Auto merge of #16779 - behnam:bidi-0.3, r=mbrubeck

[gfx] [layout] [style] Upgrade unicode-bidi to 0.3

Depends on https://github.com/servo/unicode-bidi/pull/27 , which
upgrades `unicode-bidi` crate to `0.3.0`.

Summary of changes:

* Use `unicode_bidi::Level` (instead of `u8`) in all relevant places and
replace magic computations with (inline) method calls to Level API.

* Doing so required adding `unicode-bidi` crate dependency to two more
components here: `style` and `gfx`. IMHO, totally makes sense, as
replaces local integer manipulations/checks with well-tested ones
already available in a common dependency.

---
<!-- Thank you for contributing to Servo! Please replace each `[ ]` by `[X]` when the step is complete, and replace `__` with appropriate data: -->
- [X] `./mach build -d` does not report any errors
- [X] `./mach test-tidy` does not report any errors
- [ ] These changes fix #__ (github issue number if applicable). [N/A]

<!-- Either: -->
- [ ] There are tests for these changes OR
- [X] These changes do not require tests because `unicode-bidi` has its own tests and there's no logic change in this diff.

<!-- Also, please make sure that "Allow edits from maintainers" checkbox is checked, so that we can help you if you get stuck somewhere along the way.-->

<!-- Pull requests that do not address these steps are welcome, but they will require additional verification as part of the review process. -->

<!-- Reviewable:start -->
---
This change is [<img src="https://reviewable.io/review_button.svg" height="34" align="absmiddle" alt="Reviewable"/>](https://reviewable.io/reviews/servo/servo/16779)
<!-- Reviewable:end -->

Cargo.lock

@@ -1021,6 +1021,7 @@ dependencies = [
  "style_traits 0.0.1",
  "time 0.1.37 (registry+https://github.com/rust-lang/crates.io-index)",
  "truetype 0.26.0 (registry+https://github.com/rust-lang/crates.io-index)",
+ "unicode-bidi 0.3.1 (registry+https://github.com/rust-lang/crates.io-index)",
  "unicode-script 0.1.1 (registry+https://github.com/rust-lang/crates.io-index)",
  "webrender_traits 0.39.0 (git+https://github.com/servo/webrender)",
  "xi-unicode 0.1.0 (registry+https://github.com/rust-lang/crates.io-index)",
@@ -1231,11 +1232,11 @@ dependencies = [
 
 [[package]]
 name = "idna"
-version = "0.1.1"
+version = "0.1.2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 dependencies = [
  "matches 0.1.4 (registry+https://github.com/rust-lang/crates.io-index)",
- "unicode-bidi 0.2.5 (registry+https://github.com/rust-lang/crates.io-index)",
+ "unicode-bidi 0.3.1 (registry+https://github.com/rust-lang/crates.io-index)",
  "unicode-normalization 0.1.4 (registry+https://github.com/rust-lang/crates.io-index)",
 ]
 
@@ -1407,7 +1408,7 @@ dependencies = [
  "smallvec 0.3.3 (registry+https://github.com/rust-lang/crates.io-index)",
  "style 0.0.1",
  "style_traits 0.0.1",
- "unicode-bidi 0.2.5 (registry+https://github.com/rust-lang/crates.io-index)",
+ "unicode-bidi 0.3.1 (registry+https://github.com/rust-lang/crates.io-index)",
  "unicode-script 0.1.1 (registry+https://github.com/rust-lang/crates.io-index)",
  "webrender_traits 0.39.0 (git+https://github.com/servo/webrender)",
 ]
@@ -2517,6 +2518,14 @@ dependencies = [
  "serde 0.9.15 (registry+https://github.com/rust-lang/crates.io-index)",
 ]
 
+[[package]]
+name = "serde_test"
+version = "0.9.15"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+dependencies = [
+ "serde 0.9.15 (registry+https://github.com/rust-lang/crates.io-index)",
+]
+
 [[package]]
 name = "servo"
 version = "0.0.1"
@@ -2871,6 +2880,7 @@ dependencies = [
  "style_traits 0.0.1",
  "time 0.1.37 (registry+https://github.com/rust-lang/crates.io-index)",
  "toml 0.2.1 (registry+https://github.com/rust-lang/crates.io-index)",
+ "unicode-bidi 0.3.1 (registry+https://github.com/rust-lang/crates.io-index)",
  "unicode-segmentation 1.1.0 (registry+https://github.com/rust-lang/crates.io-index)",
  "walkdir 1.0.7 (registry+https://github.com/rust-lang/crates.io-index)",
 ]
@@ -3145,10 +3155,13 @@ dependencies = [
 
