Closed Bug 1297851 Opened 9 years ago Closed 8 years ago

stylo: MaybeUnresolvedURIEquals needs to be able to unresolved URIs

Categories

(Core :: CSS Parsing and Computation, defect, P2)

defect

Tracking

()

RESOLVED FIXED

People

(Reporter: bholley, Unassigned)

References

(Blocks 1 open bug)

Details

The patch from bug 1292930 makes us claim that the URIs are different if one of them is unresolved. This is suboptimal, because it means that we think that mBinding changes much more often than it does, and generate a bunch of unnecessary ReconstructFrame hints. Comparing unresolved to unresolved is easy, because we can just compare the base URIs and the strings. Comparing unresolved to resolved is harder, because mURI stores _either_ the base URI _or_ the resolved URI, but not both. Easiest is probably just to bloat the struct and store both, which may not impact memory in practice because we already have the mURIResolved boolean, which we could eliminate if we do that.
Blocks: 1297857
Cameron, ISTR we did this at some point. Can you confirm?
Flags: needinfo?(cam)
Priority: -- → P2
Yes, this is done. URLValueData::DefinitelyEqualURIs, which gets called under CalcDifference, compares the base and specified URL strings.
Status: NEW → RESOLVED
Closed: 8 years ago
Flags: needinfo?(cam)
Resolution: --- → FIXED
You need to log in before you can comment on or make changes to this bug.