sl@0: /*
sl@0: **********************************************************************
sl@0: *   Copyright (C) 1998-2005, International Business Machines
sl@0: *   Corporation and others.  All Rights Reserved.
sl@0: **********************************************************************
sl@0: *
sl@0: * File unistr.h
sl@0: *
sl@0: * Modification History:
sl@0: *
sl@0: *   Date        Name        Description
sl@0: *   09/25/98    stephen     Creation.
sl@0: *   11/11/98    stephen     Changed per 11/9 code review.
sl@0: *   04/20/99    stephen     Overhauled per 4/16 code review.
sl@0: *   11/18/99    aliu        Made to inherit from Replaceable.  Added method
sl@0: *                           handleReplaceBetween(); other methods unchanged.
sl@0: *   06/25/01    grhoten     Remove dependency on iostream.
sl@0: ******************************************************************************
sl@0: */
sl@0: 
sl@0: #ifndef UNISTR_H
sl@0: #define UNISTR_H
sl@0: 
sl@0: /**
sl@0:  * \file 
sl@0:  * \brief C++ API: Unicode String 
sl@0:  */
sl@0: 
sl@0: #include "unicode/rep.h"
sl@0: 
sl@0: struct UConverter;          // unicode/ucnv.h
sl@0: class  StringThreadTest;
sl@0: 
sl@0: #ifndef U_COMPARE_CODE_POINT_ORDER
sl@0: /* see also ustring.h and unorm.h */
sl@0: /**
sl@0:  * Option bit for u_strCaseCompare, u_strcasecmp, unorm_compare, etc:
sl@0:  * Compare strings in code point order instead of code unit order.
sl@0:  * @stable ICU 2.2
sl@0:  */
sl@0: #define U_COMPARE_CODE_POINT_ORDER  0x8000
sl@0: #endif
sl@0: 
sl@0: #ifndef USTRING_H
sl@0: /* see ustring.h */
sl@0: U_STABLE int32_t U_EXPORT2
sl@0: u_strlen(const UChar *s);
sl@0: #endif
sl@0: 
sl@0: U_NAMESPACE_BEGIN
sl@0: 
sl@0: class Locale;               // unicode/locid.h
sl@0: class StringCharacterIterator;
sl@0: class BreakIterator;        // unicode/brkiter.h
sl@0: 
sl@0: /* The <iostream> include has been moved to unicode/ustream.h */
sl@0: 
sl@0: /**
sl@0:  * Constant to be used in the UnicodeString(char *, int32_t, EInvariant) constructor
sl@0:  * which constructs a Unicode string from an invariant-character char * string.
sl@0:  * About invariant characters see utypes.h.
sl@0:  * This constructor has no runtime dependency on conversion code and is
sl@0:  * therefore recommended over ones taking a charset name string
sl@0:  * (where the empty string "" indicates invariant-character conversion).
sl@0:  *
sl@0:  * @draft ICU 3.2
sl@0:  */
sl@0: #define US_INV UnicodeString::kInvariant
sl@0: 
sl@0: /**
sl@0:  * Unicode String literals in C++.
sl@0:  * Dependent on the platform properties, different UnicodeString
sl@0:  * constructors should be used to create a UnicodeString object from
sl@0:  * a string literal.
sl@0:  * The macros are defined for maximum performance.
sl@0:  * They work only for strings that contain "invariant characters", i.e.,
sl@0:  * only latin letters, digits, and some punctuation.
sl@0:  * See utypes.h for details.
sl@0:  *
sl@0:  * The string parameter must be a C string literal.
sl@0:  * The length of the string, not including the terminating
sl@0:  * <code>NUL</code>, must be specified as a constant.
sl@0:  * The U_STRING_DECL macro should be invoked exactly once for one
sl@0:  * such string variable before it is used.
sl@0:  * @stable ICU 2.0
sl@0:  */
sl@0: #if U_SIZEOF_WCHAR_T==U_SIZEOF_UCHAR && (U_CHARSET_FAMILY==U_ASCII_FAMILY || (U_SIZEOF_UCHAR == 2 && defined(U_WCHAR_IS_UTF16)))
sl@0: #   define UNICODE_STRING(cs, _length) UnicodeString(TRUE, (const UChar *)L ## cs, _length)
sl@0: #elif U_SIZEOF_UCHAR==1 && U_CHARSET_FAMILY==U_ASCII_FAMILY
sl@0: #   define UNICODE_STRING(cs, _length) UnicodeString(TRUE, (const UChar *)cs, _length)
sl@0: #else
sl@0: #   define UNICODE_STRING(cs, _length) UnicodeString(cs, _length, US_INV)
sl@0: #endif
sl@0: 
sl@0: /**
sl@0:  * Unicode String literals in C++.
sl@0:  * Dependent on the platform properties, different UnicodeString
sl@0:  * constructors should be used to create a UnicodeString object from
sl@0:  * a string literal.
sl@0:  * The macros are defined for improved performance.
sl@0:  * They work only for strings that contain "invariant characters", i.e.,
sl@0:  * only latin letters, digits, and some punctuation.
sl@0:  * See utypes.h for details.
sl@0:  *
sl@0:  * The string parameter must be a C string literal.
sl@0:  * @stable ICU 2.0
sl@0:  */
sl@0: #if U_SIZEOF_WCHAR_T==U_SIZEOF_UCHAR && (U_CHARSET_FAMILY==U_ASCII_FAMILY || (U_SIZEOF_UCHAR == 2 && defined(U_WCHAR_IS_UTF16)))
sl@0: #   define UNICODE_STRING_SIMPLE(cs) UnicodeString(TRUE, (const UChar *)L ## cs, -1)
sl@0: #elif U_SIZEOF_UCHAR==1 && U_CHARSET_FAMILY==U_ASCII_FAMILY
sl@0: #   define UNICODE_STRING_SIMPLE(cs) UnicodeString(TRUE, (const UChar *)cs, -1)
sl@0: #else
sl@0: #   define UNICODE_STRING_SIMPLE(cs) UnicodeString(cs, -1, US_INV)
sl@0: #endif
sl@0: 
sl@0: /**
sl@0:  * UnicodeString is a string class that stores Unicode characters directly and provides
sl@0:  * similar functionality as the Java String and StringBuffer classes.
sl@0:  * It is a concrete implementation of the abstract class Replaceable (for transliteration).
sl@0:  *
sl@0:  * The UnicodeString class is not suitable for subclassing.
sl@0:  *
sl@0:  * <p>For an overview of Unicode strings in C and C++ see the
sl@0:  * <a href="http://icu.sourceforge.net/userguide/strings.html">User Guide Strings chapter</a>.</p>
sl@0:  *
sl@0:  * <p>In ICU, a Unicode string consists of 16-bit Unicode <em>code units</em>.
sl@0:  * A Unicode character may be stored with either one code unit
sl@0:  * (the most common case) or with a matched pair of special code units
sl@0:  * ("surrogates"). The data type for code units is UChar. 
sl@0:  * For single-character handling, a Unicode character code <em>point</em> is a value
sl@0:  * in the range 0..0x10ffff. ICU uses the UChar32 type for code points.</p>
sl@0:  *
sl@0:  * <p>Indexes and offsets into and lengths of strings always count code units, not code points.
sl@0:  * This is the same as with multi-byte char* strings in traditional string handling.
sl@0:  * Operations on partial strings typically do not test for code point boundaries.
sl@0:  * If necessary, the user needs to take care of such boundaries by testing for the code unit
sl@0:  * values or by using functions like
sl@0:  * UnicodeString::getChar32Start() and UnicodeString::getChar32Limit()
sl@0:  * (or, in C, the equivalent macros U16_SET_CP_START() and U16_SET_CP_LIMIT(), see utf.h).</p>
sl@0:  *
sl@0:  * UnicodeString methods are more lenient with regard to input parameter values
sl@0:  * than other ICU APIs. In particular:
sl@0:  * - If indexes are out of bounds for a UnicodeString object
sl@0:  *   (<0 or >length()) then they are "pinned" to the nearest boundary.
sl@0:  * - If primitive string pointer values (e.g., const UChar * or char *)
sl@0:  *   for input strings are NULL, then those input string parameters are treated
sl@0:  *   as if they pointed to an empty string.
sl@0:  *   However, this is <em>not</em> the case for char * parameters for charset names
sl@0:  *   or other IDs.
sl@0:  * - Most UnicodeString methods do not take a UErrorCode parameter because
sl@0:  *   there are usually very few opportunities for failure other than a shortage
sl@0:  *   of memory, error codes in low-level C++ string methods would be inconvenient,
sl@0:  *   and the error code as the last parameter (ICU convention) would prevent
sl@0:  *   the use of default parameter values.
sl@0:  *   Instead, such methods set the UnicodeString into a "bogus" state
sl@0:  *   (see isBogus()) if an error occurs.
sl@0:  *
sl@0:  * In string comparisons, two UnicodeString objects that are both "bogus"
sl@0:  * compare equal (to be transitive and prevent endless loops in sorting),
sl@0:  * and a "bogus" string compares less than any non-"bogus" one.
sl@0:  *
sl@0:  * Const UnicodeString methods are thread-safe. Multiple threads can use
sl@0:  * const methods on the same UnicodeString object simultaneously,
sl@0:  * but non-const methods must not be called concurrently (in multiple threads)
sl@0:  * with any other (const or non-const) methods.
sl@0:  *
sl@0:  * Similarly, const UnicodeString & parameters are thread-safe.
sl@0:  * One object may be passed in as such a parameter concurrently in multiple threads.
sl@0:  * This includes the const UnicodeString & parameters for
sl@0:  * copy construction, assignment, and cloning.
sl@0:  *
sl@0:  * <p>UnicodeString uses several storage methods.
sl@0:  * String contents can be stored inside the UnicodeString object itself,
sl@0:  * in an allocated and shared buffer, or in an outside buffer that is "aliased".
sl@0:  * Most of this is done transparently, but careful aliasing in particular provides
sl@0:  * significant performance improvements.
sl@0:  * Also, the internal buffer is accessible via special functions.
sl@0:  * For details see the
sl@0:  * <a href="http://icu.sourceforge.net/userguide/strings.html">User Guide Strings chapter</a>.</p>
sl@0:  *
sl@0:  * @see utf.h
sl@0:  * @see CharacterIterator
sl@0:  * @stable ICU 2.0
sl@0:  */
sl@0: class U_COMMON_API UnicodeString : public Replaceable
sl@0: {
sl@0: public:
sl@0: 
sl@0:   /**
sl@0:    * Constant to be used in the UnicodeString(char *, int32_t, EInvariant) constructor
sl@0:    * which constructs a Unicode string from an invariant-character char * string.
sl@0:    * Use the macro US_INV instead of the full qualification for this value.
sl@0:    *
sl@0:    * @see US_INV
sl@0:    * @draft ICU 3.2
sl@0:    */
sl@0:   enum EInvariant {
sl@0:     /**
sl@0:      * @see EInvariant
sl@0:      * @draft ICU 3.2
sl@0:      */
sl@0:     kInvariant
sl@0:   };
sl@0: 
sl@0:   //========================================
sl@0:   // Read-only operations
sl@0:   //========================================
sl@0: 
sl@0:   /* Comparison - bitwise only - for international comparison use collation */
sl@0: 
sl@0:   /**
sl@0:    * Equality operator. Performs only bitwise comparison.
sl@0:    * @param text The UnicodeString to compare to this one.
sl@0:    * @return TRUE if <TT>text</TT> contains the same characters as this one,
sl@0:    * FALSE otherwise.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UBool operator== (const UnicodeString& text) const;
sl@0: 
sl@0:   /**
sl@0:    * Inequality operator. Performs only bitwise comparison.
sl@0:    * @param text The UnicodeString to compare to this one.
sl@0:    * @return FALSE if <TT>text</TT> contains the same characters as this one,
sl@0:    * TRUE otherwise.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UBool operator!= (const UnicodeString& text) const;
sl@0: 
sl@0:   /**
sl@0:    * Greater than operator. Performs only bitwise comparison.
sl@0:    * @param text The UnicodeString to compare to this one.
sl@0:    * @return TRUE if the characters in this are bitwise
sl@0:    * greater than the characters in <code>text</code>, FALSE otherwise
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UBool operator> (const UnicodeString& text) const;
sl@0: 
sl@0:   /**
sl@0:    * Less than operator. Performs only bitwise comparison.
sl@0:    * @param text The UnicodeString to compare to this one.
sl@0:    * @return TRUE if the characters in this are bitwise
sl@0:    * less than the characters in <code>text</code>, FALSE otherwise
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UBool operator< (const UnicodeString& text) const;
sl@0: 
sl@0:   /**
sl@0:    * Greater than or equal operator. Performs only bitwise comparison.
sl@0:    * @param text The UnicodeString to compare to this one.
sl@0:    * @return TRUE if the characters in this are bitwise
sl@0:    * greater than or equal to the characters in <code>text</code>, FALSE otherwise
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UBool operator>= (const UnicodeString& text) const;
sl@0: 
sl@0:   /**
sl@0:    * Less than or equal operator. Performs only bitwise comparison.
sl@0:    * @param text The UnicodeString to compare to this one.
sl@0:    * @return TRUE if the characters in this are bitwise
sl@0:    * less than or equal to the characters in <code>text</code>, FALSE otherwise
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UBool operator<= (const UnicodeString& text) const;
sl@0: 
sl@0:   /**
sl@0:    * Compare the characters bitwise in this UnicodeString to
sl@0:    * the characters in <code>text</code>.
sl@0:    * @param text The UnicodeString to compare to this one.
sl@0:    * @return The result of bitwise character comparison: 0 if this
sl@0:    * contains the same characters as <code>text</code>, -1 if the characters in
sl@0:    * this are bitwise less than the characters in <code>text</code>, +1 if the
sl@0:    * characters in this are bitwise greater than the characters
sl@0:    * in <code>text</code>.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int8_t compare(const UnicodeString& text) const;
sl@0: 
sl@0:   /**
sl@0:    * Compare the characters bitwise in the range
sl@0:    * [<TT>start</TT>, <TT>start + length</TT>) with the characters
sl@0:    * in <TT>text</TT>
sl@0:    * @param start the offset at which the compare operation begins
sl@0:    * @param length the number of characters of text to compare.
sl@0:    * @param text the other text to be compared against this string.
sl@0:    * @return The result of bitwise character comparison: 0 if this
sl@0:    * contains the same characters as <code>text</code>, -1 if the characters in
sl@0:    * this are bitwise less than the characters in <code>text</code>, +1 if the
sl@0:    * characters in this are bitwise greater than the characters
sl@0:    * in <code>text</code>.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int8_t compare(int32_t start,
sl@0:          int32_t length,
sl@0:          const UnicodeString& text) const;
sl@0: 
sl@0:   /**
sl@0:    * Compare the characters bitwise in the range
sl@0:    * [<TT>start</TT>, <TT>start + length</TT>) with the characters
sl@0:    * in <TT>srcText</TT> in the range
sl@0:    * [<TT>srcStart</TT>, <TT>srcStart + srcLength</TT>).
sl@0:    * @param start the offset at which the compare operation begins
sl@0:    * @param length the number of characters in this to compare.
sl@0:    * @param srcText the text to be compared
sl@0:    * @param srcStart the offset into <TT>srcText</TT> to start comparison
sl@0:    * @param srcLength the number of characters in <TT>src</TT> to compare
sl@0:    * @return The result of bitwise character comparison: 0 if this
sl@0:    * contains the same characters as <code>srcText</code>, -1 if the characters in
sl@0:    * this are bitwise less than the characters in <code>srcText</code>, +1 if the
sl@0:    * characters in this are bitwise greater than the characters
sl@0:    * in <code>srcText</code>.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:    inline int8_t compare(int32_t start,
sl@0:          int32_t length,
sl@0:          const UnicodeString& srcText,
sl@0:          int32_t srcStart,
sl@0:          int32_t srcLength) const;
sl@0: 
sl@0:   /**
sl@0:    * Compare the characters bitwise in this UnicodeString with the first
sl@0:    * <TT>srcLength</TT> characters in <TT>srcChars</TT>.
sl@0:    * @param srcChars The characters to compare to this UnicodeString.
sl@0:    * @param srcLength the number of characters in <TT>srcChars</TT> to compare
sl@0:    * @return The result of bitwise character comparison: 0 if this
sl@0:    * contains the same characters as <code>srcChars</code>, -1 if the characters in
sl@0:    * this are bitwise less than the characters in <code>srcChars</code>, +1 if the
sl@0:    * characters in this are bitwise greater than the characters
sl@0:    * in <code>srcChars</code>.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int8_t compare(const UChar *srcChars,
sl@0:          int32_t srcLength) const;
sl@0: 
sl@0:   /**
sl@0:    * Compare the characters bitwise in the range
sl@0:    * [<TT>start</TT>, <TT>start + length</TT>) with the first
sl@0:    * <TT>length</TT> characters in <TT>srcChars</TT>
sl@0:    * @param start the offset at which the compare operation begins
sl@0:    * @param length the number of characters to compare.
sl@0:    * @param srcChars the characters to be compared
sl@0:    * @return The result of bitwise character comparison: 0 if this
sl@0:    * contains the same characters as <code>srcChars</code>, -1 if the characters in
sl@0:    * this are bitwise less than the characters in <code>srcChars</code>, +1 if the
sl@0:    * characters in this are bitwise greater than the characters
sl@0:    * in <code>srcChars</code>.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int8_t compare(int32_t start,
sl@0:          int32_t length,
sl@0:          const UChar *srcChars) const;
sl@0: 
sl@0:   /**
sl@0:    * Compare the characters bitwise in the range
sl@0:    * [<TT>start</TT>, <TT>start + length</TT>) with the characters
sl@0:    * in <TT>srcChars</TT> in the range
sl@0:    * [<TT>srcStart</TT>, <TT>srcStart + srcLength</TT>).
sl@0:    * @param start the offset at which the compare operation begins
sl@0:    * @param length the number of characters in this to compare
sl@0:    * @param srcChars the characters to be compared
sl@0:    * @param srcStart the offset into <TT>srcChars</TT> to start comparison
sl@0:    * @param srcLength the number of characters in <TT>srcChars</TT> to compare
sl@0:    * @return The result of bitwise character comparison: 0 if this
sl@0:    * contains the same characters as <code>srcChars</code>, -1 if the characters in
sl@0:    * this are bitwise less than the characters in <code>srcChars</code>, +1 if the
sl@0:    * characters in this are bitwise greater than the characters
sl@0:    * in <code>srcChars</code>.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int8_t compare(int32_t start,
sl@0:          int32_t length,
sl@0:          const UChar *srcChars,
sl@0:          int32_t srcStart,
sl@0:          int32_t srcLength) const;
sl@0: 
sl@0:   /**
sl@0:    * Compare the characters bitwise in the range
sl@0:    * [<TT>start</TT>, <TT>limit</TT>) with the characters
sl@0:    * in <TT>srcText</TT> in the range
sl@0:    * [<TT>srcStart</TT>, <TT>srcLimit</TT>).
sl@0:    * @param start the offset at which the compare operation begins
sl@0:    * @param limit the offset immediately following the compare operation
sl@0:    * @param srcText the text to be compared
sl@0:    * @param srcStart the offset into <TT>srcText</TT> to start comparison
sl@0:    * @param srcLimit the offset into <TT>srcText</TT> to limit comparison
sl@0:    * @return The result of bitwise character comparison: 0 if this
sl@0:    * contains the same characters as <code>srcText</code>, -1 if the characters in
sl@0:    * this are bitwise less than the characters in <code>srcText</code>, +1 if the
sl@0:    * characters in this are bitwise greater than the characters
sl@0:    * in <code>srcText</code>.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int8_t compareBetween(int32_t start,
sl@0:             int32_t limit,
sl@0:             const UnicodeString& srcText,
sl@0:             int32_t srcStart,
sl@0:             int32_t srcLimit) const;
sl@0: 
sl@0:   /**
sl@0:    * Compare two Unicode strings in code point order.
sl@0:    * This is different in UTF-16 from how compare(), operator==, startsWith() etc. work
sl@0:    * if supplementary characters are present:
sl@0:    *
sl@0:    * In UTF-16, supplementary characters (with code points U+10000 and above) are
sl@0:    * stored with pairs of surrogate code units. These have values from 0xd800 to 0xdfff,
sl@0:    * which means that they compare as less than some other BMP characters like U+feff.
sl@0:    * This function compares Unicode strings in code point order.
sl@0:    * If either of the UTF-16 strings is malformed (i.e., it contains unpaired surrogates), then the result is not defined.
sl@0:    *
sl@0:    * @param text Another string to compare this one to.
sl@0:    * @return a negative/zero/positive integer corresponding to whether
sl@0:    * this string is less than/equal to/greater than the second one
sl@0:    * in code point order
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int8_t compareCodePointOrder(const UnicodeString& text) const;
sl@0: 
sl@0:   /**
sl@0:    * Compare two Unicode strings in code point order.
sl@0:    * This is different in UTF-16 from how compare(), operator==, startsWith() etc. work
sl@0:    * if supplementary characters are present:
sl@0:    *
sl@0:    * In UTF-16, supplementary characters (with code points U+10000 and above) are
sl@0:    * stored with pairs of surrogate code units. These have values from 0xd800 to 0xdfff,
sl@0:    * which means that they compare as less than some other BMP characters like U+feff.
sl@0:    * This function compares Unicode strings in code point order.
sl@0:    * If either of the UTF-16 strings is malformed (i.e., it contains unpaired surrogates), then the result is not defined.
sl@0:    *
sl@0:    * @param start The start offset in this string at which the compare operation begins.
sl@0:    * @param length The number of code units from this string to compare.
sl@0:    * @param srcText Another string to compare this one to.
sl@0:    * @return a negative/zero/positive integer corresponding to whether
sl@0:    * this string is less than/equal to/greater than the second one
sl@0:    * in code point order
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int8_t compareCodePointOrder(int32_t start,
sl@0:                                       int32_t length,
sl@0:                                       const UnicodeString& srcText) const;
sl@0: 
sl@0:   /**
sl@0:    * Compare two Unicode strings in code point order.
sl@0:    * This is different in UTF-16 from how compare(), operator==, startsWith() etc. work
sl@0:    * if supplementary characters are present:
sl@0:    *
sl@0:    * In UTF-16, supplementary characters (with code points U+10000 and above) are
sl@0:    * stored with pairs of surrogate code units. These have values from 0xd800 to 0xdfff,
sl@0:    * which means that they compare as less than some other BMP characters like U+feff.
sl@0:    * This function compares Unicode strings in code point order.
sl@0:    * If either of the UTF-16 strings is malformed (i.e., it contains unpaired surrogates), then the result is not defined.
sl@0:    *
sl@0:    * @param start The start offset in this string at which the compare operation begins.
