goofy subclass of this thing and //this// patch won't disrupt it, while a cleaner one might.)
This should unblock D9057.
Test Plan: Ran unit tests and normal lint, got accurate looking results without duplicate invocations showing up in `--trace`.
Reviewers: btrahan, joshuaspence
Reviewed By: joshuaspence
Subscribers: epriestley
Differential Revision: https://secure.phabricator.com/D9059
Summary: No subclass should need to override these methods. Additionally, none currently do.
Test Plan: N/A
Reviewers: epriestley, #blessed_reviewers
Reviewed By: epriestley, #blessed_reviewers
Subscribers: epriestley, Korvin
Differential Revision: https://secure.phabricator.com/D9035
Summary: This is a little bit tricky - if both XHPAST and PhutilXHPAST linters lint the same path then they get the same future wrapped in two different Future iterators.
Test Plan:
$ arc unit
Reviewers: epriestley
Reviewed By: epriestley
CC: aran, Korvin
Differential Revision: https://secure.phabricator.com/D5015
Summary: Also provides an example how to build custom linter using XHPAST.
Test Plan: Added debug output to `willLintPaths()`, verified that each path is parsed only once.
Reviewers: epriestley
Reviewed By: epriestley
CC: aran, Korvin
Differential Revision: https://secure.phabricator.com/D4718
