michael@0: /*
michael@0: **********************************************************************
michael@0: *   Copyright (C) 1998-2013, International Business Machines
michael@0: *   Corporation and others.  All Rights Reserved.
michael@0: **********************************************************************
michael@0: *
michael@0: * File unistr.h
michael@0: *
michael@0: * Modification History:
michael@0: *
michael@0: *   Date        Name        Description
michael@0: *   09/25/98    stephen     Creation.
michael@0: *   11/11/98    stephen     Changed per 11/9 code review.
michael@0: *   04/20/99    stephen     Overhauled per 4/16 code review.
michael@0: *   11/18/99    aliu        Made to inherit from Replaceable.  Added method
michael@0: *                           handleReplaceBetween(); other methods unchanged.
michael@0: *   06/25/01    grhoten     Remove dependency on iostream.
michael@0: ******************************************************************************
michael@0: */
michael@0: 
michael@0: #ifndef UNISTR_H
michael@0: #define UNISTR_H
michael@0: 
michael@0: /**
michael@0:  * \file 
michael@0:  * \brief C++ API: Unicode String 
michael@0:  */
michael@0: 
michael@0: #include "unicode/utypes.h"
michael@0: #include "unicode/rep.h"
michael@0: #include "unicode/std_string.h"
michael@0: #include "unicode/stringpiece.h"
michael@0: #include "unicode/bytestream.h"
michael@0: #include "unicode/ucasemap.h"
michael@0: 
michael@0: struct UConverter;          // unicode/ucnv.h
michael@0: class  StringThreadTest;
michael@0: 
michael@0: #ifndef U_COMPARE_CODE_POINT_ORDER
michael@0: /* see also ustring.h and unorm.h */
michael@0: /**
michael@0:  * Option bit for u_strCaseCompare, u_strcasecmp, unorm_compare, etc:
michael@0:  * Compare strings in code point order instead of code unit order.
michael@0:  * @stable ICU 2.2
michael@0:  */
michael@0: #define U_COMPARE_CODE_POINT_ORDER  0x8000
michael@0: #endif
michael@0: 
michael@0: #ifndef USTRING_H
michael@0: /**
michael@0:  * \ingroup ustring_ustrlen
michael@0:  */
michael@0: U_STABLE int32_t U_EXPORT2
michael@0: u_strlen(const UChar *s);
michael@0: #endif
michael@0: 
michael@0: /**
michael@0:  * \def U_STRING_CASE_MAPPER_DEFINED
michael@0:  * @internal
michael@0:  */
michael@0: #ifndef U_STRING_CASE_MAPPER_DEFINED
michael@0: #define U_STRING_CASE_MAPPER_DEFINED
michael@0: 
michael@0: /**
michael@0:  * Internal string case mapping function type.
michael@0:  * @internal
michael@0:  */
michael@0: typedef int32_t U_CALLCONV
michael@0: UStringCaseMapper(const UCaseMap *csm,
michael@0:                   UChar *dest, int32_t destCapacity,
michael@0:                   const UChar *src, int32_t srcLength,
michael@0:                   UErrorCode *pErrorCode);
michael@0: 
michael@0: #endif
michael@0: 
michael@0: U_NAMESPACE_BEGIN
michael@0: 
michael@0: class BreakIterator;        // unicode/brkiter.h
michael@0: class Locale;               // unicode/locid.h
michael@0: class StringCharacterIterator;
michael@0: class UnicodeStringAppendable;  // unicode/appendable.h
michael@0: 
michael@0: /* The <iostream> include has been moved to unicode/ustream.h */
michael@0: 
michael@0: /**
michael@0:  * Constant to be used in the UnicodeString(char *, int32_t, EInvariant) constructor
michael@0:  * which constructs a Unicode string from an invariant-character char * string.
michael@0:  * About invariant characters see utypes.h.
michael@0:  * This constructor has no runtime dependency on conversion code and is
michael@0:  * therefore recommended over ones taking a charset name string
michael@0:  * (where the empty string "" indicates invariant-character conversion).
michael@0:  *
michael@0:  * @stable ICU 3.2
michael@0:  */
michael@0: #define US_INV icu::UnicodeString::kInvariant
michael@0: 
michael@0: /**
michael@0:  * Unicode String literals in C++.
michael@0:  * Dependent on the platform properties, different UnicodeString
michael@0:  * constructors should be used to create a UnicodeString object from
michael@0:  * a string literal.
michael@0:  * The macros are defined for maximum performance.
michael@0:  * They work only for strings that contain "invariant characters", i.e.,
michael@0:  * only latin letters, digits, and some punctuation.
michael@0:  * See utypes.h for details.
michael@0:  *
michael@0:  * The string parameter must be a C string literal.
michael@0:  * The length of the string, not including the terminating
michael@0:  * <code>NUL</code>, must be specified as a constant.
michael@0:  * The U_STRING_DECL macro should be invoked exactly once for one
michael@0:  * such string variable before it is used.
michael@0:  * @stable ICU 2.0
michael@0:  */
michael@0: #if defined(U_DECLARE_UTF16)
michael@0: #   define UNICODE_STRING(cs, _length) icu::UnicodeString(TRUE, (const UChar *)U_DECLARE_UTF16(cs), _length)
michael@0: #elif U_SIZEOF_WCHAR_T==U_SIZEOF_UCHAR && (U_CHARSET_FAMILY==U_ASCII_FAMILY || (U_SIZEOF_UCHAR == 2 && defined(U_WCHAR_IS_UTF16)))
michael@0: #   define UNICODE_STRING(cs, _length) icu::UnicodeString(TRUE, (const UChar *)L ## cs, _length)
michael@0: #elif U_SIZEOF_UCHAR==1 && U_CHARSET_FAMILY==U_ASCII_FAMILY
michael@0: #   define UNICODE_STRING(cs, _length) icu::UnicodeString(TRUE, (const UChar *)cs, _length)
michael@0: #else
michael@0: #   define UNICODE_STRING(cs, _length) icu::UnicodeString(cs, _length, US_INV)
michael@0: #endif
michael@0: 
michael@0: /**
michael@0:  * Unicode String literals in C++.
michael@0:  * Dependent on the platform properties, different UnicodeString
michael@0:  * constructors should be used to create a UnicodeString object from
michael@0:  * a string literal.
michael@0:  * The macros are defined for improved performance.
michael@0:  * They work only for strings that contain "invariant characters", i.e.,
michael@0:  * only latin letters, digits, and some punctuation.
michael@0:  * See utypes.h for details.
michael@0:  *
michael@0:  * The string parameter must be a C string literal.
michael@0:  * @stable ICU 2.0
michael@0:  */
michael@0: #define UNICODE_STRING_SIMPLE(cs) UNICODE_STRING(cs, -1)
michael@0: 
michael@0: /**
michael@0:  * \def UNISTR_FROM_CHAR_EXPLICIT
michael@0:  * This can be defined to be empty or "explicit".
michael@0:  * If explicit, then the UnicodeString(UChar) and UnicodeString(UChar32)
michael@0:  * constructors are marked as explicit, preventing their inadvertent use.
michael@0:  * @stable ICU 49
michael@0:  */
michael@0: #ifndef UNISTR_FROM_CHAR_EXPLICIT
michael@0: # if defined(U_COMBINED_IMPLEMENTATION) || defined(U_COMMON_IMPLEMENTATION) || defined(U_I18N_IMPLEMENTATION) || defined(U_IO_IMPLEMENTATION)
michael@0:     // Auto-"explicit" in ICU library code.
michael@0: #   define UNISTR_FROM_CHAR_EXPLICIT explicit
michael@0: # else
michael@0:     // Empty by default for source code compatibility.
michael@0: #   define UNISTR_FROM_CHAR_EXPLICIT
michael@0: # endif
michael@0: #endif
michael@0: 
michael@0: /**
michael@0:  * \def UNISTR_FROM_STRING_EXPLICIT
michael@0:  * This can be defined to be empty or "explicit".
michael@0:  * If explicit, then the UnicodeString(const char *) and UnicodeString(const UChar *)
michael@0:  * constructors are marked as explicit, preventing their inadvertent use.
michael@0:  *
michael@0:  * In particular, this helps prevent accidentally depending on ICU conversion code
michael@0:  * by passing a string literal into an API with a const UnicodeString & parameter.
michael@0:  * @stable ICU 49
michael@0:  */
michael@0: #ifndef UNISTR_FROM_STRING_EXPLICIT
michael@0: # if defined(U_COMBINED_IMPLEMENTATION) || defined(U_COMMON_IMPLEMENTATION) || defined(U_I18N_IMPLEMENTATION) || defined(U_IO_IMPLEMENTATION)
michael@0:     // Auto-"explicit" in ICU library code.
michael@0: #   define UNISTR_FROM_STRING_EXPLICIT explicit
michael@0: # else
michael@0:     // Empty by default for source code compatibility.
michael@0: #   define UNISTR_FROM_STRING_EXPLICIT
michael@0: # endif
michael@0: #endif
michael@0: 
michael@0: /**
michael@0:  * UnicodeString is a string class that stores Unicode characters directly and provides
michael@0:  * similar functionality as the Java String and StringBuffer classes.
michael@0:  * It is a concrete implementation of the abstract class Replaceable (for transliteration).
michael@0:  *
michael@0:  * The UnicodeString class is not suitable for subclassing.
michael@0:  *
michael@0:  * <p>For an overview of Unicode strings in C and C++ see the
michael@0:  * <a href="http://icu-project.org/userguide/strings.html">User Guide Strings chapter</a>.</p>
michael@0:  *
michael@0:  * <p>In ICU, a Unicode string consists of 16-bit Unicode <em>code units</em>.
michael@0:  * A Unicode character may be stored with either one code unit
michael@0:  * (the most common case) or with a matched pair of special code units
michael@0:  * ("surrogates"). The data type for code units is UChar. 
michael@0:  * For single-character handling, a Unicode character code <em>point</em> is a value
michael@0:  * in the range 0..0x10ffff. ICU uses the UChar32 type for code points.</p>
michael@0:  *
michael@0:  * <p>Indexes and offsets into and lengths of strings always count code units, not code points.
michael@0:  * This is the same as with multi-byte char* strings in traditional string handling.
michael@0:  * Operations on partial strings typically do not test for code point boundaries.
michael@0:  * If necessary, the user needs to take care of such boundaries by testing for the code unit
michael@0:  * values or by using functions like
michael@0:  * UnicodeString::getChar32Start() and UnicodeString::getChar32Limit()
michael@0:  * (or, in C, the equivalent macros U16_SET_CP_START() and U16_SET_CP_LIMIT(), see utf.h).</p>
michael@0:  *
michael@0:  * UnicodeString methods are more lenient with regard to input parameter values
michael@0:  * than other ICU APIs. In particular:
michael@0:  * - If indexes are out of bounds for a UnicodeString object
michael@0:  *   (<0 or >length()) then they are "pinned" to the nearest boundary.
michael@0:  * - If primitive string pointer values (e.g., const UChar * or char *)
michael@0:  *   for input strings are NULL, then those input string parameters are treated
michael@0:  *   as if they pointed to an empty string.
michael@0:  *   However, this is <em>not</em> the case for char * parameters for charset names
michael@0:  *   or other IDs.
michael@0:  * - Most UnicodeString methods do not take a UErrorCode parameter because
michael@0:  *   there are usually very few opportunities for failure other than a shortage
michael@0:  *   of memory, error codes in low-level C++ string methods would be inconvenient,
michael@0:  *   and the error code as the last parameter (ICU convention) would prevent
michael@0:  *   the use of default parameter values.
michael@0:  *   Instead, such methods set the UnicodeString into a "bogus" state
michael@0:  *   (see isBogus()) if an error occurs.
michael@0:  *
michael@0:  * In string comparisons, two UnicodeString objects that are both "bogus"
michael@0:  * compare equal (to be transitive and prevent endless loops in sorting),
michael@0:  * and a "bogus" string compares less than any non-"bogus" one.
michael@0:  *
michael@0:  * Const UnicodeString methods are thread-safe. Multiple threads can use
michael@0:  * const methods on the same UnicodeString object simultaneously,
michael@0:  * but non-const methods must not be called concurrently (in multiple threads)
michael@0:  * with any other (const or non-const) methods.
michael@0:  *
michael@0:  * Similarly, const UnicodeString & parameters are thread-safe.
michael@0:  * One object may be passed in as such a parameter concurrently in multiple threads.
michael@0:  * This includes the const UnicodeString & parameters for
michael@0:  * copy construction, assignment, and cloning.
michael@0:  *
michael@0:  * <p>UnicodeString uses several storage methods.
michael@0:  * String contents can be stored inside the UnicodeString object itself,
michael@0:  * in an allocated and shared buffer, or in an outside buffer that is "aliased".
michael@0:  * Most of this is done transparently, but careful aliasing in particular provides
michael@0:  * significant performance improvements.
michael@0:  * Also, the internal buffer is accessible via special functions.
michael@0:  * For details see the
michael@0:  * <a href="http://icu-project.org/userguide/strings.html">User Guide Strings chapter</a>.</p>
michael@0:  *
michael@0:  * @see utf.h
michael@0:  * @see CharacterIterator
michael@0:  * @stable ICU 2.0
michael@0:  */
michael@0: class U_COMMON_API UnicodeString : public Replaceable
michael@0: {
michael@0: public:
michael@0: 
michael@0:   /**
michael@0:    * Constant to be used in the UnicodeString(char *, int32_t, EInvariant) constructor
michael@0:    * which constructs a Unicode string from an invariant-character char * string.
michael@0:    * Use the macro US_INV instead of the full qualification for this value.
michael@0:    *
michael@0:    * @see US_INV
michael@0:    * @stable ICU 3.2
michael@0:    */
michael@0:   enum EInvariant {
michael@0:     /**
michael@0:      * @see EInvariant
michael@0:      * @stable ICU 3.2
michael@0:      */
michael@0:     kInvariant
michael@0:   };
michael@0: 
michael@0:   //========================================
michael@0:   // Read-only operations
michael@0:   //========================================
michael@0: 
michael@0:   /* Comparison - bitwise only - for international comparison use collation */
michael@0: 
michael@0:   /**
michael@0:    * Equality operator. Performs only bitwise comparison.
michael@0:    * @param text The UnicodeString to compare to this one.
michael@0:    * @return TRUE if <TT>text</TT> contains the same characters as this one,
michael@0:    * FALSE otherwise.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UBool operator== (const UnicodeString& text) const;
michael@0: 
michael@0:   /**
michael@0:    * Inequality operator. Performs only bitwise comparison.
michael@0:    * @param text The UnicodeString to compare to this one.
michael@0:    * @return FALSE if <TT>text</TT> contains the same characters as this one,
michael@0:    * TRUE otherwise.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UBool operator!= (const UnicodeString& text) const;
michael@0: 
michael@0:   /**
michael@0:    * Greater than operator. Performs only bitwise comparison.
michael@0:    * @param text The UnicodeString to compare to this one.
michael@0:    * @return TRUE if the characters in this are bitwise
michael@0:    * greater than the characters in <code>text</code>, FALSE otherwise
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UBool operator> (const UnicodeString& text) const;
michael@0: 
michael@0:   /**
michael@0:    * Less than operator. Performs only bitwise comparison.
michael@0:    * @param text The UnicodeString to compare to this one.
michael@0:    * @return TRUE if the characters in this are bitwise
michael@0:    * less than the characters in <code>text</code>, FALSE otherwise
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UBool operator< (const UnicodeString& text) const;
michael@0: 
michael@0:   /**
michael@0:    * Greater than or equal operator. Performs only bitwise comparison.
michael@0:    * @param text The UnicodeString to compare to this one.
michael@0:    * @return TRUE if the characters in this are bitwise
michael@0:    * greater than or equal to the characters in <code>text</code>, FALSE otherwise
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UBool operator>= (const UnicodeString& text) const;
michael@0: 
michael@0:   /**
michael@0:    * Less than or equal operator. Performs only bitwise comparison.
michael@0:    * @param text The UnicodeString to compare to this one.
michael@0:    * @return TRUE if the characters in this are bitwise
michael@0:    * less than or equal to the characters in <code>text</code>, FALSE otherwise
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UBool operator<= (const UnicodeString& text) const;
michael@0: 
michael@0:   /**
michael@0:    * Compare the characters bitwise in this UnicodeString to
michael@0:    * the characters in <code>text</code>.
michael@0:    * @param text The UnicodeString to compare to this one.
michael@0:    * @return The result of bitwise character comparison: 0 if this
michael@0:    * contains the same characters as <code>text</code>, -1 if the characters in
michael@0:    * this are bitwise less than the characters in <code>text</code>, +1 if the
michael@0:    * characters in this are bitwise greater than the characters
michael@0:    * in <code>text</code>.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int8_t compare(const UnicodeString& text) const;
michael@0: 
michael@0:   /**
michael@0:    * Compare the characters bitwise in the range
michael@0:    * [<TT>start</TT>, <TT>start + length</TT>) with the characters
michael@0:    * in the <b>entire string</b> <TT>text</TT>.
michael@0:    * (The parameters "start" and "length" are not applied to the other text "text".)
michael@0:    * @param start the offset at which the compare operation begins
michael@0:    * @param length the number of characters of text to compare.
michael@0:    * @param text the other text to be compared against this string.
michael@0:    * @return The result of bitwise character comparison: 0 if this
michael@0:    * contains the same characters as <code>text</code>, -1 if the characters in
michael@0:    * this are bitwise less than the characters in <code>text</code>, +1 if the
michael@0:    * characters in this are bitwise greater than the characters
michael@0:    * in <code>text</code>.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int8_t compare(int32_t start,
michael@0:          int32_t length,
michael@0:          const UnicodeString& text) const;
michael@0: 
michael@0:   /**
michael@0:    * Compare the characters bitwise in the range
michael@0:    * [<TT>start</TT>, <TT>start + length</TT>) with the characters
michael@0:    * in <TT>srcText</TT> in the range
michael@0:    * [<TT>srcStart</TT>, <TT>srcStart + srcLength</TT>).
michael@0:    * @param start the offset at which the compare operation begins
michael@0:    * @param length the number of characters in this to compare.
michael@0:    * @param srcText the text to be compared
michael@0:    * @param srcStart the offset into <TT>srcText</TT> to start comparison
michael@0:    * @param srcLength the number of characters in <TT>src</TT> to compare
michael@0:    * @return The result of bitwise character comparison: 0 if this
michael@0:    * contains the same characters as <code>srcText</code>, -1 if the characters in
michael@0:    * this are bitwise less than the characters in <code>srcText</code>, +1 if the
michael@0:    * characters in this are bitwise greater than the characters
michael@0:    * in <code>srcText</code>.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:    inline int8_t compare(int32_t start,
michael@0:          int32_t length,
michael@0:          const UnicodeString& srcText,
michael@0:          int32_t srcStart,
michael@0:          int32_t srcLength) const;
michael@0: 
michael@0:   /**
michael@0:    * Compare the characters bitwise in this UnicodeString with the first
michael@0:    * <TT>srcLength</TT> characters in <TT>srcChars</TT>.
michael@0:    * @param srcChars The characters to compare to this UnicodeString.
michael@0:    * @param srcLength the number of characters in <TT>srcChars</TT> to compare
michael@0:    * @return The result of bitwise character comparison: 0 if this
michael@0:    * contains the same characters as <code>srcChars</code>, -1 if the characters in
michael@0:    * this are bitwise less than the characters in <code>srcChars</code>, +1 if the
michael@0:    * characters in this are bitwise greater than the characters
michael@0:    * in <code>srcChars</code>.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int8_t compare(const UChar *srcChars,
michael@0:          int32_t srcLength) const;
michael@0: 
michael@0:   /**
michael@0:    * Compare the characters bitwise in the range
michael@0:    * [<TT>start</TT>, <TT>start + length</TT>) with the first
michael@0:    * <TT>length</TT> characters in <TT>srcChars</TT>
michael@0:    * @param start the offset at which the compare operation begins
michael@0:    * @param length the number of characters to compare.