sl@0:    * @param length The number of code units from this string to compare.
sl@0:    * @param srcText Another string to compare this one to.
sl@0:    * @param srcStart The start offset in that string at which the compare operation begins.
sl@0:    * @param srcLength The number of code units from that string to compare.
sl@0:    * @return a negative/zero/positive integer corresponding to whether
sl@0:    * this string is less than/equal to/greater than the second one
sl@0:    * in code point order
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:    inline int8_t compareCodePointOrder(int32_t start,
sl@0:                                        int32_t length,
sl@0:                                        const UnicodeString& srcText,
sl@0:                                        int32_t srcStart,
sl@0:                                        int32_t srcLength) const;
sl@0: 
sl@0:   /**
sl@0:    * Compare two Unicode strings in code point order.
sl@0:    * This is different in UTF-16 from how compare(), operator==, startsWith() etc. work
sl@0:    * if supplementary characters are present:
sl@0:    *
sl@0:    * In UTF-16, supplementary characters (with code points U+10000 and above) are
sl@0:    * stored with pairs of surrogate code units. These have values from 0xd800 to 0xdfff,
sl@0:    * which means that they compare as less than some other BMP characters like U+feff.
sl@0:    * This function compares Unicode strings in code point order.
sl@0:    * If either of the UTF-16 strings is malformed (i.e., it contains unpaired surrogates), then the result is not defined.
sl@0:    *
sl@0:    * @param srcChars A pointer to another string to compare this one to.
sl@0:    * @param srcLength The number of code units from that string to compare.
sl@0:    * @return a negative/zero/positive integer corresponding to whether
sl@0:    * this string is less than/equal to/greater than the second one
sl@0:    * in code point order
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int8_t compareCodePointOrder(const UChar *srcChars,
sl@0:                                       int32_t srcLength) const;
sl@0: 
sl@0:   /**
sl@0:    * Compare two Unicode strings in code point order.
sl@0:    * This is different in UTF-16 from how compare(), operator==, startsWith() etc. work
sl@0:    * if supplementary characters are present:
sl@0:    *
sl@0:    * In UTF-16, supplementary characters (with code points U+10000 and above) are
sl@0:    * stored with pairs of surrogate code units. These have values from 0xd800 to 0xdfff,
sl@0:    * which means that they compare as less than some other BMP characters like U+feff.
sl@0:    * This function compares Unicode strings in code point order.
sl@0:    * If either of the UTF-16 strings is malformed (i.e., it contains unpaired surrogates), then the result is not defined.
sl@0:    *
sl@0:    * @param start The start offset in this string at which the compare operation begins.
sl@0:    * @param length The number of code units from this string to compare.
sl@0:    * @param srcChars A pointer to another string to compare this one to.
sl@0:    * @return a negative/zero/positive integer corresponding to whether
sl@0:    * this string is less than/equal to/greater than the second one
sl@0:    * in code point order
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int8_t compareCodePointOrder(int32_t start,
sl@0:                                       int32_t length,
sl@0:                                       const UChar *srcChars) const;
sl@0: 
sl@0:   /**
sl@0:    * Compare two Unicode strings in code point order.
sl@0:    * This is different in UTF-16 from how compare(), operator==, startsWith() etc. work
sl@0:    * if supplementary characters are present:
sl@0:    *
sl@0:    * In UTF-16, supplementary characters (with code points U+10000 and above) are
sl@0:    * stored with pairs of surrogate code units. These have values from 0xd800 to 0xdfff,
sl@0:    * which means that they compare as less than some other BMP characters like U+feff.
sl@0:    * This function compares Unicode strings in code point order.
sl@0:    * If either of the UTF-16 strings is malformed (i.e., it contains unpaired surrogates), then the result is not defined.
sl@0:    *
sl@0:    * @param start The start offset in this string at which the compare operation begins.
sl@0:    * @param length The number of code units from this string to compare.
sl@0:    * @param srcChars A pointer to another string to compare this one to.
sl@0:    * @param srcStart The start offset in that string at which the compare operation begins.
sl@0:    * @param srcLength The number of code units from that string to compare.
sl@0:    * @return a negative/zero/positive integer corresponding to whether
sl@0:    * this string is less than/equal to/greater than the second one
sl@0:    * in code point order
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int8_t compareCodePointOrder(int32_t start,
sl@0:                                       int32_t length,
sl@0:                                       const UChar *srcChars,
sl@0:                                       int32_t srcStart,
sl@0:                                       int32_t srcLength) const;
sl@0: 
sl@0:   /**
sl@0:    * Compare two Unicode strings in code point order.
sl@0:    * This is different in UTF-16 from how compare(), operator==, startsWith() etc. work
sl@0:    * if supplementary characters are present:
sl@0:    *
sl@0:    * In UTF-16, supplementary characters (with code points U+10000 and above) are
sl@0:    * stored with pairs of surrogate code units. These have values from 0xd800 to 0xdfff,
sl@0:    * which means that they compare as less than some other BMP characters like U+feff.
sl@0:    * This function compares Unicode strings in code point order.
sl@0:    * If either of the UTF-16 strings is malformed (i.e., it contains unpaired surrogates), then the result is not defined.
sl@0:    *
sl@0:    * @param start The start offset in this string at which the compare operation begins.
sl@0:    * @param limit The offset after the last code unit from this string to compare.
sl@0:    * @param srcText Another string to compare this one to.
sl@0:    * @param srcStart The start offset in that string at which the compare operation begins.
sl@0:    * @param srcLimit The offset after the last code unit from that string to compare.
sl@0:    * @return a negative/zero/positive integer corresponding to whether
sl@0:    * this string is less than/equal to/greater than the second one
sl@0:    * in code point order
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int8_t compareCodePointOrderBetween(int32_t start,
sl@0:                                              int32_t limit,
sl@0:                                              const UnicodeString& srcText,
sl@0:                                              int32_t srcStart,
sl@0:                                              int32_t srcLimit) const;
sl@0: 
sl@0:   /**
sl@0:    * Compare two strings case-insensitively using full case folding.
sl@0:    * This is equivalent to this->foldCase(options).compare(text.foldCase(options)).
sl@0:    *
sl@0:    * @param text Another string to compare this one to.
sl@0:    * @param options A bit set of options:
sl@0:    *   - U_FOLD_CASE_DEFAULT or 0 is used for default options:
sl@0:    *     Comparison in code unit order with default case folding.
sl@0:    *
sl@0:    *   - U_COMPARE_CODE_POINT_ORDER
sl@0:    *     Set to choose code point order instead of code unit order
sl@0:    *     (see u_strCompare for details).
sl@0:    *
sl@0:    *   - U_FOLD_CASE_EXCLUDE_SPECIAL_I
sl@0:    *
sl@0:    * @return A negative, zero, or positive integer indicating the comparison result.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int8_t caseCompare(const UnicodeString& text, uint32_t options) const;
sl@0: 
sl@0:   /**
sl@0:    * Compare two strings case-insensitively using full case folding.
sl@0:    * This is equivalent to this->foldCase(options).compare(srcText.foldCase(options)).
sl@0:    *
sl@0:    * @param start The start offset in this string at which the compare operation begins.
sl@0:    * @param length The number of code units from this string to compare.
sl@0:    * @param srcText Another string to compare this one to.
sl@0:    * @param options A bit set of options:
sl@0:    *   - U_FOLD_CASE_DEFAULT or 0 is used for default options:
sl@0:    *     Comparison in code unit order with default case folding.
sl@0:    *
sl@0:    *   - U_COMPARE_CODE_POINT_ORDER
sl@0:    *     Set to choose code point order instead of code unit order
sl@0:    *     (see u_strCompare for details).
sl@0:    *
sl@0:    *   - U_FOLD_CASE_EXCLUDE_SPECIAL_I
sl@0:    *
sl@0:    * @return A negative, zero, or positive integer indicating the comparison result.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int8_t caseCompare(int32_t start,
sl@0:          int32_t length,
sl@0:          const UnicodeString& srcText,
sl@0:          uint32_t options) const;
sl@0: 
sl@0:   /**
sl@0:    * Compare two strings case-insensitively using full case folding.
sl@0:    * This is equivalent to this->foldCase(options).compare(srcText.foldCase(options)).
sl@0:    *
sl@0:    * @param start The start offset in this string at which the compare operation begins.
sl@0:    * @param length The number of code units from this string to compare.
sl@0:    * @param srcText Another string to compare this one to.
sl@0:    * @param srcStart The start offset in that string at which the compare operation begins.
sl@0:    * @param srcLength The number of code units from that string to compare.
sl@0:    * @param options A bit set of options:
sl@0:    *   - U_FOLD_CASE_DEFAULT or 0 is used for default options:
sl@0:    *     Comparison in code unit order with default case folding.
sl@0:    *
sl@0:    *   - U_COMPARE_CODE_POINT_ORDER
sl@0:    *     Set to choose code point order instead of code unit order
sl@0:    *     (see u_strCompare for details).
sl@0:    *
sl@0:    *   - U_FOLD_CASE_EXCLUDE_SPECIAL_I
sl@0:    *
sl@0:    * @return A negative, zero, or positive integer indicating the comparison result.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int8_t caseCompare(int32_t start,
sl@0:          int32_t length,
sl@0:          const UnicodeString& srcText,
sl@0:          int32_t srcStart,
sl@0:          int32_t srcLength,
sl@0:          uint32_t options) const;
sl@0: 
sl@0:   /**
sl@0:    * Compare two strings case-insensitively using full case folding.
sl@0:    * This is equivalent to this->foldCase(options).compare(srcChars.foldCase(options)).
sl@0:    *
sl@0:    * @param srcChars A pointer to another string to compare this one to.
sl@0:    * @param srcLength The number of code units from that string to compare.
sl@0:    * @param options A bit set of options:
sl@0:    *   - U_FOLD_CASE_DEFAULT or 0 is used for default options:
sl@0:    *     Comparison in code unit order with default case folding.
sl@0:    *
sl@0:    *   - U_COMPARE_CODE_POINT_ORDER
sl@0:    *     Set to choose code point order instead of code unit order
sl@0:    *     (see u_strCompare for details).
sl@0:    *
sl@0:    *   - U_FOLD_CASE_EXCLUDE_SPECIAL_I
sl@0:    *
sl@0:    * @return A negative, zero, or positive integer indicating the comparison result.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int8_t caseCompare(const UChar *srcChars,
sl@0:          int32_t srcLength,
sl@0:          uint32_t options) const;
sl@0: 
sl@0:   /**
sl@0:    * Compare two strings case-insensitively using full case folding.
sl@0:    * This is equivalent to this->foldCase(options).compare(srcChars.foldCase(options)).
sl@0:    *
sl@0:    * @param start The start offset in this string at which the compare operation begins.
sl@0:    * @param length The number of code units from this string to compare.
sl@0:    * @param srcChars A pointer to another string to compare this one to.
sl@0:    * @param options A bit set of options:
sl@0:    *   - U_FOLD_CASE_DEFAULT or 0 is used for default options:
sl@0:    *     Comparison in code unit order with default case folding.
sl@0:    *
sl@0:    *   - U_COMPARE_CODE_POINT_ORDER
sl@0:    *     Set to choose code point order instead of code unit order
sl@0:    *     (see u_strCompare for details).
sl@0:    *
sl@0:    *   - U_FOLD_CASE_EXCLUDE_SPECIAL_I
sl@0:    *
sl@0:    * @return A negative, zero, or positive integer indicating the comparison result.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int8_t caseCompare(int32_t start,
sl@0:          int32_t length,
sl@0:          const UChar *srcChars,
sl@0:          uint32_t options) const;
sl@0: 
sl@0:   /**
sl@0:    * Compare two strings case-insensitively using full case folding.
sl@0:    * This is equivalent to this->foldCase(options).compare(srcChars.foldCase(options)).
sl@0:    *
sl@0:    * @param start The start offset in this string at which the compare operation begins.
sl@0:    * @param length The number of code units from this string to compare.
sl@0:    * @param srcChars A pointer to another string to compare this one to.
sl@0:    * @param srcStart The start offset in that string at which the compare operation begins.
sl@0:    * @param srcLength The number of code units from that string to compare.
sl@0:    * @param options A bit set of options:
sl@0:    *   - U_FOLD_CASE_DEFAULT or 0 is used for default options:
sl@0:    *     Comparison in code unit order with default case folding.
sl@0:    *
sl@0:    *   - U_COMPARE_CODE_POINT_ORDER
sl@0:    *     Set to choose code point order instead of code unit order
sl@0:    *     (see u_strCompare for details).
sl@0:    *
sl@0:    *   - U_FOLD_CASE_EXCLUDE_SPECIAL_I
sl@0:    *
sl@0:    * @return A negative, zero, or positive integer indicating the comparison result.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int8_t caseCompare(int32_t start,
sl@0:          int32_t length,
sl@0:          const UChar *srcChars,
sl@0:          int32_t srcStart,
sl@0:          int32_t srcLength,
sl@0:          uint32_t options) const;
sl@0: 
sl@0:   /**
sl@0:    * Compare two strings case-insensitively using full case folding.
sl@0:    * This is equivalent to this->foldCase(options).compareBetween(text.foldCase(options)).
sl@0:    *
sl@0:    * @param start The start offset in this string at which the compare operation begins.
sl@0:    * @param limit The offset after the last code unit from this string to compare.
sl@0:    * @param srcText Another string to compare this one to.
sl@0:    * @param srcStart The start offset in that string at which the compare operation begins.
sl@0:    * @param srcLimit The offset after the last code unit from that string to compare.
sl@0:    * @param options A bit set of options:
sl@0:    *   - U_FOLD_CASE_DEFAULT or 0 is used for default options:
sl@0:    *     Comparison in code unit order with default case folding.
sl@0:    *
sl@0:    *   - U_COMPARE_CODE_POINT_ORDER
sl@0:    *     Set to choose code point order instead of code unit order
sl@0:    *     (see u_strCompare for details).
sl@0:    *
sl@0:    *   - U_FOLD_CASE_EXCLUDE_SPECIAL_I
sl@0:    *
sl@0:    * @return A negative, zero, or positive integer indicating the comparison result.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int8_t caseCompareBetween(int32_t start,
sl@0:             int32_t limit,
sl@0:             const UnicodeString& srcText,
sl@0:             int32_t srcStart,
sl@0:             int32_t srcLimit,
sl@0:             uint32_t options) const;
sl@0: 
sl@0:   /**
sl@0:    * Determine if this starts with the characters in <TT>text</TT>
sl@0:    * @param text The text to match.
sl@0:    * @return TRUE if this starts with the characters in <TT>text</TT>,
sl@0:    * FALSE otherwise
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UBool startsWith(const UnicodeString& text) const;
sl@0: 
sl@0:   /**
sl@0:    * Determine if this starts with the characters in <TT>srcText</TT>
sl@0:    * in the range [<TT>srcStart</TT>, <TT>srcStart + srcLength</TT>).
sl@0:    * @param srcText The text to match.
sl@0:    * @param srcStart the offset into <TT>srcText</TT> to start matching
sl@0:    * @param srcLength the number of characters in <TT>srcText</TT> to match
sl@0:    * @return TRUE if this starts with the characters in <TT>text</TT>,
sl@0:    * FALSE otherwise
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UBool startsWith(const UnicodeString& srcText,
sl@0:             int32_t srcStart,
sl@0:             int32_t srcLength) const;
sl@0: 
sl@0:   /**
sl@0:    * Determine if this starts with the characters in <TT>srcChars</TT>
sl@0:    * @param srcChars The characters to match.
sl@0:    * @param srcLength the number of characters in <TT>srcChars</TT>
sl@0:    * @return TRUE if this starts with the characters in <TT>srcChars</TT>,
sl@0:    * FALSE otherwise
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UBool startsWith(const UChar *srcChars,
sl@0:             int32_t srcLength) const;
sl@0: 
sl@0:   /**
sl@0:    * Determine if this ends with the characters in <TT>srcChars</TT>
sl@0:    * in the range  [<TT>srcStart</TT>, <TT>srcStart + srcLength</TT>).
sl@0:    * @param srcChars The characters to match.
sl@0:    * @param srcStart the offset into <TT>srcText</TT> to start matching
sl@0:    * @param srcLength the number of characters in <TT>srcChars</TT> to match
sl@0:    * @return TRUE if this ends with the characters in <TT>srcChars</TT>, FALSE otherwise
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UBool startsWith(const UChar *srcChars,
sl@0:             int32_t srcStart,
sl@0:             int32_t srcLength) const;
sl@0: 
sl@0:   /**
sl@0:    * Determine if this ends with the characters in <TT>text</TT>
sl@0:    * @param text The text to match.
sl@0:    * @return TRUE if this ends with the characters in <TT>text</TT>,
sl@0:    * FALSE otherwise
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UBool endsWith(const UnicodeString& text) const;
sl@0: 
sl@0:   /**
sl@0:    * Determine if this ends with the characters in <TT>srcText</TT>
sl@0:    * in the range [<TT>srcStart</TT>, <TT>srcStart + srcLength</TT>).
sl@0:    * @param srcText The text to match.
sl@0:    * @param srcStart the offset into <TT>srcText</TT> to start matching
sl@0:    * @param srcLength the number of characters in <TT>srcText</TT> to match
sl@0:    * @return TRUE if this ends with the characters in <TT>text</TT>,
sl@0:    * FALSE otherwise
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UBool endsWith(const UnicodeString& srcText,
sl@0:           int32_t srcStart,
sl@0:           int32_t srcLength) const;
sl@0: 
sl@0:   /**
sl@0:    * Determine if this ends with the characters in <TT>srcChars</TT>
sl@0:    * @param srcChars The characters to match.
sl@0:    * @param srcLength the number of characters in <TT>srcChars</TT>
sl@0:    * @return TRUE if this ends with the characters in <TT>srcChars</TT>,
sl@0:    * FALSE otherwise
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UBool endsWith(const UChar *srcChars,
sl@0:           int32_t srcLength) const;
sl@0: 
sl@0:   /**
sl@0:    * Determine if this ends with the characters in <TT>srcChars</TT>
sl@0:    * in the range  [<TT>srcStart</TT>, <TT>srcStart + srcLength</TT>).
sl@0:    * @param srcChars The characters to match.
sl@0:    * @param srcStart the offset into <TT>srcText</TT> to start matching
sl@0:    * @param srcLength the number of characters in <TT>srcChars</TT> to match
sl@0:    * @return TRUE if this ends with the characters in <TT>srcChars</TT>,
sl@0:    * FALSE otherwise
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UBool endsWith(const UChar *srcChars,
sl@0:           int32_t srcStart,
sl@0:           int32_t srcLength) const;
sl@0: 
sl@0: 
sl@0:   /* Searching - bitwise only */
sl@0: 
sl@0:   /**
sl@0:    * Locate in this the first occurrence of the characters in <TT>text</TT>,
sl@0:    * using bitwise comparison.
sl@0:    * @param text The text to search for.
sl@0:    * @return The offset into this of the start of <TT>text</TT>,
sl@0:    * or -1 if not found.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int32_t indexOf(const UnicodeString& text) const;
sl@0: 
sl@0:   /**
sl@0:    * Locate in this the first occurrence of the characters in <TT>text</TT>
sl@0:    * starting at offset <TT>start</TT>, using bitwise comparison.
sl@0:    * @param text The text to search for.
sl@0:    * @param start The offset at which searching will start.
sl@0:    * @return The offset into this of the start of <TT>text</TT>,
sl@0:    * or -1 if not found.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int32_t indexOf(const UnicodeString& text,
sl@0:               int32_t start) const;
sl@0: 
sl@0:   /**
sl@0:    * Locate in this the first occurrence in the range
sl@0:    * [<TT>start</TT>, <TT>start + length</TT>) of the characters
sl@0:    * in <TT>text</TT>, using bitwise comparison.
sl@0:    * @param text The text to search for.
sl@0:    * @param start The offset at which searching will start.
sl@0:    * @param length The number of characters to search
sl@0:    * @return The offset into this of the start of <TT>text</TT>,
sl@0:    * or -1 if not found.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int32_t indexOf(const UnicodeString& text,
sl@0:               int32_t start,
sl@0:               int32_t length) const;
sl@0: 
sl@0:   /**
sl@0:    * Locate in this the first occurrence in the range
sl@0:    * [<TT>start</TT>, <TT>start + length</TT>) of the characters
sl@0:    *  in <TT>srcText</TT> in the range
sl@0:    * [<TT>srcStart</TT>, <TT>srcStart + srcLength</TT>),
sl@0:    * using bitwise comparison.
sl@0:    * @param srcText The text to search for.
sl@0:    * @param srcStart the offset into <TT>srcText</TT> at which
sl@0:    * to start matching
sl@0:    * @param srcLength the number of characters in <TT>srcText</TT> to match
sl@0:    * @param start the offset into this at which to start matching
sl@0:    * @param length the number of characters in this to search
sl@0:    * @return The offset into this of the start of <TT>text</TT>,
sl@0:    * or -1 if not found.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int32_t indexOf(const UnicodeString& srcText,
sl@0:               int32_t srcStart,
sl@0:               int32_t srcLength,
sl@0:               int32_t start,
sl@0:               int32_t length) const;
sl@0: 
sl@0:   /**
sl@0:    * Locate in this the first occurrence of the characters in
sl@0:    * <TT>srcChars</TT>
sl@0:    * starting at offset <TT>start</TT>, using bitwise comparison.
sl@0:    * @param srcChars The text to search for.
sl@0:    * @param srcLength the number of characters in <TT>srcChars</TT> to match
sl@0:    * @param start the offset into this at which to start matching
sl@0:    * @return The offset into this of the start of <TT>text</TT>,
sl@0:    * or -1 if not found.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int32_t indexOf(const UChar *srcChars,
sl@0:               int32_t srcLength,
sl@0:               int32_t start) const;
sl@0: 
sl@0:   /**
sl@0:    * Locate in this the first occurrence in the range
sl@0:    * [<TT>start</TT>, <TT>start + length</TT>) of the characters
sl@0:    * in <TT>srcChars</TT>, using bitwise comparison.
sl@0:    * @param srcChars The text to search for.
sl@0:    * @param srcLength the number of characters in <TT>srcChars</TT>
sl@0:    * @param start The offset at which searching will start.
sl@0:    * @param length The number of characters to search
sl@0:    * @return The offset into this of the start of <TT>srcChars</TT>,
sl@0:    * or -1 if not found.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int32_t indexOf(const UChar *srcChars,
sl@0:               int32_t srcLength,
sl@0:               int32_t start,
sl@0:               int32_t length) const;
sl@0: 
sl@0:   /**
sl@0:    * Locate in this the first occurrence in the range
sl@0:    * [<TT>start</TT>, <TT>start + length</TT>) of the characters
sl@0:    * in <TT>srcChars</TT> in the range
sl@0:    * [<TT>srcStart</TT>, <TT>srcStart + srcLength</TT>),
sl@0:    * using bitwise comparison.