 [[package]]
 name = "unicode-bidi"
-version = "0.2.5"
+version = "0.3.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 dependencies = [
  "matches 0.1.4 (registry+https://github.com/rust-lang/crates.io-index)",
+ "serde 0.9.15 (registry+https://github.com/rust-lang/crates.io-index)",
+ "serde_derive 0.9.15 (registry+https://github.com/rust-lang/crates.io-index)",
+ "serde_test 0.9.15 (registry+https://github.com/rust-lang/crates.io-index)",
 ]
 
 [[package]]
@@ -3194,7 +3207,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
 dependencies = [
  "encoding 0.2.33 (registry+https://github.com/rust-lang/crates.io-index)",
  "heapsize 0.3.9 (registry+https://github.com/rust-lang/crates.io-index)",
- "idna 0.1.1 (registry+https://github.com/rust-lang/crates.io-index)",
+ "idna 0.1.2 (registry+https://github.com/rust-lang/crates.io-index)",
  "matches 0.1.4 (registry+https://github.com/rust-lang/crates.io-index)",
 ]
 
@@ -3549,7 +3562,7 @@ dependencies = [
 "checksum hyper 0.10.9 (registry+https://github.com/rust-lang/crates.io-index)" = "94da93321c171e26481afeebe8288757b0501901b7c5492648163d8ec4942ec5"
 "checksum hyper-openssl 0.2.6 (registry+https://github.com/rust-lang/crates.io-index)" = "85a372eb692590b3fe014c196c30f9f52d4c42f58cd49dd94caeee1593c9cc37"
 "checksum hyper_serde 0.6.2 (registry+https://github.com/rust-lang/crates.io-index)" = "a43d985c58afed6b59991932e1d9b5f2629472849f0062c0078d82fdc0b788bb"
-"checksum idna 0.1.1 (registry+https://github.com/rust-lang/crates.io-index)" = "6ac85ec3f80c8e4e99d9325521337e14ec7555c458a14e377d189659a427f375"
+"checksum idna 0.1.2 (registry+https://github.com/rust-lang/crates.io-index)" = "2233d4940b1f19f0418c158509cd7396b8d70a5db5705ce410914dc8fa603b37"
 "checksum image 0.12.4 (registry+https://github.com/rust-lang/crates.io-index)" = "d95816db758249fe16f23a4e23f1a3a817fe11892dbfd1c5836f625324702158"
 "checksum immeta 0.3.5 (registry+https://github.com/rust-lang/crates.io-index)" = "0b9260463a221bfe3f02100c56e2d14c050d5ffe7e44a43d0a1b2b1f2b523502"
 "checksum inflate 0.1.1 (registry+https://github.com/rust-lang/crates.io-index)" = "e7e0062d2dc2f17d2f13750d95316ae8a2ff909af0fda957084f5defd87c43bb"
@@ -3645,6 +3658,7 @@ dependencies = [
 "checksum serde_codegen_internals 0.14.2 (registry+https://github.com/rust-lang/crates.io-index)" = "bc888bd283bd2420b16ad0d860e35ad8acb21941180a83a189bb2046f9d00400"
 "checksum serde_derive 0.9.15 (registry+https://github.com/rust-lang/crates.io-index)" = "978fd866f4d4872084a81ccc35e275158351d3b9fe620074e7d7504b816b74ba"
 "checksum serde_json 0.9.10 (registry+https://github.com/rust-lang/crates.io-index)" = "ad8bcf487be7d2e15d3d543f04312de991d631cfe1b43ea0ade69e6a8a5b16a1"
+"checksum serde_test 0.9.15 (registry+https://github.com/rust-lang/crates.io-index)" = "8ef7adefe71d2d715b39b93aeeee729e61dae3ac109a6e40e35e6e38317075a4"
 "checksum servo-egl 0.2.1 (registry+https://github.com/rust-lang/crates.io-index)" = "21069a884c33fe6ee596975e1f3849ed88c4ec857fbaf11d33672d8ebe051217"
 "checksum servo-fontconfig 0.2.1 (registry+https://github.com/rust-lang/crates.io-index)" = "93f799b649b4a2bf362398910eca35240704c7e765e780349b2bb1070d892262"
 "checksum servo-fontconfig-sys 4.0.3 (registry+https://github.com/rust-lang/crates.io-index)" = "6be80777ee6edecbbbf8774c76e19dddfe336256c57a4ded06d6ad3df7be358e"
@@ -3690,7 +3704,7 @@ dependencies = [
 "checksum truetype 0.26.0 (registry+https://github.com/rust-lang/crates.io-index)" = "acec30350633d6dac9dc1a625786b6cbe9150664be941aac2c35ad7199eab877"
 "checksum typeable 0.1.2 (registry+https://github.com/rust-lang/crates.io-index)" = "1410f6f91f21d1612654e7cc69193b0334f909dcf2c790c4826254fbb86f8887"
 "checksum unicase 1.4.0 (registry+https://github.com/rust-lang/crates.io-index)" = "13a5906ca2b98c799f4b1ab4557b76367ebd6ae5ef14930ec841c74aed5f3764"
-"checksum unicode-bidi 0.2.5 (registry+https://github.com/rust-lang/crates.io-index)" = "d3a078ebdd62c0e71a709c3d53d2af693fe09fe93fbff8344aebe289b78f9032"
+"checksum unicode-bidi 0.3.1 (registry+https://github.com/rust-lang/crates.io-index)" = "c44d4e7ce691e2538b886bf33669fd6da1653a12d741b9390f351955c0949c03"
 "checksum unicode-normalization 0.1.4 (registry+https://github.com/rust-lang/crates.io-index)" = "e28fa37426fceeb5cf8f41ee273faa7c82c47dc8fba5853402841e665fcd86ff"
 "checksum unicode-script 0.1.1 (registry+https://github.com/rust-lang/crates.io-index)" = "e5430ae21ef212551680d0021fc7dbd936e8b268c5ea8fdae8814e0b2496d80f"
 "checksum unicode-segmentation 1.1.0 (registry+https://github.com/rust-lang/crates.io-index)" = "18127285758f0e2c6cf325bb3f3d138a12fee27de4f23e146cd6a179f26c2cf3"