michael@0:    * @param srcChars the characters to be compared
michael@0:    * @return The result of bitwise character comparison: 0 if this
michael@0:    * contains the same characters as <code>srcChars</code>, -1 if the characters in
michael@0:    * this are bitwise less than the characters in <code>srcChars</code>, +1 if the
michael@0:    * characters in this are bitwise greater than the characters
michael@0:    * in <code>srcChars</code>.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int8_t compare(int32_t start,
michael@0:          int32_t length,
michael@0:          const UChar *srcChars) const;
michael@0: 
michael@0:   /**
michael@0:    * Compare the characters bitwise in the range
michael@0:    * [<TT>start</TT>, <TT>start + length</TT>) with the characters
michael@0:    * in <TT>srcChars</TT> in the range
michael@0:    * [<TT>srcStart</TT>, <TT>srcStart + srcLength</TT>).
michael@0:    * @param start the offset at which the compare operation begins
michael@0:    * @param length the number of characters in this to compare
michael@0:    * @param srcChars the characters to be compared
michael@0:    * @param srcStart the offset into <TT>srcChars</TT> to start comparison
michael@0:    * @param srcLength the number of characters in <TT>srcChars</TT> to compare
michael@0:    * @return The result of bitwise character comparison: 0 if this
michael@0:    * contains the same characters as <code>srcChars</code>, -1 if the characters in
michael@0:    * this are bitwise less than the characters in <code>srcChars</code>, +1 if the
michael@0:    * characters in this are bitwise greater than the characters
michael@0:    * in <code>srcChars</code>.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int8_t compare(int32_t start,
michael@0:          int32_t length,
michael@0:          const UChar *srcChars,
michael@0:          int32_t srcStart,
michael@0:          int32_t srcLength) const;
michael@0: 
michael@0:   /**
michael@0:    * Compare the characters bitwise in the range
michael@0:    * [<TT>start</TT>, <TT>limit</TT>) with the characters
michael@0:    * in <TT>srcText</TT> in the range
michael@0:    * [<TT>srcStart</TT>, <TT>srcLimit</TT>).
michael@0:    * @param start the offset at which the compare operation begins
michael@0:    * @param limit the offset immediately following the compare operation
michael@0:    * @param srcText the text to be compared
michael@0:    * @param srcStart the offset into <TT>srcText</TT> to start comparison
michael@0:    * @param srcLimit the offset into <TT>srcText</TT> to limit comparison
michael@0:    * @return The result of bitwise character comparison: 0 if this
michael@0:    * contains the same characters as <code>srcText</code>, -1 if the characters in
michael@0:    * this are bitwise less than the characters in <code>srcText</code>, +1 if the
michael@0:    * characters in this are bitwise greater than the characters
michael@0:    * in <code>srcText</code>.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int8_t compareBetween(int32_t start,
michael@0:             int32_t limit,
michael@0:             const UnicodeString& srcText,
michael@0:             int32_t srcStart,
michael@0:             int32_t srcLimit) const;
michael@0: 
michael@0:   /**
michael@0:    * Compare two Unicode strings in code point order.
michael@0:    * The result may be different from the results of compare(), operator<, etc.
michael@0:    * if supplementary characters are present:
michael@0:    *
michael@0:    * In UTF-16, supplementary characters (with code points U+10000 and above) are
michael@0:    * stored with pairs of surrogate code units. These have values from 0xd800 to 0xdfff,
michael@0:    * which means that they compare as less than some other BMP characters like U+feff.
michael@0:    * This function compares Unicode strings in code point order.
michael@0:    * If either of the UTF-16 strings is malformed (i.e., it contains unpaired surrogates), then the result is not defined.
michael@0:    *
michael@0:    * @param text Another string to compare this one to.
michael@0:    * @return a negative/zero/positive integer corresponding to whether
michael@0:    * this string is less than/equal to/greater than the second one
michael@0:    * in code point order
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int8_t compareCodePointOrder(const UnicodeString& text) const;
michael@0: 
michael@0:   /**
michael@0:    * Compare two Unicode strings in code point order.
michael@0:    * The result may be different from the results of compare(), operator<, etc.
michael@0:    * if supplementary characters are present:
michael@0:    *
michael@0:    * In UTF-16, supplementary characters (with code points U+10000 and above) are
michael@0:    * stored with pairs of surrogate code units. These have values from 0xd800 to 0xdfff,
michael@0:    * which means that they compare as less than some other BMP characters like U+feff.
michael@0:    * This function compares Unicode strings in code point order.
michael@0:    * If either of the UTF-16 strings is malformed (i.e., it contains unpaired surrogates), then the result is not defined.
michael@0:    *
michael@0:    * @param start The start offset in this string at which the compare operation begins.
michael@0:    * @param length The number of code units from this string to compare.
michael@0:    * @param srcText Another string to compare this one to.
michael@0:    * @return a negative/zero/positive integer corresponding to whether
michael@0:    * this string is less than/equal to/greater than the second one
michael@0:    * in code point order
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int8_t compareCodePointOrder(int32_t start,
michael@0:                                       int32_t length,
michael@0:                                       const UnicodeString& srcText) const;
michael@0: 
michael@0:   /**
michael@0:    * Compare two Unicode strings in code point order.
michael@0:    * The result may be different from the results of compare(), operator<, etc.
michael@0:    * if supplementary characters are present:
michael@0:    *
michael@0:    * In UTF-16, supplementary characters (with code points U+10000 and above) are
michael@0:    * stored with pairs of surrogate code units. These have values from 0xd800 to 0xdfff,
michael@0:    * which means that they compare as less than some other BMP characters like U+feff.
michael@0:    * This function compares Unicode strings in code point order.
michael@0:    * If either of the UTF-16 strings is malformed (i.e., it contains unpaired surrogates), then the result is not defined.
michael@0:    *
michael@0:    * @param start The start offset in this string at which the compare operation begins.
michael@0:    * @param length The number of code units from this string to compare.
michael@0:    * @param srcText Another string to compare this one to.
michael@0:    * @param srcStart The start offset in that string at which the compare operation begins.
michael@0:    * @param srcLength The number of code units from that string to compare.
michael@0:    * @return a negative/zero/positive integer corresponding to whether
michael@0:    * this string is less than/equal to/greater than the second one
michael@0:    * in code point order
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:    inline int8_t compareCodePointOrder(int32_t start,
michael@0:                                        int32_t length,
michael@0:                                        const UnicodeString& srcText,
michael@0:                                        int32_t srcStart,
michael@0:                                        int32_t srcLength) const;
michael@0: 
michael@0:   /**
michael@0:    * Compare two Unicode strings in code point order.
michael@0:    * The result may be different from the results of compare(), operator<, etc.
michael@0:    * if supplementary characters are present:
michael@0:    *
michael@0:    * In UTF-16, supplementary characters (with code points U+10000 and above) are
michael@0:    * stored with pairs of surrogate code units. These have values from 0xd800 to 0xdfff,
michael@0:    * which means that they compare as less than some other BMP characters like U+feff.
michael@0:    * This function compares Unicode strings in code point order.
michael@0:    * If either of the UTF-16 strings is malformed (i.e., it contains unpaired surrogates), then the result is not defined.
michael@0:    *
michael@0:    * @param srcChars A pointer to another string to compare this one to.
michael@0:    * @param srcLength The number of code units from that string to compare.
michael@0:    * @return a negative/zero/positive integer corresponding to whether
michael@0:    * this string is less than/equal to/greater than the second one
michael@0:    * in code point order
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int8_t compareCodePointOrder(const UChar *srcChars,
michael@0:                                       int32_t srcLength) const;
michael@0: 
michael@0:   /**
michael@0:    * Compare two Unicode strings in code point order.
michael@0:    * The result may be different from the results of compare(), operator<, etc.
michael@0:    * if supplementary characters are present:
michael@0:    *
michael@0:    * In UTF-16, supplementary characters (with code points U+10000 and above) are
michael@0:    * stored with pairs of surrogate code units. These have values from 0xd800 to 0xdfff,
michael@0:    * which means that they compare as less than some other BMP characters like U+feff.
michael@0:    * This function compares Unicode strings in code point order.
michael@0:    * If either of the UTF-16 strings is malformed (i.e., it contains unpaired surrogates), then the result is not defined.
michael@0:    *
michael@0:    * @param start The start offset in this string at which the compare operation begins.
michael@0:    * @param length The number of code units from this string to compare.
michael@0:    * @param srcChars A pointer to another string to compare this one to.
michael@0:    * @return a negative/zero/positive integer corresponding to whether
michael@0:    * this string is less than/equal to/greater than the second one
michael@0:    * in code point order
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int8_t compareCodePointOrder(int32_t start,
michael@0:                                       int32_t length,
michael@0:                                       const UChar *srcChars) const;
michael@0: 
michael@0:   /**
michael@0:    * Compare two Unicode strings in code point order.
michael@0:    * The result may be different from the results of compare(), operator<, etc.
michael@0:    * if supplementary characters are present:
michael@0:    *
michael@0:    * In UTF-16, supplementary characters (with code points U+10000 and above) are
michael@0:    * stored with pairs of surrogate code units. These have values from 0xd800 to 0xdfff,
michael@0:    * which means that they compare as less than some other BMP characters like U+feff.
michael@0:    * This function compares Unicode strings in code point order.
michael@0:    * If either of the UTF-16 strings is malformed (i.e., it contains unpaired surrogates), then the result is not defined.
michael@0:    *
michael@0:    * @param start The start offset in this string at which the compare operation begins.
michael@0:    * @param length The number of code units from this string to compare.
michael@0:    * @param srcChars A pointer to another string to compare this one to.
michael@0:    * @param srcStart The start offset in that string at which the compare operation begins.
michael@0:    * @param srcLength The number of code units from that string to compare.
michael@0:    * @return a negative/zero/positive integer corresponding to whether
michael@0:    * this string is less than/equal to/greater than the second one
michael@0:    * in code point order
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int8_t compareCodePointOrder(int32_t start,
michael@0:                                       int32_t length,
michael@0:                                       const UChar *srcChars,
michael@0:                                       int32_t srcStart,
michael@0:                                       int32_t srcLength) const;
michael@0: 
michael@0:   /**
michael@0:    * Compare two Unicode strings in code point order.
michael@0:    * The result may be different from the results of compare(), operator<, etc.
michael@0:    * if supplementary characters are present:
michael@0:    *
michael@0:    * In UTF-16, supplementary characters (with code points U+10000 and above) are
michael@0:    * stored with pairs of surrogate code units. These have values from 0xd800 to 0xdfff,
michael@0:    * which means that they compare as less than some other BMP characters like U+feff.
michael@0:    * This function compares Unicode strings in code point order.
michael@0:    * If either of the UTF-16 strings is malformed (i.e., it contains unpaired surrogates), then the result is not defined.
michael@0:    *
michael@0:    * @param start The start offset in this string at which the compare operation begins.
michael@0:    * @param limit The offset after the last code unit from this string to compare.
michael@0:    * @param srcText Another string to compare this one to.
michael@0:    * @param srcStart The start offset in that string at which the compare operation begins.
michael@0:    * @param srcLimit The offset after the last code unit from that string to compare.
michael@0:    * @return a negative/zero/positive integer corresponding to whether
michael@0:    * this string is less than/equal to/greater than the second one
michael@0:    * in code point order
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int8_t compareCodePointOrderBetween(int32_t start,
michael@0:                                              int32_t limit,
michael@0:                                              const UnicodeString& srcText,
michael@0:                                              int32_t srcStart,
michael@0:                                              int32_t srcLimit) const;
michael@0: 
michael@0:   /**
michael@0:    * Compare two strings case-insensitively using full case folding.
michael@0:    * This is equivalent to this->foldCase(options).compare(text.foldCase(options)).
michael@0:    *
michael@0:    * @param text Another string to compare this one to.
michael@0:    * @param options A bit set of options:
michael@0:    *   - U_FOLD_CASE_DEFAULT or 0 is used for default options:
michael@0:    *     Comparison in code unit order with default case folding.
michael@0:    *
michael@0:    *   - U_COMPARE_CODE_POINT_ORDER
michael@0:    *     Set to choose code point order instead of code unit order
michael@0:    *     (see u_strCompare for details).
michael@0:    *
michael@0:    *   - U_FOLD_CASE_EXCLUDE_SPECIAL_I
michael@0:    *
michael@0:    * @return A negative, zero, or positive integer indicating the comparison result.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int8_t caseCompare(const UnicodeString& text, uint32_t options) const;
michael@0: 
michael@0:   /**
michael@0:    * Compare two strings case-insensitively using full case folding.
michael@0:    * This is equivalent to this->foldCase(options).compare(srcText.foldCase(options)).
michael@0:    *
michael@0:    * @param start The start offset in this string at which the compare operation begins.
michael@0:    * @param length The number of code units from this string to compare.
michael@0:    * @param srcText Another string to compare this one to.
michael@0:    * @param options A bit set of options:
michael@0:    *   - U_FOLD_CASE_DEFAULT or 0 is used for default options:
michael@0:    *     Comparison in code unit order with default case folding.
michael@0:    *
michael@0:    *   - U_COMPARE_CODE_POINT_ORDER
michael@0:    *     Set to choose code point order instead of code unit order
michael@0:    *     (see u_strCompare for details).
michael@0:    *
michael@0:    *   - U_FOLD_CASE_EXCLUDE_SPECIAL_I
michael@0:    *
michael@0:    * @return A negative, zero, or positive integer indicating the comparison result.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int8_t caseCompare(int32_t start,
michael@0:          int32_t length,
michael@0:          const UnicodeString& srcText,
michael@0:          uint32_t options) const;
michael@0: 
michael@0:   /**
michael@0:    * Compare two strings case-insensitively using full case folding.
michael@0:    * This is equivalent to this->foldCase(options).compare(srcText.foldCase(options)).
michael@0:    *
michael@0:    * @param start The start offset in this string at which the compare operation begins.
michael@0:    * @param length The number of code units from this string to compare.
michael@0:    * @param srcText Another string to compare this one to.
michael@0:    * @param srcStart The start offset in that string at which the compare operation begins.
michael@0:    * @param srcLength The number of code units from that string to compare.
michael@0:    * @param options A bit set of options:
michael@0:    *   - U_FOLD_CASE_DEFAULT or 0 is used for default options:
michael@0:    *     Comparison in code unit order with default case folding.
michael@0:    *
michael@0:    *   - U_COMPARE_CODE_POINT_ORDER
michael@0:    *     Set to choose code point order instead of code unit order
michael@0:    *     (see u_strCompare for details).
michael@0:    *
michael@0:    *   - U_FOLD_CASE_EXCLUDE_SPECIAL_I
michael@0:    *
michael@0:    * @return A negative, zero, or positive integer indicating the comparison result.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int8_t caseCompare(int32_t start,
michael@0:          int32_t length,
michael@0:          const UnicodeString& srcText,
michael@0:          int32_t srcStart,
michael@0:          int32_t srcLength,
michael@0:          uint32_t options) const;
michael@0: 
michael@0:   /**
michael@0:    * Compare two strings case-insensitively using full case folding.
michael@0:    * This is equivalent to this->foldCase(options).compare(srcChars.foldCase(options)).
michael@0:    *
michael@0:    * @param srcChars A pointer to another string to compare this one to.
michael@0:    * @param srcLength The number of code units from that string to compare.
michael@0:    * @param options A bit set of options:
michael@0:    *   - U_FOLD_CASE_DEFAULT or 0 is used for default options:
michael@0:    *     Comparison in code unit order with default case folding.
michael@0:    *
michael@0:    *   - U_COMPARE_CODE_POINT_ORDER
michael@0:    *     Set to choose code point order instead of code unit order
michael@0:    *     (see u_strCompare for details).
michael@0:    *
michael@0:    *   - U_FOLD_CASE_EXCLUDE_SPECIAL_I
michael@0:    *
michael@0:    * @return A negative, zero, or positive integer indicating the comparison result.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int8_t caseCompare(const UChar *srcChars,
michael@0:          int32_t srcLength,
michael@0:          uint32_t options) const;
michael@0: 
michael@0:   /**
michael@0:    * Compare two strings case-insensitively using full case folding.
michael@0:    * This is equivalent to this->foldCase(options).compare(srcChars.foldCase(options)).
michael@0:    *
michael@0:    * @param start The start offset in this string at which the compare operation begins.
michael@0:    * @param length The number of code units from this string to compare.
michael@0:    * @param srcChars A pointer to another string to compare this one to.
michael@0:    * @param options A bit set of options:
michael@0:    *   - U_FOLD_CASE_DEFAULT or 0 is used for default options:
michael@0:    *     Comparison in code unit order with default case folding.
michael@0:    *
michael@0:    *   - U_COMPARE_CODE_POINT_ORDER
michael@0:    *     Set to choose code point order instead of code unit order
michael@0:    *     (see u_strCompare for details).
michael@0:    *
michael@0:    *   - U_FOLD_CASE_EXCLUDE_SPECIAL_I
michael@0:    *
michael@0:    * @return A negative, zero, or positive integer indicating the comparison result.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int8_t caseCompare(int32_t start,
michael@0:          int32_t length,
michael@0:          const UChar *srcChars,
michael@0:          uint32_t options) const;
michael@0: 
michael@0:   /**
michael@0:    * Compare two strings case-insensitively using full case folding.
michael@0:    * This is equivalent to this->foldCase(options).compare(srcChars.foldCase(options)).
michael@0:    *
michael@0:    * @param start The start offset in this string at which the compare operation begins.
michael@0:    * @param length The number of code units from this string to compare.
michael@0:    * @param srcChars A pointer to another string to compare this one to.
michael@0:    * @param srcStart The start offset in that string at which the compare operation begins.
michael@0:    * @param srcLength The number of code units from that string to compare.
michael@0:    * @param options A bit set of options:
michael@0:    *   - U_FOLD_CASE_DEFAULT or 0 is used for default options:
michael@0:    *     Comparison in code unit order with default case folding.
michael@0:    *
michael@0:    *   - U_COMPARE_CODE_POINT_ORDER
michael@0:    *     Set to choose code point order instead of code unit order
michael@0:    *     (see u_strCompare for details).
michael@0:    *
michael@0:    *   - U_FOLD_CASE_EXCLUDE_SPECIAL_I
michael@0:    *
michael@0:    * @return A negative, zero, or positive integer indicating the comparison result.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int8_t caseCompare(int32_t start,
michael@0:          int32_t length,
michael@0:          const UChar *srcChars,
michael@0:          int32_t srcStart,
michael@0:          int32_t srcLength,
michael@0:          uint32_t options) const;
michael@0: 
michael@0:   /**
michael@0:    * Compare two strings case-insensitively using full case folding.
michael@0:    * This is equivalent to this->foldCase(options).compareBetween(text.foldCase(options)).
michael@0:    *
michael@0:    * @param start The start offset in this string at which the compare operation begins.
michael@0:    * @param limit The offset after the last code unit from this string to compare.
michael@0:    * @param srcText Another string to compare this one to.
michael@0:    * @param srcStart The start offset in that string at which the compare operation begins.
michael@0:    * @param srcLimit The offset after the last code unit from that string to compare.
michael@0:    * @param options A bit set of options:
michael@0:    *   - U_FOLD_CASE_DEFAULT or 0 is used for default options:
michael@0:    *     Comparison in code unit order with default case folding.
michael@0:    *
michael@0:    *   - U_COMPARE_CODE_POINT_ORDER
michael@0:    *     Set to choose code point order instead of code unit order
michael@0:    *     (see u_strCompare for details).
michael@0:    *
michael@0:    *   - U_FOLD_CASE_EXCLUDE_SPECIAL_I
michael@0:    *
michael@0:    * @return A negative, zero, or positive integer indicating the comparison result.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int8_t caseCompareBetween(int32_t start,
michael@0:             int32_t limit,
michael@0:             const UnicodeString& srcText,
michael@0:             int32_t srcStart,
michael@0:             int32_t srcLimit,
michael@0:             uint32_t options) const;
michael@0: 
michael@0:   /**
michael@0:    * Determine if this starts with the characters in <TT>text</TT>
michael@0:    * @param text The text to match.
michael@0:    * @return TRUE if this starts with the characters in <TT>text</TT>,
michael@0:    * FALSE otherwise
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UBool startsWith(const UnicodeString& text) const;
michael@0: 
michael@0:   /**
michael@0:    * Determine if this starts with the characters in <TT>srcText</TT>
michael@0:    * in the range [<TT>srcStart</TT>, <TT>srcStart + srcLength</TT>).