sl@0:    * @param srcChars The text to search for.
sl@0:    * @param srcStart the offset into <TT>srcChars</TT> at which
sl@0:    * to start matching
sl@0:    * @param srcLength the number of characters in <TT>srcChars</TT> to match
sl@0:    * @param start the offset into this at which to start matching
sl@0:    * @param length the number of characters in this to search
sl@0:    * @return The offset into this of the start of <TT>text</TT>,
sl@0:    * or -1 if not found.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   int32_t indexOf(const UChar *srcChars,
sl@0:               int32_t srcStart,
sl@0:               int32_t srcLength,
sl@0:               int32_t start,
sl@0:               int32_t length) const;
sl@0: 
sl@0:   /**
sl@0:    * Locate in this the first occurrence of the BMP code point <code>c</code>,
sl@0:    * using bitwise comparison.
sl@0:    * @param c The code unit to search for.
sl@0:    * @return The offset into this of <TT>c</TT>, or -1 if not found.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int32_t indexOf(UChar c) const;
sl@0: 
sl@0:   /**
sl@0:    * Locate in this the first occurrence of the code point <TT>c</TT>,
sl@0:    * using bitwise comparison.
sl@0:    *
sl@0:    * @param c The code point to search for.
sl@0:    * @return The offset into this of <TT>c</TT>, or -1 if not found.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int32_t indexOf(UChar32 c) const;
sl@0: 
sl@0:   /**
sl@0:    * Locate in this the first occurrence of the BMP code point <code>c</code>,
sl@0:    * starting at offset <TT>start</TT>, using bitwise comparison.
sl@0:    * @param c The code unit to search for.
sl@0:    * @param start The offset at which searching will start.
sl@0:    * @return The offset into this of <TT>c</TT>, or -1 if not found.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int32_t indexOf(UChar c,
sl@0:               int32_t start) const;
sl@0: 
sl@0:   /**
sl@0:    * Locate in this the first occurrence of the code point <TT>c</TT>
sl@0:    * starting at offset <TT>start</TT>, using bitwise comparison.
sl@0:    *
sl@0:    * @param c The code point to search for.
sl@0:    * @param start The offset at which searching will start.
sl@0:    * @return The offset into this of <TT>c</TT>, or -1 if not found.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int32_t indexOf(UChar32 c,
sl@0:               int32_t start) const;
sl@0: 
sl@0:   /**
sl@0:    * Locate in this the first occurrence of the BMP code point <code>c</code>
sl@0:    * in the range [<TT>start</TT>, <TT>start + length</TT>),
sl@0:    * using bitwise comparison.
sl@0:    * @param c The code unit to search for.
sl@0:    * @param start the offset into this at which to start matching
sl@0:    * @param length the number of characters in this to search
sl@0:    * @return The offset into this of <TT>c</TT>, or -1 if not found.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int32_t indexOf(UChar c,
sl@0:               int32_t start,
sl@0:               int32_t length) const;
sl@0: 
sl@0:   /**
sl@0:    * Locate in this the first occurrence of the code point <TT>c</TT>
sl@0:    * in the range [<TT>start</TT>, <TT>start + length</TT>),
sl@0:    * using bitwise comparison.
sl@0:    *
sl@0:    * @param c The code point to search for.
sl@0:    * @param start the offset into this at which to start matching
sl@0:    * @param length the number of characters in this to search
sl@0:    * @return The offset into this of <TT>c</TT>, or -1 if not found.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int32_t indexOf(UChar32 c,
sl@0:               int32_t start,
sl@0:               int32_t length) const;
sl@0: 
sl@0:   /**
sl@0:    * Locate in this the last occurrence of the characters in <TT>text</TT>,
sl@0:    * using bitwise comparison.
sl@0:    * @param text The text to search for.
sl@0:    * @return The offset into this of the start of <TT>text</TT>,
sl@0:    * or -1 if not found.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int32_t lastIndexOf(const UnicodeString& text) const;
sl@0: 
sl@0:   /**
sl@0:    * Locate in this the last occurrence of the characters in <TT>text</TT>
sl@0:    * starting at offset <TT>start</TT>, using bitwise comparison.
sl@0:    * @param text The text to search for.
sl@0:    * @param start The offset at which searching will start.
sl@0:    * @return The offset into this of the start of <TT>text</TT>,
sl@0:    * or -1 if not found.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int32_t lastIndexOf(const UnicodeString& text,
sl@0:               int32_t start) const;
sl@0: 
sl@0:   /**
sl@0:    * Locate in this the last occurrence in the range
sl@0:    * [<TT>start</TT>, <TT>start + length</TT>) of the characters
sl@0:    * in <TT>text</TT>, using bitwise comparison.
sl@0:    * @param text The text to search for.
sl@0:    * @param start The offset at which searching will start.
sl@0:    * @param length The number of characters to search
sl@0:    * @return The offset into this of the start of <TT>text</TT>,
sl@0:    * or -1 if not found.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int32_t lastIndexOf(const UnicodeString& text,
sl@0:               int32_t start,
sl@0:               int32_t length) const;
sl@0: 
sl@0:   /**
sl@0:    * Locate in this the last occurrence in the range
sl@0:    * [<TT>start</TT>, <TT>start + length</TT>) of the characters
sl@0:    * in <TT>srcText</TT> in the range
sl@0:    * [<TT>srcStart</TT>, <TT>srcStart + srcLength</TT>),
sl@0:    * using bitwise comparison.
sl@0:    * @param srcText The text to search for.
sl@0:    * @param srcStart the offset into <TT>srcText</TT> at which
sl@0:    * to start matching
sl@0:    * @param srcLength the number of characters in <TT>srcText</TT> to match
sl@0:    * @param start the offset into this at which to start matching
sl@0:    * @param length the number of characters in this to search
sl@0:    * @return The offset into this of the start of <TT>text</TT>,
sl@0:    * or -1 if not found.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int32_t lastIndexOf(const UnicodeString& srcText,
sl@0:               int32_t srcStart,
sl@0:               int32_t srcLength,
sl@0:               int32_t start,
sl@0:               int32_t length) const;
sl@0: 
sl@0:   /**
sl@0:    * Locate in this the last occurrence of the characters in <TT>srcChars</TT>
sl@0:    * starting at offset <TT>start</TT>, using bitwise comparison.
sl@0:    * @param srcChars The text to search for.
sl@0:    * @param srcLength the number of characters in <TT>srcChars</TT> to match
sl@0:    * @param start the offset into this at which to start matching
sl@0:    * @return The offset into this of the start of <TT>text</TT>,
sl@0:    * or -1 if not found.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int32_t lastIndexOf(const UChar *srcChars,
sl@0:               int32_t srcLength,
sl@0:               int32_t start) const;
sl@0: 
sl@0:   /**
sl@0:    * Locate in this the last occurrence in the range
sl@0:    * [<TT>start</TT>, <TT>start + length</TT>) of the characters
sl@0:    * in <TT>srcChars</TT>, using bitwise comparison.
sl@0:    * @param srcChars The text to search for.
sl@0:    * @param srcLength the number of characters in <TT>srcChars</TT>
sl@0:    * @param start The offset at which searching will start.
sl@0:    * @param length The number of characters to search
sl@0:    * @return The offset into this of the start of <TT>srcChars</TT>,
sl@0:    * or -1 if not found.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int32_t lastIndexOf(const UChar *srcChars,
sl@0:               int32_t srcLength,
sl@0:               int32_t start,
sl@0:               int32_t length) const;
sl@0: 
sl@0:   /**
sl@0:    * Locate in this the last occurrence in the range
sl@0:    * [<TT>start</TT>, <TT>start + length</TT>) of the characters
sl@0:    * in <TT>srcChars</TT> in the range
sl@0:    * [<TT>srcStart</TT>, <TT>srcStart + srcLength</TT>),
sl@0:    * using bitwise comparison.
sl@0:    * @param srcChars The text to search for.
sl@0:    * @param srcStart the offset into <TT>srcChars</TT> at which
sl@0:    * to start matching
sl@0:    * @param srcLength the number of characters in <TT>srcChars</TT> to match
sl@0:    * @param start the offset into this at which to start matching
sl@0:    * @param length the number of characters in this to search
sl@0:    * @return The offset into this of the start of <TT>text</TT>,
sl@0:    * or -1 if not found.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   int32_t lastIndexOf(const UChar *srcChars,
sl@0:               int32_t srcStart,
sl@0:               int32_t srcLength,
sl@0:               int32_t start,
sl@0:               int32_t length) const;
sl@0: 
sl@0:   /**
sl@0:    * Locate in this the last occurrence of the BMP code point <code>c</code>,
sl@0:    * using bitwise comparison.
sl@0:    * @param c The code unit to search for.
sl@0:    * @return The offset into this of <TT>c</TT>, or -1 if not found.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int32_t lastIndexOf(UChar c) const;
sl@0: 
sl@0:   /**
sl@0:    * Locate in this the last occurrence of the code point <TT>c</TT>,
sl@0:    * using bitwise comparison.
sl@0:    *
sl@0:    * @param c The code point to search for.
sl@0:    * @return The offset into this of <TT>c</TT>, or -1 if not found.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int32_t lastIndexOf(UChar32 c) const;
sl@0: 
sl@0:   /**
sl@0:    * Locate in this the last occurrence of the BMP code point <code>c</code>
sl@0:    * starting at offset <TT>start</TT>, using bitwise comparison.
sl@0:    * @param c The code unit to search for.
sl@0:    * @param start The offset at which searching will start.
sl@0:    * @return The offset into this of <TT>c</TT>, or -1 if not found.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int32_t lastIndexOf(UChar c,
sl@0:               int32_t start) const;
sl@0: 
sl@0:   /**
sl@0:    * Locate in this the last occurrence of the code point <TT>c</TT>
sl@0:    * starting at offset <TT>start</TT>, using bitwise comparison.
sl@0:    *
sl@0:    * @param c The code point to search for.
sl@0:    * @param start The offset at which searching will start.
sl@0:    * @return The offset into this of <TT>c</TT>, or -1 if not found.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int32_t lastIndexOf(UChar32 c,
sl@0:               int32_t start) const;
sl@0: 
sl@0:   /**
sl@0:    * Locate in this the last occurrence of the BMP code point <code>c</code>
sl@0:    * in the range [<TT>start</TT>, <TT>start + length</TT>),
sl@0:    * using bitwise comparison.
sl@0:    * @param c The code unit to search for.
sl@0:    * @param start the offset into this at which to start matching
sl@0:    * @param length the number of characters in this to search
sl@0:    * @return The offset into this of <TT>c</TT>, or -1 if not found.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int32_t lastIndexOf(UChar c,
sl@0:               int32_t start,
sl@0:               int32_t length) const;
sl@0: 
sl@0:   /**
sl@0:    * Locate in this the last occurrence of the code point <TT>c</TT>
sl@0:    * in the range [<TT>start</TT>, <TT>start + length</TT>),
sl@0:    * using bitwise comparison.
sl@0:    *
sl@0:    * @param c The code point to search for.
sl@0:    * @param start the offset into this at which to start matching
sl@0:    * @param length the number of characters in this to search
sl@0:    * @return The offset into this of <TT>c</TT>, or -1 if not found.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int32_t lastIndexOf(UChar32 c,
sl@0:               int32_t start,
sl@0:               int32_t length) const;
sl@0: 
sl@0: 
sl@0:   /* Character access */
sl@0: 
sl@0:   /**
sl@0:    * Return the code unit at offset <tt>offset</tt>.
sl@0:    * If the offset is not valid (0..length()-1) then U+ffff is returned.
sl@0:    * @param offset a valid offset into the text
sl@0:    * @return the code unit at offset <tt>offset</tt>
sl@0:    *         or 0xffff if the offset is not valid for this string
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UChar charAt(int32_t offset) const;
sl@0: 
sl@0:   /**
sl@0:    * Return the code unit at offset <tt>offset</tt>.
sl@0:    * If the offset is not valid (0..length()-1) then U+ffff is returned.
sl@0:    * @param offset a valid offset into the text
sl@0:    * @return the code unit at offset <tt>offset</tt>
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UChar operator[] (int32_t offset) const;
sl@0: 
sl@0:   /**
sl@0:    * Return the code point that contains the code unit
sl@0:    * at offset <tt>offset</tt>.
sl@0:    * If the offset is not valid (0..length()-1) then U+ffff is returned.
sl@0:    * @param offset a valid offset into the text
sl@0:    * that indicates the text offset of any of the code units
sl@0:    * that will be assembled into a code point (21-bit value) and returned
sl@0:    * @return the code point of text at <tt>offset</tt>
sl@0:    *         or 0xffff if the offset is not valid for this string
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UChar32 char32At(int32_t offset) const;
sl@0: 
sl@0:   /**
sl@0:    * Adjust a random-access offset so that
sl@0:    * it points to the beginning of a Unicode character.
sl@0:    * The offset that is passed in points to
sl@0:    * any code unit of a code point,
sl@0:    * while the returned offset will point to the first code unit
sl@0:    * of the same code point.
sl@0:    * In UTF-16, if the input offset points to a second surrogate
sl@0:    * of a surrogate pair, then the returned offset will point
sl@0:    * to the first surrogate.
sl@0:    * @param offset a valid offset into one code point of the text
sl@0:    * @return offset of the first code unit of the same code point
sl@0:    * @see U16_SET_CP_START
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int32_t getChar32Start(int32_t offset) const;
sl@0: 
sl@0:   /**
sl@0:    * Adjust a random-access offset so that
sl@0:    * it points behind a Unicode character.
sl@0:    * The offset that is passed in points behind
sl@0:    * any code unit of a code point,
sl@0:    * while the returned offset will point behind the last code unit
sl@0:    * of the same code point.
sl@0:    * In UTF-16, if the input offset points behind the first surrogate
sl@0:    * (i.e., to the second surrogate)
sl@0:    * of a surrogate pair, then the returned offset will point
sl@0:    * behind the second surrogate (i.e., to the first surrogate).
sl@0:    * @param offset a valid offset after any code unit of a code point of the text
sl@0:    * @return offset of the first code unit after the same code point
sl@0:    * @see U16_SET_CP_LIMIT
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int32_t getChar32Limit(int32_t offset) const;
sl@0: 
sl@0:   /**
sl@0:    * Move the code unit index along the string by delta code points.
sl@0:    * Interpret the input index as a code unit-based offset into the string,
sl@0:    * move the index forward or backward by delta code points, and
sl@0:    * return the resulting index.
sl@0:    * The input index should point to the first code unit of a code point,
sl@0:    * if there is more than one.
sl@0:    *
sl@0:    * Both input and output indexes are code unit-based as for all
sl@0:    * string indexes/offsets in ICU (and other libraries, like MBCS char*).
sl@0:    * If delta<0 then the index is moved backward (toward the start of the string).
sl@0:    * If delta>0 then the index is moved forward (toward the end of the string).
sl@0:    *
sl@0:    * This behaves like CharacterIterator::move32(delta, kCurrent).
sl@0:    *
sl@0:    * Behavior for out-of-bounds indexes:
sl@0:    * <code>moveIndex32</code> pins the input index to 0..length(), i.e.,
sl@0:    * if the input index<0 then it is pinned to 0;
sl@0:    * if it is index>length() then it is pinned to length().
sl@0:    * Afterwards, the index is moved by <code>delta</code> code points
sl@0:    * forward or backward,
sl@0:    * but no further backward than to 0 and no further forward than to length().
sl@0:    * The resulting index return value will be in between 0 and length(), inclusively.
sl@0:    *
sl@0:    * Examples:
sl@0:    * <pre>
sl@0:    * // s has code points 'a' U+10000 'b' U+10ffff U+2029
sl@0:    * UnicodeString s=UNICODE_STRING("a\\U00010000b\\U0010ffff\\u2029", 31).unescape();
sl@0:    *
sl@0:    * // initial index: position of U+10000
sl@0:    * int32_t index=1;
sl@0:    *
sl@0:    * // the following examples will all result in index==4, position of U+10ffff
sl@0:    *
sl@0:    * // skip 2 code points from some position in the string
sl@0:    * index=s.moveIndex32(index, 2); // skips U+10000 and 'b'
sl@0:    *
sl@0:    * // go to the 3rd code point from the start of s (0-based)
sl@0:    * index=s.moveIndex32(0, 3); // skips 'a', U+10000, and 'b'
sl@0:    *
sl@0:    * // go to the next-to-last code point of s
sl@0:    * index=s.moveIndex32(s.length(), -2); // backward-skips U+2029 and U+10ffff
sl@0:    * </pre>
sl@0:    *
sl@0:    * @param index input code unit index
sl@0:    * @param delta (signed) code point count to move the index forward or backward
sl@0:    *        in the string
sl@0:    * @return the resulting code unit index
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   int32_t moveIndex32(int32_t index, int32_t delta) const;
sl@0: 
sl@0:   /* Substring extraction */
sl@0: 
sl@0:   /**
sl@0:    * Copy the characters in the range
sl@0:    * [<tt>start</tt>, <tt>start + length</tt>) into the array <tt>dst</tt>,
sl@0:    * beginning at <tt>dstStart</tt>.
sl@0:    * If the string aliases to <code>dst</code> itself as an external buffer,
sl@0:    * then extract() will not copy the contents.
sl@0:    *
sl@0:    * @param start offset of first character which will be copied into the array
sl@0:    * @param length the number of characters to extract
sl@0:    * @param dst array in which to copy characters.  The length of <tt>dst</tt>
sl@0:    * must be at least (<tt>dstStart + length</tt>).
sl@0:    * @param dstStart the offset in <TT>dst</TT> where the first character
sl@0:    * will be extracted
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline void extract(int32_t start,
sl@0:            int32_t length,
sl@0:            UChar *dst,
sl@0:            int32_t dstStart = 0) const;
sl@0: 
sl@0:   /**
sl@0:    * Copy the contents of the string into dest.
sl@0:    * This is a convenience function that
sl@0:    * checks if there is enough space in dest,
sl@0:    * extracts the entire string if possible,
sl@0:    * and NUL-terminates dest if possible.
sl@0:    *
sl@0:    * If the string fits into dest but cannot be NUL-terminated
sl@0:    * (length()==destCapacity) then the error code is set to U_STRING_NOT_TERMINATED_WARNING.
sl@0:    * If the string itself does not fit into dest
sl@0:    * (length()>destCapacity) then the error code is set to U_BUFFER_OVERFLOW_ERROR.
sl@0:    *
sl@0:    * If the string aliases to <code>dest</code> itself as an external buffer,
sl@0:    * then extract() will not copy the contents.
sl@0:    *
sl@0:    * @param dest Destination string buffer.
sl@0:    * @param destCapacity Number of UChars available at dest.
sl@0:    * @param errorCode ICU error code.
sl@0:    * @return length()
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   int32_t
sl@0:   extract(UChar *dest, int32_t destCapacity,
sl@0:           UErrorCode &errorCode) const;
sl@0: 
sl@0:   /**
sl@0:    * Copy the characters in the range
sl@0:    * [<tt>start</tt>, <tt>start + length</tt>) into the  UnicodeString
sl@0:    * <tt>target</tt>.
sl@0:    * @param start offset of first character which will be copied
sl@0:    * @param length the number of characters to extract
sl@0:    * @param target UnicodeString into which to copy characters.
sl@0:    * @return A reference to <TT>target</TT>
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline void extract(int32_t start,
sl@0:            int32_t length,
sl@0:            UnicodeString& target) const;
sl@0: 
sl@0:   /**
sl@0:    * Copy the characters in the range [<tt>start</tt>, <tt>limit</tt>)
sl@0:    * into the array <tt>dst</tt>, beginning at <tt>dstStart</tt>.
sl@0:    * @param start offset of first character which will be copied into the array
sl@0:    * @param limit offset immediately following the last character to be copied
sl@0:    * @param dst array in which to copy characters.  The length of <tt>dst</tt>
sl@0:    * must be at least (<tt>dstStart + (limit - start)</tt>).
sl@0:    * @param dstStart the offset in <TT>dst</TT> where the first character
sl@0:    * will be extracted
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline void extractBetween(int32_t start,
sl@0:               int32_t limit,
sl@0:               UChar *dst,
sl@0:               int32_t dstStart = 0) const;
sl@0: 
sl@0:   /**
sl@0:    * Copy the characters in the range [<tt>start</tt>, <tt>limit</tt>)
sl@0:    * into the UnicodeString <tt>target</tt>.  Replaceable API.
sl@0:    * @param start offset of first character which will be copied
sl@0:    * @param limit offset immediately following the last character to be copied
sl@0:    * @param target UnicodeString into which to copy characters.
sl@0:    * @return A reference to <TT>target</TT>
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   virtual void extractBetween(int32_t start,
sl@0:               int32_t limit,
sl@0:               UnicodeString& target) const;
sl@0: 
sl@0:   /**
sl@0:    * Copy the characters in the range 
sl@0:    * [<tt>start</TT>, <tt>start + length</TT>) into an array of characters.
sl@0:    * All characters must be invariant (see utypes.h).
sl@0:    * Use US_INV as the last, signature-distinguishing parameter.
sl@0:    *
sl@0:    * This function does not write any more than <code>targetLength</code>
sl@0:    * characters but returns the length of the entire output string
sl@0:    * so that one can allocate a larger buffer and call the function again
sl@0:    * if necessary.
sl@0:    * The output string is NUL-terminated if possible.
sl@0:    *
sl@0:    * @param start offset of first character which will be copied
sl@0:    * @param startLength the number of characters to extract
sl@0:    * @param target the target buffer for extraction, can be NULL
sl@0:    *               if targetLength is 0
sl@0:    * @param targetCapacity the length of the target buffer
sl@0:    * @param inv Signature-distinguishing paramater, use US_INV.
sl@0:    * @return the output string length, not including the terminating NUL
sl@0:    * @draft ICU 3.2
sl@0:    */
sl@0:   int32_t extract(int32_t start,
sl@0:            int32_t startLength,
sl@0:            char *target,
sl@0:            int32_t targetCapacity,
sl@0:            enum EInvariant inv) const;
sl@0: 
sl@0: #if !UCONFIG_NO_CONVERSION
sl@0: 
sl@0:   /**
sl@0:    * Copy the characters in the range
sl@0:    * [<tt>start</TT>, <tt>start + length</TT>) into an array of characters
sl@0:    * in a specified codepage.
sl@0:    * The output string is NUL-terminated.
sl@0:    *
sl@0:    * Recommendation: For invariant-character strings use
sl@0:    * extract(int32_t start, int32_t length, char *target, int32_t targetCapacity, enum EInvariant inv) const
sl@0:    * because it avoids object code dependencies of UnicodeString on
sl@0:    * the conversion code.
