Hello Searchfox Users! You may have noticed the following things
working again/better recently in Searchfox:
XPIDL:
* Clicking on symbols in the `.idl` for the current revision will once
more bring up a menu with options. There was a regression here for
a bit due to a historical lack of test coverage, but now the
regression is fixed and we have newly added test coverage in both
the "tests" repository in the searchfox repo at
https://github.com/mozsearch/mozsearch (docs on tests at
https://github.com/mozsearch/mozsearch/blob/master/docs/testing-checks.md)
but also in our production checks at
https://github.com/mozsearch/mozsearch-mozilla/tree/master/mozilla-central/checks!
* "Go to definition" for XPIDL binding symbols will take you to the
line of the `.idl` file that defines the symbol rather than the line
of the `.h` file for the C++ binding.
* For XPIDL methods, the search menu "Search for IDL method" has
gained a sibling menu option "Search for IDL C++ method" that is
specifically for the C++ binding's symbol and excluding the best
effort JS symbol which is included in "Search for IDL method". Due
to limitations in the searchfox JS analysis, for an IDL method
"doASpecificThing", the IDL analysis will link it to the JS symbol
`#doASpecificThing` which works out okay when that method name is
unique across the entire m-c codebase, but works out less well when
it's a common name.
* For XPIDL attributes, the search menu "Search for IDL attribute" has
gained 2 sibling menu options: "Search for IDL C++ getter" and
"Search for IDL C++ setter". These likewise are searches only for
the C++ symbols and excluding the JS symbols.
* Many thanks to :nika and :mccr8 for their assistance in
investigating the problem(s) and helping iterate towards an even
better UX than we had before!
Code Coverage:
* Code coverage data should be more reliably present as searchfox now
should run per-push on mozilla-central instead of based on a cron
job, allowing searchfox to only index jobs where code coverage data
is available instead of indexing DONTBUILD jobs where it is not.
This does mean that searchfox indexes for m-c no longer directly
correspond to nightly builds, but it's also likely that there will
also be little drift between what searchfox indexed and what's in a
given nightly.
* Many thanks to :marco for helping diagnose the scheduling problem
and :aki and :ahal for helping explore the solution space and then
get us to that solution quickly!
* There's still a little more work to be done here to help searchfox
only pick jobs that have coverage data available.
Things to keep in mind about Searchfox:
* Anyone can contribute! Check out the `#searchfox` channel on
https://chat.mozilla.org/ and the repo at
https://github.com/mozsearch/mozsearch!
* Searchfox is a contributor driven project. If there are features
you would like to see or rough edges that slow down your work,
please talk to your manager[1] and discuss the possibility of
finding time to potentially contribute to the project yourself if
you are interested, and/or help your manager surface the potential
efficiency improvements for you/the organization so that these can
be quantified and potentially folded into OKRs for teams with the
relevant expertise and interest.
1: While in the context of MoCo employees, manager means your manager,
for others this may be more of a metaphorical manager.
--
You received this message because you are subscribed to the Google Groups
"[email protected]" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To view this discussion on the web visit
https://groups.google.com/a/mozilla.org/d/msgid/dev-platform/da2e778a-ad84-0b67-3546-67881c591712%40asutherland.org.
