Home | History | Annotate | Download | only in text 
      1 /*
      2  * Copyright (C) 2006 Lars Knoll <lars (at) trolltech.com>
      3  * Copyright (C) 2007, 2011, 2012 Apple Inc. All rights reserved.
      4  *
      5  * This library is free software; you can redistribute it and/or
      6  * modify it under the terms of the GNU Library General Public
      7  * License as published by the Free Software Foundation; either
      8  * version 2 of the License, or (at your option) any later version.
      9  *
     10  * This library is distributed in the hope that it will be useful,
     11  * but WITHOUT ANY WARRANTY; without even the implied warranty of
     12  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
     13  * Library General Public License for more details.
     14  *
     15  * You should have received a copy of the GNU Library General Public License
     16  * along with this library; see the file COPYING.LIB.  If not, write to
     17  * the Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor,
     18  * Boston, MA 02110-1301, USA.
     19  *
     20  */
     21 
     22 #ifndef TextBreakIterator_h
     23 #define TextBreakIterator_h
     24 
     25 #include "platform/PlatformExport.h"
     26 #include "wtf/text/AtomicString.h"
     27 #include "wtf/unicode/Unicode.h"
     28 
     29 namespace WebCore {
     30 
     31 typedef icu::BreakIterator TextBreakIterator;
     32 
     33 // Note: The returned iterator is good only until you get another iterator, with the exception of acquireLineBreakIterator.
     34 
     35 // This is similar to character break iterator in most cases, but is subject to
     36 // platform UI conventions. One notable example where this can be different
     37 // from character break iterator is Thai prepend characters, see bug 24342.
     38 // Use this for insertion point and selection manipulations.
     39 PLATFORM_EXPORT TextBreakIterator* cursorMovementIterator(const UChar*, int length);
     40 
     41 PLATFORM_EXPORT TextBreakIterator* wordBreakIterator(const String&, int start, int length);
     42 PLATFORM_EXPORT TextBreakIterator* wordBreakIterator(const UChar*, int length);
     43 PLATFORM_EXPORT TextBreakIterator* acquireLineBreakIterator(const LChar*, int length, const AtomicString& locale, const UChar* priorContext, unsigned priorContextLength);
     44 PLATFORM_EXPORT TextBreakIterator* acquireLineBreakIterator(const UChar*, int length, const AtomicString& locale, const UChar* priorContext, unsigned priorContextLength);
     45 PLATFORM_EXPORT void releaseLineBreakIterator(TextBreakIterator*);
     46 PLATFORM_EXPORT TextBreakIterator* sentenceBreakIterator(const UChar*, int length);
     47 
     48 PLATFORM_EXPORT bool isWordTextBreak(TextBreakIterator*);
     49 
     50 const int TextBreakDone = -1;
     51 
     52 class PLATFORM_EXPORT LazyLineBreakIterator {
     53 public:
     54     LazyLineBreakIterator()
     55         : m_iterator(0)
     56         , m_cachedPriorContext(0)
     57         , m_cachedPriorContextLength(0)
     58     {
     59         resetPriorContext();
     60     }
     61 
     62     LazyLineBreakIterator(String string, const AtomicString& locale = AtomicString())
     63         : m_string(string)
     64         , m_locale(locale)
     65         , m_iterator(0)
     66         , m_cachedPriorContext(0)
     67         , m_cachedPriorContextLength(0)
     68     {
     69         resetPriorContext();
     70     }
     71 
     72     ~LazyLineBreakIterator()
     73     {
     74         if (m_iterator)
     75             releaseLineBreakIterator(m_iterator);
     76     }
     77 
     78     String string() const { return m_string; }
     79 
     80     UChar lastCharacter() const
     81     {
     82         COMPILE_ASSERT(WTF_ARRAY_LENGTH(m_priorContext) == 2, TextBreakIterator_unexpected_prior_context_length);
     83         return m_priorContext[1];
     84     }
     85 
     86     UChar secondToLastCharacter() const
     87     {
     88         COMPILE_ASSERT(WTF_ARRAY_LENGTH(m_priorContext) == 2, TextBreakIterator_unexpected_prior_context_length);
     89         return m_priorContext[0];
     90     }
     91 
     92     void setPriorContext(UChar last, UChar secondToLast)
     93     {
     94         COMPILE_ASSERT(WTF_ARRAY_LENGTH(m_priorContext) == 2, TextBreakIterator_unexpected_prior_context_length);
     95         m_priorContext[0] = secondToLast;
     96         m_priorContext[1] = last;
     97     }
     98 
     99     void updatePriorContext(UChar last)
    100     {
    101         COMPILE_ASSERT(WTF_ARRAY_LENGTH(m_priorContext) == 2, TextBreakIterator_unexpected_prior_context_length);
    102         m_priorContext[0] = m_priorContext[1];
    103         m_priorContext[1] = last;
    104     }
    105 
    106     void resetPriorContext()
    107     {
    108         COMPILE_ASSERT(WTF_ARRAY_LENGTH(m_priorContext) == 2, TextBreakIterator_unexpected_prior_context_length);
    109         m_priorContext[0] = 0;
    110         m_priorContext[1] = 0;
    111     }
    112 
    113     unsigned priorContextLength() const
    114     {
    115         unsigned priorContextLength = 0;
    116         COMPILE_ASSERT(WTF_ARRAY_LENGTH(m_priorContext) == 2, TextBreakIterator_unexpected_prior_context_length);
    117         if (m_priorContext[1]) {
    118             ++priorContextLength;
    119             if (m_priorContext[0])
    120                 ++priorContextLength;
    121         }
    122         return priorContextLength;
    123     }
    124 
    125     // Obtain text break iterator, possibly previously cached, where this iterator is (or has been)
    126     // initialized to use the previously stored string as the primary breaking context and using
    127     // previously stored prior context if non-empty.