sl@0:    *
sl@0:    * @param start offset of first character which will be copied
sl@0:    * @param startLength the number of characters to extract
sl@0:    * @param target the target buffer for extraction
sl@0:    * @param codepage the desired codepage for the characters.  0 has
sl@0:    * the special meaning of the default codepage
sl@0:    * If <code>codepage</code> is an empty string (<code>""</code>),
sl@0:    * then a simple conversion is performed on the codepage-invariant
sl@0:    * subset ("invariant characters") of the platform encoding. See utypes.h.
sl@0:    * If <TT>target</TT> is NULL, then the number of bytes required for
sl@0:    * <TT>target</TT> is returned. It is assumed that the target is big enough
sl@0:    * to fit all of the characters.
sl@0:    * @return the output string length, not including the terminating NUL
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int32_t extract(int32_t start,
sl@0:                  int32_t startLength,
sl@0:                  char *target,
sl@0:                  const char *codepage = 0) const;
sl@0: 
sl@0:   /**
sl@0:    * Copy the characters in the range
sl@0:    * [<tt>start</TT>, <tt>start + length</TT>) into an array of characters
sl@0:    * in a specified codepage.
sl@0:    * This function does not write any more than <code>targetLength</code>
sl@0:    * characters but returns the length of the entire output string
sl@0:    * so that one can allocate a larger buffer and call the function again
sl@0:    * if necessary.
sl@0:    * The output string is NUL-terminated if possible.
sl@0:    *
sl@0:    * Recommendation: For invariant-character strings use
sl@0:    * extract(int32_t start, int32_t length, char *target, int32_t targetCapacity, enum EInvariant inv) const
sl@0:    * because it avoids object code dependencies of UnicodeString on
sl@0:    * the conversion code.
sl@0:    *
sl@0:    * @param start offset of first character which will be copied
sl@0:    * @param startLength the number of characters to extract
sl@0:    * @param target the target buffer for extraction
sl@0:    * @param targetLength the length of the target buffer
sl@0:    * @param codepage the desired codepage for the characters.  0 has
sl@0:    * the special meaning of the default codepage
sl@0:    * If <code>codepage</code> is an empty string (<code>""</code>),
sl@0:    * then a simple conversion is performed on the codepage-invariant
sl@0:    * subset ("invariant characters") of the platform encoding. See utypes.h.
sl@0:    * If <TT>target</TT> is NULL, then the number of bytes required for
sl@0:    * <TT>target</TT> is returned.
sl@0:    * @return the output string length, not including the terminating NUL
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   int32_t extract(int32_t start,
sl@0:            int32_t startLength,
sl@0:            char *target,
sl@0:            uint32_t targetLength,
sl@0:            const char *codepage = 0) const;
sl@0: 
sl@0:   /**
sl@0:    * Convert the UnicodeString into a codepage string using an existing UConverter.
sl@0:    * The output string is NUL-terminated if possible.
sl@0:    *
sl@0:    * This function avoids the overhead of opening and closing a converter if
sl@0:    * multiple strings are extracted.
sl@0:    *
sl@0:    * @param dest destination string buffer, can be NULL if destCapacity==0
sl@0:    * @param destCapacity the number of chars available at dest
sl@0:    * @param cnv the converter object to be used (ucnv_resetFromUnicode() will be called),
sl@0:    *        or NULL for the default converter
sl@0:    * @param errorCode normal ICU error code
sl@0:    * @return the length of the output string, not counting the terminating NUL;
sl@0:    *         if the length is greater than destCapacity, then the string will not fit
sl@0:    *         and a buffer of the indicated length would need to be passed in
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   int32_t extract(char *dest, int32_t destCapacity,
sl@0:                   UConverter *cnv,
sl@0:                   UErrorCode &errorCode) const;
sl@0: 
sl@0: #endif
sl@0: 
sl@0:   /* Length operations */
sl@0: 
sl@0:   /**
sl@0:    * Return the length of the UnicodeString object.
sl@0:    * The length is the number of UChar code units are in the UnicodeString.
sl@0:    * If you want the number of code points, please use countChar32().
sl@0:    * @return the length of the UnicodeString object
sl@0:    * @see countChar32
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int32_t length(void) const;
sl@0: 
sl@0:   /**
sl@0:    * Count Unicode code points in the length UChar code units of the string.
sl@0:    * A code point may occupy either one or two UChar code units.
sl@0:    * Counting code points involves reading all code units.
sl@0:    *
sl@0:    * This functions is basically the inverse of moveIndex32().
sl@0:    *
sl@0:    * @param start the index of the first code unit to check
sl@0:    * @param length the number of UChar code units to check
sl@0:    * @return the number of code points in the specified code units
sl@0:    * @see length
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   int32_t
sl@0:   countChar32(int32_t start=0, int32_t length=INT32_MAX) const;
sl@0: 
sl@0:   /**
sl@0:    * Check if the length UChar code units of the string
sl@0:    * contain more Unicode code points than a certain number.
sl@0:    * This is more efficient than counting all code points in this part of the string
sl@0:    * and comparing that number with a threshold.
sl@0:    * This function may not need to scan the string at all if the length
sl@0:    * falls within a certain range, and
sl@0:    * never needs to count more than 'number+1' code points.
sl@0:    * Logically equivalent to (countChar32(start, length)>number).
sl@0:    * A Unicode code point may occupy either one or two UChar code units.
sl@0:    *
sl@0:    * @param start the index of the first code unit to check (0 for the entire string)
sl@0:    * @param length the number of UChar code units to check
sl@0:    *               (use INT32_MAX for the entire string; remember that start/length
sl@0:    *                values are pinned)
sl@0:    * @param number The number of code points in the (sub)string is compared against
sl@0:    *               the 'number' parameter.
sl@0:    * @return Boolean value for whether the string contains more Unicode code points
sl@0:    *         than 'number'. Same as (u_countChar32(s, length)>number).
sl@0:    * @see countChar32
sl@0:    * @see u_strHasMoreChar32Than
sl@0:    * @stable ICU 2.4
sl@0:    */
sl@0:   UBool
sl@0:   hasMoreChar32Than(int32_t start, int32_t length, int32_t number) const;
sl@0: 
sl@0:   /**
sl@0:    * Determine if this string is empty.
sl@0:    * @return TRUE if this string contains 0 characters, FALSE otherwise.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UBool isEmpty(void) const;
sl@0: 
sl@0:   /**
sl@0:    * Return the capacity of the internal buffer of the UnicodeString object.
sl@0:    * This is useful together with the getBuffer functions.
sl@0:    * See there for details.
sl@0:    *
sl@0:    * @return the number of UChars available in the internal buffer
sl@0:    * @see getBuffer
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int32_t getCapacity(void) const;
sl@0: 
sl@0:   /* Other operations */
sl@0: 
sl@0:   /**
sl@0:    * Generate a hash code for this object.
sl@0:    * @return The hash code of this UnicodeString.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline int32_t hashCode(void) const;
sl@0: 
sl@0:   /**
sl@0:    * Determine if this object contains a valid string.
sl@0:    * A bogus string has no value. It is different from an empty string.
sl@0:    * It can be used to indicate that no string value is available.
sl@0:    * getBuffer() and getTerminatedBuffer() return NULL, and
sl@0:    * length() returns 0.
sl@0:    *
sl@0:    * @return TRUE if the string is valid, FALSE otherwise
sl@0:    * @see setToBogus()
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UBool isBogus(void) const;
sl@0: 
sl@0: 
sl@0:   //========================================
sl@0:   // Write operations
sl@0:   //========================================
sl@0: 
sl@0:   /* Assignment operations */
sl@0: 
sl@0:   /**
sl@0:    * Assignment operator.  Replace the characters in this UnicodeString
sl@0:    * with the characters from <TT>srcText</TT>.
sl@0:    * @param srcText The text containing the characters to replace
sl@0:    * @return a reference to this
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   UnicodeString &operator=(const UnicodeString &srcText);
sl@0: 
sl@0:   /**
sl@0:    * Almost the same as the assignment operator.
sl@0:    * Replace the characters in this UnicodeString
sl@0:    * with the characters from <code>srcText</code>.
sl@0:    *
sl@0:    * This function works the same for all strings except for ones that
sl@0:    * are readonly aliases.
sl@0:    * Starting with ICU 2.4, the assignment operator and the copy constructor
sl@0:    * allocate a new buffer and copy the buffer contents even for readonly aliases.
sl@0:    * This function implements the old, more efficient but less safe behavior
sl@0:    * of making this string also a readonly alias to the same buffer.
sl@0:    * The fastCopyFrom function must be used only if it is known that the lifetime of
sl@0:    * this UnicodeString is at least as long as the lifetime of the aliased buffer
sl@0:    * including its contents, for example for strings from resource bundles
sl@0:    * or aliases to string contents.
sl@0:    *
sl@0:    * @param src The text containing the characters to replace.
sl@0:    * @return a reference to this
sl@0:    * @stable ICU 2.4
sl@0:    */
sl@0:   UnicodeString &fastCopyFrom(const UnicodeString &src);
sl@0: 
sl@0:   /**
sl@0:    * Assignment operator.  Replace the characters in this UnicodeString
sl@0:    * with the code unit <TT>ch</TT>.
sl@0:    * @param ch the code unit to replace
sl@0:    * @return a reference to this
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UnicodeString& operator= (UChar ch);
sl@0: 
sl@0:   /**
sl@0:    * Assignment operator.  Replace the characters in this UnicodeString
sl@0:    * with the code point <TT>ch</TT>.
sl@0:    * @param ch the code point to replace
sl@0:    * @return a reference to this
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UnicodeString& operator= (UChar32 ch);
sl@0: 
sl@0:   /**
sl@0:    * Set the text in the UnicodeString object to the characters
sl@0:    * in <TT>srcText</TT> in the range
sl@0:    * [<TT>srcStart</TT>, <TT>srcText.length()</TT>).
sl@0:    * <TT>srcText</TT> is not modified.
sl@0:    * @param srcText the source for the new characters
sl@0:    * @param srcStart the offset into <TT>srcText</TT> where new characters
sl@0:    * will be obtained
sl@0:    * @return a reference to this
sl@0:    * @stable ICU 2.2
sl@0:    */
sl@0:   inline UnicodeString& setTo(const UnicodeString& srcText,
sl@0:                int32_t srcStart);
sl@0: 
sl@0:   /**
sl@0:    * Set the text in the UnicodeString object to the characters
sl@0:    * in <TT>srcText</TT> in the range
sl@0:    * [<TT>srcStart</TT>, <TT>srcStart + srcLength</TT>).
sl@0:    * <TT>srcText</TT> is not modified.
sl@0:    * @param srcText the source for the new characters
sl@0:    * @param srcStart the offset into <TT>srcText</TT> where new characters
sl@0:    * will be obtained
sl@0:    * @param srcLength the number of characters in <TT>srcText</TT> in the
sl@0:    * replace string.
sl@0:    * @return a reference to this
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UnicodeString& setTo(const UnicodeString& srcText,
sl@0:                int32_t srcStart,
sl@0:                int32_t srcLength);
sl@0: 
sl@0:   /**
sl@0:    * Set the text in the UnicodeString object to the characters in
sl@0:    * <TT>srcText</TT>.
sl@0:    * <TT>srcText</TT> is not modified.
sl@0:    * @param srcText the source for the new characters
sl@0:    * @return a reference to this
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UnicodeString& setTo(const UnicodeString& srcText);
sl@0: 
sl@0:   /**
sl@0:    * Set the characters in the UnicodeString object to the characters
sl@0:    * in <TT>srcChars</TT>. <TT>srcChars</TT> is not modified.
sl@0:    * @param srcChars the source for the new characters
sl@0:    * @param srcLength the number of Unicode characters in srcChars.
sl@0:    * @return a reference to this
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UnicodeString& setTo(const UChar *srcChars,
sl@0:                int32_t srcLength);
sl@0: 
sl@0:   /**
sl@0:    * Set the characters in the UnicodeString object to the code unit
sl@0:    * <TT>srcChar</TT>.
sl@0:    * @param srcChar the code unit which becomes the UnicodeString's character
sl@0:    * content
sl@0:    * @return a reference to this
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   UnicodeString& setTo(UChar srcChar);
sl@0: 
sl@0:   /**
sl@0:    * Set the characters in the UnicodeString object to the code point
sl@0:    * <TT>srcChar</TT>.
sl@0:    * @param srcChar the code point which becomes the UnicodeString's character
sl@0:    * content
sl@0:    * @return a reference to this
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   UnicodeString& setTo(UChar32 srcChar);
sl@0: 
sl@0:   /**
sl@0:    * Aliasing setTo() function, analogous to the readonly-aliasing UChar* constructor.
sl@0:    * The text will be used for the UnicodeString object, but
sl@0:    * it will not be released when the UnicodeString is destroyed.
sl@0:    * This has copy-on-write semantics:
sl@0:    * When the string is modified, then the buffer is first copied into
sl@0:    * newly allocated memory.
sl@0:    * The aliased buffer is never modified.
sl@0:    * In an assignment to another UnicodeString, the text will be aliased again,
sl@0:    * so that both strings then alias the same readonly-text.
sl@0:    *
sl@0:    * @param isTerminated specifies if <code>text</code> is <code>NUL</code>-terminated.
sl@0:    *                     This must be true if <code>textLength==-1</code>.
sl@0:    * @param text The characters to alias for the UnicodeString.
sl@0:    * @param textLength The number of Unicode characters in <code>text</code> to alias.
sl@0:    *                   If -1, then this constructor will determine the length
sl@0:    *                   by calling <code>u_strlen()</code>.
sl@0:    * @return a reference to this
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   UnicodeString &setTo(UBool isTerminated,
sl@0:                        const UChar *text,
sl@0:                        int32_t textLength);
sl@0: 
sl@0:   /**
sl@0:    * Aliasing setTo() function, analogous to the writable-aliasing UChar* constructor.
sl@0:    * The text will be used for the UnicodeString object, but
sl@0:    * it will not be released when the UnicodeString is destroyed.
sl@0:    * This has write-through semantics:
sl@0:    * For as long as the capacity of the buffer is sufficient, write operations
sl@0:    * will directly affect the buffer. When more capacity is necessary, then
sl@0:    * a new buffer will be allocated and the contents copied as with regularly
sl@0:    * constructed strings.
sl@0:    * In an assignment to another UnicodeString, the buffer will be copied.
sl@0:    * The extract(UChar *dst) function detects whether the dst pointer is the same
sl@0:    * as the string buffer itself and will in this case not copy the contents.
sl@0:    *
sl@0:    * @param buffer The characters to alias for the UnicodeString.
sl@0:    * @param buffLength The number of Unicode characters in <code>buffer</code> to alias.
sl@0:    * @param buffCapacity The size of <code>buffer</code> in UChars.
sl@0:    * @return a reference to this
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   UnicodeString &setTo(UChar *buffer,
sl@0:                        int32_t buffLength,
sl@0:                        int32_t buffCapacity);
sl@0: 
sl@0:   /**
sl@0:    * Make this UnicodeString object invalid.
sl@0:    * The string will test TRUE with isBogus().
sl@0:    *
sl@0:    * A bogus string has no value. It is different from an empty string.
sl@0:    * It can be used to indicate that no string value is available.
sl@0:    * getBuffer() and getTerminatedBuffer() return NULL, and
sl@0:    * length() returns 0.
sl@0:    *
sl@0:    * This utility function is used throughout the UnicodeString
sl@0:    * implementation to indicate that a UnicodeString operation failed,
sl@0:    * and may be used in other functions,
sl@0:    * especially but not exclusively when such functions do not
sl@0:    * take a UErrorCode for simplicity.
sl@0:    *
sl@0:    * The following methods, and no others, will clear a string object's bogus flag:
sl@0:    * - remove()
sl@0:    * - remove(0, INT32_MAX)
sl@0:    * - truncate(0)
sl@0:    * - operator=() (assignment operator)
sl@0:    * - setTo(...)
sl@0:    *
sl@0:    * The simplest ways to turn a bogus string into an empty one
sl@0:    * is to use the remove() function.
sl@0:    * Examples for other functions that are equivalent to "set to empty string":
sl@0:    * \code
sl@0:    * if(s.isBogus()) {
sl@0:    *   s.remove();           // set to an empty string (remove all), or
sl@0:    *   s.remove(0, INT32_MAX); // set to an empty string (remove all), or
sl@0:    *   s.truncate(0);        // set to an empty string (complete truncation), or
sl@0:    *   s=UnicodeString();    // assign an empty string, or
sl@0:    *   s.setTo((UChar32)-1); // set to a pseudo code point that is out of range, or
sl@0:    *   static const UChar nul=0;
sl@0:    *   s.setTo(&nul, 0);     // set to an empty C Unicode string
sl@0:    * }
sl@0:    * \endcode
sl@0:    *
sl@0:    * @see isBogus()
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   void setToBogus();
sl@0: 
sl@0:   /**
sl@0:    * Set the character at the specified offset to the specified character.
sl@0:    * @param offset A valid offset into the text of the character to set
sl@0:    * @param ch The new character
sl@0:    * @return A reference to this
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   UnicodeString& setCharAt(int32_t offset,
sl@0:                UChar ch);
sl@0: 
sl@0: 
sl@0:   /* Append operations */
sl@0: 
sl@0:   /**
sl@0:    * Append operator. Append the code unit <TT>ch</TT> to the UnicodeString
sl@0:    * object.
sl@0:    * @param ch the code unit to be appended
sl@0:    * @return a reference to this
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:  inline  UnicodeString& operator+= (UChar ch);
sl@0: 
sl@0:   /**
sl@0:    * Append operator. Append the code point <TT>ch</TT> to the UnicodeString
sl@0:    * object.
sl@0:    * @param ch the code point to be appended
sl@0:    * @return a reference to this
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:  inline  UnicodeString& operator+= (UChar32 ch);
sl@0: 
sl@0:   /**
sl@0:    * Append operator. Append the characters in <TT>srcText</TT> to the
sl@0:    * UnicodeString object at offset <TT>start</TT>. <TT>srcText</TT> is
sl@0:    * not modified.
sl@0:    * @param srcText the source for the new characters
sl@0:    * @return a reference to this
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UnicodeString& operator+= (const UnicodeString& srcText);
sl@0: 
sl@0:   /**
sl@0:    * Append the characters
sl@0:    * in <TT>srcText</TT> in the range
sl@0:    * [<TT>srcStart</TT>, <TT>srcStart + srcLength</TT>) to the
sl@0:    * UnicodeString object at offset <TT>start</TT>. <TT>srcText</TT>
sl@0:    * is not modified.
sl@0:    * @param srcText the source for the new characters
sl@0:    * @param srcStart the offset into <TT>srcText</TT> where new characters
sl@0:    * will be obtained
sl@0:    * @param srcLength the number of characters in <TT>srcText</TT> in
sl@0:    * the append string
sl@0:    * @return a reference to this
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UnicodeString& append(const UnicodeString& srcText,
sl@0:             int32_t srcStart,
sl@0:             int32_t srcLength);
sl@0: 
sl@0:   /**
sl@0:    * Append the characters in <TT>srcText</TT> to the UnicodeString object at
sl@0:    * offset <TT>start</TT>. <TT>srcText</TT> is not modified.
sl@0:    * @param srcText the source for the new characters
sl@0:    * @return a reference to this
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UnicodeString& append(const UnicodeString& srcText);
sl@0: 
sl@0:   /**
sl@0:    * Append the characters in <TT>srcChars</TT> in the range
sl@0:    * [<TT>srcStart</TT>, <TT>srcStart + srcLength</TT>) to the UnicodeString
sl@0:    * object at offset
sl@0:    * <TT>start</TT>. <TT>srcChars</TT> is not modified.
sl@0:    * @param srcChars the source for the new characters
sl@0:    * @param srcStart the offset into <TT>srcChars</TT> where new characters
sl@0:    * will be obtained
sl@0:    * @param srcLength the number of characters in <TT>srcChars</TT> in
sl@0:    * the append string
sl@0:    * @return a reference to this
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UnicodeString& append(const UChar *srcChars,
sl@0:             int32_t srcStart,
sl@0:             int32_t srcLength);
sl@0: 
sl@0:   /**
sl@0:    * Append the characters in <TT>srcChars</TT> to the UnicodeString object
sl@0:    * at offset <TT>start</TT>. <TT>srcChars</TT> is not modified.
sl@0:    * @param srcChars the source for the new characters
sl@0:    * @param srcLength the number of Unicode characters in <TT>srcChars</TT>
sl@0:    * @return a reference to this
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UnicodeString& append(const UChar *srcChars,
sl@0:             int32_t srcLength);
sl@0: 
sl@0:   /**
sl@0:    * Append the code unit <TT>srcChar</TT> to the UnicodeString object.
sl@0:    * @param srcChar the code unit to append
sl@0:    * @return a reference to this
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UnicodeString& append(UChar srcChar);
sl@0: 
sl@0:   /**
sl@0:    * Append the code point <TT>srcChar</TT> to the UnicodeString object.
sl@0:    * @param srcChar the code point to append
sl@0:    * @return a reference to this
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UnicodeString& append(UChar32 srcChar);
sl@0: 
sl@0: 
sl@0:   /* Insert operations */
sl@0: 
sl@0:   /**
sl@0:    * Insert the characters in <TT>srcText</TT> in the range
sl@0:    * [<TT>srcStart</TT>, <TT>srcStart + srcLength</TT>) into the UnicodeString
sl@0:    * object at offset <TT>start</TT>. <TT>srcText</TT> is not modified.
sl@0:    * @param start the offset where the insertion begins
sl@0:    * @param srcText the source for the new characters
sl@0:    * @param srcStart the offset into <TT>srcText</TT> where new characters
sl@0:    * will be obtained
sl@0:    * @param srcLength the number of characters in <TT>srcText</TT> in
sl@0:    * the insert string
sl@0:    * @return a reference to this
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UnicodeString& insert(int32_t start,
sl@0:             const UnicodeString& srcText,
sl@0:             int32_t srcStart,
sl@0:             int32_t srcLength);
sl@0: 
sl@0:   /**
sl@0:    * Insert the characters in <TT>srcText</TT> into the UnicodeString object
sl@0:    * at offset <TT>start</TT>. <TT>srcText</TT> is not modified.
sl@0:    * @param start the offset where the insertion begins
sl@0:    * @param srcText the source for the new characters
sl@0:    * @return a reference to this
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UnicodeString& insert(int32_t start,
sl@0:             const UnicodeString& srcText);
sl@0: 
sl@0:   /**
sl@0:    * Insert the characters in <TT>srcChars</TT> in the range
sl@0:    * [<TT>srcStart</TT>, <TT>srcStart + srcLength</TT>) into the UnicodeString
sl@0:    *  object at offset <TT>start</TT>. <TT>srcChars</TT> is not modified.
