This will significantly reduce memory usage and also speed the creation
of new font families. In particular, the coverage bitmaps for the fonts
in the fallback stack will be computed once in the Zygote, rather than
separately in each app process.
Bug: 17756900
Change-Id: I66f5706bddd4658d78fe5b709f7251ca9d2ff4f8
Some fonts contain a cmap segment for char 0xffff that contains an
invalid rangeOffset. This was rejected by the existing code, which
means the font is considered to have empty Unicode coverage. This patch
just discards the invalid segment (consistent with OpenType Sanitizer),
making the custom font display.
Bug: 18106256
Change-Id: Icc8616a3030f80e62db906332be64d434ae72ea2
The old logic for fake bolding results in no fake bolding for a bold
span on a light weight (300) because the target weight (600 in this
case) didn't meet the condition. This patch fine-tunes the threshold
to enable fake bolding for this.
Bug: 17587185
Change-Id: I04abd00a74240cbed79c417f81486aa2158b2806
Fractional advance widths were causing subtle problems with text
positioning when the same text was drawn with different spans in the
hwui renderer. Quantizing the coordinates on layout (as opposed to
waiting until the renderer draws the glyphs) solves the problem.
This patch also fixes a discrepancy between x position and advance
widths when letterspacing.
Bug: 17347779
Change-Id: Ia705944047408c2839d5ad078eefd6bbec446872
This patch finds an appropriate fallback font in the case where no font
directly maps the requested character, but a font does exist for the
character's canonical decomposition. This yields correct rendering of
compatibility characters such as U+FA70.
Bug: 15816880
Bug: 16856221
Change-Id: Idff8ed6b942fec992a0815a32028b95af091d0ee
This reduces another allocation (last one?) we were doing when
fulfilling shaping requests from the cache.
Bug: 17111260
Change-Id: Ieb8ae1ccfcaacedb257e1e9263777f10623aaf98
C++ local var initialization always tricks me. Previously, Layout
didn't have a constructor, which meant that defining it on the stack
left mAdvance uninitialized. This was not an issue when we were doing
"new Layout()", since that invokes zero-initialization, but was an
issue for the skipCache path that was allocating layout on stack by
just "Layout l" instead of "Layout l = Layout()". To avoid surprises,
add a constructors that clears everything.
Also adds reset() method to reset the layout for reuse.
Change-Id: I3e02f00da9dd7d360abe13f63c310f6882292d0a
The U+20E3 COMBINING KEYCAP is used in our fonts to generate an emoji
rendering of ASCII numbers and letters through GSUB. For that to work
we need to choose the same (Emoji) font for the character coming
*before* the COMBINING KEYCAP character.
This is a special-case of a broader need to choose fonts per grapheme
cluster as opposed to per character, but for now, special-case U+20E3.
Bug: 7557244
Change-Id: I958e5a01068df8495bbb9bc3b9ed871cea1838b6
We are stack-allocating MinikinPaint objects in Minikin clients, and
without a constructor adding new members to the struct cannot be done
without updating all clients (only one right now!).
Change-Id: I4170f16498bb6b07cb795495011aca58087ed0bd
Replaces invalid unicode with replacement character U+FFFD and always
makes forward progress.
Bug: 15849380
Change-Id: Ic59ef6c64b0f5c4450bcae61597adcc269d6e7c5
When a run has no cmap coverage in any font, use the base font. Most of
the time, this will cause rendering of the .notdef glyph, which is
preferable to displaying nothing. In some cases, Harfbuzz may be able to
decompose the characters (not in the cmap) to ones that are, in which
case we'll render those, as long as they're in the base font.
Bug: 6629748
Bug: 15816880
Change-Id: Ibb1b9242c83626e0c7db363ad65ce44a967a005e
Proper Japanese layout requires sophisticated rules for spacing
punctuation, not just turning on the "palt" (proportional alternate)
feature. Until we can support the whole set, roll back palt.
Change-Id: If2359c529b70b1dd45dddc00e5f4aa1c91f8b0e9
This patch includes an implementation of grapheme cluster breaking,
which is especially useful for repositioning the cursor for left and
right arrow key presses. The implementation is closely based on Unicode
TR29, and uses the ICU grapheme cluster break property, but is tailored
to more closely match the existing implementation and expected behavior.
Part of a fix for b/15653110 Improve behavior of arrow keys in EditText
Change-Id: I8eb742f77039c9ab7b2838285018cf8a8fc88343
Fixes b/15734816 In the text "Wi-Fi", "-Fi" appears bolder than "Wi"
The problem was caused by "stickiness" in choosing fonts, where layout
would prefer using a font used for preceding characters as long as it
mapped the following characters in a run, in favor of the "best match"
rules. This patch adds a whitelist for making the stickiness more
conservative, only applying it for characters necessary for correct
shaping (ZWJ and ZWNJ in particular) and basic punctuation, where it is
desirable to match the style of the preceding text.
Change-Id: I1cf116879f074a5a71c351846707bfdd07b0d320
The simple predicate for fake bold (2 or more grades darker than
requested) was applying it to thin (100 weight) when normal was
requested. This patch tightens the predicate to also require that
the requested weight be in the bold range.