    128     TextBreakIterator* get(unsigned priorContextLength)
    129     {
    130         ASSERT(priorContextLength <= priorContextCapacity);
    131         const UChar* priorContext = priorContextLength ? &m_priorContext[priorContextCapacity - priorContextLength] : 0;
    132         if (!m_iterator) {
    133             if (m_string.is8Bit())
    134                 m_iterator = acquireLineBreakIterator(m_string.characters8(), m_string.length(), m_locale, priorContext, priorContextLength);
    135             else
    136                 m_iterator = acquireLineBreakIterator(m_string.characters16(), m_string.length(), m_locale, priorContext, priorContextLength);
    137             m_cachedPriorContext = priorContext;
    138             m_cachedPriorContextLength = priorContextLength;
    139         } else if (priorContext != m_cachedPriorContext || priorContextLength != m_cachedPriorContextLength) {
    140             this->resetStringAndReleaseIterator(m_string, m_locale);
    141             return this->get(priorContextLength);
    142         }
    143         return m_iterator;
    144     }
    145 
    146     void resetStringAndReleaseIterator(String string, const AtomicString& locale)
    147     {
    148         if (m_iterator)
    149             releaseLineBreakIterator(m_iterator);
    150 
    151         m_string = string;
    152         m_locale = locale;
    153         m_iterator = 0;
    154         m_cachedPriorContext = 0;
    155         m_cachedPriorContextLength = 0;
    156     }
    157 
    158 private:
    159     static const unsigned priorContextCapacity = 2;
    160     String m_string;
    161     AtomicString m_locale;
    162     TextBreakIterator* m_iterator;
    163     UChar m_priorContext[priorContextCapacity];
    164     const UChar* m_cachedPriorContext;
    165     unsigned m_cachedPriorContextLength;
    166 };
    167 
    168 // Iterates over "extended grapheme clusters", as defined in UAX #29.
    169 // Note that platform implementations may be less sophisticated - e.g. ICU prior to
    170 // version 4.0 only supports "legacy grapheme clusters".
    171 // Use this for general text processing, e.g. string truncation.
    172 
    173 class PLATFORM_EXPORT NonSharedCharacterBreakIterator {
    174     WTF_MAKE_NONCOPYABLE(NonSharedCharacterBreakIterator);
    175 public:
    176     explicit NonSharedCharacterBreakIterator(const String&);
    177     NonSharedCharacterBreakIterator(const UChar*, unsigned length);
    178     ~NonSharedCharacterBreakIterator();
    179 
    180     int next();
    181     int current();
    182 
    183     bool isBreak(int offset) const;
    184     int preceding(int offset) const;
    185     int following(int offset) const;
    186 
    187     bool operator!() const
    188     {
    189         return !m_is8Bit && !m_iterator;
    190     }
    191 
    192 private:
    193     void createIteratorForBuffer(const UChar*, unsigned length);
    194 
    195     unsigned clusterLengthStartingAt(unsigned offset) const
    196     {
    197         ASSERT(m_is8Bit);
    198         // The only Latin-1 Extended Grapheme Cluster is CR LF
    199         return isCRBeforeLF(offset) ? 2 : 1;
    200     }
    201 
    202     bool isCRBeforeLF(unsigned offset) const
    203     {
    204         ASSERT(m_is8Bit);
    205         return m_charaters8[offset] == '\r' && offset + 1 < m_length && m_charaters8[offset + 1] == '\n';
    206     }
    207 
    208     bool isLFAfterCR(unsigned offset) const
    209     {
    210         ASSERT(m_is8Bit);
    211         return m_charaters8[offset] == '\n' && offset >= 1 && m_charaters8[offset - 1] == '\r';
    212     }
    213 
    214     bool m_is8Bit;
    215 
    216     // For 8 bit strings, we implement the iterator ourselves.
    217     const LChar* m_charaters8;
    218     unsigned m_offset;
    219     unsigned m_length;
    220 
    221     // For 16 bit strings, we use a TextBreakIterator.
    222     TextBreakIterator* m_iterator;
    223 };
    224 
    225 // Counts the number of grapheme clusters. A surrogate pair or a sequence
    226 // of a non-combining character and following combining characters is
    227 // counted as 1 grapheme cluster.
    228 PLATFORM_EXPORT unsigned numGraphemeClusters(const String&);
    229 // Returns the number of characters which will be less than or equal to
    230 // the specified grapheme cluster length.
    231 PLATFORM_EXPORT unsigned numCharactersInGraphemeClusters(const String&, unsigned);
    232 
    233 }
    234 
    235 #endif
    236