sl@0:    * @param start the offset at which the insertion begins
sl@0:    * @param srcChars the source for the new characters
sl@0:    * @param srcStart the offset into <TT>srcChars</TT> where new characters
sl@0:    * will be obtained
sl@0:    * @param srcLength the number of characters in <TT>srcChars</TT>
sl@0:    * in the insert string
sl@0:    * @return a reference to this
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UnicodeString& insert(int32_t start,
sl@0:             const UChar *srcChars,
sl@0:             int32_t srcStart,
sl@0:             int32_t srcLength);
sl@0: 
sl@0:   /**
sl@0:    * Insert the characters in <TT>srcChars</TT> into the UnicodeString object
sl@0:    * at offset <TT>start</TT>. <TT>srcChars</TT> is not modified.
sl@0:    * @param start the offset where the insertion begins
sl@0:    * @param srcChars the source for the new characters
sl@0:    * @param srcLength the number of Unicode characters in srcChars.
sl@0:    * @return a reference to this
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UnicodeString& insert(int32_t start,
sl@0:             const UChar *srcChars,
sl@0:             int32_t srcLength);
sl@0: 
sl@0:   /**
sl@0:    * Insert the code unit <TT>srcChar</TT> into the UnicodeString object at
sl@0:    * offset <TT>start</TT>.
sl@0:    * @param start the offset at which the insertion occurs
sl@0:    * @param srcChar the code unit to insert
sl@0:    * @return a reference to this
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UnicodeString& insert(int32_t start,
sl@0:             UChar srcChar);
sl@0: 
sl@0:   /**
sl@0:    * Insert the code point <TT>srcChar</TT> into the UnicodeString object at
sl@0:    * offset <TT>start</TT>.
sl@0:    * @param start the offset at which the insertion occurs
sl@0:    * @param srcChar the code point to insert
sl@0:    * @return a reference to this
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UnicodeString& insert(int32_t start,
sl@0:             UChar32 srcChar);
sl@0: 
sl@0: 
sl@0:   /* Replace operations */
sl@0: 
sl@0:   /**
sl@0:    * Replace the characters in the range
sl@0:    * [<TT>start</TT>, <TT>start + length</TT>) with the characters in
sl@0:    * <TT>srcText</TT> in the range
sl@0:    * [<TT>srcStart</TT>, <TT>srcStart + srcLength</TT>).
sl@0:    * <TT>srcText</TT> is not modified.
sl@0:    * @param start the offset at which the replace operation begins
sl@0:    * @param length the number of characters to replace. The character at
sl@0:    * <TT>start + length</TT> is not modified.
sl@0:    * @param srcText the source for the new characters
sl@0:    * @param srcStart the offset into <TT>srcText</TT> where new characters
sl@0:    * will be obtained
sl@0:    * @param srcLength the number of characters in <TT>srcText</TT> in
sl@0:    * the replace string
sl@0:    * @return a reference to this
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   UnicodeString& replace(int32_t start,
sl@0:              int32_t length,
sl@0:              const UnicodeString& srcText,
sl@0:              int32_t srcStart,
sl@0:              int32_t srcLength);
sl@0: 
sl@0:   /**
sl@0:    * Replace the characters in the range
sl@0:    * [<TT>start</TT>, <TT>start + length</TT>)
sl@0:    * with the characters in <TT>srcText</TT>.  <TT>srcText</TT> is
sl@0:    *  not modified.
sl@0:    * @param start the offset at which the replace operation begins
sl@0:    * @param length the number of characters to replace. The character at
sl@0:    * <TT>start + length</TT> is not modified.
sl@0:    * @param srcText the source for the new characters
sl@0:    * @return a reference to this
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   UnicodeString& replace(int32_t start,
sl@0:              int32_t length,
sl@0:              const UnicodeString& srcText);
sl@0: 
sl@0:   /**
sl@0:    * Replace the characters in the range
sl@0:    * [<TT>start</TT>, <TT>start + length</TT>) with the characters in
sl@0:    * <TT>srcChars</TT> in the range
sl@0:    * [<TT>srcStart</TT>, <TT>srcStart + srcLength</TT>). <TT>srcChars</TT>
sl@0:    * is not modified.
sl@0:    * @param start the offset at which the replace operation begins
sl@0:    * @param length the number of characters to replace.  The character at
sl@0:    * <TT>start + length</TT> is not modified.
sl@0:    * @param srcChars the source for the new characters
sl@0:    * @param srcStart the offset into <TT>srcChars</TT> where new characters
sl@0:    * will be obtained
sl@0:    * @param srcLength the number of characters in <TT>srcChars</TT>
sl@0:    * in the replace string
sl@0:    * @return a reference to this
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   UnicodeString& replace(int32_t start,
sl@0:              int32_t length,
sl@0:              const UChar *srcChars,
sl@0:              int32_t srcStart,
sl@0:              int32_t srcLength);
sl@0: 
sl@0:   /**
sl@0:    * Replace the characters in the range
sl@0:    * [<TT>start</TT>, <TT>start + length</TT>) with the characters in
sl@0:    * <TT>srcChars</TT>.  <TT>srcChars</TT> is not modified.
sl@0:    * @param start the offset at which the replace operation begins
sl@0:    * @param length number of characters to replace.  The character at
sl@0:    * <TT>start + length</TT> is not modified.
sl@0:    * @param srcChars the source for the new characters
sl@0:    * @param srcLength the number of Unicode characters in srcChars
sl@0:    * @return a reference to this
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UnicodeString& replace(int32_t start,
sl@0:              int32_t length,
sl@0:              const UChar *srcChars,
sl@0:              int32_t srcLength);
sl@0: 
sl@0:   /**
sl@0:    * Replace the characters in the range
sl@0:    * [<TT>start</TT>, <TT>start + length</TT>) with the code unit
sl@0:    * <TT>srcChar</TT>.
sl@0:    * @param start the offset at which the replace operation begins
sl@0:    * @param length the number of characters to replace.  The character at
sl@0:    * <TT>start + length</TT> is not modified.
sl@0:    * @param srcChar the new code unit
sl@0:    * @return a reference to this
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UnicodeString& replace(int32_t start,
sl@0:              int32_t length,
sl@0:              UChar srcChar);
sl@0: 
sl@0:   /**
sl@0:    * Replace the characters in the range
sl@0:    * [<TT>start</TT>, <TT>start + length</TT>) with the code point
sl@0:    * <TT>srcChar</TT>.
sl@0:    * @param start the offset at which the replace operation begins
sl@0:    * @param length the number of characters to replace.  The character at
sl@0:    * <TT>start + length</TT> is not modified.
sl@0:    * @param srcChar the new code point
sl@0:    * @return a reference to this
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UnicodeString& replace(int32_t start,
sl@0:              int32_t length,
sl@0:              UChar32 srcChar);
sl@0: 
sl@0:   /**
sl@0:    * Replace the characters in the range [<TT>start</TT>, <TT>limit</TT>)
sl@0:    * with the characters in <TT>srcText</TT>. <TT>srcText</TT> is not modified.
sl@0:    * @param start the offset at which the replace operation begins
sl@0:    * @param limit the offset immediately following the replace range
sl@0:    * @param srcText the source for the new characters
sl@0:    * @return a reference to this
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UnicodeString& replaceBetween(int32_t start,
sl@0:                 int32_t limit,
sl@0:                 const UnicodeString& srcText);
sl@0: 
sl@0:   /**
sl@0:    * Replace the characters in the range [<TT>start</TT>, <TT>limit</TT>)
sl@0:    * with the characters in <TT>srcText</TT> in the range
sl@0:    * [<TT>srcStart</TT>, <TT>srcLimit</TT>). <TT>srcText</TT> is not modified.
sl@0:    * @param start the offset at which the replace operation begins
sl@0:    * @param limit the offset immediately following the replace range
sl@0:    * @param srcText the source for the new characters
sl@0:    * @param srcStart the offset into <TT>srcChars</TT> where new characters
sl@0:    * will be obtained
sl@0:    * @param srcLimit the offset immediately following the range to copy
sl@0:    * in <TT>srcText</TT>
sl@0:    * @return a reference to this
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UnicodeString& replaceBetween(int32_t start,
sl@0:                 int32_t limit,
sl@0:                 const UnicodeString& srcText,
sl@0:                 int32_t srcStart,
sl@0:                 int32_t srcLimit);
sl@0: 
sl@0:   /**
sl@0:    * Replace a substring of this object with the given text.
sl@0:    * @param start the beginning index, inclusive; <code>0 <= start
sl@0:    * <= limit</code>.
sl@0:    * @param limit the ending index, exclusive; <code>start <= limit
sl@0:    * <= length()</code>.
sl@0:    * @param text the text to replace characters <code>start</code>
sl@0:    * to <code>limit - 1</code>
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   virtual void handleReplaceBetween(int32_t start,
sl@0:                                     int32_t limit,
sl@0:                                     const UnicodeString& text);
sl@0: 
sl@0:   /**
sl@0:    * Replaceable API
sl@0:    * @return TRUE if it has MetaData
sl@0:    * @stable ICU 2.4
sl@0:    */
sl@0:   virtual UBool hasMetaData() const;
sl@0: 
sl@0:   /**
sl@0:    * Copy a substring of this object, retaining attribute (out-of-band)
sl@0:    * information.  This method is used to duplicate or reorder substrings.
sl@0:    * The destination index must not overlap the source range.
sl@0:    *
sl@0:    * @param start the beginning index, inclusive; <code>0 <= start <=
sl@0:    * limit</code>.
sl@0:    * @param limit the ending index, exclusive; <code>start <= limit <=
sl@0:    * length()</code>.
sl@0:    * @param dest the destination index.  The characters from
sl@0:    * <code>start..limit-1</code> will be copied to <code>dest</code>.
sl@0:    * Implementations of this method may assume that <code>dest <= start ||
sl@0:    * dest >= limit</code>.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   virtual void copy(int32_t start, int32_t limit, int32_t dest);
sl@0: 
sl@0:   /* Search and replace operations */
sl@0: 
sl@0:   /**
sl@0:    * Replace all occurrences of characters in oldText with the characters
sl@0:    * in newText
sl@0:    * @param oldText the text containing the search text
sl@0:    * @param newText the text containing the replacement text
sl@0:    * @return a reference to this
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UnicodeString& findAndReplace(const UnicodeString& oldText,
sl@0:                 const UnicodeString& newText);
sl@0: 
sl@0:   /**
sl@0:    * Replace all occurrences of characters in oldText with characters
sl@0:    * in newText
sl@0:    * in the range [<TT>start</TT>, <TT>start + length</TT>).
sl@0:    * @param start the start of the range in which replace will performed
sl@0:    * @param length the length of the range in which replace will be performed
sl@0:    * @param oldText the text containing the search text
sl@0:    * @param newText the text containing the replacement text
sl@0:    * @return a reference to this
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UnicodeString& findAndReplace(int32_t start,
sl@0:                 int32_t length,
sl@0:                 const UnicodeString& oldText,
sl@0:                 const UnicodeString& newText);
sl@0: 
sl@0:   /**
sl@0:    * Replace all occurrences of characters in oldText in the range
sl@0:    * [<TT>oldStart</TT>, <TT>oldStart + oldLength</TT>) with the characters
sl@0:    * in newText in the range
sl@0:    * [<TT>newStart</TT>, <TT>newStart + newLength</TT>)
sl@0:    * in the range [<TT>start</TT>, <TT>start + length</TT>).
sl@0:    * @param start the start of the range in which replace will performed
sl@0:    * @param length the length of the range in which replace will be performed
sl@0:    * @param oldText the text containing the search text
sl@0:    * @param oldStart the start of the search range in <TT>oldText</TT>
sl@0:    * @param oldLength the length of the search range in <TT>oldText</TT>
sl@0:    * @param newText the text containing the replacement text
sl@0:    * @param newStart the start of the replacement range in <TT>newText</TT>
sl@0:    * @param newLength the length of the replacement range in <TT>newText</TT>
sl@0:    * @return a reference to this
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   UnicodeString& findAndReplace(int32_t start,
sl@0:                 int32_t length,
sl@0:                 const UnicodeString& oldText,
sl@0:                 int32_t oldStart,
sl@0:                 int32_t oldLength,
sl@0:                 const UnicodeString& newText,
sl@0:                 int32_t newStart,
sl@0:                 int32_t newLength);
sl@0: 
sl@0: 
sl@0:   /* Remove operations */
sl@0: 
sl@0:   /**
sl@0:    * Remove all characters from the UnicodeString object.
sl@0:    * @return a reference to this
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UnicodeString& remove(void);
sl@0: 
sl@0:   /**
sl@0:    * Remove the characters in the range
sl@0:    * [<TT>start</TT>, <TT>start + length</TT>) from the UnicodeString object.
sl@0:    * @param start the offset of the first character to remove
sl@0:    * @param length the number of characters to remove
sl@0:    * @return a reference to this
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UnicodeString& remove(int32_t start,
sl@0:                                int32_t length = (int32_t)INT32_MAX);
sl@0: 
sl@0:   /**
sl@0:    * Remove the characters in the range
sl@0:    * [<TT>start</TT>, <TT>limit</TT>) from the UnicodeString object.
sl@0:    * @param start the offset of the first character to remove
sl@0:    * @param limit the offset immediately following the range to remove
sl@0:    * @return a reference to this
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UnicodeString& removeBetween(int32_t start,
sl@0:                                       int32_t limit = (int32_t)INT32_MAX);
sl@0: 
sl@0: 
sl@0:   /* Length operations */
sl@0: 
sl@0:   /**
sl@0:    * Pad the start of this UnicodeString with the character <TT>padChar</TT>.
sl@0:    * If the length of this UnicodeString is less than targetLength,
sl@0:    * length() - targetLength copies of padChar will be added to the
sl@0:    * beginning of this UnicodeString.
sl@0:    * @param targetLength the desired length of the string
sl@0:    * @param padChar the character to use for padding. Defaults to
sl@0:    * space (U+0020)
sl@0:    * @return TRUE if the text was padded, FALSE otherwise.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   UBool padLeading(int32_t targetLength,
sl@0:                     UChar padChar = 0x0020);
sl@0: 
sl@0:   /**
sl@0:    * Pad the end of this UnicodeString with the character <TT>padChar</TT>.
sl@0:    * If the length of this UnicodeString is less than targetLength,
sl@0:    * length() - targetLength copies of padChar will be added to the
sl@0:    * end of this UnicodeString.
sl@0:    * @param targetLength the desired length of the string
sl@0:    * @param padChar the character to use for padding. Defaults to
sl@0:    * space (U+0020)
sl@0:    * @return TRUE if the text was padded, FALSE otherwise.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   UBool padTrailing(int32_t targetLength,
sl@0:                      UChar padChar = 0x0020);
sl@0: 
sl@0:   /**
sl@0:    * Truncate this UnicodeString to the <TT>targetLength</TT>.
sl@0:    * @param targetLength the desired length of this UnicodeString.
sl@0:    * @return TRUE if the text was truncated, FALSE otherwise
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UBool truncate(int32_t targetLength);
sl@0: 
sl@0:   /**
sl@0:    * Trims leading and trailing whitespace from this UnicodeString.
sl@0:    * @return a reference to this
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   UnicodeString& trim(void);
sl@0: 
sl@0: 
sl@0:   /* Miscellaneous operations */
sl@0: 
sl@0:   /**
sl@0:    * Reverse this UnicodeString in place.
sl@0:    * @return a reference to this
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UnicodeString& reverse(void);
sl@0: 
sl@0:   /**
sl@0:    * Reverse the range [<TT>start</TT>, <TT>start + length</TT>) in
sl@0:    * this UnicodeString.
sl@0:    * @param start the start of the range to reverse
sl@0:    * @param length the number of characters to to reverse
sl@0:    * @return a reference to this
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline UnicodeString& reverse(int32_t start,
sl@0:              int32_t length);
sl@0: 
sl@0:   /**
sl@0:    * Convert the characters in this to UPPER CASE following the conventions of
sl@0:    * the default locale.
sl@0:    * @return A reference to this.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   UnicodeString& toUpper(void);
sl@0: 
sl@0:   /**
sl@0:    * Convert the characters in this to UPPER CASE following the conventions of
sl@0:    * a specific locale.
sl@0:    * @param locale The locale containing the conventions to use.
sl@0:    * @return A reference to this.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   UnicodeString& toUpper(const Locale& locale);
sl@0: 
sl@0:   /**
sl@0:    * Convert the characters in this to lower case following the conventions of
sl@0:    * the default locale.
sl@0:    * @return A reference to this.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   UnicodeString& toLower(void);
sl@0: 
sl@0:   /**
sl@0:    * Convert the characters in this to lower case following the conventions of
sl@0:    * a specific locale.
sl@0:    * @param locale The locale containing the conventions to use.
sl@0:    * @return A reference to this.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   UnicodeString& toLower(const Locale& locale);
sl@0: 
sl@0: #if !UCONFIG_NO_BREAK_ITERATION
sl@0: 
sl@0:   /**
sl@0:    * Titlecase this string, convenience function using the default locale.
sl@0:    *
sl@0:    * Casing is locale-dependent and context-sensitive.
sl@0:    * Titlecasing uses a break iterator to find the first characters of words
sl@0:    * that are to be titlecased. It titlecases those characters and lowercases
sl@0:    * all others.
sl@0:    *
sl@0:    * The titlecase break iterator can be provided to customize for arbitrary
sl@0:    * styles, using rules and dictionaries beyond the standard iterators.
sl@0:    * It may be more efficient to always provide an iterator to avoid
sl@0:    * opening and closing one for each string.
sl@0:    * The standard titlecase iterator for the root locale implements the
sl@0:    * algorithm of Unicode TR 21.
sl@0:    *
sl@0:    * This function uses only the first() and next() methods of the
sl@0:    * provided break iterator.
sl@0:    *
sl@0:    * @param titleIter A break iterator to find the first characters of words
sl@0:    *                  that are to be titlecased.
sl@0:    *                  If none is provided (0), then a standard titlecase
sl@0:    *                  break iterator is opened.
sl@0:    *                  Otherwise the provided iterator is set to the string's text.
sl@0:    * @return A reference to this.
sl@0:    * @stable ICU 2.1
sl@0:    */
sl@0:   UnicodeString &toTitle(BreakIterator *titleIter);
sl@0: 
sl@0:   /**
sl@0:    * Titlecase this string.
sl@0:    *
sl@0:    * Casing is locale-dependent and context-sensitive.
sl@0:    * Titlecasing uses a break iterator to find the first characters of words
sl@0:    * that are to be titlecased. It titlecases those characters and lowercases
sl@0:    * all others.
sl@0:    *
sl@0:    * The titlecase break iterator can be provided to customize for arbitrary
sl@0:    * styles, using rules and dictionaries beyond the standard iterators.
sl@0:    * It may be more efficient to always provide an iterator to avoid
sl@0:    * opening and closing one for each string.
sl@0:    * The standard titlecase iterator for the root locale implements the
sl@0:    * algorithm of Unicode TR 21.
sl@0:    *
sl@0:    * This function uses only the first() and next() methods of the
sl@0:    * provided break iterator.
sl@0:    *
sl@0:    * @param titleIter A break iterator to find the first characters of words
sl@0:    *                  that are to be titlecased.
sl@0:    *                  If none is provided (0), then a standard titlecase
sl@0:    *                  break iterator is opened.
sl@0:    *                  Otherwise the provided iterator is set to the string's text.
sl@0:    * @param locale    The locale to consider.
sl@0:    * @return A reference to this.
sl@0:    * @stable ICU 2.1
sl@0:    */
sl@0:   UnicodeString &toTitle(BreakIterator *titleIter, const Locale &locale);
sl@0: 
sl@0: #endif
sl@0: 
sl@0:   /**
sl@0:    * Case-fold the characters in this string.
sl@0:    * Case-folding is locale-independent and not context-sensitive,
sl@0:    * but there is an option for whether to include or exclude mappings for dotted I
sl@0:    * and dotless i that are marked with 'I' in CaseFolding.txt.
sl@0:    * The result may be longer or shorter than the original.
sl@0:    *
sl@0:    * @param options Either U_FOLD_CASE_DEFAULT or U_FOLD_CASE_EXCLUDE_SPECIAL_I
sl@0:    * @return A reference to this.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   UnicodeString &foldCase(uint32_t options=0 /*U_FOLD_CASE_DEFAULT*/);
sl@0: 
sl@0:   //========================================
sl@0:   // Access to the internal buffer
sl@0:   //========================================
sl@0: 
sl@0:   /**
sl@0:    * Get a read/write pointer to the internal buffer.
sl@0:    * The buffer is guaranteed to be large enough for at least minCapacity UChars,
sl@0:    * writable, and is still owned by the UnicodeString object.
sl@0:    * Calls to getBuffer(minCapacity) must not be nested, and
sl@0:    * must be matched with calls to releaseBuffer(newLength).
sl@0:    * If the string buffer was read-only or shared,
sl@0:    * then it will be reallocated and copied.
sl@0:    *
sl@0:    * An attempted nested call will return 0, and will not further modify the
sl@0:    * state of the UnicodeString object.
sl@0:    * It also returns 0 if the string is bogus.
sl@0:    *
sl@0:    * The actual capacity of the string buffer may be larger than minCapacity.
sl@0:    * getCapacity() returns the actual capacity.
sl@0:    * For many operations, the full capacity should be used to avoid reallocations.
sl@0:    *
sl@0:    * While the buffer is "open" between getBuffer(minCapacity)
sl@0:    * and releaseBuffer(newLength), the following applies:
sl@0:    * - The string length is set to 0.
sl@0:    * - Any read API call on the UnicodeString object will behave like on a 0-length string.
sl@0:    * - Any write API call on the UnicodeString object is disallowed and will have no effect.
sl@0:    * - You can read from and write to the returned buffer.
sl@0:    * - The previous string contents will still be in the buffer;
sl@0:    *   if you want to use it, then you need to call length() before getBuffer(minCapacity).
sl@0:    *   If the length() was greater than minCapacity, then any contents after minCapacity
sl@0:    *   may be lost.
sl@0:    *   The buffer contents is not NUL-terminated by getBuffer().
sl@0:    *   If length()<getCapacity() then you can terminate it by writing a NUL
sl@0:    *   at index length().
sl@0:    * - You must call releaseBuffer(newLength) before and in order to
sl@0:    *   return to normal UnicodeString operation.
sl@0:    *
sl@0:    * @param minCapacity the minimum number of UChars that are to be available
sl@0:    *        in the buffer, starting at the returned pointer;
sl@0:    *        default to the current string capacity if minCapacity==-1
sl@0:    * @return a writable pointer to the internal string buffer,
sl@0:    *         or 0 if an error occurs (nested calls, out of memory)
sl@0:    *
sl@0:    * @see releaseBuffer
sl@0:    * @see getTerminatedBuffer()
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   UChar *getBuffer(int32_t minCapacity);
sl@0: 
sl@0:   /**
sl@0:    * Release a read/write buffer on a UnicodeString object with an
sl@0:    * "open" getBuffer(minCapacity).
