1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202
// Copyright 2015 The Servo Project Developers. See the
// COPYRIGHT file at the top-level directory of this distribution.
//
// Licensed under the Apache License, Version 2.0 <LICENSE-APACHE or
// http://www.apache.org/licenses/LICENSE-2.0> or the MIT license
// <LICENSE-MIT or http://opensource.org/licenses/MIT>, at your
// option. This file may not be copied, modified, or distributed
// except according to those terms.
//! This crate implements the [Unicode Bidirectional Algorithm][tr9] for display of mixed
//! right-to-left and left-to-right text. It is written in safe Rust, compatible with the
//! current stable release.
//!
//! ## Example
//!
//! ```rust
//! # #[cfg(feature = "hardcoded-data")] {
//! use unicode_bidi::BidiInfo;
//!
//! // This example text is defined using `concat!` because some browsers
//! // and text editors have trouble displaying bidi strings.
//! let text = concat![
//! "א",
//! "ב",
//! "ג",
//! "a",
//! "b",
//! "c",
//! ];
//!
//! // Resolve embedding levels within the text. Pass `None` to detect the
//! // paragraph level automatically.
//! let bidi_info = BidiInfo::new(&text, None);
//!
//! // This paragraph has embedding level 1 because its first strong character is RTL.
//! assert_eq!(bidi_info.paragraphs.len(), 1);
//! let para = &bidi_info.paragraphs[0];
//! assert_eq!(para.level.number(), 1);
//! assert_eq!(para.level.is_rtl(), true);
//!
//! // Re-ordering is done after wrapping each paragraph into a sequence of
//! // lines. For this example, I'll just use a single line that spans the
//! // entire paragraph.
//! let line = para.range.clone();
//!
//! let display = bidi_info.reorder_line(para, line);
//! assert_eq!(display, concat![
//! "a",
//! "b",
//! "c",
//! "ג",
//! "ב",
//! "א",
//! ]);
//! # } // feature = "hardcoded-data"
//! ```
//!
//! # Features
//!
//! - `std`: Enabled by default, but can be disabled to make `unicode_bidi`
//! `#![no_std]` + `alloc` compatible.
//! - `hardcoded-data`: Enabled by default. Includes hardcoded Unicode bidi data and more convenient APIs.
//! - `serde`: Adds [`serde::Serialize`] and [`serde::Deserialize`]
//! implementations to relevant types.
//!
//! [tr9]: <http://www.unicode.org/reports/tr9/>
#![no_std]
// We need to link to std to make doc tests work on older Rust versions
#[cfg(feature = "std")]
extern crate std;
#[macro_use]
extern crate alloc;
pub mod data_source;
pub mod deprecated;
pub mod format_chars;
pub mod level;
mod char_data;
mod explicit;
mod implicit;
mod prepare;
pub use crate::char_data::{BidiClass, UNICODE_VERSION};
pub use crate::data_source::BidiDataSource;
pub use crate::level::{Level, LTR_LEVEL, RTL_LEVEL};
pub use crate::prepare::LevelRun;
#[cfg(feature = "hardcoded-data")]
pub use crate::char_data::{bidi_class, HardcodedBidiData};
use alloc::borrow::Cow;
use alloc::string::String;
use alloc::vec::Vec;
use core::cmp;
use core::iter::repeat;
use core::ops::Range;
use crate::format_chars as chars;
use crate::BidiClass::*;
#[derive(PartialEq, Debug)]
pub enum Direction {
Ltr,
Rtl,
Mixed,
}
/// Bidi information about a single paragraph
#[derive(Debug, PartialEq)]
pub struct ParagraphInfo {
/// The paragraphs boundaries within the text, as byte indices.
///
/// TODO: Shrink this to only include the starting index?
pub range: Range<usize>,
/// The paragraph embedding level.
///
/// <http://www.unicode.org/reports/tr9/#BD4>
pub level: Level,
}
impl ParagraphInfo {
/// Gets the length of the paragraph in the source text.
pub fn len(&self) -> usize {
self.range.end - self.range.start
}
}
/// Initial bidi information of the text.
///
/// Contains the text paragraphs and `BidiClass` of its characters.
#[derive(PartialEq, Debug)]
pub struct InitialInfo<'text> {
/// The text
pub text: &'text str,
/// The BidiClass of the character at each byte in the text.
/// If a character is multiple bytes, its class will appear multiple times in the vector.
pub original_classes: Vec<BidiClass>,
/// The boundaries and level of each paragraph within the text.
pub paragraphs: Vec<ParagraphInfo>,
}
impl<'text> InitialInfo<'text> {
/// Find the paragraphs and BidiClasses in a string of text.
///
/// <http://www.unicode.org/reports/tr9/#The_Paragraph_Level>
///
/// Also sets the class for each First Strong Isolate initiator (FSI) to LRI or RLI if a strong
/// character is found before the matching PDI. If no strong character is found, the class will
/// remain FSI, and it's up to later stages to treat these as LRI when needed.
///
/// The `hardcoded-data` Cargo feature (enabled by default) must be enabled to use this.