michael@0:    * @param srcText The text to match.
michael@0:    * @param srcStart the offset into <TT>srcText</TT> to start matching
michael@0:    * @param srcLength the number of characters in <TT>srcText</TT> to match
michael@0:    * @return TRUE if this starts with the characters in <TT>text</TT>,
michael@0:    * FALSE otherwise
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UBool startsWith(const UnicodeString& srcText,
michael@0:             int32_t srcStart,
michael@0:             int32_t srcLength) const;
michael@0: 
michael@0:   /**
michael@0:    * Determine if this starts with the characters in <TT>srcChars</TT>
michael@0:    * @param srcChars The characters to match.
michael@0:    * @param srcLength the number of characters in <TT>srcChars</TT>
michael@0:    * @return TRUE if this starts with the characters in <TT>srcChars</TT>,
michael@0:    * FALSE otherwise
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UBool startsWith(const UChar *srcChars,
michael@0:             int32_t srcLength) const;
michael@0: 
michael@0:   /**
michael@0:    * Determine if this ends with the characters in <TT>srcChars</TT>
michael@0:    * in the range  [<TT>srcStart</TT>, <TT>srcStart + srcLength</TT>).
michael@0:    * @param srcChars The characters to match.
michael@0:    * @param srcStart the offset into <TT>srcText</TT> to start matching
michael@0:    * @param srcLength the number of characters in <TT>srcChars</TT> to match
michael@0:    * @return TRUE if this ends with the characters in <TT>srcChars</TT>, FALSE otherwise
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UBool startsWith(const UChar *srcChars,
michael@0:             int32_t srcStart,
michael@0:             int32_t srcLength) const;
michael@0: 
michael@0:   /**
michael@0:    * Determine if this ends with the characters in <TT>text</TT>
michael@0:    * @param text The text to match.
michael@0:    * @return TRUE if this ends with the characters in <TT>text</TT>,
michael@0:    * FALSE otherwise
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UBool endsWith(const UnicodeString& text) const;
michael@0: 
michael@0:   /**
michael@0:    * Determine if this ends with the characters in <TT>srcText</TT>
michael@0:    * in the range [<TT>srcStart</TT>, <TT>srcStart + srcLength</TT>).
michael@0:    * @param srcText The text to match.
michael@0:    * @param srcStart the offset into <TT>srcText</TT> to start matching
michael@0:    * @param srcLength the number of characters in <TT>srcText</TT> to match
michael@0:    * @return TRUE if this ends with the characters in <TT>text</TT>,
michael@0:    * FALSE otherwise
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UBool endsWith(const UnicodeString& srcText,
michael@0:           int32_t srcStart,
michael@0:           int32_t srcLength) const;
michael@0: 
michael@0:   /**
michael@0:    * Determine if this ends with the characters in <TT>srcChars</TT>
michael@0:    * @param srcChars The characters to match.
michael@0:    * @param srcLength the number of characters in <TT>srcChars</TT>
michael@0:    * @return TRUE if this ends with the characters in <TT>srcChars</TT>,
michael@0:    * FALSE otherwise
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UBool endsWith(const UChar *srcChars,
michael@0:           int32_t srcLength) const;
michael@0: 
michael@0:   /**
michael@0:    * Determine if this ends with the characters in <TT>srcChars</TT>
michael@0:    * in the range  [<TT>srcStart</TT>, <TT>srcStart + srcLength</TT>).
michael@0:    * @param srcChars The characters to match.
michael@0:    * @param srcStart the offset into <TT>srcText</TT> to start matching
michael@0:    * @param srcLength the number of characters in <TT>srcChars</TT> to match
michael@0:    * @return TRUE if this ends with the characters in <TT>srcChars</TT>,
michael@0:    * FALSE otherwise
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UBool endsWith(const UChar *srcChars,
michael@0:           int32_t srcStart,
michael@0:           int32_t srcLength) const;
michael@0: 
michael@0: 
michael@0:   /* Searching - bitwise only */
michael@0: 
michael@0:   /**
michael@0:    * Locate in this the first occurrence of the characters in <TT>text</TT>,
michael@0:    * using bitwise comparison.
michael@0:    * @param text The text to search for.
michael@0:    * @return The offset into this of the start of <TT>text</TT>,
michael@0:    * or -1 if not found.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int32_t indexOf(const UnicodeString& text) const;
michael@0: 
michael@0:   /**
michael@0:    * Locate in this the first occurrence of the characters in <TT>text</TT>
michael@0:    * starting at offset <TT>start</TT>, using bitwise comparison.
michael@0:    * @param text The text to search for.
michael@0:    * @param start The offset at which searching will start.
michael@0:    * @return The offset into this of the start of <TT>text</TT>,
michael@0:    * or -1 if not found.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int32_t indexOf(const UnicodeString& text,
michael@0:               int32_t start) const;
michael@0: 
michael@0:   /**
michael@0:    * Locate in this the first occurrence in the range
michael@0:    * [<TT>start</TT>, <TT>start + length</TT>) of the characters
michael@0:    * in <TT>text</TT>, using bitwise comparison.
michael@0:    * @param text The text to search for.
michael@0:    * @param start The offset at which searching will start.
michael@0:    * @param length The number of characters to search
michael@0:    * @return The offset into this of the start of <TT>text</TT>,
michael@0:    * or -1 if not found.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int32_t indexOf(const UnicodeString& text,
michael@0:               int32_t start,
michael@0:               int32_t length) const;
michael@0: 
michael@0:   /**
michael@0:    * Locate in this the first occurrence in the range
michael@0:    * [<TT>start</TT>, <TT>start + length</TT>) of the characters
michael@0:    *  in <TT>srcText</TT> in the range
michael@0:    * [<TT>srcStart</TT>, <TT>srcStart + srcLength</TT>),
michael@0:    * using bitwise comparison.
michael@0:    * @param srcText The text to search for.
michael@0:    * @param srcStart the offset into <TT>srcText</TT> at which
michael@0:    * to start matching
michael@0:    * @param srcLength the number of characters in <TT>srcText</TT> to match
michael@0:    * @param start the offset into this at which to start matching
michael@0:    * @param length the number of characters in this to search
michael@0:    * @return The offset into this of the start of <TT>text</TT>,
michael@0:    * or -1 if not found.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int32_t indexOf(const UnicodeString& srcText,
michael@0:               int32_t srcStart,
michael@0:               int32_t srcLength,
michael@0:               int32_t start,
michael@0:               int32_t length) const;
michael@0: 
michael@0:   /**
michael@0:    * Locate in this the first occurrence of the characters in
michael@0:    * <TT>srcChars</TT>
michael@0:    * starting at offset <TT>start</TT>, using bitwise comparison.
michael@0:    * @param srcChars The text to search for.
michael@0:    * @param srcLength the number of characters in <TT>srcChars</TT> to match
michael@0:    * @param start the offset into this at which to start matching
michael@0:    * @return The offset into this of the start of <TT>text</TT>,
michael@0:    * or -1 if not found.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int32_t indexOf(const UChar *srcChars,
michael@0:               int32_t srcLength,
michael@0:               int32_t start) const;
michael@0: 
michael@0:   /**
michael@0:    * Locate in this the first occurrence in the range
michael@0:    * [<TT>start</TT>, <TT>start + length</TT>) of the characters
michael@0:    * in <TT>srcChars</TT>, using bitwise comparison.
michael@0:    * @param srcChars The text to search for.
michael@0:    * @param srcLength the number of characters in <TT>srcChars</TT>
michael@0:    * @param start The offset at which searching will start.
michael@0:    * @param length The number of characters to search
michael@0:    * @return The offset into this of the start of <TT>srcChars</TT>,
michael@0:    * or -1 if not found.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int32_t indexOf(const UChar *srcChars,
michael@0:               int32_t srcLength,
michael@0:               int32_t start,
michael@0:               int32_t length) const;
michael@0: 
michael@0:   /**
michael@0:    * Locate in this the first occurrence in the range
michael@0:    * [<TT>start</TT>, <TT>start + length</TT>) of the characters
michael@0:    * in <TT>srcChars</TT> in the range
michael@0:    * [<TT>srcStart</TT>, <TT>srcStart + srcLength</TT>),
michael@0:    * using bitwise comparison.
michael@0:    * @param srcChars The text to search for.
michael@0:    * @param srcStart the offset into <TT>srcChars</TT> at which
michael@0:    * to start matching
michael@0:    * @param srcLength the number of characters in <TT>srcChars</TT> to match
michael@0:    * @param start the offset into this at which to start matching
michael@0:    * @param length the number of characters in this to search
michael@0:    * @return The offset into this of the start of <TT>text</TT>,
michael@0:    * or -1 if not found.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   int32_t indexOf(const UChar *srcChars,
michael@0:               int32_t srcStart,
michael@0:               int32_t srcLength,
michael@0:               int32_t start,
michael@0:               int32_t length) const;
michael@0: 
michael@0:   /**
michael@0:    * Locate in this the first occurrence of the BMP code point <code>c</code>,
michael@0:    * using bitwise comparison.
michael@0:    * @param c The code unit to search for.
michael@0:    * @return The offset into this of <TT>c</TT>, or -1 if not found.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int32_t indexOf(UChar c) const;
michael@0: 
michael@0:   /**
michael@0:    * Locate in this the first occurrence of the code point <TT>c</TT>,
michael@0:    * using bitwise comparison.
michael@0:    *
michael@0:    * @param c The code point to search for.
michael@0:    * @return The offset into this of <TT>c</TT>, or -1 if not found.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int32_t indexOf(UChar32 c) const;
michael@0: 
michael@0:   /**
michael@0:    * Locate in this the first occurrence of the BMP code point <code>c</code>,
michael@0:    * starting at offset <TT>start</TT>, using bitwise comparison.
michael@0:    * @param c The code unit to search for.
michael@0:    * @param start The offset at which searching will start.
michael@0:    * @return The offset into this of <TT>c</TT>, or -1 if not found.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int32_t indexOf(UChar c,
michael@0:               int32_t start) const;
michael@0: 
michael@0:   /**
michael@0:    * Locate in this the first occurrence of the code point <TT>c</TT>
michael@0:    * starting at offset <TT>start</TT>, using bitwise comparison.
michael@0:    *
michael@0:    * @param c The code point to search for.
michael@0:    * @param start The offset at which searching will start.
michael@0:    * @return The offset into this of <TT>c</TT>, or -1 if not found.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int32_t indexOf(UChar32 c,
michael@0:               int32_t start) const;
michael@0: 
michael@0:   /**
michael@0:    * Locate in this the first occurrence of the BMP code point <code>c</code>
michael@0:    * in the range [<TT>start</TT>, <TT>start + length</TT>),
michael@0:    * using bitwise comparison.
michael@0:    * @param c The code unit to search for.
michael@0:    * @param start the offset into this at which to start matching
michael@0:    * @param length the number of characters in this to search
michael@0:    * @return The offset into this of <TT>c</TT>, or -1 if not found.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int32_t indexOf(UChar c,
michael@0:               int32_t start,
michael@0:               int32_t length) const;
michael@0: 
michael@0:   /**
michael@0:    * Locate in this the first occurrence of the code point <TT>c</TT>
michael@0:    * in the range [<TT>start</TT>, <TT>start + length</TT>),
michael@0:    * using bitwise comparison.
michael@0:    *
michael@0:    * @param c The code point to search for.
michael@0:    * @param start the offset into this at which to start matching
michael@0:    * @param length the number of characters in this to search
michael@0:    * @return The offset into this of <TT>c</TT>, or -1 if not found.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int32_t indexOf(UChar32 c,
michael@0:               int32_t start,
michael@0:               int32_t length) const;
michael@0: 
michael@0:   /**
michael@0:    * Locate in this the last occurrence of the characters in <TT>text</TT>,
michael@0:    * using bitwise comparison.
michael@0:    * @param text The text to search for.
michael@0:    * @return The offset into this of the start of <TT>text</TT>,
michael@0:    * or -1 if not found.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int32_t lastIndexOf(const UnicodeString& text) const;
michael@0: 
michael@0:   /**
michael@0:    * Locate in this the last occurrence of the characters in <TT>text</TT>
michael@0:    * starting at offset <TT>start</TT>, using bitwise comparison.
michael@0:    * @param text The text to search for.
michael@0:    * @param start The offset at which searching will start.
michael@0:    * @return The offset into this of the start of <TT>text</TT>,
michael@0:    * or -1 if not found.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int32_t lastIndexOf(const UnicodeString& text,
michael@0:               int32_t start) const;
michael@0: 
michael@0:   /**
michael@0:    * Locate in this the last occurrence in the range
michael@0:    * [<TT>start</TT>, <TT>start + length</TT>) of the characters
michael@0:    * in <TT>text</TT>, using bitwise comparison.
michael@0:    * @param text The text to search for.
michael@0:    * @param start The offset at which searching will start.
michael@0:    * @param length The number of characters to search
michael@0:    * @return The offset into this of the start of <TT>text</TT>,
michael@0:    * or -1 if not found.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int32_t lastIndexOf(const UnicodeString& text,
michael@0:               int32_t start,
michael@0:               int32_t length) const;
michael@0: 
michael@0:   /**
michael@0:    * Locate in this the last occurrence in the range
michael@0:    * [<TT>start</TT>, <TT>start + length</TT>) of the characters
michael@0:    * in <TT>srcText</TT> in the range
michael@0:    * [<TT>srcStart</TT>, <TT>srcStart + srcLength</TT>),
michael@0:    * using bitwise comparison.
michael@0:    * @param srcText The text to search for.
michael@0:    * @param srcStart the offset into <TT>srcText</TT> at which
michael@0:    * to start matching
michael@0:    * @param srcLength the number of characters in <TT>srcText</TT> to match
michael@0:    * @param start the offset into this at which to start matching
michael@0:    * @param length the number of characters in this to search
michael@0:    * @return The offset into this of the start of <TT>text</TT>,
michael@0:    * or -1 if not found.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int32_t lastIndexOf(const UnicodeString& srcText,
michael@0:               int32_t srcStart,
michael@0:               int32_t srcLength,
michael@0:               int32_t start,
michael@0:               int32_t length) const;
michael@0: 
michael@0:   /**
michael@0:    * Locate in this the last occurrence of the characters in <TT>srcChars</TT>
michael@0:    * starting at offset <TT>start</TT>, using bitwise comparison.
michael@0:    * @param srcChars The text to search for.
michael@0:    * @param srcLength the number of characters in <TT>srcChars</TT> to match
michael@0:    * @param start the offset into this at which to start matching
michael@0:    * @return The offset into this of the start of <TT>text</TT>,
michael@0:    * or -1 if not found.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int32_t lastIndexOf(const UChar *srcChars,
michael@0:               int32_t srcLength,
michael@0:               int32_t start) const;
michael@0: 
michael@0:   /**
michael@0:    * Locate in this the last occurrence in the range
michael@0:    * [<TT>start</TT>, <TT>start + length</TT>) of the characters
michael@0:    * in <TT>srcChars</TT>, using bitwise comparison.
michael@0:    * @param srcChars The text to search for.
michael@0:    * @param srcLength the number of characters in <TT>srcChars</TT>
michael@0:    * @param start The offset at which searching will start.
michael@0:    * @param length The number of characters to search
michael@0:    * @return The offset into this of the start of <TT>srcChars</TT>,
michael@0:    * or -1 if not found.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int32_t lastIndexOf(const UChar *srcChars,
michael@0:               int32_t srcLength,
michael@0:               int32_t start,
michael@0:               int32_t length) const;
michael@0: 
michael@0:   /**
michael@0:    * Locate in this the last occurrence in the range
michael@0:    * [<TT>start</TT>, <TT>start + length</TT>) of the characters
michael@0:    * in <TT>srcChars</TT> in the range
michael@0:    * [<TT>srcStart</TT>, <TT>srcStart + srcLength</TT>),
michael@0:    * using bitwise comparison.
michael@0:    * @param srcChars The text to search for.
michael@0:    * @param srcStart the offset into <TT>srcChars</TT> at which
michael@0:    * to start matching
michael@0:    * @param srcLength the number of characters in <TT>srcChars</TT> to match
michael@0:    * @param start the offset into this at which to start matching
michael@0:    * @param length the number of characters in this to search
michael@0:    * @return The offset into this of the start of <TT>text</TT>,
michael@0:    * or -1 if not found.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   int32_t lastIndexOf(const UChar *srcChars,
michael@0:               int32_t srcStart,
michael@0:               int32_t srcLength,
michael@0:               int32_t start,
michael@0:               int32_t length) const;
michael@0: 
michael@0:   /**
michael@0:    * Locate in this the last occurrence of the BMP code point <code>c</code>,
michael@0:    * using bitwise comparison.
michael@0:    * @param c The code unit to search for.
michael@0:    * @return The offset into this of <TT>c</TT>, or -1 if not found.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int32_t lastIndexOf(UChar c) const;
michael@0: 
michael@0:   /**
michael@0:    * Locate in this the last occurrence of the code point <TT>c</TT>,
michael@0:    * using bitwise comparison.
michael@0:    *
michael@0:    * @param c The code point to search for.
michael@0:    * @return The offset into this of <TT>c</TT>, or -1 if not found.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int32_t lastIndexOf(UChar32 c) const;
michael@0: 
michael@0:   /**
michael@0:    * Locate in this the last occurrence of the BMP code point <code>c</code>
michael@0:    * starting at offset <TT>start</TT>, using bitwise comparison.
michael@0:    * @param c The code unit to search for.
michael@0:    * @param start The offset at which searching will start.
michael@0:    * @return The offset into this of <TT>c</TT>, or -1 if not found.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int32_t lastIndexOf(UChar c,
michael@0:               int32_t start) const;
michael@0: 
michael@0:   /**
michael@0:    * Locate in this the last occurrence of the code point <TT>c</TT>
michael@0:    * starting at offset <TT>start</TT>, using bitwise comparison.
michael@0:    *
michael@0:    * @param c The code point to search for.
michael@0:    * @param start The offset at which searching will start.
michael@0:    * @return The offset into this of <TT>c</TT>, or -1 if not found.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int32_t lastIndexOf(UChar32 c,
michael@0:               int32_t start) const;
michael@0: 
michael@0:   /**
michael@0:    * Locate in this the last occurrence of the BMP code point <code>c</code>
michael@0:    * in the range [<TT>start</TT>, <TT>start + length</TT>),
michael@0:    * using bitwise comparison.
michael@0:    * @param c The code unit to search for.
michael@0:    * @param start the offset into this at which to start matching
michael@0:    * @param length the number of characters in this to search
michael@0:    * @return The offset into this of <TT>c</TT>, or -1 if not found.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int32_t lastIndexOf(UChar c,
michael@0:               int32_t start,
michael@0:               int32_t length) const;
michael@0: 
michael@0:   /**
michael@0:    * Locate in this the last occurrence of the code point <TT>c</TT>
michael@0:    * in the range [<TT>start</TT>, <TT>start + length</TT>),
michael@0:    * using bitwise comparison.
michael@0:    *
michael@0:    * @param c The code point to search for.
michael@0:    * @param start the offset into this at which to start matching
michael@0:    * @param length the number of characters in this to search
michael@0:    * @return The offset into this of <TT>c</TT>, or -1 if not found.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int32_t lastIndexOf(UChar32 c,
michael@0:               int32_t start,
michael@0:               int32_t length) const;
michael@0: 
michael@0: 
michael@0:   /* Character access */
michael@0: 
michael@0:   /**
michael@0:    * Return the code unit at offset <tt>offset</tt>.
michael@0:    * If the offset is not valid (0..length()-1) then U+ffff is returned.
michael@0:    * @param offset a valid offset into the text
michael@0:    * @return the code unit at offset <tt>offset</tt>
michael@0:    *         or 0xffff if the offset is not valid for this string
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UChar charAt(int32_t offset) const;
michael@0: 
michael@0:   /**
michael@0:    * Return the code unit at offset <tt>offset</tt>.
michael@0:    * If the offset is not valid (0..length()-1) then U+ffff is returned.
michael@0:    * @param offset a valid offset into the text
michael@0:    * @return the code unit at offset <tt>offset</tt>
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UChar operator[] (int32_t offset) const;
michael@0: 
michael@0:   /**
michael@0:    * Return the code point that contains the code unit
michael@0:    * at offset <tt>offset</tt>.