components/gfx/Cargo.toml

@@ -37,6 +37,7 @@ smallvec = "0.3"
 style = {path = "../style"}
 style_traits = {path = "../style_traits"}
 time = "0.1.12"
+unicode-bidi = {version = "0.3", features = ["with_serde"]}
 unicode-script = {version = "0.1", features = ["harfbuzz"]}
 webrender_traits = {git = "https://github.com/servo/webrender", features = ["ipc"]}
 xi-unicode = "0.1.0"

components/gfx/lib.rs

@@ -69,6 +69,7 @@ extern crate smallvec;
 extern crate style;
 extern crate style_traits;
 extern crate time;
+extern crate unicode_bidi;
 extern crate unicode_script;
 extern crate webrender_traits;
 extern crate xi_unicode;

components/gfx/text/text_run.rs

@@ -13,6 +13,7 @@ use std::slice::Iter;
 use std::sync::Arc;
 use style::str::char_is_whitespace;
 use text::glyph::{ByteIndex, GlyphStore};
+use unicode_bidi as bidi;
 use webrender_traits;
 use xi_unicode::LineBreakIterator;
 
@@ -32,7 +33,7 @@ pub struct TextRun {
     pub font_key: webrender_traits::FontKey,
     /// The glyph runs that make up this text run.
     pub glyphs: Arc<Vec<GlyphRun>>,
-    pub bidi_level: u8,
+    pub bidi_level: bidi::Level,
     pub extra_word_spacing: Au,
 }
 
@@ -179,7 +180,7 @@ impl<'a> Iterator for CharacterSliceIterator<'a> {
 }
 
 impl<'a> TextRun {
-    pub fn new(font: &mut Font, text: String, options: &ShapingOptions, bidi_level: u8) -> TextRun {
+    pub fn new(font: &mut Font, text: String, options: &ShapingOptions, bidi_level: bidi::Level) -> TextRun {
         let glyphs = TextRun::break_and_shape(font, &text, options);
         TextRun {
             text: Arc::new(text),
@@ -340,7 +341,7 @@ impl<'a> TextRun {
     pub fn natural_word_slices_in_visual_order(&'a self, range: &Range<ByteIndex>)
                                         -> NaturalWordSliceIterator<'a> {
         // Iterate in reverse order if bidi level is RTL.
-        let reverse = self.bidi_level % 2 == 1;
+        let reverse = self.bidi_level.is_rtl();
 