#[cfg_attr(feature = "flame_it", flamer::flame)]
#[cfg(feature = "hardcoded-data")]
pub fn new(text: &str, default_para_level: Option<Level>) -> InitialInfo<'_> {
Self::new_with_data_source(&HardcodedBidiData, text, default_para_level)
}
/// Find the paragraphs and BidiClasses in a string of text, with a custom [`BidiDataSource`]
/// for Bidi data. If you just wish to use the hardcoded Bidi data, please use [`InitialInfo::new()`]
/// instead (enabled with tbe default `hardcoded-data` Cargo feature)
///
/// <http://www.unicode.org/reports/tr9/#The_Paragraph_Level>
///
/// Also sets the class for each First Strong Isolate initiator (FSI) to LRI or RLI if a strong
/// character is found before the matching PDI. If no strong character is found, the class will
/// remain FSI, and it's up to later stages to treat these as LRI when needed.
#[cfg_attr(feature = "flame_it", flamer::flame)]
pub fn new_with_data_source<'a, D: BidiDataSource>(
data_source: &D,
text: &'a str,
default_para_level: Option<Level>,
) -> InitialInfo<'a> {
let mut original_classes = Vec::with_capacity(text.len());
// The stack contains the starting byte index for each nested isolate we're inside.
let mut isolate_stack = Vec::new();
let mut paragraphs = Vec::new();
let mut para_start = 0;
let mut para_level = default_para_level;
#[cfg(feature = "flame_it")]
flame::start("InitialInfo::new(): iter text.char_indices()");
for (i, c) in text.char_indices() {
let class = data_source.bidi_class(c);
#[cfg(feature = "flame_it")]
flame::start("original_classes.extend()");
original_classes.extend(repeat(class).take(c.len_utf8()));
#[cfg(feature = "flame_it")]
flame::end("original_classes.extend()");
match class {
B => {
// P1. Split the text into separate paragraphs. The paragraph separator is kept
// with the previous paragraph.
let para_end = i + c.len_utf8();
paragraphs.push(ParagraphInfo {
range: para_start..para_end,
// P3. If no character is found in p2, set the paragraph level to zero.
level: para_level.unwrap_or(LTR_LEVEL),
});
// Reset state for the start of the next paragraph.
para_start = para_end;
// TODO: Support defaulting to direction of previous paragraph
//
// <http://www.unicode.org/reports/tr9/#HL1>
para_level = default_para_level;
isolate_stack.clear();
}
L | R | AL => {
match isolate_stack.last() {
Some(&start) => {
if original_classes[start] == FSI {
// X5c. If the first strong character between FSI and its matching
// PDI is R or AL, treat it as RLI. Otherwise, treat it as LRI.
for j in 0..chars::FSI.len_utf8() {
original_classes[start + j] =
if class == L { LRI } else { RLI };
}
}
}
None => {
if para_level.is_none() {
// P2. Find the first character of type L, AL, or R, while skipping
// any characters between an isolate initiator and its matching
// PDI.
para_level = Some(if class != L { RTL_LEVEL } else { LTR_LEVEL });
}
}
}
}
RLI | LRI | FSI => {
isolate_stack.push(i);
}
PDI => {
isolate_stack.pop();
}
_ => {}
}
}
if para_start < text.len() {
paragraphs.push(ParagraphInfo {
range: para_start..text.len(),
level: para_level.unwrap_or(LTR_LEVEL),
});
}
assert_eq!(original_classes.len(), text.len());
#[cfg(feature = "flame_it")]
flame::end("InitialInfo::new(): iter text.char_indices()");
InitialInfo {
text,
original_classes,
paragraphs,
}
}
}
/// Bidi information of the text.
///
/// The `original_classes` and `levels` vectors are indexed by byte offsets into the text. If a
/// character is multiple bytes wide, then its class and level will appear multiple times in these
/// vectors.
// TODO: Impl `struct StringProperty<T> { values: Vec<T> }` and use instead of Vec<T>
#[derive(Debug, PartialEq)]
pub struct BidiInfo<'text> {
/// The text
pub text: &'text str,
/// The BidiClass of the character at each byte in the text.
pub original_classes: Vec<BidiClass>,
/// The directional embedding level of each byte in the text.
pub levels: Vec<Level>,
/// The boundaries and paragraph embedding level of each paragraph within the text.
///
/// TODO: Use SmallVec or similar to avoid overhead when there are only one or two paragraphs?
/// Or just don't include the first paragraph, which always starts at 0?
pub paragraphs: Vec<ParagraphInfo>,
}
impl<'text> BidiInfo<'text> {
/// Split the text into paragraphs and determine the bidi embedding levels for each paragraph.
///
///
/// The `hardcoded-data` Cargo feature (enabled by default) must be enabled to use this.
///
/// TODO: In early steps, check for special cases that allow later steps to be skipped. like
/// text that is entirely LTR. See the `nsBidi` class from Gecko for comparison.