michael@0:    * If the offset is not valid (0..length()-1) then U+ffff is returned.
michael@0:    * @param offset a valid offset into the text
michael@0:    * that indicates the text offset of any of the code units
michael@0:    * that will be assembled into a code point (21-bit value) and returned
michael@0:    * @return the code point of text at <tt>offset</tt>
michael@0:    *         or 0xffff if the offset is not valid for this string
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   UChar32 char32At(int32_t offset) const;
michael@0: 
michael@0:   /**
michael@0:    * Adjust a random-access offset so that
michael@0:    * it points to the beginning of a Unicode character.
michael@0:    * The offset that is passed in points to
michael@0:    * any code unit of a code point,
michael@0:    * while the returned offset will point to the first code unit
michael@0:    * of the same code point.
michael@0:    * In UTF-16, if the input offset points to a second surrogate
michael@0:    * of a surrogate pair, then the returned offset will point
michael@0:    * to the first surrogate.
michael@0:    * @param offset a valid offset into one code point of the text
michael@0:    * @return offset of the first code unit of the same code point
michael@0:    * @see U16_SET_CP_START
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   int32_t getChar32Start(int32_t offset) const;
michael@0: 
michael@0:   /**
michael@0:    * Adjust a random-access offset so that
michael@0:    * it points behind a Unicode character.
michael@0:    * The offset that is passed in points behind
michael@0:    * any code unit of a code point,
michael@0:    * while the returned offset will point behind the last code unit
michael@0:    * of the same code point.
michael@0:    * In UTF-16, if the input offset points behind the first surrogate
michael@0:    * (i.e., to the second surrogate)
michael@0:    * of a surrogate pair, then the returned offset will point
michael@0:    * behind the second surrogate (i.e., to the first surrogate).
michael@0:    * @param offset a valid offset after any code unit of a code point of the text
michael@0:    * @return offset of the first code unit after the same code point
michael@0:    * @see U16_SET_CP_LIMIT
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   int32_t getChar32Limit(int32_t offset) const;
michael@0: 
michael@0:   /**
michael@0:    * Move the code unit index along the string by delta code points.
michael@0:    * Interpret the input index as a code unit-based offset into the string,
michael@0:    * move the index forward or backward by delta code points, and
michael@0:    * return the resulting index.
michael@0:    * The input index should point to the first code unit of a code point,
michael@0:    * if there is more than one.
michael@0:    *
michael@0:    * Both input and output indexes are code unit-based as for all
michael@0:    * string indexes/offsets in ICU (and other libraries, like MBCS char*).
michael@0:    * If delta<0 then the index is moved backward (toward the start of the string).
michael@0:    * If delta>0 then the index is moved forward (toward the end of the string).
michael@0:    *
michael@0:    * This behaves like CharacterIterator::move32(delta, kCurrent).
michael@0:    *
michael@0:    * Behavior for out-of-bounds indexes:
michael@0:    * <code>moveIndex32</code> pins the input index to 0..length(), i.e.,
michael@0:    * if the input index<0 then it is pinned to 0;
michael@0:    * if it is index>length() then it is pinned to length().
michael@0:    * Afterwards, the index is moved by <code>delta</code> code points
michael@0:    * forward or backward,
michael@0:    * but no further backward than to 0 and no further forward than to length().
michael@0:    * The resulting index return value will be in between 0 and length(), inclusively.
michael@0:    *
michael@0:    * Examples:
michael@0:    * <pre>
michael@0:    * // s has code points 'a' U+10000 'b' U+10ffff U+2029
michael@0:    * UnicodeString s=UNICODE_STRING("a\\U00010000b\\U0010ffff\\u2029", 31).unescape();
michael@0:    *
michael@0:    * // initial index: position of U+10000
michael@0:    * int32_t index=1;
michael@0:    *
michael@0:    * // the following examples will all result in index==4, position of U+10ffff
michael@0:    *
michael@0:    * // skip 2 code points from some position in the string
michael@0:    * index=s.moveIndex32(index, 2); // skips U+10000 and 'b'
michael@0:    *
michael@0:    * // go to the 3rd code point from the start of s (0-based)
michael@0:    * index=s.moveIndex32(0, 3); // skips 'a', U+10000, and 'b'
michael@0:    *
michael@0:    * // go to the next-to-last code point of s
michael@0:    * index=s.moveIndex32(s.length(), -2); // backward-skips U+2029 and U+10ffff
michael@0:    * </pre>
michael@0:    *
michael@0:    * @param index input code unit index
michael@0:    * @param delta (signed) code point count to move the index forward or backward
michael@0:    *        in the string
michael@0:    * @return the resulting code unit index
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   int32_t moveIndex32(int32_t index, int32_t delta) const;
michael@0: 
michael@0:   /* Substring extraction */
michael@0: 
michael@0:   /**
michael@0:    * Copy the characters in the range
michael@0:    * [<tt>start</tt>, <tt>start + length</tt>) into the array <tt>dst</tt>,
michael@0:    * beginning at <tt>dstStart</tt>.
michael@0:    * If the string aliases to <code>dst</code> itself as an external buffer,
michael@0:    * then extract() will not copy the contents.
michael@0:    *
michael@0:    * @param start offset of first character which will be copied into the array
michael@0:    * @param length the number of characters to extract
michael@0:    * @param dst array in which to copy characters.  The length of <tt>dst</tt>
michael@0:    * must be at least (<tt>dstStart + length</tt>).
michael@0:    * @param dstStart the offset in <TT>dst</TT> where the first character
michael@0:    * will be extracted
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline void extract(int32_t start,
michael@0:            int32_t length,
michael@0:            UChar *dst,
michael@0:            int32_t dstStart = 0) const;
michael@0: 
michael@0:   /**
michael@0:    * Copy the contents of the string into dest.
michael@0:    * This is a convenience function that
michael@0:    * checks if there is enough space in dest,
michael@0:    * extracts the entire string if possible,
michael@0:    * and NUL-terminates dest if possible.
michael@0:    *
michael@0:    * If the string fits into dest but cannot be NUL-terminated
michael@0:    * (length()==destCapacity) then the error code is set to U_STRING_NOT_TERMINATED_WARNING.
michael@0:    * If the string itself does not fit into dest
michael@0:    * (length()>destCapacity) then the error code is set to U_BUFFER_OVERFLOW_ERROR.
michael@0:    *
michael@0:    * If the string aliases to <code>dest</code> itself as an external buffer,
michael@0:    * then extract() will not copy the contents.
michael@0:    *
michael@0:    * @param dest Destination string buffer.
michael@0:    * @param destCapacity Number of UChars available at dest.
michael@0:    * @param errorCode ICU error code.
michael@0:    * @return length()
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   int32_t
michael@0:   extract(UChar *dest, int32_t destCapacity,
michael@0:           UErrorCode &errorCode) const;
michael@0: 
michael@0:   /**
michael@0:    * Copy the characters in the range
michael@0:    * [<tt>start</tt>, <tt>start + length</tt>) into the  UnicodeString
michael@0:    * <tt>target</tt>.
michael@0:    * @param start offset of first character which will be copied
michael@0:    * @param length the number of characters to extract
michael@0:    * @param target UnicodeString into which to copy characters.
michael@0:    * @return A reference to <TT>target</TT>
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline void extract(int32_t start,
michael@0:            int32_t length,
michael@0:            UnicodeString& target) const;
michael@0: 
michael@0:   /**
michael@0:    * Copy the characters in the range [<tt>start</tt>, <tt>limit</tt>)
michael@0:    * into the array <tt>dst</tt>, beginning at <tt>dstStart</tt>.
michael@0:    * @param start offset of first character which will be copied into the array
michael@0:    * @param limit offset immediately following the last character to be copied
michael@0:    * @param dst array in which to copy characters.  The length of <tt>dst</tt>
michael@0:    * must be at least (<tt>dstStart + (limit - start)</tt>).
michael@0:    * @param dstStart the offset in <TT>dst</TT> where the first character
michael@0:    * will be extracted
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline void extractBetween(int32_t start,
michael@0:               int32_t limit,
michael@0:               UChar *dst,
michael@0:               int32_t dstStart = 0) const;
michael@0: 
michael@0:   /**
michael@0:    * Copy the characters in the range [<tt>start</tt>, <tt>limit</tt>)
michael@0:    * into the UnicodeString <tt>target</tt>.  Replaceable API.
michael@0:    * @param start offset of first character which will be copied
michael@0:    * @param limit offset immediately following the last character to be copied
michael@0:    * @param target UnicodeString into which to copy characters.
michael@0:    * @return A reference to <TT>target</TT>
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   virtual void extractBetween(int32_t start,
michael@0:               int32_t limit,
michael@0:               UnicodeString& target) const;
michael@0: 
michael@0:   /**
michael@0:    * Copy the characters in the range 
michael@0:    * [<tt>start</TT>, <tt>start + length</TT>) into an array of characters.
michael@0:    * All characters must be invariant (see utypes.h).
michael@0:    * Use US_INV as the last, signature-distinguishing parameter.
michael@0:    *
michael@0:    * This function does not write any more than <code>targetLength</code>
michael@0:    * characters but returns the length of the entire output string
michael@0:    * so that one can allocate a larger buffer and call the function again
michael@0:    * if necessary.
michael@0:    * The output string is NUL-terminated if possible.
michael@0:    *
michael@0:    * @param start offset of first character which will be copied
michael@0:    * @param startLength the number of characters to extract
michael@0:    * @param target the target buffer for extraction, can be NULL
michael@0:    *               if targetLength is 0
michael@0:    * @param targetCapacity the length of the target buffer
michael@0:    * @param inv Signature-distinguishing paramater, use US_INV.
michael@0:    * @return the output string length, not including the terminating NUL
michael@0:    * @stable ICU 3.2
michael@0:    */
michael@0:   int32_t extract(int32_t start,
michael@0:            int32_t startLength,
michael@0:            char *target,
michael@0:            int32_t targetCapacity,
michael@0:            enum EInvariant inv) const;
michael@0: 
michael@0: #if U_CHARSET_IS_UTF8 || !UCONFIG_NO_CONVERSION
michael@0: 
michael@0:   /**
michael@0:    * Copy the characters in the range
michael@0:    * [<tt>start</TT>, <tt>start + length</TT>) into an array of characters
michael@0:    * in the platform's default codepage.
michael@0:    * This function does not write any more than <code>targetLength</code>
michael@0:    * characters but returns the length of the entire output string
michael@0:    * so that one can allocate a larger buffer and call the function again
michael@0:    * if necessary.
michael@0:    * The output string is NUL-terminated if possible.
michael@0:    *
michael@0:    * @param start offset of first character which will be copied
michael@0:    * @param startLength the number of characters to extract
michael@0:    * @param target the target buffer for extraction
michael@0:    * @param targetLength the length of the target buffer
michael@0:    * If <TT>target</TT> is NULL, then the number of bytes required for
michael@0:    * <TT>target</TT> is returned.
michael@0:    * @return the output string length, not including the terminating NUL
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   int32_t extract(int32_t start,
michael@0:            int32_t startLength,
michael@0:            char *target,
michael@0:            uint32_t targetLength) const;
michael@0: 
michael@0: #endif
michael@0: 
michael@0: #if !UCONFIG_NO_CONVERSION
michael@0: 
michael@0:   /**
michael@0:    * Copy the characters in the range
michael@0:    * [<tt>start</TT>, <tt>start + length</TT>) into an array of characters
michael@0:    * in a specified codepage.
michael@0:    * The output string is NUL-terminated.
michael@0:    *
michael@0:    * Recommendation: For invariant-character strings use
michael@0:    * extract(int32_t start, int32_t length, char *target, int32_t targetCapacity, enum EInvariant inv) const
michael@0:    * because it avoids object code dependencies of UnicodeString on
michael@0:    * the conversion code.
michael@0:    *
michael@0:    * @param start offset of first character which will be copied
michael@0:    * @param startLength the number of characters to extract
michael@0:    * @param target the target buffer for extraction
michael@0:    * @param codepage the desired codepage for the characters.  0 has
michael@0:    * the special meaning of the default codepage
michael@0:    * If <code>codepage</code> is an empty string (<code>""</code>),
michael@0:    * then a simple conversion is performed on the codepage-invariant
michael@0:    * subset ("invariant characters") of the platform encoding. See utypes.h.
michael@0:    * If <TT>target</TT> is NULL, then the number of bytes required for
michael@0:    * <TT>target</TT> is returned. It is assumed that the target is big enough
michael@0:    * to fit all of the characters.
michael@0:    * @return the output string length, not including the terminating NUL
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int32_t extract(int32_t start,
michael@0:                  int32_t startLength,
michael@0:                  char *target,
michael@0:                  const char *codepage = 0) const;
michael@0: 
michael@0:   /**
michael@0:    * Copy the characters in the range
michael@0:    * [<tt>start</TT>, <tt>start + length</TT>) into an array of characters
michael@0:    * in a specified codepage.
michael@0:    * This function does not write any more than <code>targetLength</code>
michael@0:    * characters but returns the length of the entire output string
michael@0:    * so that one can allocate a larger buffer and call the function again
michael@0:    * if necessary.
michael@0:    * The output string is NUL-terminated if possible.
michael@0:    *
michael@0:    * Recommendation: For invariant-character strings use
michael@0:    * extract(int32_t start, int32_t length, char *target, int32_t targetCapacity, enum EInvariant inv) const
michael@0:    * because it avoids object code dependencies of UnicodeString on
michael@0:    * the conversion code.
michael@0:    *
michael@0:    * @param start offset of first character which will be copied
michael@0:    * @param startLength the number of characters to extract
michael@0:    * @param target the target buffer for extraction
michael@0:    * @param targetLength the length of the target buffer
michael@0:    * @param codepage the desired codepage for the characters.  0 has
michael@0:    * the special meaning of the default codepage
michael@0:    * If <code>codepage</code> is an empty string (<code>""</code>),
michael@0:    * then a simple conversion is performed on the codepage-invariant
michael@0:    * subset ("invariant characters") of the platform encoding. See utypes.h.
michael@0:    * If <TT>target</TT> is NULL, then the number of bytes required for
michael@0:    * <TT>target</TT> is returned.
michael@0:    * @return the output string length, not including the terminating NUL
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   int32_t extract(int32_t start,
michael@0:            int32_t startLength,
michael@0:            char *target,
michael@0:            uint32_t targetLength,
michael@0:            const char *codepage) const;
michael@0: 
michael@0:   /**
michael@0:    * Convert the UnicodeString into a codepage string using an existing UConverter.
michael@0:    * The output string is NUL-terminated if possible.
michael@0:    *
michael@0:    * This function avoids the overhead of opening and closing a converter if
michael@0:    * multiple strings are extracted.
michael@0:    *
michael@0:    * @param dest destination string buffer, can be NULL if destCapacity==0
michael@0:    * @param destCapacity the number of chars available at dest
michael@0:    * @param cnv the converter object to be used (ucnv_resetFromUnicode() will be called),
michael@0:    *        or NULL for the default converter
michael@0:    * @param errorCode normal ICU error code
michael@0:    * @return the length of the output string, not counting the terminating NUL;
michael@0:    *         if the length is greater than destCapacity, then the string will not fit
michael@0:    *         and a buffer of the indicated length would need to be passed in
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   int32_t extract(char *dest, int32_t destCapacity,
michael@0:                   UConverter *cnv,
michael@0:                   UErrorCode &errorCode) const;
michael@0: 
michael@0: #endif
michael@0: 
michael@0:   /**
michael@0:    * Create a temporary substring for the specified range.
michael@0:    * Unlike the substring constructor and setTo() functions,
michael@0:    * the object returned here will be a read-only alias (using getBuffer())
michael@0:    * rather than copying the text.
michael@0:    * As a result, this substring operation is much faster but requires
michael@0:    * that the original string not be modified or deleted during the lifetime
michael@0:    * of the returned substring object.
michael@0:    * @param start offset of the first character visible in the substring
michael@0:    * @param length length of the substring
michael@0:    * @return a read-only alias UnicodeString object for the substring
michael@0:    * @stable ICU 4.4
michael@0:    */
michael@0:   UnicodeString tempSubString(int32_t start=0, int32_t length=INT32_MAX) const;
michael@0: 
michael@0:   /**
michael@0:    * Create a temporary substring for the specified range.
michael@0:    * Same as tempSubString(start, length) except that the substring range
michael@0:    * is specified as a (start, limit) pair (with an exclusive limit index)
michael@0:    * rather than a (start, length) pair.
michael@0:    * @param start offset of the first character visible in the substring
michael@0:    * @param limit offset immediately following the last character visible in the substring
michael@0:    * @return a read-only alias UnicodeString object for the substring
michael@0:    * @stable ICU 4.4
michael@0:    */
michael@0:   inline UnicodeString tempSubStringBetween(int32_t start, int32_t limit=INT32_MAX) const;
michael@0: 
michael@0:   /**
michael@0:    * Convert the UnicodeString to UTF-8 and write the result
michael@0:    * to a ByteSink. This is called by toUTF8String().
michael@0:    * Unpaired surrogates are replaced with U+FFFD.
michael@0:    * Calls u_strToUTF8WithSub().
michael@0:    *
michael@0:    * @param sink A ByteSink to which the UTF-8 version of the string is written.
michael@0:    *             sink.Flush() is called at the end.
michael@0:    * @stable ICU 4.2
michael@0:    * @see toUTF8String
michael@0:    */
michael@0:   void toUTF8(ByteSink &sink) const;
michael@0: 
michael@0: #if U_HAVE_STD_STRING
michael@0: 
michael@0:   /**
michael@0:    * Convert the UnicodeString to UTF-8 and append the result
michael@0:    * to a standard string.
michael@0:    * Unpaired surrogates are replaced with U+FFFD.
michael@0:    * Calls toUTF8().
michael@0:    *
michael@0:    * @param result A standard string (or a compatible object)
michael@0:    *        to which the UTF-8 version of the string is appended.
michael@0:    * @return The string object.
michael@0:    * @stable ICU 4.2
michael@0:    * @see toUTF8
michael@0:    */
michael@0:   template<typename StringClass>
michael@0:   StringClass &toUTF8String(StringClass &result) const {
michael@0:     StringByteSink<StringClass> sbs(&result);
michael@0:     toUTF8(sbs);
michael@0:     return result;
michael@0:   }
michael@0: 
michael@0: #endif
michael@0: 
michael@0:   /**
michael@0:    * Convert the UnicodeString to UTF-32.
michael@0:    * Unpaired surrogates are replaced with U+FFFD.
michael@0:    * Calls u_strToUTF32WithSub().
michael@0:    *
michael@0:    * @param utf32 destination string buffer, can be NULL if capacity==0
michael@0:    * @param capacity the number of UChar32s available at utf32
michael@0:    * @param errorCode Standard ICU error code. Its input value must
michael@0:    *                  pass the U_SUCCESS() test, or else the function returns
michael@0:    *                  immediately. Check for U_FAILURE() on output or use with
michael@0:    *                  function chaining. (See User Guide for details.)
michael@0:    * @return The length of the UTF-32 string.
michael@0:    * @see fromUTF32
michael@0:    * @stable ICU 4.2
michael@0:    */
michael@0:   int32_t toUTF32(UChar32 *utf32, int32_t capacity, UErrorCode &errorCode) const;
michael@0: 
michael@0:   /* Length operations */
michael@0: 
michael@0:   /**
michael@0:    * Return the length of the UnicodeString object.
michael@0:    * The length is the number of UChar code units are in the UnicodeString.
michael@0:    * If you want the number of code points, please use countChar32().
michael@0:    * @return the length of the UnicodeString object
michael@0:    * @see countChar32
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int32_t length(void) const;
michael@0: 
michael@0:   /**
michael@0:    * Count Unicode code points in the length UChar code units of the string.
michael@0:    * A code point may occupy either one or two UChar code units.
michael@0:    * Counting code points involves reading all code units.
michael@0:    *
michael@0:    * This functions is basically the inverse of moveIndex32().
michael@0:    *
michael@0:    * @param start the index of the first code unit to check
michael@0:    * @param length the number of UChar code units to check
michael@0:    * @return the number of code points in the specified code units
michael@0:    * @see length
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   int32_t
michael@0:   countChar32(int32_t start=0, int32_t length=INT32_MAX) const;
michael@0: 
michael@0:   /**
michael@0:    * Check if the length UChar code units of the string
michael@0:    * contain more Unicode code points than a certain number.
