OLD | NEW |
| (Empty) |
1 // Copyright (c) 2010 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. | |
4 | |
5 #ifndef BASE_I18N_WORD_ITERATOR_H_ | |
6 #define BASE_I18N_WORD_ITERATOR_H_ | |
7 #pragma once | |
8 | |
9 #include <vector> | |
10 | |
11 #include "base/basictypes.h" | |
12 #include "base/string16.h" | |
13 | |
14 // The WordIterator class iterates through the words and word breaks | |
15 // in a UTF-16 string. | |
16 // | |
17 // It provides two modes, BREAK_WORD and BREAK_LINE, which modify how | |
18 // trailing non-word characters are aggregated into the returned word. | |
19 // | |
20 // Under BREAK_WORD mode (more common), the non-word characters are | |
21 // not included with a returned word (e.g. in the UTF-16 equivalent of | |
22 // the string " foo bar! ", the word breaks are at the periods in | |
23 // ". .foo. .bar.!. ."). | |
24 // | |
25 // Under BREAK_LINE mode (less common), the non-word characters are | |
26 // included in the word, breaking only when a space-equivalent character | |
27 // is encountered (e.g. in the UTF16-equivalent of the string " foo bar! ", | |
28 // the word breaks are at the periods in ". .foo .bar! ."). | |
29 // | |
30 // To extract the words from a string, move a BREAK_WORD WordIterator | |
31 // through the string and test whether IsWord() is true. E.g., | |
32 // WordIterator iter(&str, WordIterator::BREAK_WORD); | |
33 // if (!iter.Init()) return false; | |
34 // while (iter.Advance()) { | |
35 // if (iter.IsWord()) { | |
36 // // region [iter.prev(),iter.pos()) contains a word. | |
37 // VLOG(1) << "word: " << iter.GetWord(); | |
38 // } | |
39 // } | |
40 | |
41 | |
42 class WordIterator { | |
43 public: | |
44 enum BreakType { | |
45 BREAK_WORD, | |
46 BREAK_LINE | |
47 }; | |
48 | |
49 // Requires |str| to live as long as the WordIterator does. | |
50 WordIterator(const string16* str, BreakType break_type); | |
51 ~WordIterator(); | |
52 | |
53 // Init() must be called before any of the iterators are valid. | |
54 // Returns false if ICU failed to initialize. | |
55 bool Init(); | |
56 | |
57 // Return the current break position within the string, | |
58 // or WordIterator::npos when done. | |
59 size_t pos() const { return pos_; } | |
60 // Return the value of pos() returned before Advance() was last called. | |
61 size_t prev() const { return prev_; } | |
62 | |
63 // Advance to the next break. Returns false if we've run past the end of | |
64 // the string. (Note that the very last "word break" is after the final | |
65 // character in the string, and when we advance to that position it's the | |
66 // last time Advance() returns true.) | |
67 bool Advance(); | |
68 | |
69 // Returns true if the break we just hit is the end of a word. | |
70 // (Otherwise, the break iterator just skipped over e.g. whitespace | |
71 // or punctuation.) | |
72 bool IsWord() const; | |
73 | |
74 // Return the word between prev() and pos(). | |
75 // Advance() must have been called successfully at least once | |
76 // for pos() to have advanced to somewhere useful. | |
77 string16 GetWord() const; | |
78 | |
79 private: | |
80 // ICU iterator, avoiding ICU ubrk.h dependence. | |
81 // This is actually an ICU UBreakiterator* type, which turns out to be | |
82 // a typedef for a void* in the ICU headers. Using void* directly prevents | |
83 // callers from needing access to the ICU public headers directory. | |
84 void* iter_; | |
85 | |
86 // The string we're iterating over. | |
87 const string16* string_; | |
88 | |
89 // The breaking style (word/line). | |
90 BreakType break_type_; | |
91 | |
92 // Previous and current iterator positions. | |
93 size_t prev_, pos_; | |
94 | |
95 DISALLOW_COPY_AND_ASSIGN(WordIterator); | |
96 }; | |
97 | |
98 #endif // BASE_I18N_WORD_ITERATOR_H__ | |
OLD | NEW |