8c3112733a
Sync with https://github.com/tc39/proposal-intl-segmenter/pull/55 and https://github.com/tc39/test262/pull/1994 Bug: v8:8588 Change-Id: Ifc46779f156ea986e3dbe6a632a781bb7134112b Reviewed-on: https://chromium-review.googlesource.com/c/1374997 Reviewed-by: Sathya Gunasekaran <gsathya@chromium.org> Commit-Queue: Frank Tang <ftang@chromium.org> Cr-Commit-Position: refs/heads/master@{#58301}
46 lines
2.7 KiB
JavaScript
46 lines
2.7 KiB
JavaScript
// Copyright 2018 the V8 project authors. All rights reserved.
|
||
// Use of this source code is governed by a BSD-style license that can be
|
||
// found in the LICENSE file.
|
||
|
||
// Flags: --harmony-intl-segmenter
|
||
|
||
const seg = new Intl.Segmenter([], {granularity: "word"})
|
||
for (const text of [
|
||
"Hello world!", // English
|
||
" Hello world! ", // English with space before/after
|
||
" Hello world? Foo bar!", // English
|
||
"Jedovatou mambu objevila žena v zahrádkářské kolonii.", // Czech
|
||
"Việt Nam: Nhất thể hóa sẽ khác Trung Quốc?", // Vietnamese
|
||
"Σοβαρές ενστάσεις Κομισιόν για τον προϋπολογισμό της Ιταλίας", // Greek
|
||
"Решение Индии о покупке российских С-400 расценили как вызов США", // Russian
|
||
"הרופא שהציל נשים והנערה ששועבדה ע", // Hebrew,
|
||
"ترامب للملك سلمان: أنا جاد للغاية.. عليك دفع المزيد", // Arabic
|
||
"भारत की एस 400 मिसाइल के मुकाबले पाक की थाड, जानें कौन कितना ताकतवर", // Hindi
|
||
"ரெட் அலர்ட் எச்சரிக்கை; புதுச்சேரியில் நாளை அரசு விடுமுறை!", // Tamil
|
||
"'ఉత్తర్వులు అందే వరకు ఓటర్ల తుది జాబితాను వెబ్సైట్లో పెట్టవద్దు'", // Telugu
|
||
"台北》抹黑柯P失敗？朱學恒酸：姚文智氣pupu嗆大老闆", // Chinese
|
||
"วัดไทรตีระฆังเบาลงช่วงเข้าพรรษา เจ้าอาวาสเผยคนร้องเรียนรับผลกรรมแล้ว", // Thai
|
||
"九州北部の一部が暴風域に入りました(日直予報士 2018年10月06日) - 日本気象協会 tenki.jp", // Japanese
|
||
"법원 “다스 지분 처분권·수익권 모두 MB가 보유”", // Korean
|
||
]) {
|
||
let segments = [];
|
||
// Create another %SegmentIterator% to compare with result from the one that
|
||
// created in the for of loop.
|
||
let iter = seg.segment(text);
|
||
let prev = 0;
|
||
for (const v of seg.segment(text)) {
|
||
assertTrue(["word", "none"].includes(v.breakType), v.breakType);
|
||
assertEquals("string", typeof v.segment);
|
||
assertTrue(v.segment.length > 0);
|
||
segments.push(v.segment);
|
||
|
||
// manually advance the iter.
|
||
assertFalse(iter.following());
|
||
assertEquals(iter.breakType, v.breakType);
|
||
assertEquals(text.substring(prev, iter.index), v.segment);
|
||
prev = iter.index;
|
||
}
|
||
assertTrue(iter.following());
|
||
assertEquals(text, segments.join(''));
|
||
}
|