-
Notifications
You must be signed in to change notification settings - Fork 2.2k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Include Hiragana and Katakana code blocks for local glyph generation #8365
Conversation
edc1b8b
to
62d98b9
Compare
}); | ||
}); | ||
|
||
test('GlyphManager generates Hiragan PBF locally', (t) => { |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Typo Hiragan
const manager = new GlyphManager((url) => ({url}), 'sans-serif'); | ||
manager.setURL('https://localhost/fonts/v1/{fontstack}/{range}.pbf'); | ||
|
||
//Hiragan letter te |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Same typo
//Hiragan letter te | ||
manager.getGlyphs({'Arial Unicode MS': [0x3066]}, (err, glyphs) => { | ||
t.ifError(err); | ||
t.equal(glyphs['Arial Unicode MS'][0x3066].metrics.advance, 24); |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I'm not sure I fully understand what the tests do. I get that you're stubbing the draw
method of the static TinySDF
object. But how does this test ensure that TinySDF.draw
was called as opposed to going through the regular pathway? In either case, wouldn't we expect the glyph to be 24px wide?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
GlyphManager#loadGlyphRange
would be the other code path, and that is not stubbed, so the test will error out if it tries to go down that path. The metrics
test is to ensure that the code path called _tinySDF
successfully and returned a StyleGlyph
successfully.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Ah ok, that makes sense.
62d98b9
to
d65b4fa
Compare
* [docs] adds docs-wide search component (#8349) * update dr-ui, mr-ui * add search component * adjust columns to prevent overlap * v1.1.0-beta.1 (#8353) * Update CHANGELOG.md for v1.1.0-beta.1 release * Bump package.json version to 1.1.0-beta.1 * [docs] updates to search component and layout (#8362) * do not index /api sidebar * exclude code snippets in examples from being indexed * Update docs-page-shell * update dr-ui, react, react-dom * adjust columns to fit search and navigation better * [docs] Replace example with a third-party raster source (#8274) * remove Add a raster tile source example * revert "remove Add a raster tile source example" This reverts commit ea8bf28. * add third-party tiles to map-tiles example * fix attribution * rm unused var * Add mapbox-gl-framerate to plugins list (#8356) * remove old plugins page and update reference to the maintained version (#8369) * Add link to iOS SDK documentation (#8359) * add ios sdk link per comment from stephen g * add new link from minh and reword * Include Hiragana and Katakana code blocks for local glyph generation (#8365) (#8378) * document the tile size drawn with showTileBoundaries (#8379) * use browser `Cache` interface to cache tiles (#8363) Recent pricing changes introduced a `sku=` query parameter that changes with every map load. This defeats the browser's ability to cache these tiles. We're working around that by implementing our own caching with the new `Cache` api. - skips caching tiles that expire soon - only caches mapbox tiles (no 3rd party tiles, styles, etc because the browser should cache these fine) - does not work in IE, Safari Browser tests can be found in debug/cache_api.html mapboxgl.clearStorage added to provide a way to clear the cache. * fix cache support check for enforcing size limit This was throwing errors in browsers that don't support the cache api but wasn't previously noticed because it was not breaking anything. * Fix background rotation hovering (#8367) * update changelog * [email protected] (#8297) * [email protected] * add one more file to v4.4.0 * fix Response construction in Edge, fix #8391 (#8392) (#8394) Edge does not support constructing responses with readable streams. * v1.1.0-beta.2 (#8395) * Add sdk support spec section for text-radial-offset (#8384) (#8401) This PR also assigns SDK versions for the `text-variable-anchor` property (both `text-variable-anchor` and `text-radial-offset` are the parts of the variable text placement feature). * fix #8411, fetch in case of cache error (#8415) (#8417) * fix #8411, fetch in case of cache error * Use warnOnce * v1.1.0 & [email protected] (#8408) * Update style-spec version & changelog for [email protected] - Update style-spec/CHANGELOG.md to add bugfix items - Bump style-spec/package.json version to 13.7.1 (patch) * Update mapbox-gl package version to 1.1.0 for final release * Add missing style-spec changelog entry * Fix yarn.lock not installing under CI by auto-updating dependencies * manually strip urls for faster cache matching (#8434) (#8449) * update examples to use mapbox-gl-geocoder v4.4.1 (#8451) * set antialias to true for custom layers and extrusion examples (#8474) * swap map-tiles example to use https (#8482) * fix race condition between caching and aborting (#8472) fix #8470, continuous memory usage growth * fix #8481, unbounded memory growth in IE (#8490) Cancellable requests to workers should never call back after they have been cancelled. * v1.1.1 * [docs] fix SDK support for symbol-sort-key (#8501)
This PR adds support for locally generating glyphs in the Hiragana and Katakana code blocks. This allows layers with mixed japaense syllabary to be able to use a single font. Without this change - kanji would use the locally generated glyphs but hiragana and katakana would use glyphs based on the style.
The
localIdeographFontFamily
map option that is used to control this behavior is now not the best name given that Hiragana and Katakana are not ideographic. The name should be changed in a future API breaking release.[ ] post benchmark scorescc @1ec5 @chloekraw @tristen