sl@0:    * This function must be called in a matched pair with getBuffer(minCapacity).
sl@0:    * releaseBuffer(newLength) must be called if and only if a getBuffer(minCapacity) is "open".
sl@0:    *
sl@0:    * It will set the string length to newLength, at most to the current capacity.
sl@0:    * If newLength==-1 then it will set the length according to the
sl@0:    * first NUL in the buffer, or to the capacity if there is no NUL.
sl@0:    *
sl@0:    * After calling releaseBuffer(newLength) the UnicodeString is back to normal operation.
sl@0:    *
sl@0:    * @param newLength the new length of the UnicodeString object;
sl@0:    *        defaults to the current capacity if newLength is greater than that;
sl@0:    *        if newLength==-1, it defaults to u_strlen(buffer) but not more than
sl@0:    *        the current capacity of the string
sl@0:    *
sl@0:    * @see getBuffer(int32_t minCapacity)
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   void releaseBuffer(int32_t newLength=-1);
sl@0: 
sl@0:   /**
sl@0:    * Get a read-only pointer to the internal buffer.
sl@0:    * This can be called at any time on a valid UnicodeString.
sl@0:    *
sl@0:    * It returns 0 if the string is bogus, or
sl@0:    * during an "open" getBuffer(minCapacity).
sl@0:    *
sl@0:    * It can be called as many times as desired.
sl@0:    * The pointer that it returns will remain valid until the UnicodeString object is modified,
sl@0:    * at which time the pointer is semantically invalidated and must not be used any more.
sl@0:    *
sl@0:    * The capacity of the buffer can be determined with getCapacity().
sl@0:    * The part after length() may or may not be initialized and valid,
sl@0:    * depending on the history of the UnicodeString object.
sl@0:    *
sl@0:    * The buffer contents is (probably) not NUL-terminated.
sl@0:    * You can check if it is with
sl@0:    * <code>(s.length()<s.getCapacity() && buffer[s.length()]==0)</code>.
sl@0:    * (See getTerminatedBuffer().)
sl@0:    *
sl@0:    * The buffer may reside in read-only memory. Its contents must not
sl@0:    * be modified.
sl@0:    *
sl@0:    * @return a read-only pointer to the internal string buffer,
sl@0:    *         or 0 if the string is empty or bogus
sl@0:    *
sl@0:    * @see getBuffer(int32_t minCapacity)
sl@0:    * @see getTerminatedBuffer()
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   inline const UChar *getBuffer() const;
sl@0: 
sl@0:   /**
sl@0:    * Get a read-only pointer to the internal buffer,
sl@0:    * making sure that it is NUL-terminated.
sl@0:    * This can be called at any time on a valid UnicodeString.
sl@0:    *
sl@0:    * It returns 0 if the string is bogus, or
sl@0:    * during an "open" getBuffer(minCapacity), or if the buffer cannot
sl@0:    * be NUL-terminated (because memory allocation failed).
sl@0:    *
sl@0:    * It can be called as many times as desired.
sl@0:    * The pointer that it returns will remain valid until the UnicodeString object is modified,
sl@0:    * at which time the pointer is semantically invalidated and must not be used any more.
sl@0:    *
sl@0:    * The capacity of the buffer can be determined with getCapacity().
sl@0:    * The part after length()+1 may or may not be initialized and valid,
sl@0:    * depending on the history of the UnicodeString object.
sl@0:    *
sl@0:    * The buffer contents is guaranteed to be NUL-terminated.
sl@0:    * getTerminatedBuffer() may reallocate the buffer if a terminating NUL
sl@0:    * is written.
sl@0:    * For this reason, this function is not const, unlike getBuffer().
sl@0:    * Note that a UnicodeString may also contain NUL characters as part of its contents.
sl@0:    *
sl@0:    * The buffer may reside in read-only memory. Its contents must not
sl@0:    * be modified.
sl@0:    *
sl@0:    * @return a read-only pointer to the internal string buffer,
sl@0:    *         or 0 if the string is empty or bogus
sl@0:    *
sl@0:    * @see getBuffer(int32_t minCapacity)
sl@0:    * @see getBuffer()
sl@0:    * @stable ICU 2.2
sl@0:    */
sl@0:   inline const UChar *getTerminatedBuffer();
sl@0: 
sl@0:   //========================================
sl@0:   // Constructors
sl@0:   //========================================
sl@0: 
sl@0:   /** Construct an empty UnicodeString.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   UnicodeString();
sl@0: 
sl@0:   /**
sl@0:    * Construct a UnicodeString with capacity to hold <TT>capacity</TT> UChars
sl@0:    * @param capacity the number of UChars this UnicodeString should hold
sl@0:    * before a resize is necessary; if count is greater than 0 and count
sl@0:    * code points c take up more space than capacity, then capacity is adjusted
sl@0:    * accordingly.
sl@0:    * @param c is used to initially fill the string
sl@0:    * @param count specifies how many code points c are to be written in the
sl@0:    *              string
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   UnicodeString(int32_t capacity, UChar32 c, int32_t count);
sl@0: 
sl@0:   /**
sl@0:    * Single UChar (code unit) constructor.
sl@0:    * @param ch the character to place in the UnicodeString
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   UnicodeString(UChar ch);
sl@0: 
sl@0:   /**
sl@0:    * Single UChar32 (code point) constructor.
sl@0:    * @param ch the character to place in the UnicodeString
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   UnicodeString(UChar32 ch);
sl@0: 
sl@0:   /**
sl@0:    * UChar* constructor.
sl@0:    * @param text The characters to place in the UnicodeString.  <TT>text</TT>
sl@0:    * must be NULL (U+0000) terminated.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   UnicodeString(const UChar *text);
sl@0: 
sl@0:   /**
sl@0:    * UChar* constructor.
sl@0:    * @param text The characters to place in the UnicodeString.
sl@0:    * @param textLength The number of Unicode characters in <TT>text</TT>
sl@0:    * to copy.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   UnicodeString(const UChar *text,
sl@0:         int32_t textLength);
sl@0: 
sl@0:   /**
sl@0:    * Readonly-aliasing UChar* constructor.
sl@0:    * The text will be used for the UnicodeString object, but
sl@0:    * it will not be released when the UnicodeString is destroyed.
sl@0:    * This has copy-on-write semantics:
sl@0:    * When the string is modified, then the buffer is first copied into
sl@0:    * newly allocated memory.
sl@0:    * The aliased buffer is never modified.
sl@0:    * In an assignment to another UnicodeString, the text will be aliased again,
sl@0:    * so that both strings then alias the same readonly-text.
sl@0:    *
sl@0:    * @param isTerminated specifies if <code>text</code> is <code>NUL</code>-terminated.
sl@0:    *                     This must be true if <code>textLength==-1</code>.
sl@0:    * @param text The characters to alias for the UnicodeString.
sl@0:    * @param textLength The number of Unicode characters in <code>text</code> to alias.
sl@0:    *                   If -1, then this constructor will determine the length
sl@0:    *                   by calling <code>u_strlen()</code>.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   UnicodeString(UBool isTerminated,
sl@0:                 const UChar *text,
sl@0:                 int32_t textLength);
sl@0: 
sl@0:   /**
sl@0:    * Writable-aliasing UChar* constructor.
sl@0:    * The text will be used for the UnicodeString object, but
sl@0:    * it will not be released when the UnicodeString is destroyed.
sl@0:    * This has write-through semantics:
sl@0:    * For as long as the capacity of the buffer is sufficient, write operations
sl@0:    * will directly affect the buffer. When more capacity is necessary, then
sl@0:    * a new buffer will be allocated and the contents copied as with regularly
sl@0:    * constructed strings.
sl@0:    * In an assignment to another UnicodeString, the buffer will be copied.
sl@0:    * The extract(UChar *dst) function detects whether the dst pointer is the same
sl@0:    * as the string buffer itself and will in this case not copy the contents.
sl@0:    *
sl@0:    * @param buffer The characters to alias for the UnicodeString.
sl@0:    * @param buffLength The number of Unicode characters in <code>buffer</code> to alias.
sl@0:    * @param buffCapacity The size of <code>buffer</code> in UChars.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   UnicodeString(UChar *buffer, int32_t buffLength, int32_t buffCapacity);
sl@0: 
sl@0: #if !UCONFIG_NO_CONVERSION
sl@0: 
sl@0:   /**
sl@0:    * char* constructor.
sl@0:    * @param codepageData an array of bytes, null-terminated
sl@0:    * @param codepage the encoding of <TT>codepageData</TT>.  The special
sl@0:    * value 0 for <TT>codepage</TT> indicates that the text is in the
sl@0:    * platform's default codepage.
sl@0:    *
sl@0:    * If <code>codepage</code> is an empty string (<code>""</code>),
sl@0:    * then a simple conversion is performed on the codepage-invariant
sl@0:    * subset ("invariant characters") of the platform encoding. See utypes.h.
sl@0:    * Recommendation: For invariant-character strings use the constructor
sl@0:    * UnicodeString(const char *src, int32_t length, enum EInvariant inv)
sl@0:    * because it avoids object code dependencies of UnicodeString on
sl@0:    * the conversion code.
sl@0:    *
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   UnicodeString(const char *codepageData,
sl@0:         const char *codepage = 0);
sl@0: 
sl@0:   /**
sl@0:    * char* constructor.
sl@0:    * @param codepageData an array of bytes.
sl@0:    * @param dataLength The number of bytes in <TT>codepageData</TT>.
sl@0:    * @param codepage the encoding of <TT>codepageData</TT>.  The special
sl@0:    * value 0 for <TT>codepage</TT> indicates that the text is in the
sl@0:    * platform's default codepage.
sl@0:    * If <code>codepage</code> is an empty string (<code>""</code>),
sl@0:    * then a simple conversion is performed on the codepage-invariant
sl@0:    * subset ("invariant characters") of the platform encoding. See utypes.h.
sl@0:    * Recommendation: For invariant-character strings use the constructor
sl@0:    * UnicodeString(const char *src, int32_t length, enum EInvariant inv)
sl@0:    * because it avoids object code dependencies of UnicodeString on
sl@0:    * the conversion code.
sl@0:    *
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   UnicodeString(const char *codepageData,
sl@0:         int32_t dataLength,
sl@0:         const char *codepage = 0);
sl@0: 
sl@0:   /**
sl@0:    * char * / UConverter constructor.
sl@0:    * This constructor uses an existing UConverter object to
sl@0:    * convert the codepage string to Unicode and construct a UnicodeString
sl@0:    * from that.
sl@0:    *
sl@0:    * The converter is reset at first.
sl@0:    * If the error code indicates a failure before this constructor is called,
sl@0:    * or if an error occurs during conversion or construction,
sl@0:    * then the string will be bogus.
sl@0:    *
sl@0:    * This function avoids the overhead of opening and closing a converter if
sl@0:    * multiple strings are constructed.
sl@0:    *
sl@0:    * @param src input codepage string
sl@0:    * @param srcLength length of the input string, can be -1 for NUL-terminated strings
sl@0:    * @param cnv converter object (ucnv_resetToUnicode() will be called),
sl@0:    *        can be NULL for the default converter
sl@0:    * @param errorCode normal ICU error code
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   UnicodeString(
sl@0:         const char *src, int32_t srcLength,
sl@0:         UConverter *cnv,
sl@0:         UErrorCode &errorCode);
sl@0: 
sl@0: #endif
sl@0: 
sl@0:   /**
sl@0:    * Constructs a Unicode string from an invariant-character char * string.
sl@0:    * About invariant characters see utypes.h.
sl@0:    * This constructor has no runtime dependency on conversion code and is
sl@0:    * therefore recommended over ones taking a charset name string
sl@0:    * (where the empty string "" indicates invariant-character conversion).
sl@0:    *
sl@0:    * Use the macro US_INV as the third, signature-distinguishing parameter.
sl@0:    *
sl@0:    * For example:
sl@0:    * \code
sl@0:    * void fn(const char *s) {
sl@0:    *   UnicodeString ustr(s, -1, US_INV);
sl@0:    *   // use ustr ...
sl@0:    * }
sl@0:    * \endcode
sl@0:    *
sl@0:    * @param src String using only invariant characters.
sl@0:    * @param length Length of src, or -1 if NUL-terminated.
sl@0:    * @param inv Signature-distinguishing paramater, use US_INV.
sl@0:    *
sl@0:    * @see US_INV
sl@0:    * @draft ICU 3.2
sl@0:    */
sl@0:   UnicodeString(const char *src, int32_t length, enum EInvariant inv);
sl@0: 
sl@0: 
sl@0:   /**
sl@0:    * Copy constructor.
sl@0:    * @param that The UnicodeString object to copy.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   UnicodeString(const UnicodeString& that);
sl@0: 
sl@0:   /**
sl@0:    * 'Substring' constructor from tail of source string.
sl@0:    * @param src The UnicodeString object to copy.
sl@0:    * @param srcStart The offset into <tt>src</tt> at which to start copying.
sl@0:    * @stable ICU 2.2
sl@0:    */
sl@0:   UnicodeString(const UnicodeString& src, int32_t srcStart);
sl@0: 
sl@0:   /**
sl@0:    * 'Substring' constructor from subrange of source string.
sl@0:    * @param src The UnicodeString object to copy.
sl@0:    * @param srcStart The offset into <tt>src</tt> at which to start copying.
sl@0:    * @param srcLength The number of characters from <tt>src</tt> to copy.
sl@0:    * @stable ICU 2.2
sl@0:    */
sl@0:   UnicodeString(const UnicodeString& src, int32_t srcStart, int32_t srcLength);
sl@0: 
sl@0:   /**
sl@0:    * Clone this object, an instance of a subclass of Replaceable.
sl@0:    * Clones can be used concurrently in multiple threads.
sl@0:    * If a subclass does not implement clone(), or if an error occurs,
sl@0:    * then NULL is returned.
sl@0:    * The clone functions in all subclasses return a pointer to a Replaceable
sl@0:    * because some compilers do not support covariant (same-as-this)
sl@0:    * return types; cast to the appropriate subclass if necessary.
sl@0:    * The caller must delete the clone.
sl@0:    *
sl@0:    * @return a clone of this object
sl@0:    *
sl@0:    * @see Replaceable::clone
sl@0:    * @see getDynamicClassID
sl@0:    * @stable ICU 2.6
sl@0:    */
sl@0:   virtual Replaceable *clone() const;
sl@0: 
sl@0:   /** Destructor.
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   virtual ~UnicodeString();
sl@0: 
sl@0: 
sl@0:   /* Miscellaneous operations */
sl@0: 
sl@0:   /**
sl@0:    * Unescape a string of characters and return a string containing
sl@0:    * the result.  The following escape sequences are recognized:
sl@0:    *
sl@0:    * \\uhhhh       4 hex digits; h in [0-9A-Fa-f]
sl@0:    * \\Uhhhhhhhh   8 hex digits
sl@0:    * \\xhh         1-2 hex digits
sl@0:    * \\ooo         1-3 octal digits; o in [0-7]
sl@0:    * \\cX          control-X; X is masked with 0x1F
sl@0:    *
sl@0:    * as well as the standard ANSI C escapes:
sl@0:    *
sl@0:    * \\a => U+0007, \\b => U+0008, \\t => U+0009, \\n => U+000A,
sl@0:    * \\v => U+000B, \\f => U+000C, \\r => U+000D, \\e => U+001B,
sl@0:    * \\" => U+0022, \\' => U+0027, \\? => U+003F, \\\\ => U+005C
sl@0:    *
sl@0:    * Anything else following a backslash is generically escaped.  For
sl@0:    * example, "[a\\-z]" returns "[a-z]".
sl@0:    *
sl@0:    * If an escape sequence is ill-formed, this method returns an empty
sl@0:    * string.  An example of an ill-formed sequence is "\\u" followed by
sl@0:    * fewer than 4 hex digits.
sl@0:    *
sl@0:    * This function is similar to u_unescape() but not identical to it.
sl@0:    * The latter takes a source char*, so it does escape recognition
sl@0:    * and also invariant conversion.
sl@0:    *
sl@0:    * @return a string with backslash escapes interpreted, or an
sl@0:    * empty string on error.
sl@0:    * @see UnicodeString#unescapeAt()
sl@0:    * @see u_unescape()
sl@0:    * @see u_unescapeAt()
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   UnicodeString unescape() const;
sl@0: 
sl@0:   /**
sl@0:    * Unescape a single escape sequence and return the represented
sl@0:    * character.  See unescape() for a listing of the recognized escape
sl@0:    * sequences.  The character at offset-1 is assumed (without
sl@0:    * checking) to be a backslash.  If the escape sequence is
sl@0:    * ill-formed, or the offset is out of range, (UChar32)0xFFFFFFFF is
sl@0:    * returned.
sl@0:    *
sl@0:    * @param offset an input output parameter.  On input, it is the
sl@0:    * offset into this string where the escape sequence is located,
sl@0:    * after the initial backslash.  On output, it is advanced after the
sl@0:    * last character parsed.  On error, it is not advanced at all.
sl@0:    * @return the character represented by the escape sequence at
sl@0:    * offset, or (UChar32)0xFFFFFFFF on error.
sl@0:    * @see UnicodeString#unescape()
sl@0:    * @see u_unescape()
sl@0:    * @see u_unescapeAt()
sl@0:    * @stable ICU 2.0
sl@0:    */
sl@0:   UChar32 unescapeAt(int32_t &offset) const;
sl@0: 
sl@0:   /**
sl@0:    * ICU "poor man's RTTI", returns a UClassID for this class.
sl@0:    *
sl@0:    * @stable ICU 2.2
sl@0:    */
sl@0:   static UClassID U_EXPORT2 getStaticClassID();
sl@0: 
sl@0:   /**
sl@0:    * ICU "poor man's RTTI", returns a UClassID for the actual class.
sl@0:    *
sl@0:    * @stable ICU 2.2
sl@0:    */
sl@0:   virtual UClassID getDynamicClassID() const;
sl@0: 
sl@0:   //========================================
sl@0:   // Implementation methods
sl@0:   //========================================
sl@0: 
sl@0: protected:
sl@0:   /**
sl@0:    * Implement Replaceable::getLength() (see jitterbug 1027).
sl@0:    * @stable ICU 2.4
sl@0:    */
sl@0:   virtual int32_t getLength() const;
sl@0: 
sl@0:   /**
sl@0:    * The change in Replaceable to use virtual getCharAt() allows
sl@0:    * UnicodeString::charAt() to be inline again (see jitterbug 709).
sl@0:    * @stable ICU 2.4
sl@0:    */
sl@0:   virtual UChar getCharAt(int32_t offset) const;
sl@0: 
sl@0:   /**
sl@0:    * The change in Replaceable to use virtual getChar32At() allows
sl@0:    * UnicodeString::char32At() to be inline again (see jitterbug 709).
sl@0:    * @stable ICU 2.4
sl@0:    */
sl@0:   virtual UChar32 getChar32At(int32_t offset) const;
sl@0: 
sl@0: private:
sl@0: 
sl@0:   inline int8_t
sl@0:   doCompare(int32_t start,
sl@0:            int32_t length,
sl@0:            const UnicodeString& srcText,
sl@0:            int32_t srcStart,
sl@0:            int32_t srcLength) const;
sl@0: 
sl@0:   int8_t doCompare(int32_t start,
sl@0:            int32_t length,
sl@0:            const UChar *srcChars,
sl@0:            int32_t srcStart,
sl@0:            int32_t srcLength) const;
sl@0: 
sl@0:   inline int8_t
sl@0:   doCompareCodePointOrder(int32_t start,
sl@0:                           int32_t length,
sl@0:                           const UnicodeString& srcText,
sl@0:                           int32_t srcStart,
sl@0:                           int32_t srcLength) const;
sl@0: 
sl@0:   int8_t doCompareCodePointOrder(int32_t start,
sl@0:                                  int32_t length,
sl@0:                                  const UChar *srcChars,
sl@0:                                  int32_t srcStart,
sl@0:                                  int32_t srcLength) const;
sl@0: 
sl@0:   inline int8_t
sl@0:   doCaseCompare(int32_t start,
sl@0:                 int32_t length,
sl@0:                 const UnicodeString &srcText,
sl@0:                 int32_t srcStart,
sl@0:                 int32_t srcLength,
sl@0:                 uint32_t options) const;
sl@0: 
sl@0:   int8_t
sl@0:   doCaseCompare(int32_t start,
sl@0:                 int32_t length,
sl@0:                 const UChar *srcChars,
sl@0:                 int32_t srcStart,
sl@0:                 int32_t srcLength,
sl@0:                 uint32_t options) const;
sl@0: 
sl@0:   int32_t doIndexOf(UChar c,
sl@0:             int32_t start,
sl@0:             int32_t length) const;
sl@0: 
sl@0:   int32_t doIndexOf(UChar32 c,
sl@0:                         int32_t start,
sl@0:                         int32_t length) const;
sl@0: 
sl@0:   int32_t doLastIndexOf(UChar c,
sl@0:                 int32_t start,
sl@0:                 int32_t length) const;
sl@0: 
sl@0:   int32_t doLastIndexOf(UChar32 c,
sl@0:                             int32_t start,
sl@0:                             int32_t length) const;
sl@0: 
sl@0:   void doExtract(int32_t start,
sl@0:          int32_t length,
sl@0:          UChar *dst,
sl@0:          int32_t dstStart) const;
sl@0: 
sl@0:   inline void doExtract(int32_t start,
sl@0:          int32_t length,
sl@0:          UnicodeString& target) const;
sl@0: 
sl@0:   inline UChar doCharAt(int32_t offset)  const;
sl@0: 
sl@0:   UnicodeString& doReplace(int32_t start,
sl@0:                int32_t length,
sl@0:                const UnicodeString& srcText,
sl@0:                int32_t srcStart,
sl@0:                int32_t srcLength);
sl@0: 
sl@0:   UnicodeString& doReplace(int32_t start,
sl@0:                int32_t length,
sl@0:                const UChar *srcChars,
sl@0:                int32_t srcStart,
sl@0:                int32_t srcLength);
sl@0: 
sl@0:   UnicodeString& doReverse(int32_t start,
sl@0:                int32_t length);
sl@0: 
sl@0:   // calculate hash code
sl@0:   int32_t doHashCode(void) const;
sl@0: 
sl@0:   // get pointer to start of array
sl@0:   inline UChar* getArrayStart(void);
sl@0:   inline const UChar* getArrayStart(void) const;
sl@0: 
sl@0:   // allocate the array; result may be fStackBuffer
sl@0:   // sets refCount to 1 if appropriate
sl@0:   // sets fArray, fCapacity, and fFlags
sl@0:   // returns boolean for success or failure
sl@0:   UBool allocate(int32_t capacity);
sl@0: 
sl@0:   // release the array if owned
sl@0:   void releaseArray(void);
sl@0: 
sl@0:   // turn a bogus string into an empty one
sl@0:   void unBogus();
sl@0: 
sl@0:   // implements assigment operator, copy constructor, and fastCopyFrom()
sl@0:   UnicodeString &copyFrom(const UnicodeString &src, UBool fastCopy=FALSE);
sl@0: 
sl@0:   // Pin start and limit to acceptable values.
