Roll src/third_party/WebKit 06cb9e9:a978ee5 (svn 202558:202559)
[chromium-blink-merge.git] / base / i18n / break_iterator.cc
blobbc20fff928a6c77d1360446cfc95251e4134bd78
1 // Copyright (c) 2011 The Chromium Authors. All rights reserved.
2 // Use of this source code is governed by a BSD-style license that can be
3 // found in the LICENSE file.
5 #include "base/i18n/break_iterator.h"
7 #include "base/logging.h"
8 #include "third_party/icu/source/common/unicode/ubrk.h"
9 #include "third_party/icu/source/common/unicode/uchar.h"
10 #include "third_party/icu/source/common/unicode/ustring.h"
12 namespace base {
13 namespace i18n {
15 const size_t npos = static_cast<size_t>(-1);
17 BreakIterator::BreakIterator(const StringPiece16& str, BreakType break_type)
18 : iter_(NULL),
19 string_(str),
20 break_type_(break_type),
21 prev_(npos),
22 pos_(0) {
25 BreakIterator::BreakIterator(const StringPiece16& str, const string16& rules)
26 : iter_(NULL),
27 string_(str),
28 rules_(rules),
29 break_type_(RULE_BASED),
30 prev_(npos),
31 pos_(0) {
34 BreakIterator::~BreakIterator() {
35 if (iter_)
36 ubrk_close(static_cast<UBreakIterator*>(iter_));
39 bool BreakIterator::Init() {
40 UErrorCode status = U_ZERO_ERROR;
41 UParseError parse_error;
42 UBreakIteratorType break_type;
43 switch (break_type_) {
44 case BREAK_CHARACTER:
45 break_type = UBRK_CHARACTER;
46 break;
47 case BREAK_WORD:
48 break_type = UBRK_WORD;
49 break;
50 case BREAK_LINE:
51 case BREAK_NEWLINE:
52 case RULE_BASED: // (Keep compiler happy, break_type not used in this case)
53 break_type = UBRK_LINE;
54 break;
55 default:
56 NOTREACHED() << "invalid break_type_";
57 return false;
59 if (break_type_ == RULE_BASED) {
60 iter_ = ubrk_openRules(rules_.c_str(),
61 static_cast<int32_t>(rules_.length()),
62 string_.data(),
63 static_cast<int32_t>(string_.size()),
64 &parse_error,
65 &status);
66 if (U_FAILURE(status)) {
67 NOTREACHED() << "ubrk_openRules failed to parse rule string at line "
68 << parse_error.line << ", offset " << parse_error.offset;
70 } else {
71 iter_ = ubrk_open(break_type,
72 NULL,
73 string_.data(),
74 static_cast<int32_t>(string_.size()),
75 &status);
76 if (U_FAILURE(status)) {
77 NOTREACHED() << "ubrk_open failed for type " << break_type
78 << " with error " << status;
82 if (U_FAILURE(status)) {
83 return false;
86 // Move the iterator to the beginning of the string.
87 ubrk_first(static_cast<UBreakIterator*>(iter_));
88 return true;
91 bool BreakIterator::Advance() {
92 int32_t pos;
93 int32_t status;
94 prev_ = pos_;
95 switch (break_type_) {
96 case BREAK_CHARACTER:
97 case BREAK_WORD:
98 case BREAK_LINE:
99 case RULE_BASED:
100 pos = ubrk_next(static_cast<UBreakIterator*>(iter_));
101 if (pos == UBRK_DONE) {
102 pos_ = npos;
103 return false;
105 pos_ = static_cast<size_t>(pos);
106 return true;
107 case BREAK_NEWLINE:
108 do {
109 pos = ubrk_next(static_cast<UBreakIterator*>(iter_));
110 if (pos == UBRK_DONE)
111 break;
112 pos_ = static_cast<size_t>(pos);
113 status = ubrk_getRuleStatus(static_cast<UBreakIterator*>(iter_));
114 } while (status >= UBRK_LINE_SOFT && status < UBRK_LINE_SOFT_LIMIT);
115 if (pos == UBRK_DONE && prev_ == pos_) {
116 pos_ = npos;
117 return false;
119 return true;
120 default:
121 NOTREACHED() << "invalid break_type_";
122 return false;
126 bool BreakIterator::SetText(const base::char16* text, const size_t length) {
127 UErrorCode status = U_ZERO_ERROR;
128 ubrk_setText(static_cast<UBreakIterator*>(iter_),
129 text, length, &status);
130 pos_ = 0; // implicit when ubrk_setText is done
131 prev_ = npos;
132 if (U_FAILURE(status)) {
133 NOTREACHED() << "ubrk_setText failed";
134 return false;
136 string_ = StringPiece16(text, length);
137 return true;
140 bool BreakIterator::IsWord() const {
141 return GetWordBreakStatus() == IS_WORD_BREAK;
144 BreakIterator::WordBreakStatus BreakIterator::GetWordBreakStatus() const {
145 int32_t status = ubrk_getRuleStatus(static_cast<UBreakIterator*>(iter_));
146 if (break_type_ != BREAK_WORD && break_type_ != RULE_BASED)
147 return IS_LINE_OR_CHAR_BREAK;
148 return status == UBRK_WORD_NONE ? IS_SKIPPABLE_WORD : IS_WORD_BREAK;
151 bool BreakIterator::IsEndOfWord(size_t position) const {
152 if (break_type_ != BREAK_WORD && break_type_ != RULE_BASED)
153 return false;
155 UBreakIterator* iter = static_cast<UBreakIterator*>(iter_);
156 UBool boundary = ubrk_isBoundary(iter, static_cast<int32_t>(position));
157 int32_t status = ubrk_getRuleStatus(iter);
158 return (!!boundary && status != UBRK_WORD_NONE);
161 bool BreakIterator::IsStartOfWord(size_t position) const {
162 if (break_type_ != BREAK_WORD && break_type_ != RULE_BASED)
163 return false;
165 UBreakIterator* iter = static_cast<UBreakIterator*>(iter_);
166 UBool boundary = ubrk_isBoundary(iter, static_cast<int32_t>(position));
167 ubrk_next(iter);
168 int32_t next_status = ubrk_getRuleStatus(iter);
169 return (!!boundary && next_status != UBRK_WORD_NONE);
172 bool BreakIterator::IsGraphemeBoundary(size_t position) const {
173 if (break_type_ != BREAK_CHARACTER)
174 return false;
176 UBreakIterator* iter = static_cast<UBreakIterator*>(iter_);
177 return !!ubrk_isBoundary(iter, static_cast<int32_t>(position));
180 string16 BreakIterator::GetString() const {
181 return GetStringPiece().as_string();
184 StringPiece16 BreakIterator::GetStringPiece() const {
185 DCHECK(prev_ != npos && pos_ != npos);
186 return string_.substr(prev_, pos_ - prev_);
189 } // namespace i18n
190 } // namespace base