///
/// TODO: Support auto-RTL base direction
#[cfg_attr(feature = "flame_it", flamer::flame)]
#[cfg(feature = "hardcoded-data")]
pub fn new(text: &str, default_para_level: Option<Level>) -> BidiInfo<'_> {
Self::new_with_data_source(&HardcodedBidiData, text, default_para_level)
}
/// Split the text into paragraphs and determine the bidi embedding levels for each paragraph, with a custom [`BidiDataSource`]
/// for Bidi data. If you just wish to use the hardcoded Bidi data, please use [`BidiInfo::new()`]
/// instead (enabled with tbe default `hardcoded-data` Cargo feature).
///
/// TODO: In early steps, check for special cases that allow later steps to be skipped. like
/// text that is entirely LTR. See the `nsBidi` class from Gecko for comparison.
///
/// TODO: Support auto-RTL base direction
#[cfg_attr(feature = "flame_it", flamer::flame)]
pub fn new_with_data_source<'a, D: BidiDataSource>(
data_source: &D,
text: &'a str,
default_para_level: Option<Level>,
) -> BidiInfo<'a> {
let InitialInfo {
original_classes,
paragraphs,
..
} = InitialInfo::new_with_data_source(data_source, text, default_para_level);
let mut levels = Vec::<Level>::with_capacity(text.len());
let mut processing_classes = original_classes.clone();
for para in ¶graphs {
let text = &text[para.range.clone()];
let original_classes = &original_classes[para.range.clone()];
let processing_classes = &mut processing_classes[para.range.clone()];
let new_len = levels.len() + para.range.len();
levels.resize(new_len, para.level);
let levels = &mut levels[para.range.clone()];
explicit::compute(
text,
para.level,
original_classes,
levels,
processing_classes,
);
let sequences = prepare::isolating_run_sequences(para.level, original_classes, levels);
for sequence in &sequences {
implicit::resolve_weak(text, sequence, processing_classes);
implicit::resolve_neutral(
text,
data_source,
sequence,
levels,
original_classes,
processing_classes,
);
}
implicit::resolve_levels(processing_classes, levels);
assign_levels_to_removed_chars(para.level, original_classes, levels);
}
BidiInfo {
text,
original_classes,
paragraphs,
levels,
}
}
/// Re-order a line based on resolved levels and return only the embedding levels, one `Level`
/// per *byte*.
#[cfg_attr(feature = "flame_it", flamer::flame)]
pub fn reordered_levels(&self, para: &ParagraphInfo, line: Range<usize>) -> Vec<Level> {
let (levels, _) = self.visual_runs(para, line);
levels
}
/// Re-order a line based on resolved levels and return only the embedding levels, one `Level`
/// per *character*.
#[cfg_attr(feature = "flame_it", flamer::flame)]
pub fn reordered_levels_per_char(
&self,
para: &ParagraphInfo,
line: Range<usize>,
) -> Vec<Level> {
let levels = self.reordered_levels(para, line);
self.text.char_indices().map(|(i, _)| levels[i]).collect()
}
/// Re-order a line based on resolved levels and return the line in display order.
#[cfg_attr(feature = "flame_it", flamer::flame)]
pub fn reorder_line(&self, para: &ParagraphInfo, line: Range<usize>) -> Cow<'text, str> {
let (levels, runs) = self.visual_runs(para, line.clone());
// If all isolating run sequences are LTR, no reordering is needed
if runs.iter().all(|run| levels[run.start].is_ltr()) {
return self.text[line].into();
}
let mut result = String::with_capacity(line.len());
for run in runs {
if levels[run.start].is_rtl() {
result.extend(self.text[run].chars().rev());
} else {
result.push_str(&self.text[run]);
}
}
result.into()
}
/// Reorders pre-calculated levels of a sequence of characters.
///
/// NOTE: This is a convenience method that does not use a `Paragraph` object. It is
/// intended to be used when an application has determined the levels of the objects (character sequences)
/// and just needs to have them reordered.
///
/// the index map will result in `indexMap[visualIndex]==logicalIndex`.
///
/// This only runs [Rule L2](http://www.unicode.org/reports/tr9/#L2) as it does not have
/// information about the actual text.
///
/// Furthermore, if `levels` is an array that is aligned with code units, bytes within a codepoint may be
/// reversed. You may need to fix up the map to deal with this. Alternatively, only pass in arrays where each `Level`
/// is for a single code point.