         let index = if reverse {
             match self.index_of_first_glyph_run_containing(range.end() - ByteIndex(1)) {

components/layout/Cargo.toml

@@ -43,6 +43,6 @@ servo_url = {path = "../url"}
 smallvec = "0.3"
 style = {path = "../style"}
 style_traits = {path = "../style_traits"}
-unicode-bidi = "0.2"
+unicode-bidi = {version = "0.3", features = ["with_serde"]}
 unicode-script = {version = "0.1", features = ["harfbuzz"]}
 webrender_traits = {git = "https://github.com/servo/webrender", features = ["ipc"]}

components/layout/inline.rs

@@ -36,7 +36,7 @@ use style::logical_geometry::{LogicalRect, LogicalSize, WritingMode};
 use style::properties::{longhands, ServoComputedValues};
 use style::servo::restyle_damage::{BUBBLE_ISIZES, REFLOW, REFLOW_OUT_OF_FLOW, REPOSITION, RESOLVE_GENERATED_CONTENT};
 use text;
-use unicode_bidi;
+use unicode_bidi as bidi;
 
 /// `Line`s are represented as offsets into the child list, rather than
 /// as an object that "owns" fragments. Choosing a different set of line
@@ -95,7 +95,7 @@ pub struct Line {
     /// The bidirectional embedding level runs for this line, in visual order.
     ///
     /// Can be set to `None` if the line is 100% left-to-right.
-    pub visual_runs: Option<Vec<(Range<FragmentIndex>, u8)>>,
+    pub visual_runs: Option<Vec<(Range<FragmentIndex>, bidi::Level)>>,
 
     /// The bounds are the exact position and extents of the line with respect
     /// to the parent box.
@@ -293,22 +293,25 @@ impl LineBreaker {
         // (because we split fragments on level run boundaries during flow
         // construction), so we can build a level array with just one entry per
         // fragment.
-        let levels: Vec<u8> = self.new_fragments.iter().map(|fragment| match fragment.specific {
+        let levels: Vec<bidi::Level> = self.new_fragments.iter().map(
-            SpecificFragmentInfo::ScannedText(ref info) => info.run.bidi_level,
+            |fragment| match fragment.specific {
-            _ => para_level
+                SpecificFragmentInfo::ScannedText(ref info) => info.run.bidi_level,
-        }).collect();
+                _ => para_level
+            }
+        ).collect();
 
         let mut lines = mem::replace(&mut self.lines, Vec::new());
 
         // If everything is LTR, don't bother with reordering.
-        let has_rtl = levels.iter().cloned().any(unicode_bidi::is_rtl);
+        if bidi::level::has_rtl(&levels) {
-
-        if has_rtl {
             // Compute and store the visual ordering of the fragments within the
             // line.
             for line in &mut lines {
                 let range = line.range.begin().to_usize()..line.range.end().to_usize();
-                let runs = unicode_bidi::visual_runs(range, &levels);
+                // FIXME: Update to use BidiInfo::visual_runs, as this algorithm needs access to
+                // the original text and original BidiClass of its characters.
+                #[allow(deprecated)]
+                let runs = bidi::deprecated::visual_runs(range, &levels);
                 line.visual_runs = Some(runs.iter().map(|run| {
                     let start = FragmentIndex(run.start as isize);
                     let len = FragmentIndex(run.len() as isize);
@@ -957,11 +960,11 @@ impl InlineFlow {
             let (range, level) = match line.visual_runs {
                 Some(ref runs) if is_ltr => runs[run_idx],
                 Some(ref runs) => runs[run_count - run_idx - 1], // reverse order for RTL runs
-                None => (line.range, 0)
+                None => (line.range, bidi::Level::ltr())
             };
             // If the bidi embedding direction is opposite the layout direction, lay out this
             // run in reverse order.
-            let reverse = unicode_bidi::is_ltr(level) != is_ltr;
+            let reverse = level.is_ltr() != is_ltr;
             let fragment_indices = if reverse {
                 (range.end().get() - 1..range.begin().get() - 1).step_by(-1)
             } else {

components/layout/text.rs

@@ -28,7 +28,7 @@ use style::computed_values::{word_break, white_space};
 use style::logical_geometry::{LogicalSize, WritingMode};
 use style::properties::ServoComputedValues;
 use style::properties::style_structs;
-use unicode_bidi::{is_rtl, process_text};
+use unicode_bidi as bidi;
 use unicode_script::{Script, get_script};
 