michael@0:    * This is more efficient than counting all code points in this part of the string
michael@0:    * and comparing that number with a threshold.
michael@0:    * This function may not need to scan the string at all if the length
michael@0:    * falls within a certain range, and
michael@0:    * never needs to count more than 'number+1' code points.
michael@0:    * Logically equivalent to (countChar32(start, length)>number).
michael@0:    * A Unicode code point may occupy either one or two UChar code units.
michael@0:    *
michael@0:    * @param start the index of the first code unit to check (0 for the entire string)
michael@0:    * @param length the number of UChar code units to check
michael@0:    *               (use INT32_MAX for the entire string; remember that start/length
michael@0:    *                values are pinned)
michael@0:    * @param number The number of code points in the (sub)string is compared against
michael@0:    *               the 'number' parameter.
michael@0:    * @return Boolean value for whether the string contains more Unicode code points
michael@0:    *         than 'number'. Same as (u_countChar32(s, length)>number).
michael@0:    * @see countChar32
michael@0:    * @see u_strHasMoreChar32Than
michael@0:    * @stable ICU 2.4
michael@0:    */
michael@0:   UBool
michael@0:   hasMoreChar32Than(int32_t start, int32_t length, int32_t number) const;
michael@0: 
michael@0:   /**
michael@0:    * Determine if this string is empty.
michael@0:    * @return TRUE if this string contains 0 characters, FALSE otherwise.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UBool isEmpty(void) const;
michael@0: 
michael@0:   /**
michael@0:    * Return the capacity of the internal buffer of the UnicodeString object.
michael@0:    * This is useful together with the getBuffer functions.
michael@0:    * See there for details.
michael@0:    *
michael@0:    * @return the number of UChars available in the internal buffer
michael@0:    * @see getBuffer
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int32_t getCapacity(void) const;
michael@0: 
michael@0:   /* Other operations */
michael@0: 
michael@0:   /**
michael@0:    * Generate a hash code for this object.
michael@0:    * @return The hash code of this UnicodeString.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline int32_t hashCode(void) const;
michael@0: 
michael@0:   /**
michael@0:    * Determine if this object contains a valid string.
michael@0:    * A bogus string has no value. It is different from an empty string,
michael@0:    * although in both cases isEmpty() returns TRUE and length() returns 0.
michael@0:    * setToBogus() and isBogus() can be used to indicate that no string value is available.
michael@0:    * For a bogus string, getBuffer() and getTerminatedBuffer() return NULL, and
michael@0:    * length() returns 0.
michael@0:    *
michael@0:    * @return TRUE if the string is bogus/invalid, FALSE otherwise
michael@0:    * @see setToBogus()
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UBool isBogus(void) const;
michael@0: 
michael@0: 
michael@0:   //========================================
michael@0:   // Write operations
michael@0:   //========================================
michael@0: 
michael@0:   /* Assignment operations */
michael@0: 
michael@0:   /**
michael@0:    * Assignment operator.  Replace the characters in this UnicodeString
michael@0:    * with the characters from <TT>srcText</TT>.
michael@0:    * @param srcText The text containing the characters to replace
michael@0:    * @return a reference to this
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   UnicodeString &operator=(const UnicodeString &srcText);
michael@0: 
michael@0:   /**
michael@0:    * Almost the same as the assignment operator.
michael@0:    * Replace the characters in this UnicodeString
michael@0:    * with the characters from <code>srcText</code>.
michael@0:    *
michael@0:    * This function works the same as the assignment operator
michael@0:    * for all strings except for ones that are readonly aliases.
michael@0:    *
michael@0:    * Starting with ICU 2.4, the assignment operator and the copy constructor
michael@0:    * allocate a new buffer and copy the buffer contents even for readonly aliases.
michael@0:    * This function implements the old, more efficient but less safe behavior
michael@0:    * of making this string also a readonly alias to the same buffer.
michael@0:    *
michael@0:    * The fastCopyFrom function must be used only if it is known that the lifetime of
michael@0:    * this UnicodeString does not exceed the lifetime of the aliased buffer
michael@0:    * including its contents, for example for strings from resource bundles
michael@0:    * or aliases to string constants.
michael@0:    *
michael@0:    * @param src The text containing the characters to replace.
michael@0:    * @return a reference to this
michael@0:    * @stable ICU 2.4
michael@0:    */
michael@0:   UnicodeString &fastCopyFrom(const UnicodeString &src);
michael@0: 
michael@0:   /**
michael@0:    * Assignment operator.  Replace the characters in this UnicodeString
michael@0:    * with the code unit <TT>ch</TT>.
michael@0:    * @param ch the code unit to replace
michael@0:    * @return a reference to this
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UnicodeString& operator= (UChar ch);
michael@0: 
michael@0:   /**
michael@0:    * Assignment operator.  Replace the characters in this UnicodeString
michael@0:    * with the code point <TT>ch</TT>.
michael@0:    * @param ch the code point to replace
michael@0:    * @return a reference to this
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UnicodeString& operator= (UChar32 ch);
michael@0: 
michael@0:   /**
michael@0:    * Set the text in the UnicodeString object to the characters
michael@0:    * in <TT>srcText</TT> in the range
michael@0:    * [<TT>srcStart</TT>, <TT>srcText.length()</TT>).
michael@0:    * <TT>srcText</TT> is not modified.
michael@0:    * @param srcText the source for the new characters
michael@0:    * @param srcStart the offset into <TT>srcText</TT> where new characters
michael@0:    * will be obtained
michael@0:    * @return a reference to this
michael@0:    * @stable ICU 2.2
michael@0:    */
michael@0:   inline UnicodeString& setTo(const UnicodeString& srcText,
michael@0:                int32_t srcStart);
michael@0: 
michael@0:   /**
michael@0:    * Set the text in the UnicodeString object to the characters
michael@0:    * in <TT>srcText</TT> in the range
michael@0:    * [<TT>srcStart</TT>, <TT>srcStart + srcLength</TT>).
michael@0:    * <TT>srcText</TT> is not modified.
michael@0:    * @param srcText the source for the new characters
michael@0:    * @param srcStart the offset into <TT>srcText</TT> where new characters
michael@0:    * will be obtained
michael@0:    * @param srcLength the number of characters in <TT>srcText</TT> in the
michael@0:    * replace string.
michael@0:    * @return a reference to this
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UnicodeString& setTo(const UnicodeString& srcText,
michael@0:                int32_t srcStart,
michael@0:                int32_t srcLength);
michael@0: 
michael@0:   /**
michael@0:    * Set the text in the UnicodeString object to the characters in
michael@0:    * <TT>srcText</TT>.
michael@0:    * <TT>srcText</TT> is not modified.
michael@0:    * @param srcText the source for the new characters
michael@0:    * @return a reference to this
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UnicodeString& setTo(const UnicodeString& srcText);
michael@0: 
michael@0:   /**
michael@0:    * Set the characters in the UnicodeString object to the characters
michael@0:    * in <TT>srcChars</TT>. <TT>srcChars</TT> is not modified.
michael@0:    * @param srcChars the source for the new characters
michael@0:    * @param srcLength the number of Unicode characters in srcChars.
michael@0:    * @return a reference to this
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UnicodeString& setTo(const UChar *srcChars,
michael@0:                int32_t srcLength);
michael@0: 
michael@0:   /**
michael@0:    * Set the characters in the UnicodeString object to the code unit
michael@0:    * <TT>srcChar</TT>.
michael@0:    * @param srcChar the code unit which becomes the UnicodeString's character
michael@0:    * content
michael@0:    * @return a reference to this
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   UnicodeString& setTo(UChar srcChar);
michael@0: 
michael@0:   /**
michael@0:    * Set the characters in the UnicodeString object to the code point
michael@0:    * <TT>srcChar</TT>.
michael@0:    * @param srcChar the code point which becomes the UnicodeString's character
michael@0:    * content
michael@0:    * @return a reference to this
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   UnicodeString& setTo(UChar32 srcChar);
michael@0: 
michael@0:   /**
michael@0:    * Aliasing setTo() function, analogous to the readonly-aliasing UChar* constructor.
michael@0:    * The text will be used for the UnicodeString object, but
michael@0:    * it will not be released when the UnicodeString is destroyed.
michael@0:    * This has copy-on-write semantics:
michael@0:    * When the string is modified, then the buffer is first copied into
michael@0:    * newly allocated memory.
michael@0:    * The aliased buffer is never modified.
michael@0:    *
michael@0:    * In an assignment to another UnicodeString, when using the copy constructor
michael@0:    * or the assignment operator, the text will be copied.
michael@0:    * When using fastCopyFrom(), the text will be aliased again,
michael@0:    * so that both strings then alias the same readonly-text.
michael@0:    *
michael@0:    * @param isTerminated specifies if <code>text</code> is <code>NUL</code>-terminated.
michael@0:    *                     This must be true if <code>textLength==-1</code>.
michael@0:    * @param text The characters to alias for the UnicodeString.
michael@0:    * @param textLength The number of Unicode characters in <code>text</code> to alias.
michael@0:    *                   If -1, then this constructor will determine the length
michael@0:    *                   by calling <code>u_strlen()</code>.
michael@0:    * @return a reference to this
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   UnicodeString &setTo(UBool isTerminated,
michael@0:                        const UChar *text,
michael@0:                        int32_t textLength);
michael@0: 
michael@0:   /**
michael@0:    * Aliasing setTo() function, analogous to the writable-aliasing UChar* constructor.
michael@0:    * The text will be used for the UnicodeString object, but
michael@0:    * it will not be released when the UnicodeString is destroyed.
michael@0:    * This has write-through semantics:
michael@0:    * For as long as the capacity of the buffer is sufficient, write operations
michael@0:    * will directly affect the buffer. When more capacity is necessary, then
michael@0:    * a new buffer will be allocated and the contents copied as with regularly
michael@0:    * constructed strings.
michael@0:    * In an assignment to another UnicodeString, the buffer will be copied.
michael@0:    * The extract(UChar *dst) function detects whether the dst pointer is the same
michael@0:    * as the string buffer itself and will in this case not copy the contents.
michael@0:    *
michael@0:    * @param buffer The characters to alias for the UnicodeString.
michael@0:    * @param buffLength The number of Unicode characters in <code>buffer</code> to alias.
michael@0:    * @param buffCapacity The size of <code>buffer</code> in UChars.
michael@0:    * @return a reference to this
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   UnicodeString &setTo(UChar *buffer,
michael@0:                        int32_t buffLength,
michael@0:                        int32_t buffCapacity);
michael@0: 
michael@0:   /**
michael@0:    * Make this UnicodeString object invalid.
michael@0:    * The string will test TRUE with isBogus().
michael@0:    *
michael@0:    * A bogus string has no value. It is different from an empty string.
michael@0:    * It can be used to indicate that no string value is available.
michael@0:    * getBuffer() and getTerminatedBuffer() return NULL, and
michael@0:    * length() returns 0.
michael@0:    *
michael@0:    * This utility function is used throughout the UnicodeString
michael@0:    * implementation to indicate that a UnicodeString operation failed,
michael@0:    * and may be used in other functions,
michael@0:    * especially but not exclusively when such functions do not
michael@0:    * take a UErrorCode for simplicity.
michael@0:    *
michael@0:    * The following methods, and no others, will clear a string object's bogus flag:
michael@0:    * - remove()
michael@0:    * - remove(0, INT32_MAX)
michael@0:    * - truncate(0)
michael@0:    * - operator=() (assignment operator)
michael@0:    * - setTo(...)
michael@0:    *
michael@0:    * The simplest ways to turn a bogus string into an empty one
michael@0:    * is to use the remove() function.
michael@0:    * Examples for other functions that are equivalent to "set to empty string":
michael@0:    * \code
michael@0:    * if(s.isBogus()) {
michael@0:    *   s.remove();           // set to an empty string (remove all), or
michael@0:    *   s.remove(0, INT32_MAX); // set to an empty string (remove all), or
michael@0:    *   s.truncate(0);        // set to an empty string (complete truncation), or
michael@0:    *   s=UnicodeString();    // assign an empty string, or
michael@0:    *   s.setTo((UChar32)-1); // set to a pseudo code point that is out of range, or
michael@0:    *   static const UChar nul=0;
michael@0:    *   s.setTo(&nul, 0);     // set to an empty C Unicode string
michael@0:    * }
michael@0:    * \endcode
michael@0:    *
michael@0:    * @see isBogus()
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   void setToBogus();
michael@0: 
michael@0:   /**
michael@0:    * Set the character at the specified offset to the specified character.
michael@0:    * @param offset A valid offset into the text of the character to set
michael@0:    * @param ch The new character
michael@0:    * @return A reference to this
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   UnicodeString& setCharAt(int32_t offset,
michael@0:                UChar ch);
michael@0: 
michael@0: 
michael@0:   /* Append operations */
michael@0: 
michael@0:   /**
michael@0:    * Append operator. Append the code unit <TT>ch</TT> to the UnicodeString
michael@0:    * object.
michael@0:    * @param ch the code unit to be appended
michael@0:    * @return a reference to this
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:  inline  UnicodeString& operator+= (UChar ch);
michael@0: 
michael@0:   /**
michael@0:    * Append operator. Append the code point <TT>ch</TT> to the UnicodeString
michael@0:    * object.
michael@0:    * @param ch the code point to be appended
michael@0:    * @return a reference to this
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:  inline  UnicodeString& operator+= (UChar32 ch);
michael@0: 
michael@0:   /**
michael@0:    * Append operator. Append the characters in <TT>srcText</TT> to the
michael@0:    * UnicodeString object. <TT>srcText</TT> is not modified.
michael@0:    * @param srcText the source for the new characters
michael@0:    * @return a reference to this
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UnicodeString& operator+= (const UnicodeString& srcText);
michael@0: 
michael@0:   /**
michael@0:    * Append the characters
michael@0:    * in <TT>srcText</TT> in the range
michael@0:    * [<TT>srcStart</TT>, <TT>srcStart + srcLength</TT>) to the
michael@0:    * UnicodeString object at offset <TT>start</TT>. <TT>srcText</TT>
michael@0:    * is not modified.
michael@0:    * @param srcText the source for the new characters
michael@0:    * @param srcStart the offset into <TT>srcText</TT> where new characters
michael@0:    * will be obtained
michael@0:    * @param srcLength the number of characters in <TT>srcText</TT> in
michael@0:    * the append string
michael@0:    * @return a reference to this
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UnicodeString& append(const UnicodeString& srcText,
michael@0:             int32_t srcStart,
michael@0:             int32_t srcLength);
michael@0: 
michael@0:   /**
michael@0:    * Append the characters in <TT>srcText</TT> to the UnicodeString object.
michael@0:    * <TT>srcText</TT> is not modified.
michael@0:    * @param srcText the source for the new characters
michael@0:    * @return a reference to this
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UnicodeString& append(const UnicodeString& srcText);
michael@0: 
michael@0:   /**
michael@0:    * Append the characters in <TT>srcChars</TT> in the range
michael@0:    * [<TT>srcStart</TT>, <TT>srcStart + srcLength</TT>) to the UnicodeString
michael@0:    * object at offset
michael@0:    * <TT>start</TT>. <TT>srcChars</TT> is not modified.
michael@0:    * @param srcChars the source for the new characters
michael@0:    * @param srcStart the offset into <TT>srcChars</TT> where new characters
michael@0:    * will be obtained
michael@0:    * @param srcLength the number of characters in <TT>srcChars</TT> in
michael@0:    *                  the append string; can be -1 if <TT>srcChars</TT> is NUL-terminated
michael@0:    * @return a reference to this
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UnicodeString& append(const UChar *srcChars,
michael@0:             int32_t srcStart,
michael@0:             int32_t srcLength);
michael@0: 
michael@0:   /**
michael@0:    * Append the characters in <TT>srcChars</TT> to the UnicodeString object
michael@0:    * at offset <TT>start</TT>. <TT>srcChars</TT> is not modified.
michael@0:    * @param srcChars the source for the new characters
michael@0:    * @param srcLength the number of Unicode characters in <TT>srcChars</TT>;
michael@0:    *                  can be -1 if <TT>srcChars</TT> is NUL-terminated
michael@0:    * @return a reference to this
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UnicodeString& append(const UChar *srcChars,
michael@0:             int32_t srcLength);
michael@0: 
michael@0:   /**
michael@0:    * Append the code unit <TT>srcChar</TT> to the UnicodeString object.
michael@0:    * @param srcChar the code unit to append
michael@0:    * @return a reference to this
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UnicodeString& append(UChar srcChar);
michael@0: 
michael@0:   /**
michael@0:    * Append the code point <TT>srcChar</TT> to the UnicodeString object.
michael@0:    * @param srcChar the code point to append
michael@0:    * @return a reference to this
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   UnicodeString& append(UChar32 srcChar);
michael@0: 
michael@0: 
michael@0:   /* Insert operations */
michael@0: 
michael@0:   /**
michael@0:    * Insert the characters in <TT>srcText</TT> in the range
michael@0:    * [<TT>srcStart</TT>, <TT>srcStart + srcLength</TT>) into the UnicodeString
michael@0:    * object at offset <TT>start</TT>. <TT>srcText</TT> is not modified.
michael@0:    * @param start the offset where the insertion begins
michael@0:    * @param srcText the source for the new characters
michael@0:    * @param srcStart the offset into <TT>srcText</TT> where new characters
michael@0:    * will be obtained
michael@0:    * @param srcLength the number of characters in <TT>srcText</TT> in
michael@0:    * the insert string
michael@0:    * @return a reference to this
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UnicodeString& insert(int32_t start,
michael@0:             const UnicodeString& srcText,
michael@0:             int32_t srcStart,
michael@0:             int32_t srcLength);
michael@0: 
michael@0:   /**
michael@0:    * Insert the characters in <TT>srcText</TT> into the UnicodeString object
michael@0:    * at offset <TT>start</TT>. <TT>srcText</TT> is not modified.
michael@0:    * @param start the offset where the insertion begins
michael@0:    * @param srcText the source for the new characters
michael@0:    * @return a reference to this
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UnicodeString& insert(int32_t start,
michael@0:             const UnicodeString& srcText);
michael@0: 
michael@0:   /**
michael@0:    * Insert the characters in <TT>srcChars</TT> in the range
michael@0:    * [<TT>srcStart</TT>, <TT>srcStart + srcLength</TT>) into the UnicodeString
michael@0:    *  object at offset <TT>start</TT>. <TT>srcChars</TT> is not modified.
michael@0:    * @param start the offset at which the insertion begins
michael@0:    * @param srcChars the source for the new characters
michael@0:    * @param srcStart the offset into <TT>srcChars</TT> where new characters
michael@0:    * will be obtained
michael@0:    * @param srcLength the number of characters in <TT>srcChars</TT>
michael@0:    * in the insert string
michael@0:    * @return a reference to this
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UnicodeString& insert(int32_t start,
michael@0:             const UChar *srcChars,
michael@0:             int32_t srcStart,
michael@0:             int32_t srcLength);
michael@0: 
michael@0:   /**
michael@0:    * Insert the characters in <TT>srcChars</TT> into the UnicodeString object
michael@0:    * at offset <TT>start</TT>. <TT>srcChars</TT> is not modified.
michael@0:    * @param start the offset where the insertion begins
michael@0:    * @param srcChars the source for the new characters
michael@0:    * @param srcLength the number of Unicode characters in srcChars.
michael@0:    * @return a reference to this
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UnicodeString& insert(int32_t start,
michael@0:             const UChar *srcChars,
michael@0:             int32_t srcLength);
michael@0: 
michael@0:   /**
michael@0:    * Insert the code unit <TT>srcChar</TT> into the UnicodeString object at
michael@0:    * offset <TT>start</TT>.
michael@0:    * @param start the offset at which the insertion occurs
michael@0:    * @param srcChar the code unit to insert
michael@0:    * @return a reference to this
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UnicodeString& insert(int32_t start,
michael@0:             UChar srcChar);
michael@0: 
michael@0:   /**
michael@0:    * Insert the code point <TT>srcChar</TT> into the UnicodeString object at
michael@0:    * offset <TT>start</TT>.