///
///
/// # # Example
/// ```
/// use unicode_bidi::BidiInfo;
/// use unicode_bidi::Level;
///
/// let l0 = Level::from(0);
/// let l1 = Level::from(1);
/// let l2 = Level::from(2);
///
/// let levels = vec![l0, l0, l0, l0];
/// let index_map = BidiInfo::reorder_visual(&levels);
/// assert_eq!(levels.len(), index_map.len());
/// assert_eq!(index_map, [0, 1, 2, 3]);
///
/// let levels: Vec<Level> = vec![l0, l0, l0, l1, l1, l1, l2, l2];
/// let index_map = BidiInfo::reorder_visual(&levels);
/// assert_eq!(levels.len(), index_map.len());
/// assert_eq!(index_map, [0, 1, 2, 6, 7, 5, 4, 3]);
/// ```
pub fn reorder_visual(levels: &[Level]) -> Vec<usize> {
// Gets the next range of characters after start_index with a level greater
// than or equal to `max`
fn next_range(levels: &[level::Level], mut start_index: usize, max: Level) -> Range<usize> {
if levels.is_empty() || start_index >= levels.len() {
return start_index..start_index;
}
while let Some(l) = levels.get(start_index) {
if *l >= max {
break;
}
start_index += 1;
}
if levels.get(start_index).is_none() {
// If at the end of the array, adding one will
// produce an out-of-range end element
return start_index..start_index;
}
let mut end_index = start_index + 1;
while let Some(l) = levels.get(end_index) {
if *l < max {
return start_index..end_index;
}
end_index += 1;
}
start_index..end_index
}
// This implementation is similar to the L2 implementation in `visual_runs()`
// but it cannot benefit from a precalculated LevelRun vector so needs to be different.
if levels.is_empty() {
return vec![];
}
// Get the min and max levels
let (mut min, mut max) = levels
.iter()
.fold((levels[0], levels[0]), |(min, max), &l| {
(cmp::min(min, l), cmp::max(max, l))
});
// Initialize an index map
let mut result: Vec<usize> = (0..levels.len()).collect();
if min == max && min.is_ltr() {
// Everything is LTR and at the same level, do nothing
return result;
}
// Stop at the lowest *odd* level, since everything below that
// is LTR and does not need further reordering
min = min.new_lowest_ge_rtl().expect("Level error");
// For each max level, take all contiguous chunks of
// levels ≥ max and reverse them
//
// We can do this check with the original levels instead of checking reorderings because all
// prior reorderings will have been for contiguous chunks of levels >> max, which will
// be a subset of these chunks anyway.
while min <= max {
let mut range = 0..0;
loop {
range = next_range(levels, range.end, max);
result[range.clone()].reverse();
if range.end >= levels.len() {
break;
}
}
max.lower(1).expect("Level error");
}
result
}
/// Find the level runs within a line and return them in visual order.
///
/// `line` is a range of bytes indices within `levels`.
///
/// <http://www.unicode.org/reports/tr9/#Reordering_Resolved_Levels>
#[cfg_attr(feature = "flame_it", flamer::flame)]
pub fn visual_runs(
&self,
para: &ParagraphInfo,
line: Range<usize>,
) -> (Vec<Level>, Vec<LevelRun>) {
assert!(line.start <= self.levels.len());
assert!(line.end <= self.levels.len());
let mut levels = self.levels.clone();
let line_classes = &self.original_classes[line.clone()];
let line_levels = &mut levels[line.clone()];
// Reset some whitespace chars to paragraph level.
// <http://www.unicode.org/reports/tr9/#L1>
let line_str: &str = &self.text[line.clone()];
let mut reset_from: Option<usize> = Some(0);
let mut reset_to: Option<usize> = None;
let mut prev_level = para.level;
for (i, c) in line_str.char_indices() {
match line_classes[i] {
// Segment separator, Paragraph separator
B | S => {
assert_eq!(reset_to, None);
reset_to = Some(i + c.len_utf8());
if reset_from == None {
reset_from = Some(i);
}
}
// Whitespace, isolate formatting
WS | FSI | LRI | RLI | PDI => {
if reset_from == None {
reset_from = Some(i);
}
}
// <https://www.unicode.org/reports/tr9/#Retaining_Explicit_Formatting_Characters>
// same as above + set the level
RLE | LRE | RLO | LRO | PDF | BN => {
if reset_from == None {
reset_from = Some(i);
}
// also set the level to previous
line_levels[i] = prev_level;
}
_ => {
reset_from = None;
}
}
if let (Some(from), Some(to)) = (reset_from, reset_to) {
for level in &mut line_levels[from..to] {
*level = para.level;
}
reset_from = None;
reset_to = None;
}
prev_level = line_levels[i];
}
if let Some(from) = reset_from {
for level in &mut line_levels[from..] {
*level = para.level;
}
}
// Find consecutive level runs.
let mut runs = Vec::new();
let mut start = line.start;
let mut run_level = levels[start];
let mut min_level = run_level;
let mut max_level = run_level;
for (i, &new_level) in levels.iter().enumerate().take(line.end).skip(start + 1) {
if new_level != run_level {
// End of the previous run, start of a new one.
runs.push(start..i);
start = i;
run_level = new_level;
min_level = cmp::min(run_level, min_level);
max_level = cmp::max(run_level, max_level);
}
}
runs.push(start..line.end);
let run_count = runs.len();
// Re-order the odd runs.
// <http://www.unicode.org/reports/tr9/#L2>
// Stop at the lowest *odd* level.
min_level = min_level.new_lowest_ge_rtl().expect("Level error");
// This loop goes through contiguous chunks of level runs that have a level
// ≥ max_level and reverses their contents, reducing max_level by 1 each time.