 /// Returns the concatenated text of a list of unscanned text fragments.
@@ -74,10 +74,10 @@ impl TextRunScanner {
         // Calculate bidi embedding levels, so we can split bidirectional fragments for reordering.
         let text = text(&fragments);
         let para_level = fragments.front().unwrap().style.writing_mode.to_bidi_level();
-        let bidi_info = process_text(&text, Some(para_level));
+        let bidi_info = bidi::BidiInfo::new(&text, Some(para_level));
 
         // Optimization: If all the text is LTR, don't bother splitting on bidi levels.
-        let bidi_levels = if bidi_info.levels.iter().cloned().any(is_rtl) {
+        let bidi_levels = if bidi_info.has_rtl() {
             Some(&bidi_info.levels[..])
         } else {
             None
@@ -126,7 +126,7 @@ impl TextRunScanner {
                            font_context: &mut FontContext,
                            out_fragments: &mut Vec<Fragment>,
                            paragraph_bytes_processed: &mut usize,
-                           bidi_levels: Option<&[u8]>,
+                           bidi_levels: Option<&[bidi::Level]>,
                            mut last_whitespace: bool)
                            -> bool {
         debug!("TextRunScanner: flushing {} fragments in range", self.clump.len());
@@ -211,7 +211,7 @@ impl TextRunScanner {
 
                     let bidi_level = match bidi_levels {
                         Some(levels) => levels[*paragraph_bytes_processed],
-                        None => 0
+                        None => bidi::Level::ltr(),
                     };
 
                     // Break the run if the new character has a different explicit script than the
@@ -310,7 +310,7 @@ impl TextRunScanner {
             run_info_list.into_iter().map(|run_info| {
                 let mut options = options;
                 options.script = run_info.script;
-                if is_rtl(run_info.bidi_level) {
+                if run_info.bidi_level.is_rtl() {
                     options.flags.insert(RTL_FLAG);
                 }
                 let mut font = fontgroup.fonts.get(run_info.font_index).unwrap().borrow_mut();
@@ -522,7 +522,7 @@ struct RunInfo {
     /// The index of the applicable font in the font group.
     font_index: usize,
     /// The bidirection embedding level of this text run.
-    bidi_level: u8,
+    bidi_level: bidi::Level,
     /// The Unicode script property of this text run.
     script: Script,
 }
@@ -533,7 +533,7 @@ impl RunInfo {
             text: String::new(),
             insertion_point: None,
             font_index: 0,
-            bidi_level: 0,
+            bidi_level: bidi::Level::ltr(),
             script: Script::Common,
         }
     }

components/style/Cargo.toml

@@ -67,6 +67,7 @@ style_derive = {path = "../style_derive"}
 style_traits = {path = "../style_traits"}
 servo_url = {path = "../url", optional = true}
 time = "0.1"
+unicode-bidi = {version = "0.3", features = ["with_serde"]}
 unicode-segmentation = "1.0"
 
 [target.'cfg(windows)'.dependencies]

components/style/lib.rs

@@ -83,6 +83,7 @@ extern crate style_derive;
 #[macro_use]
 extern crate style_traits;
 extern crate time;
+extern crate unicode_bidi;
 #[allow(unused_extern_crates)]
 extern crate unicode_segmentation;
 

components/style/logical_geometry.rs

@@ -10,6 +10,7 @@ use euclid::side_offsets::SideOffsets2D;
 use std::cmp::{max, min};
 use std::fmt::{self, Debug, Error, Formatter};
 use std::ops::{Add, Sub};
+use unicode_bidi as bidi;
 
 pub enum BlockFlowDirection {
     TopToBottom,
@@ -131,9 +132,13 @@ impl WritingMode {
     #[inline]
     /// The default bidirectional embedding level for this writing mode.
     ///
-    /// Returns 0 if the mode is LTR, or 1 otherwise.
+    /// Returns bidi level 0 if the mode is LTR, or 1 otherwise.
-    pub fn to_bidi_level(&self) -> u8 {
+    pub fn to_bidi_level(&self) -> bidi::Level {
-        !self.is_bidi_ltr() as u8
+        if self.is_bidi_ltr() {
+            bidi::Level::ltr()
+        } else {
+            bidi::Level::rtl()
+        }
     }
 }