michael@0:    * @param start the offset at which the insertion occurs
michael@0:    * @param srcChar the code point to insert
michael@0:    * @return a reference to this
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UnicodeString& insert(int32_t start,
michael@0:             UChar32 srcChar);
michael@0: 
michael@0: 
michael@0:   /* Replace operations */
michael@0: 
michael@0:   /**
michael@0:    * Replace the characters in the range
michael@0:    * [<TT>start</TT>, <TT>start + length</TT>) with the characters in
michael@0:    * <TT>srcText</TT> in the range
michael@0:    * [<TT>srcStart</TT>, <TT>srcStart + srcLength</TT>).
michael@0:    * <TT>srcText</TT> is not modified.
michael@0:    * @param start the offset at which the replace operation begins
michael@0:    * @param length the number of characters to replace. The character at
michael@0:    * <TT>start + length</TT> is not modified.
michael@0:    * @param srcText the source for the new characters
michael@0:    * @param srcStart the offset into <TT>srcText</TT> where new characters
michael@0:    * will be obtained
michael@0:    * @param srcLength the number of characters in <TT>srcText</TT> in
michael@0:    * the replace string
michael@0:    * @return a reference to this
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   UnicodeString& replace(int32_t start,
michael@0:              int32_t length,
michael@0:              const UnicodeString& srcText,
michael@0:              int32_t srcStart,
michael@0:              int32_t srcLength);
michael@0: 
michael@0:   /**
michael@0:    * Replace the characters in the range
michael@0:    * [<TT>start</TT>, <TT>start + length</TT>)
michael@0:    * with the characters in <TT>srcText</TT>.  <TT>srcText</TT> is
michael@0:    *  not modified.
michael@0:    * @param start the offset at which the replace operation begins
michael@0:    * @param length the number of characters to replace. The character at
michael@0:    * <TT>start + length</TT> is not modified.
michael@0:    * @param srcText the source for the new characters
michael@0:    * @return a reference to this
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   UnicodeString& replace(int32_t start,
michael@0:              int32_t length,
michael@0:              const UnicodeString& srcText);
michael@0: 
michael@0:   /**
michael@0:    * Replace the characters in the range
michael@0:    * [<TT>start</TT>, <TT>start + length</TT>) with the characters in
michael@0:    * <TT>srcChars</TT> in the range
michael@0:    * [<TT>srcStart</TT>, <TT>srcStart + srcLength</TT>). <TT>srcChars</TT>
michael@0:    * is not modified.
michael@0:    * @param start the offset at which the replace operation begins
michael@0:    * @param length the number of characters to replace.  The character at
michael@0:    * <TT>start + length</TT> is not modified.
michael@0:    * @param srcChars the source for the new characters
michael@0:    * @param srcStart the offset into <TT>srcChars</TT> where new characters
michael@0:    * will be obtained
michael@0:    * @param srcLength the number of characters in <TT>srcChars</TT>
michael@0:    * in the replace string
michael@0:    * @return a reference to this
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   UnicodeString& replace(int32_t start,
michael@0:              int32_t length,
michael@0:              const UChar *srcChars,
michael@0:              int32_t srcStart,
michael@0:              int32_t srcLength);
michael@0: 
michael@0:   /**
michael@0:    * Replace the characters in the range
michael@0:    * [<TT>start</TT>, <TT>start + length</TT>) with the characters in
michael@0:    * <TT>srcChars</TT>.  <TT>srcChars</TT> is not modified.
michael@0:    * @param start the offset at which the replace operation begins
michael@0:    * @param length number of characters to replace.  The character at
michael@0:    * <TT>start + length</TT> is not modified.
michael@0:    * @param srcChars the source for the new characters
michael@0:    * @param srcLength the number of Unicode characters in srcChars
michael@0:    * @return a reference to this
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UnicodeString& replace(int32_t start,
michael@0:              int32_t length,
michael@0:              const UChar *srcChars,
michael@0:              int32_t srcLength);
michael@0: 
michael@0:   /**
michael@0:    * Replace the characters in the range
michael@0:    * [<TT>start</TT>, <TT>start + length</TT>) with the code unit
michael@0:    * <TT>srcChar</TT>.
michael@0:    * @param start the offset at which the replace operation begins
michael@0:    * @param length the number of characters to replace.  The character at
michael@0:    * <TT>start + length</TT> is not modified.
michael@0:    * @param srcChar the new code unit
michael@0:    * @return a reference to this
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UnicodeString& replace(int32_t start,
michael@0:              int32_t length,
michael@0:              UChar srcChar);
michael@0: 
michael@0:   /**
michael@0:    * Replace the characters in the range
michael@0:    * [<TT>start</TT>, <TT>start + length</TT>) with the code point
michael@0:    * <TT>srcChar</TT>.
michael@0:    * @param start the offset at which the replace operation begins
michael@0:    * @param length the number of characters to replace.  The character at
michael@0:    * <TT>start + length</TT> is not modified.
michael@0:    * @param srcChar the new code point
michael@0:    * @return a reference to this
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   UnicodeString& replace(int32_t start, int32_t length, UChar32 srcChar);
michael@0: 
michael@0:   /**
michael@0:    * Replace the characters in the range [<TT>start</TT>, <TT>limit</TT>)
michael@0:    * with the characters in <TT>srcText</TT>. <TT>srcText</TT> is not modified.
michael@0:    * @param start the offset at which the replace operation begins
michael@0:    * @param limit the offset immediately following the replace range
michael@0:    * @param srcText the source for the new characters
michael@0:    * @return a reference to this
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UnicodeString& replaceBetween(int32_t start,
michael@0:                 int32_t limit,
michael@0:                 const UnicodeString& srcText);
michael@0: 
michael@0:   /**
michael@0:    * Replace the characters in the range [<TT>start</TT>, <TT>limit</TT>)
michael@0:    * with the characters in <TT>srcText</TT> in the range
michael@0:    * [<TT>srcStart</TT>, <TT>srcLimit</TT>). <TT>srcText</TT> is not modified.
michael@0:    * @param start the offset at which the replace operation begins
michael@0:    * @param limit the offset immediately following the replace range
michael@0:    * @param srcText the source for the new characters
michael@0:    * @param srcStart the offset into <TT>srcChars</TT> where new characters
michael@0:    * will be obtained
michael@0:    * @param srcLimit the offset immediately following the range to copy
michael@0:    * in <TT>srcText</TT>
michael@0:    * @return a reference to this
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UnicodeString& replaceBetween(int32_t start,
michael@0:                 int32_t limit,
michael@0:                 const UnicodeString& srcText,
michael@0:                 int32_t srcStart,
michael@0:                 int32_t srcLimit);
michael@0: 
michael@0:   /**
michael@0:    * Replace a substring of this object with the given text.
michael@0:    * @param start the beginning index, inclusive; <code>0 <= start
michael@0:    * <= limit</code>.
michael@0:    * @param limit the ending index, exclusive; <code>start <= limit
michael@0:    * <= length()</code>.
michael@0:    * @param text the text to replace characters <code>start</code>
michael@0:    * to <code>limit - 1</code>
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   virtual void handleReplaceBetween(int32_t start,
michael@0:                                     int32_t limit,
michael@0:                                     const UnicodeString& text);
michael@0: 
michael@0:   /**
michael@0:    * Replaceable API
michael@0:    * @return TRUE if it has MetaData
michael@0:    * @stable ICU 2.4
michael@0:    */
michael@0:   virtual UBool hasMetaData() const;
michael@0: 
michael@0:   /**
michael@0:    * Copy a substring of this object, retaining attribute (out-of-band)
michael@0:    * information.  This method is used to duplicate or reorder substrings.
michael@0:    * The destination index must not overlap the source range.
michael@0:    *
michael@0:    * @param start the beginning index, inclusive; <code>0 <= start <=
michael@0:    * limit</code>.
michael@0:    * @param limit the ending index, exclusive; <code>start <= limit <=
michael@0:    * length()</code>.
michael@0:    * @param dest the destination index.  The characters from
michael@0:    * <code>start..limit-1</code> will be copied to <code>dest</code>.
michael@0:    * Implementations of this method may assume that <code>dest <= start ||
michael@0:    * dest >= limit</code>.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   virtual void copy(int32_t start, int32_t limit, int32_t dest);
michael@0: 
michael@0:   /* Search and replace operations */
michael@0: 
michael@0:   /**
michael@0:    * Replace all occurrences of characters in oldText with the characters
michael@0:    * in newText
michael@0:    * @param oldText the text containing the search text
michael@0:    * @param newText the text containing the replacement text
michael@0:    * @return a reference to this
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UnicodeString& findAndReplace(const UnicodeString& oldText,
michael@0:                 const UnicodeString& newText);
michael@0: 
michael@0:   /**
michael@0:    * Replace all occurrences of characters in oldText with characters
michael@0:    * in newText
michael@0:    * in the range [<TT>start</TT>, <TT>start + length</TT>).
michael@0:    * @param start the start of the range in which replace will performed
michael@0:    * @param length the length of the range in which replace will be performed
michael@0:    * @param oldText the text containing the search text
michael@0:    * @param newText the text containing the replacement text
michael@0:    * @return a reference to this
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UnicodeString& findAndReplace(int32_t start,
michael@0:                 int32_t length,
michael@0:                 const UnicodeString& oldText,
michael@0:                 const UnicodeString& newText);
michael@0: 
michael@0:   /**
michael@0:    * Replace all occurrences of characters in oldText in the range
michael@0:    * [<TT>oldStart</TT>, <TT>oldStart + oldLength</TT>) with the characters
michael@0:    * in newText in the range
michael@0:    * [<TT>newStart</TT>, <TT>newStart + newLength</TT>)
michael@0:    * in the range [<TT>start</TT>, <TT>start + length</TT>).
michael@0:    * @param start the start of the range in which replace will performed
michael@0:    * @param length the length of the range in which replace will be performed
michael@0:    * @param oldText the text containing the search text
michael@0:    * @param oldStart the start of the search range in <TT>oldText</TT>
michael@0:    * @param oldLength the length of the search range in <TT>oldText</TT>
michael@0:    * @param newText the text containing the replacement text
michael@0:    * @param newStart the start of the replacement range in <TT>newText</TT>
michael@0:    * @param newLength the length of the replacement range in <TT>newText</TT>
michael@0:    * @return a reference to this
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   UnicodeString& findAndReplace(int32_t start,
michael@0:                 int32_t length,
michael@0:                 const UnicodeString& oldText,
michael@0:                 int32_t oldStart,
michael@0:                 int32_t oldLength,
michael@0:                 const UnicodeString& newText,
michael@0:                 int32_t newStart,
michael@0:                 int32_t newLength);
michael@0: 
michael@0: 
michael@0:   /* Remove operations */
michael@0: 
michael@0:   /**
michael@0:    * Remove all characters from the UnicodeString object.
michael@0:    * @return a reference to this
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UnicodeString& remove(void);
michael@0: 
michael@0:   /**
michael@0:    * Remove the characters in the range
michael@0:    * [<TT>start</TT>, <TT>start + length</TT>) from the UnicodeString object.
michael@0:    * @param start the offset of the first character to remove
michael@0:    * @param length the number of characters to remove
michael@0:    * @return a reference to this
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UnicodeString& remove(int32_t start,
michael@0:                                int32_t length = (int32_t)INT32_MAX);
michael@0: 
michael@0:   /**
michael@0:    * Remove the characters in the range
michael@0:    * [<TT>start</TT>, <TT>limit</TT>) from the UnicodeString object.
michael@0:    * @param start the offset of the first character to remove
michael@0:    * @param limit the offset immediately following the range to remove
michael@0:    * @return a reference to this
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UnicodeString& removeBetween(int32_t start,
michael@0:                                       int32_t limit = (int32_t)INT32_MAX);
michael@0: 
michael@0:   /**
michael@0:    * Retain only the characters in the range
michael@0:    * [<code>start</code>, <code>limit</code>) from the UnicodeString object.
michael@0:    * Removes characters before <code>start</code> and at and after <code>limit</code>.
michael@0:    * @param start the offset of the first character to retain
michael@0:    * @param limit the offset immediately following the range to retain
michael@0:    * @return a reference to this
michael@0:    * @stable ICU 4.4
michael@0:    */
michael@0:   inline UnicodeString &retainBetween(int32_t start, int32_t limit = INT32_MAX);
michael@0: 
michael@0:   /* Length operations */
michael@0: 
michael@0:   /**
michael@0:    * Pad the start of this UnicodeString with the character <TT>padChar</TT>.
michael@0:    * If the length of this UnicodeString is less than targetLength,
michael@0:    * length() - targetLength copies of padChar will be added to the
michael@0:    * beginning of this UnicodeString.
michael@0:    * @param targetLength the desired length of the string
michael@0:    * @param padChar the character to use for padding. Defaults to
michael@0:    * space (U+0020)
michael@0:    * @return TRUE if the text was padded, FALSE otherwise.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   UBool padLeading(int32_t targetLength,
michael@0:                     UChar padChar = 0x0020);
michael@0: 
michael@0:   /**
michael@0:    * Pad the end of this UnicodeString with the character <TT>padChar</TT>.
michael@0:    * If the length of this UnicodeString is less than targetLength,
michael@0:    * length() - targetLength copies of padChar will be added to the
michael@0:    * end of this UnicodeString.
michael@0:    * @param targetLength the desired length of the string
michael@0:    * @param padChar the character to use for padding. Defaults to
michael@0:    * space (U+0020)
michael@0:    * @return TRUE if the text was padded, FALSE otherwise.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   UBool padTrailing(int32_t targetLength,
michael@0:                      UChar padChar = 0x0020);
michael@0: 
michael@0:   /**
michael@0:    * Truncate this UnicodeString to the <TT>targetLength</TT>.
michael@0:    * @param targetLength the desired length of this UnicodeString.
michael@0:    * @return TRUE if the text was truncated, FALSE otherwise
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UBool truncate(int32_t targetLength);
michael@0: 
michael@0:   /**
michael@0:    * Trims leading and trailing whitespace from this UnicodeString.
michael@0:    * @return a reference to this
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   UnicodeString& trim(void);
michael@0: 
michael@0: 
michael@0:   /* Miscellaneous operations */
michael@0: 
michael@0:   /**
michael@0:    * Reverse this UnicodeString in place.
michael@0:    * @return a reference to this
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UnicodeString& reverse(void);
michael@0: 
michael@0:   /**
michael@0:    * Reverse the range [<TT>start</TT>, <TT>start + length</TT>) in
michael@0:    * this UnicodeString.
michael@0:    * @param start the start of the range to reverse
michael@0:    * @param length the number of characters to to reverse
michael@0:    * @return a reference to this
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UnicodeString& reverse(int32_t start,
michael@0:              int32_t length);
michael@0: 
michael@0:   /**
michael@0:    * Convert the characters in this to UPPER CASE following the conventions of
michael@0:    * the default locale.
michael@0:    * @return A reference to this.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   UnicodeString& toUpper(void);
michael@0: 
michael@0:   /**
michael@0:    * Convert the characters in this to UPPER CASE following the conventions of
michael@0:    * a specific locale.
michael@0:    * @param locale The locale containing the conventions to use.
michael@0:    * @return A reference to this.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   UnicodeString& toUpper(const Locale& locale);
michael@0: 
michael@0:   /**
michael@0:    * Convert the characters in this to lower case following the conventions of
michael@0:    * the default locale.
michael@0:    * @return A reference to this.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   UnicodeString& toLower(void);
michael@0: 
michael@0:   /**
michael@0:    * Convert the characters in this to lower case following the conventions of
michael@0:    * a specific locale.
michael@0:    * @param locale The locale containing the conventions to use.
michael@0:    * @return A reference to this.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   UnicodeString& toLower(const Locale& locale);
michael@0: 
michael@0: #if !UCONFIG_NO_BREAK_ITERATION
michael@0: 
michael@0:   /**
michael@0:    * Titlecase this string, convenience function using the default locale.
michael@0:    *
michael@0:    * Casing is locale-dependent and context-sensitive.
michael@0:    * Titlecasing uses a break iterator to find the first characters of words
michael@0:    * that are to be titlecased. It titlecases those characters and lowercases
michael@0:    * all others.
michael@0:    *
michael@0:    * The titlecase break iterator can be provided to customize for arbitrary
michael@0:    * styles, using rules and dictionaries beyond the standard iterators.
michael@0:    * It may be more efficient to always provide an iterator to avoid
michael@0:    * opening and closing one for each string.
michael@0:    * The standard titlecase iterator for the root locale implements the
michael@0:    * algorithm of Unicode TR 21.
michael@0:    *
michael@0:    * This function uses only the setText(), first() and next() methods of the
michael@0:    * provided break iterator.
michael@0:    *
michael@0:    * @param titleIter A break iterator to find the first characters of words
michael@0:    *                  that are to be titlecased.
michael@0:    *                  If none is provided (0), then a standard titlecase
michael@0:    *                  break iterator is opened.
michael@0:    *                  Otherwise the provided iterator is set to the string's text.
michael@0:    * @return A reference to this.
michael@0:    * @stable ICU 2.1
michael@0:    */
michael@0:   UnicodeString &toTitle(BreakIterator *titleIter);
michael@0: 
michael@0:   /**
michael@0:    * Titlecase this string.
michael@0:    *
michael@0:    * Casing is locale-dependent and context-sensitive.
michael@0:    * Titlecasing uses a break iterator to find the first characters of words
michael@0:    * that are to be titlecased. It titlecases those characters and lowercases
michael@0:    * all others.
michael@0:    *
michael@0:    * The titlecase break iterator can be provided to customize for arbitrary
michael@0:    * styles, using rules and dictionaries beyond the standard iterators.
michael@0:    * It may be more efficient to always provide an iterator to avoid
michael@0:    * opening and closing one for each string.
michael@0:    * The standard titlecase iterator for the root locale implements the
michael@0:    * algorithm of Unicode TR 21.
michael@0:    *
michael@0:    * This function uses only the setText(), first() and next() methods of the
michael@0:    * provided break iterator.
michael@0:    *
michael@0:    * @param titleIter A break iterator to find the first characters of words
michael@0:    *                  that are to be titlecased.
michael@0:    *                  If none is provided (0), then a standard titlecase
michael@0:    *                  break iterator is opened.
michael@0:    *                  Otherwise the provided iterator is set to the string's text.
michael@0:    * @param locale    The locale to consider.
michael@0:    * @return A reference to this.
michael@0:    * @stable ICU 2.1
michael@0:    */
michael@0:   UnicodeString &toTitle(BreakIterator *titleIter, const Locale &locale);
michael@0: 
michael@0:   /**
michael@0:    * Titlecase this string, with options.
michael@0:    *
michael@0:    * Casing is locale-dependent and context-sensitive.
michael@0:    * Titlecasing uses a break iterator to find the first characters of words
michael@0:    * that are to be titlecased. It titlecases those characters and lowercases
michael@0:    * all others. (This can be modified with options.)
michael@0:    *
michael@0:    * The titlecase break iterator can be provided to customize for arbitrary
michael@0:    * styles, using rules and dictionaries beyond the standard iterators.
michael@0:    * It may be more efficient to always provide an iterator to avoid
michael@0:    * opening and closing one for each string.
michael@0:    * The standard titlecase iterator for the root locale implements the
michael@0:    * algorithm of Unicode TR 21.
michael@0:    *
michael@0:    * This function uses only the setText(), first() and next() methods of the
michael@0:    * provided break iterator.
michael@0:    *
michael@0:    * @param titleIter A break iterator to find the first characters of words
michael@0:    *                  that are to be titlecased.
michael@0:    *                  If none is provided (0), then a standard titlecase
michael@0:    *                  break iterator is opened.
michael@0:    *                  Otherwise the provided iterator is set to the string's text.
michael@0:    * @param locale    The locale to consider.
michael@0:    * @param options Options bit set, see ucasemap_open().
michael@0:    * @return A reference to this.
michael@0:    * @see U_TITLECASE_NO_LOWERCASE
michael@0:    * @see U_TITLECASE_NO_BREAK_ADJUSTMENT
michael@0:    * @see ucasemap_open
michael@0:    * @stable ICU 3.8
michael@0:    */
michael@0:   UnicodeString &toTitle(BreakIterator *titleIter, const Locale &locale, uint32_t options);
michael@0: 
michael@0: #endif
michael@0: 
michael@0:   /**
michael@0:    * Case-folds the characters in this string.