//
// It can do this check with the original levels instead of checking reorderings because all
// prior reorderings will have been for contiguous chunks of levels >> max, which will
// be a subset of these chunks anyway.
while max_level >= min_level {
// Look for the start of a sequence of consecutive runs of max_level or higher.
let mut seq_start = 0;
while seq_start < run_count {
if self.levels[runs[seq_start].start] < max_level {
seq_start += 1;
continue;
}
// Found the start of a sequence. Now find the end.
let mut seq_end = seq_start + 1;
while seq_end < run_count {
if self.levels[runs[seq_end].start] < max_level {
break;
}
seq_end += 1;
}
// Reverse the runs within this sequence.
runs[seq_start..seq_end].reverse();
seq_start = seq_end;
}
max_level
.lower(1)
.expect("Lowering embedding level below zero");
}
(levels, runs)
}
/// If processed text has any computed RTL levels
///
/// This information is usually used to skip re-ordering of text when no RTL level is present
#[inline]
pub fn has_rtl(&self) -> bool {
level::has_rtl(&self.levels)
}
}
/// Contains a reference of `BidiInfo` and one of its `paragraphs`.
/// And it supports all operation in the `Paragraph` that needs also its
/// `BidiInfo` such as `direction`.
#[derive(Debug)]
pub struct Paragraph<'a, 'text> {
pub info: &'a BidiInfo<'text>,
pub para: &'a ParagraphInfo,
}
impl<'a, 'text> Paragraph<'a, 'text> {
pub fn new(info: &'a BidiInfo<'text>, para: &'a ParagraphInfo) -> Paragraph<'a, 'text> {
Paragraph { info, para }
}
/// Returns if the paragraph is Left direction, right direction or mixed.
pub fn direction(&self) -> Direction {
let mut ltr = false;
let mut rtl = false;
for i in self.para.range.clone() {
if self.info.levels[i].is_ltr() {
ltr = true;
}
if self.info.levels[i].is_rtl() {
rtl = true;
}
}
if ltr && rtl {
return Direction::Mixed;
}
if ltr {
return Direction::Ltr;
}
Direction::Rtl
}
/// Returns the `Level` of a certain character in the paragraph.
pub fn level_at(&self, pos: usize) -> Level {
let actual_position = self.para.range.start + pos;
self.info.levels[actual_position]
}
}
/// Assign levels to characters removed by rule X9.
///
/// The levels assigned to these characters are not specified by the algorithm. This function
/// assigns each one the level of the previous character, to avoid breaking level runs.
#[cfg_attr(feature = "flame_it", flamer::flame)]
fn assign_levels_to_removed_chars(para_level: Level, classes: &[BidiClass], levels: &mut [Level]) {
for i in 0..levels.len() {
if prepare::removed_by_x9(classes[i]) {
levels[i] = if i > 0 { levels[i - 1] } else { para_level };
}
}
}
#[cfg(test)]
#[cfg(feature = "hardcoded-data")]
mod tests {
use super::*;
#[test]
fn test_initial_text_info() {
let text = "a1";
assert_eq!(
InitialInfo::new(text, None),
InitialInfo {
text,
original_classes: vec![L, EN],
paragraphs: vec![ParagraphInfo {
range: 0..2,
level: LTR_LEVEL,
},],
}
);
let text = "غ א";
assert_eq!(
InitialInfo::new(text, None),
InitialInfo {
text,
original_classes: vec![AL, AL, WS, R, R],
paragraphs: vec![ParagraphInfo {
range: 0..5,
level: RTL_LEVEL,
},],
}
);
let text = "a\u{2029}b";
assert_eq!(
InitialInfo::new(text, None),
InitialInfo {
text,
original_classes: vec![L, B, B, B, L],
paragraphs: vec![
ParagraphInfo {
range: 0..4,
level: LTR_LEVEL,
},
ParagraphInfo {
range: 4..5,
level: LTR_LEVEL,
},
],
}
);
let text = format!("{}א{}a", chars::FSI, chars::PDI);
assert_eq!(
InitialInfo::new(&text, None),
InitialInfo {
text: &text,
original_classes: vec![RLI, RLI, RLI, R, R, PDI, PDI, PDI, L],
paragraphs: vec![ParagraphInfo {
range: 0..9,
level: LTR_LEVEL,
},],