Fix for bug 15588352 "sans-serif-thin doesn't work on lockscreen"
Change-Id: Id9988bd149a9c8a7c943e3b221f7fb4b37fb6ddb
Fix for bug 15570313 "Missing text on nonexistent font file"
This patch makes sure that the lastChar and mInstances arrays are in
sync with each other even when a FontFamily being added has no valid
fonts in it. Previously, when they got out of sync, unicode coverage
calculation would be wrong, resulting in missing text.
Change-Id: I69c727ef69e2c61e2b2d6b81d5a28c806327f865
This patch adds support for computing when fake bold and fake italics
are needed (because the styles are requested but not provided by the
matching FontFamily), and providing them as part of the layout result.
Part of the fix for bug 15436379 Fake bold doesn't fully work (Minikin)
Change-Id: I180c034b559837943673b5c272c8e890178dff0d
We want to test configurations where the Noto Japanese font will have
its "palt" feature (to select tighter spacing in kana) will be enabled
for framework but not WebView or Chrome rendering of Japanese text. This
patch simply hardcodes this feature on.
This is also a first step towards more general setting of OpenType
features. The hardcoded feature list will grow into one set by
parameters which will eventually be plumbed up to Java.
Change-Id: Ie284e0487a1434155c8ac1cb68ddc4fc4b3c018a
This patch adds a method to retrieve the base font from a
FontCollection, which is useful when querying global font metrics.
Part of the fix for bug 15467288 "Inconsistent line heights on
Minikin builds"
Change-Id: I268ae5128d0852a020d746bc22af81fc1a623228
This patch completes support for adding context for complex script
layout, for example when a string with joins straddles two spans.
Part of the fix for 15431028: "Properly support context for joining
scripts (Minikin)"
Change-Id: I65b0833be92eb477aa531bbef0ac6eddeb3a962a
Change internal plumbing of paint flags (including CssParse) to uint32_t
consistently, to match the type used in the client. This will probably
prevent compiler warnings. Also renames "float" to "double" to avoid
confusion about precision.
Change-Id: I80374712c4067ca9e7711cc2d4ec33c440ab9c7c
Adds pseudo-css properties for scaleX and skewX, as well as paint flags,
and plumb them through to the MinikinPaint abstraction and to Harfbuzz,
to support nontrivial scale and stretch of text.
This is the Minikin part of the fix for bug 15186705 "Usability of the
suggestion strip in recent OTA's is severely reduced"
Change-Id: Ifa60355e086e4691ff92c5d50d84eb7cea0fea95
This is a fix for bug 15185229 ZWJ not working in Sinhala and Kannada.
Indic fonts (unlike Arabic) require the entire string, including ZWJ,
to be passed to Harfbuzz; it's not enough for the ZWJ to be present in
the context. The solution is to be "sticky" in font itemization,
continuing to use the same font as long as it has Unicode coverage.
Change-Id: I7673bc56fbda09f1e1a4582e8d88342343b706f1
This patch adds a "lang" pseudo-CSS property and uses it both to select
an appropriate font and control the "locl" OpenType feature to get the
most appropriate rendering for the langauge and script. In addition,
the "-minikin-variant" property selects between "compact" and "elegant"
variants of a font, as the former is needed for vertically cramped
spaces.
This is part of the fix for bug 15179652 "Japanese font isn't shown on
LMP".
Change-Id: I7fab23c12d4c797a6d339a16e497b79a3afe9df1
The context start offset wasn't being taken into account for
accumulating the advance values, leading in some cases to array index
overflow.
This is a fix for bug 15327918 "SIGSEGV in
android::MinikinFontSkia::GetSkTypeface()"
Change-Id: I9b646785724c9b72d862b822cd84661c106fbe52
This is a fix for bug 15252902 "Crash observed on keep launch or
existing youtube app after playing video". It was doing the test for a
null font after trying to resolve the font in a cache, which caused a
crash when there was no font for the run. This patch just tests before
cache lookup.
Change-Id: Iee41f7ce6b69cb09438462b6aaa916f242da7b77
This patch adds caching for both layouts and for HarfBuzz face objects.
The granularity of the cache for layouts is words, so it splits the
input string at word boundaries (using a heuristic). There are is also
some refactoring to reduce the amount of allocation and copying, and
movement towards properly supporting contexts.
The size of the caches is a fixed number of entries; thus, it is
possible to consume a large amount of memory by filling the cache with
lots of large strings. This should be refined towards a scheme that
bounds the total memory used by the cache.
This patch fixes bug 15237293 "Regression: Measure performance is
significantly slower with minikin".
Change-Id: Ie8176857e2d78656ce5479a7c04969819ef2718d
This is a fix for bug 15130102 "Language name for Hebrew displayed the
wrong way around on keyboard".
This patch extends the previous BiDi support (when the direction for the
entire string was given by the caller) to run the BiDi algorithm
(provided by ICU) over the string to break it into BiDi runs. Thus, it
handles mixed LTR and RTL strings in a single layout, and also respects
heuristics for inferring the paragraph direction from the string.
Change-Id: Ia4b869de3c139c5a7d16b8ce7766870b98a815ea
(cherry picked from commit 4b3a941128454e55893d65433a835e78a9e9781d)