michael@0:    *
michael@0:    * Case-folding is locale-independent and not context-sensitive,
michael@0:    * but there is an option for whether to include or exclude mappings for dotted I
michael@0:    * and dotless i that are marked with 'T' in CaseFolding.txt.
michael@0:    *
michael@0:    * The result may be longer or shorter than the original.
michael@0:    *
michael@0:    * @param options Either U_FOLD_CASE_DEFAULT or U_FOLD_CASE_EXCLUDE_SPECIAL_I
michael@0:    * @return A reference to this.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   UnicodeString &foldCase(uint32_t options=0 /*U_FOLD_CASE_DEFAULT*/);
michael@0: 
michael@0:   //========================================
michael@0:   // Access to the internal buffer
michael@0:   //========================================
michael@0: 
michael@0:   /**
michael@0:    * Get a read/write pointer to the internal buffer.
michael@0:    * The buffer is guaranteed to be large enough for at least minCapacity UChars,
michael@0:    * writable, and is still owned by the UnicodeString object.
michael@0:    * Calls to getBuffer(minCapacity) must not be nested, and
michael@0:    * must be matched with calls to releaseBuffer(newLength).
michael@0:    * If the string buffer was read-only or shared,
michael@0:    * then it will be reallocated and copied.
michael@0:    *
michael@0:    * An attempted nested call will return 0, and will not further modify the
michael@0:    * state of the UnicodeString object.
michael@0:    * It also returns 0 if the string is bogus.
michael@0:    *
michael@0:    * The actual capacity of the string buffer may be larger than minCapacity.
michael@0:    * getCapacity() returns the actual capacity.
michael@0:    * For many operations, the full capacity should be used to avoid reallocations.
michael@0:    *
michael@0:    * While the buffer is "open" between getBuffer(minCapacity)
michael@0:    * and releaseBuffer(newLength), the following applies:
michael@0:    * - The string length is set to 0.
michael@0:    * - Any read API call on the UnicodeString object will behave like on a 0-length string.
michael@0:    * - Any write API call on the UnicodeString object is disallowed and will have no effect.
michael@0:    * - You can read from and write to the returned buffer.
michael@0:    * - The previous string contents will still be in the buffer;
michael@0:    *   if you want to use it, then you need to call length() before getBuffer(minCapacity).
michael@0:    *   If the length() was greater than minCapacity, then any contents after minCapacity
michael@0:    *   may be lost.
michael@0:    *   The buffer contents is not NUL-terminated by getBuffer().
michael@0:    *   If length()<getCapacity() then you can terminate it by writing a NUL
michael@0:    *   at index length().
michael@0:    * - You must call releaseBuffer(newLength) before and in order to
michael@0:    *   return to normal UnicodeString operation.
michael@0:    *
michael@0:    * @param minCapacity the minimum number of UChars that are to be available
michael@0:    *        in the buffer, starting at the returned pointer;
michael@0:    *        default to the current string capacity if minCapacity==-1
michael@0:    * @return a writable pointer to the internal string buffer,
michael@0:    *         or 0 if an error occurs (nested calls, out of memory)
michael@0:    *
michael@0:    * @see releaseBuffer
michael@0:    * @see getTerminatedBuffer()
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   UChar *getBuffer(int32_t minCapacity);
michael@0: 
michael@0:   /**
michael@0:    * Release a read/write buffer on a UnicodeString object with an
michael@0:    * "open" getBuffer(minCapacity).
michael@0:    * This function must be called in a matched pair with getBuffer(minCapacity).
michael@0:    * releaseBuffer(newLength) must be called if and only if a getBuffer(minCapacity) is "open".
michael@0:    *
michael@0:    * It will set the string length to newLength, at most to the current capacity.
michael@0:    * If newLength==-1 then it will set the length according to the
michael@0:    * first NUL in the buffer, or to the capacity if there is no NUL.
michael@0:    *
michael@0:    * After calling releaseBuffer(newLength) the UnicodeString is back to normal operation.
michael@0:    *
michael@0:    * @param newLength the new length of the UnicodeString object;
michael@0:    *        defaults to the current capacity if newLength is greater than that;
michael@0:    *        if newLength==-1, it defaults to u_strlen(buffer) but not more than
michael@0:    *        the current capacity of the string
michael@0:    *
michael@0:    * @see getBuffer(int32_t minCapacity)
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   void releaseBuffer(int32_t newLength=-1);
michael@0: 
michael@0:   /**
michael@0:    * Get a read-only pointer to the internal buffer.
michael@0:    * This can be called at any time on a valid UnicodeString.
michael@0:    *
michael@0:    * It returns 0 if the string is bogus, or
michael@0:    * during an "open" getBuffer(minCapacity).
michael@0:    *
michael@0:    * It can be called as many times as desired.
michael@0:    * The pointer that it returns will remain valid until the UnicodeString object is modified,
michael@0:    * at which time the pointer is semantically invalidated and must not be used any more.
michael@0:    *
michael@0:    * The capacity of the buffer can be determined with getCapacity().
michael@0:    * The part after length() may or may not be initialized and valid,
michael@0:    * depending on the history of the UnicodeString object.
michael@0:    *
michael@0:    * The buffer contents is (probably) not NUL-terminated.
michael@0:    * You can check if it is with
michael@0:    * <code>(s.length()<s.getCapacity() && buffer[s.length()]==0)</code>.
michael@0:    * (See getTerminatedBuffer().)
michael@0:    *
michael@0:    * The buffer may reside in read-only memory. Its contents must not
michael@0:    * be modified.
michael@0:    *
michael@0:    * @return a read-only pointer to the internal string buffer,
michael@0:    *         or 0 if the string is empty or bogus
michael@0:    *
michael@0:    * @see getBuffer(int32_t minCapacity)
michael@0:    * @see getTerminatedBuffer()
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline const UChar *getBuffer() const;
michael@0: 
michael@0:   /**
michael@0:    * Get a read-only pointer to the internal buffer,
michael@0:    * making sure that it is NUL-terminated.
michael@0:    * This can be called at any time on a valid UnicodeString.
michael@0:    *
michael@0:    * It returns 0 if the string is bogus, or
michael@0:    * during an "open" getBuffer(minCapacity), or if the buffer cannot
michael@0:    * be NUL-terminated (because memory allocation failed).
michael@0:    *
michael@0:    * It can be called as many times as desired.
michael@0:    * The pointer that it returns will remain valid until the UnicodeString object is modified,
michael@0:    * at which time the pointer is semantically invalidated and must not be used any more.
michael@0:    *
michael@0:    * The capacity of the buffer can be determined with getCapacity().
michael@0:    * The part after length()+1 may or may not be initialized and valid,
michael@0:    * depending on the history of the UnicodeString object.
michael@0:    *
michael@0:    * The buffer contents is guaranteed to be NUL-terminated.
michael@0:    * getTerminatedBuffer() may reallocate the buffer if a terminating NUL
michael@0:    * is written.
michael@0:    * For this reason, this function is not const, unlike getBuffer().
michael@0:    * Note that a UnicodeString may also contain NUL characters as part of its contents.
michael@0:    *
michael@0:    * The buffer may reside in read-only memory. Its contents must not
michael@0:    * be modified.
michael@0:    *
michael@0:    * @return a read-only pointer to the internal string buffer,
michael@0:    *         or 0 if the string is empty or bogus
michael@0:    *
michael@0:    * @see getBuffer(int32_t minCapacity)
michael@0:    * @see getBuffer()
michael@0:    * @stable ICU 2.2
michael@0:    */
michael@0:   const UChar *getTerminatedBuffer();
michael@0: 
michael@0:   //========================================
michael@0:   // Constructors
michael@0:   //========================================
michael@0: 
michael@0:   /** Construct an empty UnicodeString.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   inline UnicodeString();
michael@0: 
michael@0:   /**
michael@0:    * Construct a UnicodeString with capacity to hold <TT>capacity</TT> UChars
michael@0:    * @param capacity the number of UChars this UnicodeString should hold
michael@0:    * before a resize is necessary; if count is greater than 0 and count
michael@0:    * code points c take up more space than capacity, then capacity is adjusted
michael@0:    * accordingly.
michael@0:    * @param c is used to initially fill the string
michael@0:    * @param count specifies how many code points c are to be written in the
michael@0:    *              string
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   UnicodeString(int32_t capacity, UChar32 c, int32_t count);
michael@0: 
michael@0:   /**
michael@0:    * Single UChar (code unit) constructor.
michael@0:    *
michael@0:    * It is recommended to mark this constructor "explicit" by
michael@0:    * <code>-DUNISTR_FROM_CHAR_EXPLICIT=explicit</code>
michael@0:    * on the compiler command line or similar.
michael@0:    * @param ch the character to place in the UnicodeString
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   UNISTR_FROM_CHAR_EXPLICIT UnicodeString(UChar ch);
michael@0: 
michael@0:   /**
michael@0:    * Single UChar32 (code point) constructor.
michael@0:    *
michael@0:    * It is recommended to mark this constructor "explicit" by
michael@0:    * <code>-DUNISTR_FROM_CHAR_EXPLICIT=explicit</code>
michael@0:    * on the compiler command line or similar.
michael@0:    * @param ch the character to place in the UnicodeString
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   UNISTR_FROM_CHAR_EXPLICIT UnicodeString(UChar32 ch);
michael@0: 
michael@0:   /**
michael@0:    * UChar* constructor.
michael@0:    *
michael@0:    * It is recommended to mark this constructor "explicit" by
michael@0:    * <code>-DUNISTR_FROM_STRING_EXPLICIT=explicit</code>
michael@0:    * on the compiler command line or similar.
michael@0:    * @param text The characters to place in the UnicodeString.  <TT>text</TT>
michael@0:    * must be NULL (U+0000) terminated.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   UNISTR_FROM_STRING_EXPLICIT UnicodeString(const UChar *text);
michael@0: 
michael@0:   /**
michael@0:    * UChar* constructor.
michael@0:    * @param text The characters to place in the UnicodeString.
michael@0:    * @param textLength The number of Unicode characters in <TT>text</TT>
michael@0:    * to copy.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   UnicodeString(const UChar *text,
michael@0:         int32_t textLength);
michael@0: 
michael@0:   /**
michael@0:    * Readonly-aliasing UChar* constructor.
michael@0:    * The text will be used for the UnicodeString object, but
michael@0:    * it will not be released when the UnicodeString is destroyed.
michael@0:    * This has copy-on-write semantics:
michael@0:    * When the string is modified, then the buffer is first copied into
michael@0:    * newly allocated memory.
michael@0:    * The aliased buffer is never modified.
michael@0:    *
michael@0:    * In an assignment to another UnicodeString, when using the copy constructor
michael@0:    * or the assignment operator, the text will be copied.
michael@0:    * When using fastCopyFrom(), the text will be aliased again,
michael@0:    * so that both strings then alias the same readonly-text.
michael@0:    *
michael@0:    * @param isTerminated specifies if <code>text</code> is <code>NUL</code>-terminated.
michael@0:    *                     This must be true if <code>textLength==-1</code>.
michael@0:    * @param text The characters to alias for the UnicodeString.
michael@0:    * @param textLength The number of Unicode characters in <code>text</code> to alias.
michael@0:    *                   If -1, then this constructor will determine the length
michael@0:    *                   by calling <code>u_strlen()</code>.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   UnicodeString(UBool isTerminated,
michael@0:                 const UChar *text,
michael@0:                 int32_t textLength);
michael@0: 
michael@0:   /**
michael@0:    * Writable-aliasing UChar* constructor.
michael@0:    * The text will be used for the UnicodeString object, but
michael@0:    * it will not be released when the UnicodeString is destroyed.
michael@0:    * This has write-through semantics:
michael@0:    * For as long as the capacity of the buffer is sufficient, write operations
michael@0:    * will directly affect the buffer. When more capacity is necessary, then
michael@0:    * a new buffer will be allocated and the contents copied as with regularly
michael@0:    * constructed strings.
michael@0:    * In an assignment to another UnicodeString, the buffer will be copied.
michael@0:    * The extract(UChar *dst) function detects whether the dst pointer is the same
michael@0:    * as the string buffer itself and will in this case not copy the contents.
michael@0:    *
michael@0:    * @param buffer The characters to alias for the UnicodeString.
michael@0:    * @param buffLength The number of Unicode characters in <code>buffer</code> to alias.
michael@0:    * @param buffCapacity The size of <code>buffer</code> in UChars.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   UnicodeString(UChar *buffer, int32_t buffLength, int32_t buffCapacity);
michael@0: 
michael@0: #if U_CHARSET_IS_UTF8 || !UCONFIG_NO_CONVERSION
michael@0: 
michael@0:   /**
michael@0:    * char* constructor.
michael@0:    * Uses the default converter (and thus depends on the ICU conversion code)
michael@0:    * unless U_CHARSET_IS_UTF8 is set to 1.
michael@0:    *
michael@0:    * For ASCII (really "invariant character") strings it is more efficient to use
michael@0:    * the constructor that takes a US_INV (for its enum EInvariant).
michael@0:    * For ASCII (invariant-character) string literals, see UNICODE_STRING and
michael@0:    * UNICODE_STRING_SIMPLE.
michael@0:    *
michael@0:    * It is recommended to mark this constructor "explicit" by
michael@0:    * <code>-DUNISTR_FROM_STRING_EXPLICIT=explicit</code>
michael@0:    * on the compiler command line or similar.
michael@0:    * @param codepageData an array of bytes, null-terminated,
michael@0:    *                     in the platform's default codepage.
michael@0:    * @stable ICU 2.0
michael@0:    * @see UNICODE_STRING
michael@0:    * @see UNICODE_STRING_SIMPLE
michael@0:    */
michael@0:   UNISTR_FROM_STRING_EXPLICIT UnicodeString(const char *codepageData);
michael@0: 
michael@0:   /**
michael@0:    * char* constructor.
michael@0:    * Uses the default converter (and thus depends on the ICU conversion code)
michael@0:    * unless U_CHARSET_IS_UTF8 is set to 1.
michael@0:    * @param codepageData an array of bytes in the platform's default codepage.
michael@0:    * @param dataLength The number of bytes in <TT>codepageData</TT>.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   UnicodeString(const char *codepageData, int32_t dataLength);
michael@0: 
michael@0: #endif
michael@0: 
michael@0: #if !UCONFIG_NO_CONVERSION
michael@0: 
michael@0:   /**
michael@0:    * char* constructor.
michael@0:    * @param codepageData an array of bytes, null-terminated
michael@0:    * @param codepage the encoding of <TT>codepageData</TT>.  The special
michael@0:    * value 0 for <TT>codepage</TT> indicates that the text is in the
michael@0:    * platform's default codepage.
michael@0:    *
michael@0:    * If <code>codepage</code> is an empty string (<code>""</code>),
michael@0:    * then a simple conversion is performed on the codepage-invariant
michael@0:    * subset ("invariant characters") of the platform encoding. See utypes.h.
michael@0:    * Recommendation: For invariant-character strings use the constructor
michael@0:    * UnicodeString(const char *src, int32_t length, enum EInvariant inv)
michael@0:    * because it avoids object code dependencies of UnicodeString on
michael@0:    * the conversion code.
michael@0:    *
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   UnicodeString(const char *codepageData, const char *codepage);
michael@0: 
michael@0:   /**
michael@0:    * char* constructor.
michael@0:    * @param codepageData an array of bytes.
michael@0:    * @param dataLength The number of bytes in <TT>codepageData</TT>.
michael@0:    * @param codepage the encoding of <TT>codepageData</TT>.  The special
michael@0:    * value 0 for <TT>codepage</TT> indicates that the text is in the
michael@0:    * platform's default codepage.
michael@0:    * If <code>codepage</code> is an empty string (<code>""</code>),
michael@0:    * then a simple conversion is performed on the codepage-invariant
michael@0:    * subset ("invariant characters") of the platform encoding. See utypes.h.
michael@0:    * Recommendation: For invariant-character strings use the constructor
michael@0:    * UnicodeString(const char *src, int32_t length, enum EInvariant inv)
michael@0:    * because it avoids object code dependencies of UnicodeString on
michael@0:    * the conversion code.
michael@0:    *
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   UnicodeString(const char *codepageData, int32_t dataLength, const char *codepage);
michael@0: 
michael@0:   /**
michael@0:    * char * / UConverter constructor.
michael@0:    * This constructor uses an existing UConverter object to
michael@0:    * convert the codepage string to Unicode and construct a UnicodeString
michael@0:    * from that.
michael@0:    *
michael@0:    * The converter is reset at first.
michael@0:    * If the error code indicates a failure before this constructor is called,
michael@0:    * or if an error occurs during conversion or construction,
michael@0:    * then the string will be bogus.
michael@0:    *
michael@0:    * This function avoids the overhead of opening and closing a converter if
michael@0:    * multiple strings are constructed.
michael@0:    *
michael@0:    * @param src input codepage string
michael@0:    * @param srcLength length of the input string, can be -1 for NUL-terminated strings
michael@0:    * @param cnv converter object (ucnv_resetToUnicode() will be called),
michael@0:    *        can be NULL for the default converter
michael@0:    * @param errorCode normal ICU error code
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   UnicodeString(
michael@0:         const char *src, int32_t srcLength,
michael@0:         UConverter *cnv,
michael@0:         UErrorCode &errorCode);
michael@0: 
michael@0: #endif
michael@0: 
michael@0:   /**
michael@0:    * Constructs a Unicode string from an invariant-character char * string.
michael@0:    * About invariant characters see utypes.h.
michael@0:    * This constructor has no runtime dependency on conversion code and is
michael@0:    * therefore recommended over ones taking a charset name string
michael@0:    * (where the empty string "" indicates invariant-character conversion).
michael@0:    *
michael@0:    * Use the macro US_INV as the third, signature-distinguishing parameter.
michael@0:    *
michael@0:    * For example:
michael@0:    * \code
michael@0:    * void fn(const char *s) {
michael@0:    *   UnicodeString ustr(s, -1, US_INV);
michael@0:    *   // use ustr ...
michael@0:    * }
michael@0:    * \endcode
michael@0:    *
michael@0:    * @param src String using only invariant characters.
michael@0:    * @param length Length of src, or -1 if NUL-terminated.
michael@0:    * @param inv Signature-distinguishing paramater, use US_INV.
michael@0:    *
michael@0:    * @see US_INV
michael@0:    * @stable ICU 3.2
michael@0:    */
michael@0:   UnicodeString(const char *src, int32_t length, enum EInvariant inv);
michael@0: 
michael@0: 
michael@0:   /**
michael@0:    * Copy constructor.
michael@0:    * @param that The UnicodeString object to copy.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   UnicodeString(const UnicodeString& that);
michael@0: 
michael@0:   /**
michael@0:    * 'Substring' constructor from tail of source string.
michael@0:    * @param src The UnicodeString object to copy.
michael@0:    * @param srcStart The offset into <tt>src</tt> at which to start copying.
michael@0:    * @stable ICU 2.2
michael@0:    */
michael@0:   UnicodeString(const UnicodeString& src, int32_t srcStart);
michael@0: 
michael@0:   /**
michael@0:    * 'Substring' constructor from subrange of source string.
michael@0:    * @param src The UnicodeString object to copy.
michael@0:    * @param srcStart The offset into <tt>src</tt> at which to start copying.
michael@0:    * @param srcLength The number of characters from <tt>src</tt> to copy.
michael@0:    * @stable ICU 2.2
michael@0:    */
michael@0:   UnicodeString(const UnicodeString& src, int32_t srcStart, int32_t srcLength);
michael@0: 
michael@0:   /**
michael@0:    * Clone this object, an instance of a subclass of Replaceable.
michael@0:    * Clones can be used concurrently in multiple threads.
michael@0:    * If a subclass does not implement clone(), or if an error occurs,
michael@0:    * then NULL is returned.
michael@0:    * The clone functions in all subclasses return a pointer to a Replaceable
michael@0:    * because some compilers do not support covariant (same-as-this)
michael@0:    * return types; cast to the appropriate subclass if necessary.
michael@0:    * The caller must delete the clone.
michael@0:    *
michael@0:    * @return a clone of this object
michael@0:    *
michael@0:    * @see Replaceable::clone
michael@0:    * @see getDynamicClassID
michael@0:    * @stable ICU 2.6
michael@0:    */
michael@0:   virtual Replaceable *clone() const;
michael@0: 
michael@0:   /** Destructor.