}
);
}
#[test]
#[cfg(feature = "hardcoded-data")]
fn test_process_text() {
let text = "abc123";
assert_eq!(
BidiInfo::new(text, Some(LTR_LEVEL)),
BidiInfo {
text,
levels: Level::vec(&[0, 0, 0, 0, 0, 0]),
original_classes: vec![L, L, L, EN, EN, EN],
paragraphs: vec![ParagraphInfo {
range: 0..6,
level: LTR_LEVEL,
},],
}
);
let text = "abc אבג";
assert_eq!(
BidiInfo::new(text, Some(LTR_LEVEL)),
BidiInfo {
text,
levels: Level::vec(&[0, 0, 0, 0, 1, 1, 1, 1, 1, 1]),
original_classes: vec![L, L, L, WS, R, R, R, R, R, R],
paragraphs: vec![ParagraphInfo {
range: 0..10,
level: LTR_LEVEL,
},],
}
);
assert_eq!(
BidiInfo::new(text, Some(RTL_LEVEL)),
BidiInfo {
text,
levels: Level::vec(&[2, 2, 2, 1, 1, 1, 1, 1, 1, 1]),
original_classes: vec![L, L, L, WS, R, R, R, R, R, R],
paragraphs: vec![ParagraphInfo {
range: 0..10,
level: RTL_LEVEL,
},],
}
);
let text = "אבג abc";
assert_eq!(
BidiInfo::new(text, Some(LTR_LEVEL)),
BidiInfo {
text,
levels: Level::vec(&[1, 1, 1, 1, 1, 1, 0, 0, 0, 0]),
original_classes: vec![R, R, R, R, R, R, WS, L, L, L],
paragraphs: vec![ParagraphInfo {
range: 0..10,
level: LTR_LEVEL,
},],
}
);
assert_eq!(
BidiInfo::new(text, None),
BidiInfo {
text,
levels: Level::vec(&[1, 1, 1, 1, 1, 1, 1, 2, 2, 2]),
original_classes: vec![R, R, R, R, R, R, WS, L, L, L],
paragraphs: vec![ParagraphInfo {
range: 0..10,
level: RTL_LEVEL,
},],
}
);
let text = "غ2ظ א2ג";
assert_eq!(
BidiInfo::new(text, Some(LTR_LEVEL)),
BidiInfo {
text,
levels: Level::vec(&[1, 1, 2, 1, 1, 1, 1, 1, 2, 1, 1]),
original_classes: vec![AL, AL, EN, AL, AL, WS, R, R, EN, R, R],
paragraphs: vec![ParagraphInfo {
range: 0..11,
level: LTR_LEVEL,
},],
}
);
let text = "a א.\nג";
assert_eq!(
BidiInfo::new(text, None),
BidiInfo {
text,
original_classes: vec![L, WS, R, R, CS, B, R, R],
levels: Level::vec(&[0, 0, 1, 1, 0, 0, 1, 1]),
paragraphs: vec![
ParagraphInfo {
range: 0..6,
level: LTR_LEVEL,
},
ParagraphInfo {
range: 6..8,
level: RTL_LEVEL,
},
],
}
);
// BidiTest:69635 (AL ET EN)
let bidi_info = BidiInfo::new("\u{060B}\u{20CF}\u{06F9}", None);
assert_eq!(bidi_info.original_classes, vec![AL, AL, ET, ET, ET, EN, EN]);
}
#[test]
#[cfg(feature = "hardcoded-data")]
fn test_bidi_info_has_rtl() {
// ASCII only
assert_eq!(BidiInfo::new("123", None).has_rtl(), false);
assert_eq!(BidiInfo::new("123", Some(LTR_LEVEL)).has_rtl(), false);
assert_eq!(BidiInfo::new("123", Some(RTL_LEVEL)).has_rtl(), false);
assert_eq!(BidiInfo::new("abc", None).has_rtl(), false);
assert_eq!(BidiInfo::new("abc", Some(LTR_LEVEL)).has_rtl(), false);
assert_eq!(BidiInfo::new("abc", Some(RTL_LEVEL)).has_rtl(), false);
assert_eq!(BidiInfo::new("abc 123", None).has_rtl(), false);
assert_eq!(BidiInfo::new("abc\n123", None).has_rtl(), false);
// With Hebrew
assert_eq!(BidiInfo::new("אבּג", None).has_rtl(), true);
assert_eq!(BidiInfo::new("אבּג", Some(LTR_LEVEL)).has_rtl(), true);
assert_eq!(BidiInfo::new("אבּג", Some(RTL_LEVEL)).has_rtl(), true);
assert_eq!(BidiInfo::new("abc אבּג", None).has_rtl(), true);
assert_eq!(BidiInfo::new("abc\nאבּג", None).has_rtl(), true);
assert_eq!(BidiInfo::new("אבּג abc", None).has_rtl(), true);
assert_eq!(BidiInfo::new("אבּג\nabc", None).has_rtl(), true);
assert_eq!(BidiInfo::new("אבּג 123", None).has_rtl(), true);
assert_eq!(BidiInfo::new("אבּג\n123", None).has_rtl(), true);
}
#[cfg(feature = "hardcoded-data")]
fn reorder_paras(text: &str) -> Vec<Cow<'_, str>> {
let bidi_info = BidiInfo::new(text, None);
bidi_info
.paragraphs
.iter()
.map(|para| bidi_info.reorder_line(para, para.range.clone()))
.collect()
}
#[test]
#[cfg(feature = "hardcoded-data")]
fn test_reorder_line() {
// Bidi_Class: L L L B L L L B L L L
