- On some gl implementations the uniform locations are shifted after linking the program a second time, resulting in errors when using any uniform. On some implementations they are actually doubled. Re-binding the uniforms after selectively binding the vertex attributes prevents this.
(cherry picked from commit 4a0e025)
Mentioned MGLComputedShapeSource in a few more places. Updated symbol style layer documentation to refer to aftermarket expression functions. Copyedited some heatmap style layer documentation.
Prevents querying a FeatureIndex built against a separate set of data, which can lead to invalid index exceptions.
The GeometryTileWorker 'data' member can still change independently of the data in the feature index, at the time 'setData' is called. The GeometryTileWorker maintains ownership of its local data (which may be used to re-parse) and clones the data for use by the FeatureIndex in the foreground.
* [ios, macos] Eviscerate mbgl expression to Foundation JSON object conversion
* [ios, macos] Remove literal special case
MGLJSONObjectFromMBGLValue() is just a literal translation from C++ types to Foundation types; no need for expression-specific literal wrapping.
* [ios, macos] Implemented RGB(A) operators
* [ios, macos] More robust color conversion
* [macos] Create calibrated RGB from expression on Sierra
* [macos] Calibrate colors in expressions at runtime
* [ios, macos] Special-case number conversion without fallbacks
* [ios, macos] Test mgl_number expressions
* [ios, macos] Updated tests for aftermarket expressions
* [ios, macos] Relax roundtripping requirement for serialize()
* [ios, macos] Fixed interpolation expression tests
* [ios, macos] Removed dead code
* [ios, macos] Fixed color conversion on Sierra
* [ios, macos] Introduced more ergonomic expression concatenation syntax
* [ios, macos] Fixed mgl_join:
* [ios, macos] Simplified expression document headings
* [ios, macos] Documented mgl_join:
* [ios, macos] Convert length operator to length: for strings
Only works for literal strings.
* [ios, macos] Implemented type conversion using CAST()
* [ios, macos] Aftermarket let expressions
* [ios, macos] Refactored aftermarket expression functions
Refactored the installation of aftermarket expression functions to use macros. It is no longer necessary to handwrite the type encoding of a function. Also added aftermarket functions for interpolating and stepping.
* [ios, macos] Updated documentation, tests, demo apps
* [ios, macos] Added generic expression function
* [ios, macos] Add MGL_MATCH function placeholder.
* [ios, macos] Add MGL_SWITCH expression operator.
* [ios, macos] Add mgl_coalesce: as expression function.
* [ios, macos] Update style documentation.
* [ios, macos] Add conventional custom function support.
* [ios, macos] Updated example code
* [ios, macos] Add mgl_coalesce conventional custom function support.
* [ios, macos] Add aftermarket function to 'has' operator.
* [ios, macos] Add documentation for lookup and feature operators.
* [ios, macos] Documented simple lookup
* [ios, macos] Renamed, reversed has expression
Renamed mgl_hasProperty:properties: to mgl_does:have: for better readability and consistency with the conventional mgl_has: function. Documented both forms of mgl_has:.
* [ios, macos] Restored OEM conditionals where available
This is the preferred syntax for simple conditionals on iOS 9 and above, because you can inline the predicate instead of wrapping it in a constant value expression, which means you can write a conditional in a single format string.
* [ios, macos] Update style docs.
Co-authored-by: Fabian Guerra <fabian.guerra@mapbox.com>
Second half of fix for issue #11538.
If a global placement took place between the time a tile's non-symbol layout updated and the time new symbol buckets arrived, the tile's new FeatureIndex would be committed, but the global CollisionIndex would be generated against the old symbolBuckets. The mismatch could cause the CollisionIndex to contain indices that didn't exist in the new FeatureIndex, and attempting to access them would generate an out-of-bounds exception.
First half of fix for issue #11538.
Testing `if (pendingData)` didn't work if there _was_ a data update, but the update was "the data for this tile is now null". In that case, the tile's FeatureIndex would be updated, but the tile's data member would remain unchanged.
In the case of a tile's data being deleted, the matching FeatureIndex would have an empty set of bucketLayerIDs, but the _old_ data would still be in place for querying, and the symbolBuckets might not be updated yet (pending onPlacement). In this case `bucketLayerIDs.at(indexedFeature.bucketName)` could throw an out-of-range exception.