sl@0:   inline void pinIndex(int32_t& start) const;
sl@0:   inline void pinIndices(int32_t& start,
sl@0:                          int32_t& length) const;
sl@0: 
sl@0: #if !UCONFIG_NO_CONVERSION
sl@0: 
sl@0:   /* Internal extract() using UConverter. */
sl@0:   int32_t doExtract(int32_t start, int32_t length,
sl@0:                     char *dest, int32_t destCapacity,
sl@0:                     UConverter *cnv,
sl@0:                     UErrorCode &errorCode) const;
sl@0: 
sl@0:   /*
sl@0:    * Real constructor for converting from codepage data.
sl@0:    * It assumes that it is called with !fRefCounted.
sl@0:    *
sl@0:    * If <code>codepage==0</code>, then the default converter
sl@0:    * is used for the platform encoding.
sl@0:    * If <code>codepage</code> is an empty string (<code>""</code>),
sl@0:    * then a simple conversion is performed on the codepage-invariant
sl@0:    * subset ("invariant characters") of the platform encoding. See utypes.h.
sl@0:    */
sl@0:   void doCodepageCreate(const char *codepageData,
sl@0:                         int32_t dataLength,
sl@0:                         const char *codepage);
sl@0: 
sl@0:   /*
sl@0:    * Worker function for creating a UnicodeString from
sl@0:    * a codepage string using a UConverter.
sl@0:    */
sl@0:   void
sl@0:   doCodepageCreate(const char *codepageData,
sl@0:                    int32_t dataLength,
sl@0:                    UConverter *converter,
sl@0:                    UErrorCode &status);
sl@0: 
sl@0: #endif
sl@0: 
sl@0:   /*
sl@0:    * This function is called when write access to the array
sl@0:    * is necessary.
sl@0:    *
sl@0:    * We need to make a copy of the array if
sl@0:    * the buffer is read-only, or
sl@0:    * the buffer is refCounted (shared), and refCount>1, or
sl@0:    * the buffer is too small.
sl@0:    *
sl@0:    * Return FALSE if memory could not be allocated.
sl@0:    */
sl@0:   UBool cloneArrayIfNeeded(int32_t newCapacity = -1,
sl@0:                             int32_t growCapacity = -1,
sl@0:                             UBool doCopyArray = TRUE,
sl@0:                             int32_t **pBufferToDelete = 0,
sl@0:                             UBool forceClone = FALSE);
sl@0: 
sl@0:   // common function for case mappings
sl@0:   UnicodeString &
sl@0:   caseMap(BreakIterator *titleIter,
sl@0:           const char *locale,
sl@0:           uint32_t options,
sl@0:           int32_t toWhichCase);
sl@0: 
sl@0:   // ref counting
sl@0:   void addRef(void);
sl@0:   int32_t removeRef(void);
sl@0:   int32_t refCount(void) const;
sl@0: 
sl@0:   // constants
sl@0:   enum {
sl@0:     US_STACKBUF_SIZE=7, // Size of stack buffer for small strings
sl@0:     kInvalidUChar=0xffff, // invalid UChar index
sl@0:     kGrowSize=128, // grow size for this buffer
sl@0:     kInvalidHashCode=0, // invalid hash code
sl@0:     kEmptyHashCode=1, // hash code for empty string
sl@0: 
sl@0:     // bit flag values for fFlags
sl@0:     kIsBogus=1,         // this string is bogus, i.e., not valid or NULL
sl@0:     kUsingStackBuffer=2,// fArray==fStackBuffer
sl@0:     kRefCounted=4,      // there is a refCount field before the characters in fArray
sl@0:     kBufferIsReadonly=8,// do not write to this buffer
sl@0:     kOpenGetBuffer=16,  // getBuffer(minCapacity) was called (is "open"),
sl@0:                         // and releaseBuffer(newLength) must be called
sl@0: 
sl@0:     // combined values for convenience
sl@0:     kShortString=kUsingStackBuffer,
sl@0:     kLongString=kRefCounted,
sl@0:     kReadonlyAlias=kBufferIsReadonly,
sl@0:     kWritableAlias=0
sl@0:   };
sl@0: 
sl@0:   friend class StringCharacterIterator;
sl@0:   friend class StringThreadTest;
sl@0: 
sl@0:   /*
sl@0:    * The following are all the class fields that are stored
sl@0:    * in each UnicodeString object.
sl@0:    * Note that UnicodeString has virtual functions,
sl@0:    * therefore there is an implicit vtable pointer
sl@0:    * as the first real field.
sl@0:    * The fields should be aligned such that no padding is
sl@0:    * necessary, mostly by having larger types first.
sl@0:    * On 32-bit machines, the size should be 32 bytes,
sl@0:    * on 64-bit machines (8-byte pointers), it should be 40 bytes.
sl@0:    */
sl@0:   // (implicit) *vtable;
sl@0:   int32_t   fLength;        // number of characters in fArray
sl@0:   int32_t   fCapacity;      // sizeof fArray
sl@0:   UChar     *fArray;        // the Unicode data
sl@0:   uint16_t  fFlags;         // bit flags: see constants above
sl@0:   UChar     fStackBuffer [ US_STACKBUF_SIZE ]; // buffer for small strings
sl@0: 
sl@0: };
sl@0: 
sl@0: /**
sl@0:  * Create a new UnicodeString with the concatenation of two others.
sl@0:  *
sl@0:  * @param s1 The first string to be copied to the new one.
sl@0:  * @param s2 The second string to be copied to the new one, after s1.
sl@0:  * @return UnicodeString(s1).append(s2)
sl@0:  * @stable ICU 2.8
sl@0:  */
sl@0: U_COMMON_API UnicodeString U_EXPORT2
sl@0: operator+ (const UnicodeString &s1, const UnicodeString &s2);
sl@0: 
sl@0: U_NAMESPACE_END
sl@0: 
sl@0: // inline implementations -------------------------------------------------- ***
sl@0: 
sl@0: //========================================
sl@0: // Array copying
sl@0: //========================================
sl@0: /**
sl@0:  * Copy an array of UnicodeString OBJECTS (not pointers).
sl@0:  * @internal
sl@0:  */
sl@0: inline void
sl@0: uprv_arrayCopy(const U_NAMESPACE_QUALIFIER UnicodeString *src, U_NAMESPACE_QUALIFIER UnicodeString *dst, int32_t count)
sl@0: { while(count-- > 0) *dst++ = *src++; }
sl@0: 
sl@0: /**
sl@0:  * Copy an array of UnicodeString OBJECTS (not pointers).
sl@0:  * @internal
sl@0:  */
sl@0: inline void
sl@0: uprv_arrayCopy(const U_NAMESPACE_QUALIFIER UnicodeString *src, int32_t srcStart,
sl@0:         U_NAMESPACE_QUALIFIER UnicodeString *dst, int32_t dstStart, int32_t count)
sl@0: { uprv_arrayCopy(src+srcStart, dst+dstStart, count); }
sl@0: 
sl@0: U_NAMESPACE_BEGIN
sl@0: 
sl@0: //========================================
sl@0: // Inline members
sl@0: //========================================
sl@0: 
sl@0: //========================================
sl@0: // Privates
sl@0: //========================================
sl@0: 
sl@0: inline void
sl@0: UnicodeString::pinIndex(int32_t& start) const
sl@0: {
sl@0:   // pin index
sl@0:   if(start < 0) {
sl@0:     start = 0;
sl@0:   } else if(start > fLength) {
sl@0:     start = fLength;
sl@0:   }
sl@0: }
sl@0: 
sl@0: inline void
sl@0: UnicodeString::pinIndices(int32_t& start,
sl@0:                           int32_t& _length) const
sl@0: {
sl@0:   // pin indices
sl@0:   if(start < 0) {
sl@0:     start = 0;
sl@0:   } else if(start > fLength) {
sl@0:     start = fLength;
sl@0:   }
sl@0:   if(_length < 0) {
sl@0:     _length = 0;
sl@0:   } else if(_length > (fLength - start)) {
sl@0:     _length = (fLength - start);
sl@0:   }
sl@0: }
sl@0: 
sl@0: inline UChar*
sl@0: UnicodeString::getArrayStart()
sl@0: { return fArray; }
sl@0: 
sl@0: inline const UChar*
sl@0: UnicodeString::getArrayStart() const
sl@0: { return fArray; }
sl@0: 
sl@0: //========================================
sl@0: // Read-only implementation methods
sl@0: //========================================
sl@0: inline int32_t
sl@0: UnicodeString::length() const
sl@0: { return fLength; }
sl@0: 
sl@0: inline int32_t
sl@0: UnicodeString::getCapacity() const
sl@0: { return fCapacity; }
sl@0: 
sl@0: inline int32_t
sl@0: UnicodeString::hashCode() const
sl@0: { return doHashCode(); }
sl@0: 
sl@0: inline UBool
sl@0: UnicodeString::isBogus() const
sl@0: { return (UBool)(fFlags & kIsBogus); }
sl@0: 
sl@0: inline const UChar *
sl@0: UnicodeString::getBuffer() const {
sl@0:   if(!(fFlags&(kIsBogus|kOpenGetBuffer))) {
sl@0:     return fArray;
sl@0:   } else {
sl@0:     return 0;
sl@0:   }
sl@0: }
sl@0: 
sl@0: //========================================
sl@0: // Read-only alias methods
sl@0: //========================================
sl@0: inline int8_t
sl@0: UnicodeString::doCompare(int32_t start,
sl@0:               int32_t length,
sl@0:               const UnicodeString& srcText,
sl@0:               int32_t srcStart,
sl@0:               int32_t srcLength) const
sl@0: {
sl@0:   if(srcText.isBogus()) {
sl@0:     return (int8_t)!isBogus(); // 0 if both are bogus, 1 otherwise
sl@0:   } else {
sl@0:     srcText.pinIndices(srcStart, srcLength);
sl@0:     return doCompare(start, length, srcText.fArray, srcStart, srcLength);
sl@0:   }
sl@0: }
sl@0: 
sl@0: inline UBool
sl@0: UnicodeString::operator== (const UnicodeString& text) const
sl@0: {
sl@0:   if(isBogus()) {
sl@0:     return text.isBogus();
sl@0:   } else {
sl@0:     return
sl@0:       !text.isBogus() &&
sl@0:       fLength == text.fLength &&
sl@0:       doCompare(0, fLength, text, 0, text.fLength) == 0;
sl@0:   }
sl@0: }
sl@0: 
sl@0: inline UBool
sl@0: UnicodeString::operator!= (const UnicodeString& text) const
sl@0: { return (! operator==(text)); }
sl@0: 
sl@0: inline UBool
sl@0: UnicodeString::operator> (const UnicodeString& text) const
sl@0: { return doCompare(0, fLength, text, 0, text.fLength) == 1; }
sl@0: 
sl@0: inline UBool
sl@0: UnicodeString::operator< (const UnicodeString& text) const
sl@0: { return doCompare(0, fLength, text, 0, text.fLength) == -1; }
sl@0: 
sl@0: inline UBool
sl@0: UnicodeString::operator>= (const UnicodeString& text) const
sl@0: { return doCompare(0, fLength, text, 0, text.fLength) != -1; }
sl@0: 
sl@0: inline UBool
sl@0: UnicodeString::operator<= (const UnicodeString& text) const
sl@0: { return doCompare(0, fLength, text, 0, text.fLength) != 1; }
sl@0: 
sl@0: inline int8_t
sl@0: UnicodeString::compare(const UnicodeString& text) const
sl@0: { return doCompare(0, fLength, text, 0, text.fLength); }
sl@0: 
sl@0: inline int8_t
sl@0: UnicodeString::compare(int32_t start,
sl@0:                int32_t _length,
sl@0:                const UnicodeString& srcText) const
sl@0: { return doCompare(start, _length, srcText, 0, srcText.fLength); }
sl@0: 
sl@0: inline int8_t
sl@0: UnicodeString::compare(const UChar *srcChars,
sl@0:                int32_t srcLength) const
sl@0: { return doCompare(0, fLength, srcChars, 0, srcLength); }
sl@0: 
sl@0: inline int8_t
sl@0: UnicodeString::compare(int32_t start,
sl@0:                int32_t _length,
sl@0:                const UnicodeString& srcText,
sl@0:                int32_t srcStart,
sl@0:                int32_t srcLength) const
sl@0: { return doCompare(start, _length, srcText, srcStart, srcLength); }
sl@0: 
sl@0: inline int8_t
sl@0: UnicodeString::compare(int32_t start,
sl@0:                int32_t _length,
sl@0:                const UChar *srcChars) const
sl@0: { return doCompare(start, _length, srcChars, 0, _length); }
sl@0: 
sl@0: inline int8_t
sl@0: UnicodeString::compare(int32_t start,
sl@0:                int32_t _length,
sl@0:                const UChar *srcChars,
sl@0:                int32_t srcStart,
sl@0:                int32_t srcLength) const
sl@0: { return doCompare(start, _length, srcChars, srcStart, srcLength); }
sl@0: 
sl@0: inline int8_t
sl@0: UnicodeString::compareBetween(int32_t start,
sl@0:                   int32_t limit,
sl@0:                   const UnicodeString& srcText,
sl@0:                   int32_t srcStart,
sl@0:                   int32_t srcLimit) const
sl@0: { return doCompare(start, limit - start,
sl@0:            srcText, srcStart, srcLimit - srcStart); }
sl@0: 
sl@0: inline int8_t
sl@0: UnicodeString::doCompareCodePointOrder(int32_t start,
sl@0:                                        int32_t length,
sl@0:                                        const UnicodeString& srcText,
sl@0:                                        int32_t srcStart,
sl@0:                                        int32_t srcLength) const
sl@0: {
sl@0:   if(srcText.isBogus()) {
sl@0:     return (int8_t)!isBogus(); // 0 if both are bogus, 1 otherwise
sl@0:   } else {
sl@0:     srcText.pinIndices(srcStart, srcLength);
sl@0:     return doCompareCodePointOrder(start, length, srcText.fArray, srcStart, srcLength);
sl@0:   }
sl@0: }
sl@0: 
sl@0: inline int8_t
sl@0: UnicodeString::compareCodePointOrder(const UnicodeString& text) const
sl@0: { return doCompareCodePointOrder(0, fLength, text, 0, text.fLength); }
sl@0: 
sl@0: inline int8_t
sl@0: UnicodeString::compareCodePointOrder(int32_t start,
sl@0:                                      int32_t _length,
sl@0:                                      const UnicodeString& srcText) const
sl@0: { return doCompareCodePointOrder(start, _length, srcText, 0, srcText.fLength); }
sl@0: 
sl@0: inline int8_t
sl@0: UnicodeString::compareCodePointOrder(const UChar *srcChars,
sl@0:                                      int32_t srcLength) const
sl@0: { return doCompareCodePointOrder(0, fLength, srcChars, 0, srcLength); }
sl@0: 
sl@0: inline int8_t
sl@0: UnicodeString::compareCodePointOrder(int32_t start,
sl@0:                                      int32_t _length,
sl@0:                                      const UnicodeString& srcText,
sl@0:                                      int32_t srcStart,
sl@0:                                      int32_t srcLength) const
sl@0: { return doCompareCodePointOrder(start, _length, srcText, srcStart, srcLength); }
sl@0: 
sl@0: inline int8_t
sl@0: UnicodeString::compareCodePointOrder(int32_t start,
sl@0:                                      int32_t _length,
sl@0:                                      const UChar *srcChars) const
sl@0: { return doCompareCodePointOrder(start, _length, srcChars, 0, _length); }
sl@0: 
sl@0: inline int8_t
sl@0: UnicodeString::compareCodePointOrder(int32_t start,
sl@0:                                      int32_t _length,
sl@0:                                      const UChar *srcChars,
sl@0:                                      int32_t srcStart,
sl@0:                                      int32_t srcLength) const
sl@0: { return doCompareCodePointOrder(start, _length, srcChars, srcStart, srcLength); }
sl@0: 
sl@0: inline int8_t
sl@0: UnicodeString::compareCodePointOrderBetween(int32_t start,
sl@0:                                             int32_t limit,
sl@0:                                             const UnicodeString& srcText,
sl@0:                                             int32_t srcStart,
sl@0:                                             int32_t srcLimit) const
sl@0: { return doCompareCodePointOrder(start, limit - start,
sl@0:            srcText, srcStart, srcLimit - srcStart); }
sl@0: 
sl@0: inline int8_t
sl@0: UnicodeString::doCaseCompare(int32_t start,
sl@0:                              int32_t length,
sl@0:                              const UnicodeString &srcText,
sl@0:                              int32_t srcStart,
sl@0:                              int32_t srcLength,
sl@0:                              uint32_t options) const
sl@0: {
sl@0:   if(srcText.isBogus()) {
sl@0:     return (int8_t)!isBogus(); // 0 if both are bogus, 1 otherwise
sl@0:   } else {
sl@0:     srcText.pinIndices(srcStart, srcLength);
sl@0:     return doCaseCompare(start, length, srcText.fArray, srcStart, srcLength, options);
sl@0:   }
sl@0: }
sl@0: 
sl@0: inline int8_t
sl@0: UnicodeString::caseCompare(const UnicodeString &text, uint32_t options) const {
sl@0:   return doCaseCompare(0, fLength, text, 0, text.fLength, options);
sl@0: }
sl@0: 
sl@0: inline int8_t
sl@0: UnicodeString::caseCompare(int32_t start,
sl@0:                            int32_t _length,
sl@0:                            const UnicodeString &srcText,
sl@0:                            uint32_t options) const {
sl@0:   return doCaseCompare(start, _length, srcText, 0, srcText.fLength, options);
sl@0: }
sl@0: 
sl@0: inline int8_t
sl@0: UnicodeString::caseCompare(const UChar *srcChars,
sl@0:                            int32_t srcLength,
sl@0:                            uint32_t options) const {
sl@0:   return doCaseCompare(0, fLength, srcChars, 0, srcLength, options);
sl@0: }
sl@0: 
sl@0: inline int8_t
sl@0: UnicodeString::caseCompare(int32_t start,
sl@0:                            int32_t _length,
sl@0:                            const UnicodeString &srcText,
sl@0:                            int32_t srcStart,
sl@0:                            int32_t srcLength,
sl@0:                            uint32_t options) const {
sl@0:   return doCaseCompare(start, _length, srcText, srcStart, srcLength, options);
sl@0: }
sl@0: 
sl@0: inline int8_t
sl@0: UnicodeString::caseCompare(int32_t start,
sl@0:                            int32_t _length,
sl@0:                            const UChar *srcChars,
sl@0:                            uint32_t options) const {
sl@0:   return doCaseCompare(start, _length, srcChars, 0, _length, options);
sl@0: }
sl@0: 
sl@0: inline int8_t
sl@0: UnicodeString::caseCompare(int32_t start,
sl@0:                            int32_t _length,
sl@0:                            const UChar *srcChars,
sl@0:                            int32_t srcStart,
sl@0:                            int32_t srcLength,
sl@0:                            uint32_t options) const {
sl@0:   return doCaseCompare(start, _length, srcChars, srcStart, srcLength, options);
sl@0: }
sl@0: 
sl@0: inline int8_t
sl@0: UnicodeString::caseCompareBetween(int32_t start,
sl@0:                                   int32_t limit,
sl@0:                                   const UnicodeString &srcText,
sl@0:                                   int32_t srcStart,
sl@0:                                   int32_t srcLimit,
sl@0:                                   uint32_t options) const {
sl@0:   return doCaseCompare(start, limit - start, srcText, srcStart, srcLimit - srcStart, options);
sl@0: }
sl@0: 
sl@0: inline int32_t
sl@0: UnicodeString::indexOf(const UnicodeString& srcText,
sl@0:                int32_t srcStart,
sl@0:                int32_t srcLength,
sl@0:                int32_t start,
sl@0:                int32_t _length) const
sl@0: {
sl@0:   if(!srcText.isBogus()) {
sl@0:     srcText.pinIndices(srcStart, srcLength);
sl@0:     if(srcLength > 0) {
sl@0:       return indexOf(srcText.getArrayStart(), srcStart, srcLength, start, _length);
sl@0:     }
sl@0:   }
sl@0:   return -1;
sl@0: }
sl@0: 
sl@0: inline int32_t
sl@0: UnicodeString::indexOf(const UnicodeString& text) const
sl@0: { return indexOf(text, 0, text.fLength, 0, fLength); }
sl@0: 
sl@0: inline int32_t
sl@0: UnicodeString::indexOf(const UnicodeString& text,
sl@0:                int32_t start) const {
sl@0:   pinIndex(start);
sl@0:   return indexOf(text, 0, text.fLength, start, fLength - start);
sl@0: }
sl@0: 
sl@0: inline int32_t
sl@0: UnicodeString::indexOf(const UnicodeString& text,
sl@0:                int32_t start,
sl@0:                int32_t _length) const
sl@0: { return indexOf(text, 0, text.fLength, start, _length); }
sl@0: 
sl@0: inline int32_t
sl@0: UnicodeString::indexOf(const UChar *srcChars,
sl@0:                int32_t srcLength,
sl@0:                int32_t start) const {
sl@0:   pinIndex(start);
sl@0:   return indexOf(srcChars, 0, srcLength, start, fLength - start);
sl@0: }
sl@0: 