assert_eq!(
reorder_paras("abc\ndef\nghi"),
vec!["abc\n", "def\n", "ghi"]
);
// Bidi_Class: L L EN B L L EN B L L EN
assert_eq!(
reorder_paras("ab1\nde2\ngh3"),
vec!["ab1\n", "de2\n", "gh3"]
);
// Bidi_Class: L L L B AL AL AL
assert_eq!(reorder_paras("abc\nابج"), vec!["abc\n", "جبا"]);
// Bidi_Class: AL AL AL B L L L
assert_eq!(reorder_paras("ابج\nabc"), vec!["\nجبا", "abc"]);
assert_eq!(reorder_paras("1.-2"), vec!["1.-2"]);
assert_eq!(reorder_paras("1-.2"), vec!["1-.2"]);
assert_eq!(reorder_paras("abc אבג"), vec!["abc גבא"]);
// Numbers being weak LTR characters, cannot reorder strong RTL
assert_eq!(reorder_paras("123 אבג"), vec!["גבא 123"]);
assert_eq!(reorder_paras("abc\u{202A}def"), vec!["abc\u{202A}def"]);
assert_eq!(
reorder_paras("abc\u{202A}def\u{202C}ghi"),
vec!["abc\u{202A}def\u{202C}ghi"]
);
assert_eq!(
reorder_paras("abc\u{2066}def\u{2069}ghi"),
vec!["abc\u{2066}def\u{2069}ghi"]
);
// Testing for RLE Character
assert_eq!(
reorder_paras("\u{202B}abc אבג\u{202C}"),
vec!["\u{202B}\u{202C}גבא abc"]
);
// Testing neutral characters
assert_eq!(reorder_paras("אבג? אבג"), vec!["גבא ?גבא"]);
// Testing neutral characters with special case
assert_eq!(reorder_paras("A אבג?"), vec!["A גבא?"]);
// Testing neutral characters with Implicit RTL Marker
assert_eq!(reorder_paras("A אבג?\u{200F}"), vec!["A \u{200F}?גבא"]);
assert_eq!(reorder_paras("אבג abc"), vec!["abc גבא"]);
assert_eq!(
reorder_paras("abc\u{2067}.-\u{2069}ghi"),
vec!["abc\u{2067}-.\u{2069}ghi"]
);
assert_eq!(
reorder_paras("Hello, \u{2068}\u{202E}world\u{202C}\u{2069}!"),
vec!["Hello, \u{2068}\u{202E}\u{202C}dlrow\u{2069}!"]
);
// With mirrorable characters in RTL run
assert_eq!(reorder_paras("א(ב)ג."), vec![".ג)ב(א"]);
// With mirrorable characters on level boundry
assert_eq!(reorder_paras("אב(גד[&ef].)gh"), vec!["gh).]ef&[דג(בא"]);
}
fn reordered_levels_for_paras(text: &str) -> Vec<Vec<Level>> {
let bidi_info = BidiInfo::new(text, None);
bidi_info
.paragraphs
.iter()
.map(|para| bidi_info.reordered_levels(para, para.range.clone()))
.collect()
}
fn reordered_levels_per_char_for_paras(text: &str) -> Vec<Vec<Level>> {
let bidi_info = BidiInfo::new(text, None);
bidi_info
.paragraphs
.iter()
.map(|para| bidi_info.reordered_levels_per_char(para, para.range.clone()))
.collect()
}
#[test]
#[cfg(feature = "hardcoded-data")]
fn test_reordered_levels() {
// BidiTest:946 (LRI PDI)
let text = "\u{2067}\u{2069}";
assert_eq!(
reordered_levels_for_paras(text),
vec![Level::vec(&[0, 0, 0, 0, 0, 0])]
);
assert_eq!(
reordered_levels_per_char_for_paras(text),
vec![Level::vec(&[0, 0])]
);
let text = "aa טֶ";
let bidi_info = BidiInfo::new(text, None);
assert_eq!(
bidi_info.reordered_levels(&bidi_info.paragraphs[0], 3..7),
Level::vec(&[0, 0, 0, 1, 1, 1, 1]),
)
/* TODO
/// BidiTest:69635 (AL ET EN)
let text = "\u{060B}\u{20CF}\u{06F9}";
assert_eq!(
reordered_levels_for_paras(text),
vec![Level::vec(&[1, 1, 1, 1, 1, 2, 2])]
);
assert_eq!(
reordered_levels_per_char_for_paras(text),
vec![Level::vec(&[1, 1, 2])]
);
*/
/* TODO
// BidiTest:291284 (AN RLI PDF R)
assert_eq!(
reordered_levels_per_char_for_paras("\u{0605}\u{2067}\u{202C}\u{0590}"),
vec![&["2", "0", "x", "1"]]
);
*/
}
#[test]
fn test_paragraph_info_len() {
let text = "hello world";
let bidi_info = BidiInfo::new(text, None);
assert_eq!(bidi_info.paragraphs.len(), 1);
assert_eq!(bidi_info.paragraphs[0].len(), text.len());
let text2 = "How are you";
let whole_text = format!("{}\n{}", text, text2);
let bidi_info = BidiInfo::new(&whole_text, None);
assert_eq!(bidi_info.paragraphs.len(), 2);
// The first paragraph include the paragraph separator.
// TODO: investigate if the paragraph separator character
// should not be part of any paragraph.
assert_eq!(bidi_info.paragraphs[0].len(), text.len() + 1);
assert_eq!(bidi_info.paragraphs[1].len(), text2.len());
}
#[test]
fn test_direction() {
let ltr_text = "hello world";
let rtl_text = "أهلا بكم";
let all_paragraphs = format!("{}\n{}\n{}{}", ltr_text, rtl_text, ltr_text, rtl_text);
let bidi_info = BidiInfo::new(&all_paragraphs, None);
assert_eq!(bidi_info.paragraphs.len(), 3);
let p_ltr = Paragraph::new(&bidi_info, &bidi_info.paragraphs[0]);
let p_rtl = Paragraph::new(&bidi_info, &bidi_info.paragraphs[1]);
let p_mixed = Paragraph::new(&bidi_info, &bidi_info.paragraphs[2]);
assert_eq!(p_ltr.direction(), Direction::Ltr);
assert_eq!(p_rtl.direction(), Direction::Rtl);
assert_eq!(p_mixed.direction(), Direction::Mixed);
}
#[test]
fn test_edge_cases_direction() {
// No paragraphs for empty text.
let empty = "";
let bidi_info = BidiInfo::new(empty, Option::from(RTL_LEVEL));
assert_eq!(bidi_info.paragraphs.len(), 0);
// The paragraph separator will take the value of the default direction
// which is left to right.
let empty = "\n";
let bidi_info = BidiInfo::new(empty, None);
assert_eq!(bidi_info.paragraphs.len(), 1);
let p = Paragraph::new(&bidi_info, &bidi_info.paragraphs[0]);
assert_eq!(p.direction(), Direction::Ltr);
// The paragraph separator will take the value of the given initial direction
// which is left to right.
let empty = "\n";
let bidi_info = BidiInfo::new(empty, Option::from(LTR_LEVEL));
assert_eq!(bidi_info.paragraphs.len(), 1);
let p = Paragraph::new(&bidi_info, &bidi_info.paragraphs[0]);
assert_eq!(p.direction(), Direction::Ltr);
// The paragraph separator will take the value of the given initial direction
// which is right to left.
let empty = "\n";
let bidi_info = BidiInfo::new(empty, Option::from(RTL_LEVEL));
assert_eq!(bidi_info.paragraphs.len(), 1);
let p = Paragraph::new(&bidi_info, &bidi_info.paragraphs[0]);
assert_eq!(p.direction(), Direction::Rtl);
}
#[test]
fn test_level_at() {
let ltr_text = "hello world";
let rtl_text = "أهلا بكم";
let all_paragraphs = format!("{}\n{}\n{}{}", ltr_text, rtl_text, ltr_text, rtl_text);
let bidi_info = BidiInfo::new(&all_paragraphs, None);
assert_eq!(bidi_info.paragraphs.len(), 3);
let p_ltr = Paragraph::new(&bidi_info, &bidi_info.paragraphs[0]);
let p_rtl = Paragraph::new(&bidi_info, &bidi_info.paragraphs[1]);
let p_mixed = Paragraph::new(&bidi_info, &bidi_info.paragraphs[2]);
assert_eq!(p_ltr.level_at(0), LTR_LEVEL);
assert_eq!(p_rtl.level_at(0), RTL_LEVEL);
assert_eq!(p_mixed.level_at(0), LTR_LEVEL);
assert_eq!(p_mixed.info.levels.len(), 54);
assert_eq!(p_mixed.para.range.start, 28);
assert_eq!(p_mixed.level_at(ltr_text.len()), RTL_LEVEL);
}
}
#[cfg(all(feature = "serde", feature = "hardcoded-data", test))]
mod serde_tests {
use super::*;
use serde_test::{assert_tokens, Token};
#[test]
fn test_levels() {
let text = "abc אבג";
let bidi_info = BidiInfo::new(text, None);
let levels = bidi_info.levels;
assert_eq!(text.as_bytes().len(), 10);
assert_eq!(levels.len(), 10);
assert_tokens(
&levels,
&[
Token::Seq { len: Some(10) },
Token::NewtypeStruct { name: "Level" },
Token::U8(0),
Token::NewtypeStruct { name: "Level" },
Token::U8(0),
Token::NewtypeStruct { name: "Level" },
Token::U8(0),
Token::NewtypeStruct { name: "Level" },
Token::U8(0),
Token::NewtypeStruct { name: "Level" },
Token::U8(1),
Token::NewtypeStruct { name: "Level" },
Token::U8(1),
Token::NewtypeStruct { name: "Level" },
Token::U8(1),
Token::NewtypeStruct { name: "Level" },
Token::U8(1),
Token::NewtypeStruct { name: "Level" },
Token::U8(1),
Token::NewtypeStruct { name: "Level" },
Token::U8(1),
Token::SeqEnd,
],
);
}
}