sl@0: inline int32_t
sl@0: UnicodeString::indexOf(const UChar *srcChars,
sl@0:                int32_t srcLength,
sl@0:                int32_t start,
sl@0:                int32_t _length) const
sl@0: { return indexOf(srcChars, 0, srcLength, start, _length); }
sl@0: 
sl@0: inline int32_t
sl@0: UnicodeString::indexOf(UChar c,
sl@0:                int32_t start,
sl@0:                int32_t _length) const
sl@0: { return doIndexOf(c, start, _length); }
sl@0: 
sl@0: inline int32_t
sl@0: UnicodeString::indexOf(UChar32 c,
sl@0:                int32_t start,
sl@0:                int32_t _length) const
sl@0: { return doIndexOf(c, start, _length); }
sl@0: 
sl@0: inline int32_t
sl@0: UnicodeString::indexOf(UChar c) const
sl@0: { return doIndexOf(c, 0, fLength); }
sl@0: 
sl@0: inline int32_t
sl@0: UnicodeString::indexOf(UChar32 c) const
sl@0: { return indexOf(c, 0, fLength); }
sl@0: 
sl@0: inline int32_t
sl@0: UnicodeString::indexOf(UChar c,
sl@0:                int32_t start) const {
sl@0:   pinIndex(start);
sl@0:   return doIndexOf(c, start, fLength - start);
sl@0: }
sl@0: 
sl@0: inline int32_t
sl@0: UnicodeString::indexOf(UChar32 c,
sl@0:                int32_t start) const {
sl@0:   pinIndex(start);
sl@0:   return indexOf(c, start, fLength - start);
sl@0: }
sl@0: 
sl@0: inline int32_t
sl@0: UnicodeString::lastIndexOf(const UChar *srcChars,
sl@0:                int32_t srcLength,
sl@0:                int32_t start,
sl@0:                int32_t _length) const
sl@0: { return lastIndexOf(srcChars, 0, srcLength, start, _length); }
sl@0: 
sl@0: inline int32_t
sl@0: UnicodeString::lastIndexOf(const UChar *srcChars,
sl@0:                int32_t srcLength,
sl@0:                int32_t start) const {
sl@0:   pinIndex(start);
sl@0:   return lastIndexOf(srcChars, 0, srcLength, start, fLength - start);
sl@0: }
sl@0: 
sl@0: inline int32_t
sl@0: UnicodeString::lastIndexOf(const UnicodeString& srcText,
sl@0:                int32_t srcStart,
sl@0:                int32_t srcLength,
sl@0:                int32_t start,
sl@0:                int32_t _length) const
sl@0: {
sl@0:   if(!srcText.isBogus()) {
sl@0:     srcText.pinIndices(srcStart, srcLength);
sl@0:     if(srcLength > 0) {
sl@0:       return lastIndexOf(srcText.getArrayStart(), srcStart, srcLength, start, _length);
sl@0:     }
sl@0:   }
sl@0:   return -1;
sl@0: }
sl@0: 
sl@0: inline int32_t
sl@0: UnicodeString::lastIndexOf(const UnicodeString& text,
sl@0:                int32_t start,
sl@0:                int32_t _length) const
sl@0: { return lastIndexOf(text, 0, text.fLength, start, _length); }
sl@0: 
sl@0: inline int32_t
sl@0: UnicodeString::lastIndexOf(const UnicodeString& text,
sl@0:                int32_t start) const {
sl@0:   pinIndex(start);
sl@0:   return lastIndexOf(text, 0, text.fLength, start, fLength - start);
sl@0: }
sl@0: 
sl@0: inline int32_t
sl@0: UnicodeString::lastIndexOf(const UnicodeString& text) const
sl@0: { return lastIndexOf(text, 0, text.fLength, 0, fLength); }
sl@0: 
sl@0: inline int32_t
sl@0: UnicodeString::lastIndexOf(UChar c,
sl@0:                int32_t start,
sl@0:                int32_t _length) const
sl@0: { return doLastIndexOf(c, start, _length); }
sl@0: 
sl@0: inline int32_t
sl@0: UnicodeString::lastIndexOf(UChar32 c,
sl@0:                int32_t start,
sl@0:                int32_t _length) const {
sl@0:   return doLastIndexOf(c, start, _length);
sl@0: }
sl@0: 
sl@0: inline int32_t
sl@0: UnicodeString::lastIndexOf(UChar c) const
sl@0: { return doLastIndexOf(c, 0, fLength); }
sl@0: 
sl@0: inline int32_t
sl@0: UnicodeString::lastIndexOf(UChar32 c) const {
sl@0:   return lastIndexOf(c, 0, fLength);
sl@0: }
sl@0: 
sl@0: inline int32_t
sl@0: UnicodeString::lastIndexOf(UChar c,
sl@0:                int32_t start) const {
sl@0:   pinIndex(start);
sl@0:   return doLastIndexOf(c, start, fLength - start);
sl@0: }
sl@0: 
sl@0: inline int32_t
sl@0: UnicodeString::lastIndexOf(UChar32 c,
sl@0:                int32_t start) const {
sl@0:   pinIndex(start);
sl@0:   return lastIndexOf(c, start, fLength - start);
sl@0: }
sl@0: 
sl@0: inline UBool
sl@0: UnicodeString::startsWith(const UnicodeString& text) const
sl@0: { return compare(0, text.fLength, text, 0, text.fLength) == 0; }
sl@0: 
sl@0: inline UBool
sl@0: UnicodeString::startsWith(const UnicodeString& srcText,
sl@0:               int32_t srcStart,
sl@0:               int32_t srcLength) const
sl@0: { return doCompare(0, srcLength, srcText, srcStart, srcLength) == 0; }
sl@0: 
sl@0: inline UBool
sl@0: UnicodeString::startsWith(const UChar *srcChars,
sl@0:               int32_t srcLength) const
sl@0: { return doCompare(0, srcLength, srcChars, 0, srcLength) == 0; }
sl@0: 
sl@0: inline UBool
sl@0: UnicodeString::startsWith(const UChar *srcChars,
sl@0:               int32_t srcStart,
sl@0:               int32_t srcLength) const
sl@0: { return doCompare(0, srcLength, srcChars, srcStart, srcLength) == 0;}
sl@0: 
sl@0: inline UBool
sl@0: UnicodeString::endsWith(const UnicodeString& text) const
sl@0: { return doCompare(fLength - text.fLength, text.fLength,
sl@0:            text, 0, text.fLength) == 0; }
sl@0: 
sl@0: inline UBool
sl@0: UnicodeString::endsWith(const UnicodeString& srcText,
sl@0:             int32_t srcStart,
sl@0:             int32_t srcLength) const {
sl@0:   srcText.pinIndices(srcStart, srcLength);
sl@0:   return doCompare(fLength - srcLength, srcLength,
sl@0:                    srcText, srcStart, srcLength) == 0;
sl@0: }
sl@0: 
sl@0: inline UBool
sl@0: UnicodeString::endsWith(const UChar *srcChars,
sl@0:             int32_t srcLength) const {
sl@0:   if(srcLength < 0) {
sl@0:     srcLength = u_strlen(srcChars);
sl@0:   }
sl@0:   return doCompare(fLength - srcLength, srcLength,
sl@0:                    srcChars, 0, srcLength) == 0;
sl@0: }
sl@0: 
sl@0: inline UBool
sl@0: UnicodeString::endsWith(const UChar *srcChars,
sl@0:             int32_t srcStart,
sl@0:             int32_t srcLength) const {
sl@0:   if(srcLength < 0) {
sl@0:     srcLength = u_strlen(srcChars + srcStart);
sl@0:   }
sl@0:   return doCompare(fLength - srcLength, srcLength,
sl@0:                    srcChars, srcStart, srcLength) == 0;
sl@0: }
sl@0: 
sl@0: //========================================
sl@0: // replace
sl@0: //========================================
sl@0: inline UnicodeString&
sl@0: UnicodeString::replace(int32_t start,
sl@0:                int32_t _length,
sl@0:                const UnicodeString& srcText)
sl@0: { return doReplace(start, _length, srcText, 0, srcText.fLength); }
sl@0: 
sl@0: inline UnicodeString&
sl@0: UnicodeString::replace(int32_t start,
sl@0:                int32_t _length,
sl@0:                const UnicodeString& srcText,
sl@0:                int32_t srcStart,
sl@0:                int32_t srcLength)
sl@0: { return doReplace(start, _length, srcText, srcStart, srcLength); }
sl@0: 
sl@0: inline UnicodeString&
sl@0: UnicodeString::replace(int32_t start,
sl@0:                int32_t _length,
sl@0:                const UChar *srcChars,
sl@0:                int32_t srcLength)
sl@0: { return doReplace(start, _length, srcChars, 0, srcLength); }
sl@0: 
sl@0: inline UnicodeString&
sl@0: UnicodeString::replace(int32_t start,
sl@0:                int32_t _length,
sl@0:                const UChar *srcChars,
sl@0:                int32_t srcStart,
sl@0:                int32_t srcLength)
sl@0: { return doReplace(start, _length, srcChars, srcStart, srcLength); }
sl@0: 
sl@0: inline UnicodeString&
sl@0: UnicodeString::replace(int32_t start,
sl@0:                int32_t _length,
sl@0:                UChar srcChar)
sl@0: { return doReplace(start, _length, &srcChar, 0, 1); }
sl@0: 
sl@0: inline UnicodeString&
sl@0: UnicodeString::replace(int32_t start,
sl@0:                int32_t _length,
sl@0:                UChar32 srcChar) {
sl@0:   UChar buffer[U16_MAX_LENGTH];
sl@0:   int32_t count = 0;
sl@0:   UBool isError = FALSE;
sl@0:   U16_APPEND(buffer, count, U16_MAX_LENGTH, srcChar, isError);
sl@0:   return doReplace(start, _length, buffer, 0, count);
sl@0: }
sl@0: 
sl@0: inline UnicodeString&
sl@0: UnicodeString::replaceBetween(int32_t start,
sl@0:                   int32_t limit,
sl@0:                   const UnicodeString& srcText)
sl@0: { return doReplace(start, limit - start, srcText, 0, srcText.fLength); }
sl@0: 
sl@0: inline UnicodeString&
sl@0: UnicodeString::replaceBetween(int32_t start,
sl@0:                   int32_t limit,
sl@0:                   const UnicodeString& srcText,
sl@0:                   int32_t srcStart,
sl@0:                   int32_t srcLimit)
sl@0: { return doReplace(start, limit - start, srcText, srcStart, srcLimit - srcStart); }
sl@0: 
sl@0: inline UnicodeString&
sl@0: UnicodeString::findAndReplace(const UnicodeString& oldText,
sl@0:                   const UnicodeString& newText)
sl@0: { return findAndReplace(0, fLength, oldText, 0, oldText.fLength,
sl@0:             newText, 0, newText.fLength); }
sl@0: 
sl@0: inline UnicodeString&
sl@0: UnicodeString::findAndReplace(int32_t start,
sl@0:                   int32_t _length,
sl@0:                   const UnicodeString& oldText,
sl@0:                   const UnicodeString& newText)
sl@0: { return findAndReplace(start, _length, oldText, 0, oldText.fLength,
sl@0:             newText, 0, newText.fLength); }
sl@0: 
sl@0: // ============================
sl@0: // extract
sl@0: // ============================
sl@0: inline void
sl@0: UnicodeString::doExtract(int32_t start,
sl@0:              int32_t _length,
sl@0:              UnicodeString& target) const
sl@0: { target.replace(0, target.fLength, *this, start, _length); }
sl@0: 
sl@0: inline void
sl@0: UnicodeString::extract(int32_t start,
sl@0:                int32_t _length,
sl@0:                UChar *target,
sl@0:                int32_t targetStart) const
sl@0: { doExtract(start, _length, target, targetStart); }
sl@0: 
sl@0: inline void
sl@0: UnicodeString::extract(int32_t start,
sl@0:                int32_t _length,
sl@0:                UnicodeString& target) const
sl@0: { doExtract(start, _length, target); }
sl@0: 
sl@0: #if !UCONFIG_NO_CONVERSION
sl@0: 
sl@0: inline int32_t
sl@0: UnicodeString::extract(int32_t start,
sl@0:                int32_t _length,
sl@0:                char *dst,
sl@0:                const char *codepage) const
sl@0: 
sl@0: {
sl@0:   // This dstSize value will be checked explicitly
sl@0:   return extract(start, _length, dst, dst!=0 ? 0xffffffff : 0, codepage);
sl@0: }
sl@0: 
sl@0: #endif
sl@0: 
sl@0: inline void
sl@0: UnicodeString::extractBetween(int32_t start,
sl@0:                   int32_t limit,
sl@0:                   UChar *dst,
sl@0:                   int32_t dstStart) const {
sl@0:   pinIndex(start);
sl@0:   pinIndex(limit);
sl@0:   doExtract(start, limit - start, dst, dstStart);
sl@0: }
sl@0: 
sl@0: inline UChar
sl@0: UnicodeString::doCharAt(int32_t offset) const
sl@0: {
sl@0:   if((uint32_t)offset < (uint32_t)fLength) {
sl@0:     return fArray[offset];
sl@0:   } else {
sl@0:     return kInvalidUChar;
sl@0:   }
sl@0: }
sl@0: 
sl@0: inline UChar
sl@0: UnicodeString::charAt(int32_t offset) const
sl@0: { return doCharAt(offset); }
sl@0: 
sl@0: inline UChar
sl@0: UnicodeString::operator[] (int32_t offset) const
sl@0: { return doCharAt(offset); }
sl@0: 
sl@0: inline UChar32
sl@0: UnicodeString::char32At(int32_t offset) const
sl@0: {
sl@0:   if((uint32_t)offset < (uint32_t)fLength) {
sl@0:     UChar32 c;
sl@0:     U16_GET(fArray, 0, offset, fLength, c);
sl@0:     return c;
sl@0:   } else {
sl@0:     return kInvalidUChar;
sl@0:   }
sl@0: }
sl@0: 
sl@0: inline int32_t
sl@0: UnicodeString::getChar32Start(int32_t offset) const {
sl@0:   if((uint32_t)offset < (uint32_t)fLength) {
sl@0:     U16_SET_CP_START(fArray, 0, offset);
sl@0:     return offset;
sl@0:   } else {
sl@0:     return 0;
sl@0:   }
sl@0: }
sl@0: 
sl@0: inline int32_t
sl@0: UnicodeString::getChar32Limit(int32_t offset) const {
sl@0:   if((uint32_t)offset < (uint32_t)fLength) {
sl@0:     U16_SET_CP_LIMIT(fArray, 0, offset, fLength);
sl@0:     return offset;
sl@0:   } else {
sl@0:     return fLength;
sl@0:   }
sl@0: }
sl@0: 
sl@0: inline UBool
sl@0: UnicodeString::isEmpty() const {
sl@0:   return fLength == 0;
sl@0: }
sl@0: 
sl@0: //========================================
sl@0: // Write implementation methods
sl@0: //========================================
sl@0: inline const UChar *
sl@0: UnicodeString::getTerminatedBuffer() {
sl@0:   if(fFlags&(kIsBogus|kOpenGetBuffer)) {
sl@0:     return 0;
sl@0:   } else if(fLength<fCapacity && fArray[fLength]==0) {
sl@0:     return fArray;
sl@0:   } else if(cloneArrayIfNeeded(fLength+1)) {
sl@0:     fArray[fLength]=0;
sl@0:     return fArray;
sl@0:   } else {
sl@0:     return 0;
sl@0:   }
sl@0: }
sl@0: 
sl@0: inline UnicodeString&
sl@0: UnicodeString::operator= (UChar ch)
sl@0: { return doReplace(0, fLength, &ch, 0, 1); }
sl@0: 
sl@0: inline UnicodeString&
sl@0: UnicodeString::operator= (UChar32 ch)
sl@0: { return replace(0, fLength, ch); }
sl@0: 
sl@0: inline UnicodeString&
sl@0: UnicodeString::setTo(const UnicodeString& srcText,
sl@0:              int32_t srcStart,
sl@0:              int32_t srcLength)
sl@0: {
sl@0:   unBogus();
sl@0:   return doReplace(0, fLength, srcText, srcStart, srcLength);
sl@0: }
sl@0: 
sl@0: inline UnicodeString&
sl@0: UnicodeString::setTo(const UnicodeString& srcText,
sl@0:              int32_t srcStart)
sl@0: {
sl@0:   unBogus();
sl@0:   srcText.pinIndex(srcStart);
sl@0:   return doReplace(0, fLength, srcText, srcStart, srcText.fLength - srcStart);
sl@0: }
sl@0: 
sl@0: inline UnicodeString&
sl@0: UnicodeString::setTo(const UnicodeString& srcText)
sl@0: {
sl@0:   unBogus();
sl@0:   return doReplace(0, fLength, srcText, 0, srcText.fLength);
sl@0: }
sl@0: 
sl@0: inline UnicodeString&
sl@0: UnicodeString::setTo(const UChar *srcChars,
sl@0:              int32_t srcLength)
sl@0: {
sl@0:   unBogus();
sl@0:   return doReplace(0, fLength, srcChars, 0, srcLength);
sl@0: }
sl@0: 
sl@0: inline UnicodeString&
sl@0: UnicodeString::setTo(UChar srcChar)
sl@0: {
sl@0:   unBogus();
sl@0:   return doReplace(0, fLength, &srcChar, 0, 1);
sl@0: }
sl@0: 
sl@0: inline UnicodeString&
sl@0: UnicodeString::setTo(UChar32 srcChar)
sl@0: {
sl@0:   unBogus();
sl@0:   return replace(0, fLength, srcChar);
sl@0: }
sl@0: 
sl@0: inline UnicodeString&
sl@0: UnicodeString::operator+= (UChar ch)
sl@0: { return doReplace(fLength, 0, &ch, 0, 1); }
sl@0: 
sl@0: inline UnicodeString&
sl@0: UnicodeString::operator+= (UChar32 ch) {
sl@0:   UChar buffer[U16_MAX_LENGTH];
sl@0:   int32_t _length = 0;
sl@0:   UBool isError = FALSE;
sl@0:   U16_APPEND(buffer, _length, U16_MAX_LENGTH, ch, isError);
sl@0:   return doReplace(fLength, 0, buffer, 0, _length);
sl@0: }
sl@0: 
sl@0: inline UnicodeString&
sl@0: UnicodeString::operator+= (const UnicodeString& srcText)
sl@0: { return doReplace(fLength, 0, srcText, 0, srcText.fLength); }
sl@0: 
sl@0: inline UnicodeString&
sl@0: UnicodeString::append(const UnicodeString& srcText,
sl@0:               int32_t srcStart,
sl@0:               int32_t srcLength)
sl@0: { return doReplace(fLength, 0, srcText, srcStart, srcLength); }
sl@0: 
sl@0: inline UnicodeString&
sl@0: UnicodeString::append(const UnicodeString& srcText)
sl@0: { return doReplace(fLength, 0, srcText, 0, srcText.fLength); }
sl@0: 
sl@0: inline UnicodeString&
sl@0: UnicodeString::append(const UChar *srcChars,
sl@0:               int32_t srcStart,
sl@0:               int32_t srcLength)
sl@0: { return doReplace(fLength, 0, srcChars, srcStart, srcLength); }
sl@0: 
sl@0: inline UnicodeString&
sl@0: UnicodeString::append(const UChar *srcChars,
sl@0:               int32_t srcLength)
sl@0: { return doReplace(fLength, 0, srcChars, 0, srcLength); }
sl@0: 
sl@0: inline UnicodeString&
sl@0: UnicodeString::append(UChar srcChar)
sl@0: { return doReplace(fLength, 0, &srcChar, 0, 1); }
sl@0: 
sl@0: inline UnicodeString&
sl@0: UnicodeString::append(UChar32 srcChar) {
sl@0:   UChar buffer[U16_MAX_LENGTH];
sl@0:   int32_t _length = 0;
sl@0:   UBool isError = FALSE;
sl@0:   U16_APPEND(buffer, _length, U16_MAX_LENGTH, srcChar, isError);
sl@0:   return doReplace(fLength, 0, buffer, 0, _length);
sl@0: }
sl@0: 
sl@0: inline UnicodeString&
sl@0: UnicodeString::insert(int32_t start,
sl@0:               const UnicodeString& srcText,
sl@0:               int32_t srcStart,
sl@0:               int32_t srcLength)
sl@0: { return doReplace(start, 0, srcText, srcStart, srcLength); }
sl@0: 
sl@0: inline UnicodeString&
sl@0: UnicodeString::insert(int32_t start,
sl@0:               const UnicodeString& srcText)
sl@0: { return doReplace(start, 0, srcText, 0, srcText.fLength); }
sl@0: 
sl@0: inline UnicodeString&
sl@0: UnicodeString::insert(int32_t start,
sl@0:               const UChar *srcChars,
sl@0:               int32_t srcStart,
sl@0:               int32_t srcLength)
sl@0: { return doReplace(start, 0, srcChars, srcStart, srcLength); }
sl@0: 
sl@0: inline UnicodeString&
sl@0: UnicodeString::insert(int32_t start,
sl@0:               const UChar *srcChars,
sl@0:               int32_t srcLength)
sl@0: { return doReplace(start, 0, srcChars, 0, srcLength); }
sl@0: 
sl@0: inline UnicodeString&
sl@0: UnicodeString::insert(int32_t start,
sl@0:               UChar srcChar)
sl@0: { return doReplace(start, 0, &srcChar, 0, 1); }
sl@0: 
sl@0: inline UnicodeString&
sl@0: UnicodeString::insert(int32_t start,
sl@0:               UChar32 srcChar)
sl@0: { return replace(start, 0, srcChar); }
sl@0: 
sl@0: 
sl@0: inline UnicodeString&
sl@0: UnicodeString::remove()
sl@0: {
sl@0:   // remove() of a bogus string makes the string empty and non-bogus
sl@0:   if(isBogus()) {
sl@0:     unBogus();
sl@0:   } else {
sl@0:     fLength = 0;
sl@0:   }
sl@0:   return *this;
sl@0: }
sl@0: 
sl@0: inline UnicodeString&
sl@0: UnicodeString::remove(int32_t start,
sl@0:              int32_t _length)
sl@0: {
sl@0:   if(start <= 0 && _length == INT32_MAX) {
sl@0:     // remove(guaranteed everything) of a bogus string makes the string empty and non-bogus
sl@0:     return remove();
sl@0:   } else {
sl@0:     return doReplace(start, _length, NULL, 0, 0);
sl@0:   }
sl@0: }
sl@0: 
sl@0: inline UnicodeString&
sl@0: UnicodeString::removeBetween(int32_t start,
sl@0:                 int32_t limit)
sl@0: { return doReplace(start, limit - start, NULL, 0, 0); }
sl@0: 
sl@0: inline UBool
sl@0: UnicodeString::truncate(int32_t targetLength)
sl@0: {
sl@0:   if(isBogus() && targetLength == 0) {
sl@0:     // truncate(0) of a bogus string makes the string empty and non-bogus
sl@0:     unBogus();
sl@0:     return FALSE;
sl@0:   } else if((uint32_t)targetLength < (uint32_t)fLength) {
sl@0:     fLength = targetLength;
sl@0:     return TRUE;
sl@0:   } else {
sl@0:     return FALSE;
sl@0:   }
sl@0: }
sl@0: 
sl@0: inline UnicodeString&
sl@0: UnicodeString::reverse()
sl@0: { return doReverse(0, fLength); }
sl@0: 
sl@0: inline UnicodeString&
sl@0: UnicodeString::reverse(int32_t start,
sl@0:                int32_t _length)
sl@0: { return doReverse(start, _length); }
sl@0: 
sl@0: U_NAMESPACE_END
sl@0: 
sl@0: #endif