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   virtual ~UnicodeString();
michael@0: 
michael@0:   /**
michael@0:    * Create a UnicodeString from a UTF-8 string.
michael@0:    * Illegal input is replaced with U+FFFD. Otherwise, errors result in a bogus string.
michael@0:    * Calls u_strFromUTF8WithSub().
michael@0:    *
michael@0:    * @param utf8 UTF-8 input string.
michael@0:    *             Note that a StringPiece can be implicitly constructed
michael@0:    *             from a std::string or a NUL-terminated const char * string.
michael@0:    * @return A UnicodeString with equivalent UTF-16 contents.
michael@0:    * @see toUTF8
michael@0:    * @see toUTF8String
michael@0:    * @stable ICU 4.2
michael@0:    */
michael@0:   static UnicodeString fromUTF8(const StringPiece &utf8);
michael@0: 
michael@0:   /**
michael@0:    * Create a UnicodeString from a UTF-32 string.
michael@0:    * Illegal input is replaced with U+FFFD. Otherwise, errors result in a bogus string.
michael@0:    * Calls u_strFromUTF32WithSub().
michael@0:    *
michael@0:    * @param utf32 UTF-32 input string. Must not be NULL.
michael@0:    * @param length Length of the input string, or -1 if NUL-terminated.
michael@0:    * @return A UnicodeString with equivalent UTF-16 contents.
michael@0:    * @see toUTF32
michael@0:    * @stable ICU 4.2
michael@0:    */
michael@0:   static UnicodeString fromUTF32(const UChar32 *utf32, int32_t length);
michael@0: 
michael@0:   /* Miscellaneous operations */
michael@0: 
michael@0:   /**
michael@0:    * Unescape a string of characters and return a string containing
michael@0:    * the result.  The following escape sequences are recognized:
michael@0:    *
michael@0:    * \\uhhhh       4 hex digits; h in [0-9A-Fa-f]
michael@0:    * \\Uhhhhhhhh   8 hex digits
michael@0:    * \\xhh         1-2 hex digits
michael@0:    * \\ooo         1-3 octal digits; o in [0-7]
michael@0:    * \\cX          control-X; X is masked with 0x1F
michael@0:    *
michael@0:    * as well as the standard ANSI C escapes:
michael@0:    *
michael@0:    * \\a => U+0007, \\b => U+0008, \\t => U+0009, \\n => U+000A,
michael@0:    * \\v => U+000B, \\f => U+000C, \\r => U+000D, \\e => U+001B,
michael@0:    * \\&quot; => U+0022, \\' => U+0027, \\? => U+003F, \\\\ => U+005C
michael@0:    *
michael@0:    * Anything else following a backslash is generically escaped.  For
michael@0:    * example, "[a\\-z]" returns "[a-z]".
michael@0:    *
michael@0:    * If an escape sequence is ill-formed, this method returns an empty
michael@0:    * string.  An example of an ill-formed sequence is "\\u" followed by
michael@0:    * fewer than 4 hex digits.
michael@0:    *
michael@0:    * This function is similar to u_unescape() but not identical to it.
michael@0:    * The latter takes a source char*, so it does escape recognition
michael@0:    * and also invariant conversion.
michael@0:    *
michael@0:    * @return a string with backslash escapes interpreted, or an
michael@0:    * empty string on error.
michael@0:    * @see UnicodeString#unescapeAt()
michael@0:    * @see u_unescape()
michael@0:    * @see u_unescapeAt()
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   UnicodeString unescape() const;
michael@0: 
michael@0:   /**
michael@0:    * Unescape a single escape sequence and return the represented
michael@0:    * character.  See unescape() for a listing of the recognized escape
michael@0:    * sequences.  The character at offset-1 is assumed (without
michael@0:    * checking) to be a backslash.  If the escape sequence is
michael@0:    * ill-formed, or the offset is out of range, U_SENTINEL=-1 is
michael@0:    * returned.
michael@0:    *
michael@0:    * @param offset an input output parameter.  On input, it is the
michael@0:    * offset into this string where the escape sequence is located,
michael@0:    * after the initial backslash.  On output, it is advanced after the
michael@0:    * last character parsed.  On error, it is not advanced at all.
michael@0:    * @return the character represented by the escape sequence at
michael@0:    * offset, or U_SENTINEL=-1 on error.
michael@0:    * @see UnicodeString#unescape()
michael@0:    * @see u_unescape()
michael@0:    * @see u_unescapeAt()
michael@0:    * @stable ICU 2.0
michael@0:    */
michael@0:   UChar32 unescapeAt(int32_t &offset) const;
michael@0: 
michael@0:   /**
michael@0:    * ICU "poor man's RTTI", returns a UClassID for this class.
michael@0:    *
michael@0:    * @stable ICU 2.2
michael@0:    */
michael@0:   static UClassID U_EXPORT2 getStaticClassID();
michael@0: 
michael@0:   /**
michael@0:    * ICU "poor man's RTTI", returns a UClassID for the actual class.
michael@0:    *
michael@0:    * @stable ICU 2.2
michael@0:    */
michael@0:   virtual UClassID getDynamicClassID() const;
michael@0: 
michael@0:   //========================================
michael@0:   // Implementation methods
michael@0:   //========================================
michael@0: 
michael@0: protected:
michael@0:   /**
michael@0:    * Implement Replaceable::getLength() (see jitterbug 1027).
michael@0:    * @stable ICU 2.4
michael@0:    */
michael@0:   virtual int32_t getLength() const;
michael@0: 
michael@0:   /**
michael@0:    * The change in Replaceable to use virtual getCharAt() allows
michael@0:    * UnicodeString::charAt() to be inline again (see jitterbug 709).
michael@0:    * @stable ICU 2.4
michael@0:    */
michael@0:   virtual UChar getCharAt(int32_t offset) const;
michael@0: 
michael@0:   /**
michael@0:    * The change in Replaceable to use virtual getChar32At() allows
michael@0:    * UnicodeString::char32At() to be inline again (see jitterbug 709).
michael@0:    * @stable ICU 2.4
michael@0:    */
michael@0:   virtual UChar32 getChar32At(int32_t offset) const;
michael@0: 
michael@0: private:
michael@0:   // For char* constructors. Could be made public.
michael@0:   UnicodeString &setToUTF8(const StringPiece &utf8);
michael@0:   // For extract(char*).
michael@0:   // We could make a toUTF8(target, capacity, errorCode) public but not
michael@0:   // this version: New API will be cleaner if we make callers create substrings
michael@0:   // rather than having start+length on every method,
michael@0:   // and it should take a UErrorCode&.
michael@0:   int32_t
michael@0:   toUTF8(int32_t start, int32_t len,
michael@0:          char *target, int32_t capacity) const;
michael@0: 
michael@0:   /**
michael@0:    * Internal string contents comparison, called by operator==.
michael@0:    * Requires: this & text not bogus and have same lengths.
michael@0:    */
michael@0:   UBool doEquals(const UnicodeString &text, int32_t len) const;
michael@0: 
michael@0:   inline int8_t
michael@0:   doCompare(int32_t start,
michael@0:            int32_t length,
michael@0:            const UnicodeString& srcText,
michael@0:            int32_t srcStart,
michael@0:            int32_t srcLength) const;
michael@0: 
michael@0:   int8_t doCompare(int32_t start,
michael@0:            int32_t length,
michael@0:            const UChar *srcChars,
michael@0:            int32_t srcStart,
michael@0:            int32_t srcLength) const;
michael@0: 
michael@0:   inline int8_t
michael@0:   doCompareCodePointOrder(int32_t start,
michael@0:                           int32_t length,
michael@0:                           const UnicodeString& srcText,
michael@0:                           int32_t srcStart,
michael@0:                           int32_t srcLength) const;
michael@0: 
michael@0:   int8_t doCompareCodePointOrder(int32_t start,
michael@0:                                  int32_t length,
michael@0:                                  const UChar *srcChars,
michael@0:                                  int32_t srcStart,
michael@0:                                  int32_t srcLength) const;
michael@0: 
michael@0:   inline int8_t
michael@0:   doCaseCompare(int32_t start,
michael@0:                 int32_t length,
michael@0:                 const UnicodeString &srcText,
michael@0:                 int32_t srcStart,
michael@0:                 int32_t srcLength,
michael@0:                 uint32_t options) const;
michael@0: 
michael@0:   int8_t
michael@0:   doCaseCompare(int32_t start,
michael@0:                 int32_t length,
michael@0:                 const UChar *srcChars,
michael@0:                 int32_t srcStart,
michael@0:                 int32_t srcLength,
michael@0:                 uint32_t options) const;
michael@0: 
michael@0:   int32_t doIndexOf(UChar c,
michael@0:             int32_t start,
michael@0:             int32_t length) const;
michael@0: 
michael@0:   int32_t doIndexOf(UChar32 c,
michael@0:                         int32_t start,
michael@0:                         int32_t length) const;
michael@0: 
michael@0:   int32_t doLastIndexOf(UChar c,
michael@0:                 int32_t start,
michael@0:                 int32_t length) const;
michael@0: 
michael@0:   int32_t doLastIndexOf(UChar32 c,
michael@0:                             int32_t start,
michael@0:                             int32_t length) const;
michael@0: 
michael@0:   void doExtract(int32_t start,
michael@0:          int32_t length,
michael@0:          UChar *dst,
michael@0:          int32_t dstStart) const;
michael@0: 
michael@0:   inline void doExtract(int32_t start,
michael@0:          int32_t length,
michael@0:          UnicodeString& target) const;
michael@0: 
michael@0:   inline UChar doCharAt(int32_t offset)  const;
michael@0: 
michael@0:   UnicodeString& doReplace(int32_t start,
michael@0:                int32_t length,
michael@0:                const UnicodeString& srcText,
michael@0:                int32_t srcStart,
michael@0:                int32_t srcLength);
michael@0: 
michael@0:   UnicodeString& doReplace(int32_t start,
michael@0:                int32_t length,
michael@0:                const UChar *srcChars,
michael@0:                int32_t srcStart,
michael@0:                int32_t srcLength);
michael@0: 
michael@0:   UnicodeString& doReverse(int32_t start,
michael@0:                int32_t length);
michael@0: 
michael@0:   // calculate hash code
michael@0:   int32_t doHashCode(void) const;
michael@0: 
michael@0:   // get pointer to start of array
michael@0:   // these do not check for kOpenGetBuffer, unlike the public getBuffer() function
michael@0:   inline UChar* getArrayStart(void);
michael@0:   inline const UChar* getArrayStart(void) const;
michael@0: 
michael@0:   // A UnicodeString object (not necessarily its current buffer)
michael@0:   // is writable unless it isBogus() or it has an "open" getBuffer(minCapacity).
michael@0:   inline UBool isWritable() const;
michael@0: 
michael@0:   // Is the current buffer writable?
michael@0:   inline UBool isBufferWritable() const;
michael@0: 
michael@0:   // None of the following does releaseArray().
michael@0:   inline void setLength(int32_t len);        // sets only fShortLength and fLength
michael@0:   inline void setToEmpty();                  // sets fFlags=kShortString
michael@0:   inline void setArray(UChar *array, int32_t len, int32_t capacity); // does not set fFlags
michael@0: 
michael@0:   // allocate the array; result may be fStackBuffer
michael@0:   // sets refCount to 1 if appropriate
michael@0:   // sets fArray, fCapacity, and fFlags
michael@0:   // returns boolean for success or failure
michael@0:   UBool allocate(int32_t capacity);
michael@0: 
michael@0:   // release the array if owned
michael@0:   void releaseArray(void);
michael@0: 
michael@0:   // turn a bogus string into an empty one
michael@0:   void unBogus();
michael@0: 
michael@0:   // implements assigment operator, copy constructor, and fastCopyFrom()
michael@0:   UnicodeString &copyFrom(const UnicodeString &src, UBool fastCopy=FALSE);
michael@0: 
michael@0:   // Pin start and limit to acceptable values.
michael@0:   inline void pinIndex(int32_t& start) const;
michael@0:   inline void pinIndices(int32_t& start,
michael@0:                          int32_t& length) const;
michael@0: 
michael@0: #if !UCONFIG_NO_CONVERSION
michael@0: 
michael@0:   /* Internal extract() using UConverter. */
michael@0:   int32_t doExtract(int32_t start, int32_t length,
michael@0:                     char *dest, int32_t destCapacity,
michael@0:                     UConverter *cnv,
michael@0:                     UErrorCode &errorCode) const;
michael@0: 
michael@0:   /*
michael@0:    * Real constructor for converting from codepage data.
michael@0:    * It assumes that it is called with !fRefCounted.
michael@0:    *
michael@0:    * If <code>codepage==0</code>, then the default converter
michael@0:    * is used for the platform encoding.
michael@0:    * If <code>codepage</code> is an empty string (<code>""</code>),
michael@0:    * then a simple conversion is performed on the codepage-invariant
michael@0:    * subset ("invariant characters") of the platform encoding. See utypes.h.
michael@0:    */
michael@0:   void doCodepageCreate(const char *codepageData,
michael@0:                         int32_t dataLength,
michael@0:                         const char *codepage);
michael@0: 
michael@0:   /*
michael@0:    * Worker function for creating a UnicodeString from
michael@0:    * a codepage string using a UConverter.
michael@0:    */
michael@0:   void
michael@0:   doCodepageCreate(const char *codepageData,
michael@0:                    int32_t dataLength,
michael@0:                    UConverter *converter,
michael@0:                    UErrorCode &status);
michael@0: 
michael@0: #endif
michael@0: 
michael@0:   /*
michael@0:    * This function is called when write access to the array
michael@0:    * is necessary.
michael@0:    *
michael@0:    * We need to make a copy of the array if
michael@0:    * the buffer is read-only, or
michael@0:    * the buffer is refCounted (shared), and refCount>1, or
michael@0:    * the buffer is too small.
michael@0:    *
michael@0:    * Return FALSE if memory could not be allocated.
michael@0:    */
michael@0:   UBool cloneArrayIfNeeded(int32_t newCapacity = -1,
michael@0:                             int32_t growCapacity = -1,
michael@0:                             UBool doCopyArray = TRUE,
michael@0:                             int32_t **pBufferToDelete = 0,
michael@0:                             UBool forceClone = FALSE);
michael@0: 
michael@0:   /**
michael@0:    * Common function for UnicodeString case mappings.
michael@0:    * The stringCaseMapper has the same type UStringCaseMapper
michael@0:    * as in ustr_imp.h for ustrcase_map().
michael@0:    */
michael@0:   UnicodeString &
michael@0:   caseMap(const UCaseMap *csm, UStringCaseMapper *stringCaseMapper);
michael@0: 
michael@0:   // ref counting
michael@0:   void addRef(void);
michael@0:   int32_t removeRef(void);
michael@0:   int32_t refCount(void) const;
michael@0: 
michael@0:   // constants
michael@0:   enum {
michael@0:     // Set the stack buffer size so that sizeof(UnicodeString) is,
michael@0:     // naturally (without padding), a multiple of sizeof(pointer).
michael@0:     US_STACKBUF_SIZE= sizeof(void *)==4 ? 13 : 15, // Size of stack buffer for short strings
michael@0:     kInvalidUChar=0xffff, // invalid UChar index
michael@0:     kGrowSize=128, // grow size for this buffer
michael@0:     kInvalidHashCode=0, // invalid hash code
michael@0:     kEmptyHashCode=1, // hash code for empty string
michael@0: 
michael@0:     // bit flag values for fFlags
michael@0:     kIsBogus=1,         // this string is bogus, i.e., not valid or NULL
michael@0:     kUsingStackBuffer=2,// using fUnion.fStackBuffer instead of fUnion.fFields
michael@0:     kRefCounted=4,      // there is a refCount field before the characters in fArray
michael@0:     kBufferIsReadonly=8,// do not write to this buffer
michael@0:     kOpenGetBuffer=16,  // getBuffer(minCapacity) was called (is "open"),
michael@0:                         // and releaseBuffer(newLength) must be called
michael@0: 
michael@0:     // combined values for convenience
michael@0:     kShortString=kUsingStackBuffer,
michael@0:     kLongString=kRefCounted,
michael@0:     kReadonlyAlias=kBufferIsReadonly,
michael@0:     kWritableAlias=0
michael@0:   };
michael@0: 
michael@0:   friend class StringThreadTest;
michael@0:   friend class UnicodeStringAppendable;
michael@0: 
michael@0:   union StackBufferOrFields;        // forward declaration necessary before friend declaration
michael@0:   friend union StackBufferOrFields; // make US_STACKBUF_SIZE visible inside fUnion
michael@0: 
michael@0:   /*
michael@0:    * The following are all the class fields that are stored
michael@0:    * in each UnicodeString object.
michael@0:    * Note that UnicodeString has virtual functions,
michael@0:    * therefore there is an implicit vtable pointer
michael@0:    * as the first real field.
michael@0:    * The fields should be aligned such that no padding is necessary.
michael@0:    * On 32-bit machines, the size should be 32 bytes,
michael@0:    * on 64-bit machines (8-byte pointers), it should be 40 bytes.
michael@0:    *
michael@0:    * We use a hack to achieve this.
michael@0:    *
michael@0:    * With at least some compilers, each of the following is forced to
michael@0:    * a multiple of sizeof(pointer) [the largest field base unit here is a data pointer],
michael@0:    * rounded up with additional padding if the fields do not already fit that requirement:
michael@0:    * - sizeof(class UnicodeString)
michael@0:    * - offsetof(UnicodeString, fUnion)
michael@0:    * - sizeof(fUnion)
michael@0:    * - sizeof(fFields)
michael@0:    *
michael@0:    * In order to avoid padding, we make sizeof(fStackBuffer)=16 (=8 UChars)
michael@0:    * which is at least as large as sizeof(fFields) on 32-bit and 64-bit machines.
michael@0:    * (Padding at the end of fFields is ok:
michael@0:    * As long as there is no padding after fStackBuffer, it is not wasted space.)
michael@0:    *
michael@0:    * We further assume that the compiler does not reorder the fields,
michael@0:    * so that fRestOfStackBuffer (which holds a few more UChars) immediately follows after fUnion,
michael@0:    * with at most some padding (but no other field) in between.
michael@0:    * (Padding there would be wasted space, but functionally harmless.)
michael@0:    *
michael@0:    * We use a few more sizeof(pointer)'s chunks of space with
michael@0:    * fRestOfStackBuffer, fShortLength and fFlags,
michael@0:    * to get up exactly to the intended sizeof(UnicodeString).
michael@0:    */
michael@0:   // (implicit) *vtable;
michael@0:   union StackBufferOrFields {
michael@0:     // fStackBuffer is used iff (fFlags&kUsingStackBuffer)
michael@0:     // else fFields is used
michael@0:     UChar fStackBuffer[8];  // buffer for short strings, together with fRestOfStackBuffer
michael@0:     struct {
michael@0:       UChar   *fArray;    // the Unicode data
michael@0:       int32_t fCapacity;  // capacity of fArray (in UChars)
michael@0:       int32_t fLength;    // number of characters in fArray if >127; else undefined
michael@0:     } fFields;
michael@0:   } fUnion;
michael@0:   UChar fRestOfStackBuffer[US_STACKBUF_SIZE-8];
michael@0:   int8_t fShortLength;  // 0..127: length  <0: real length is in fUnion.fFields.fLength
michael@0:   uint8_t fFlags;       // bit flags: see constants above
michael@0: };
michael@0: 
michael@0: /**
michael@0:  * Create a new UnicodeString with the concatenation of two others.
michael@0:  *
michael@0:  * @param s1 The first string to be copied to the new one.
michael@0:  * @param s2 The second string to be copied to the new one, after s1.
michael@0:  * @return UnicodeString(s1).append(s2)
michael@0:  * @stable ICU 2.8
michael@0:  */
michael@0: U_COMMON_API UnicodeString U_EXPORT2
michael@0: operator+ (const UnicodeString &s1, const UnicodeString &s2);
michael@0: 
michael@0: //========================================
michael@0: // Inline members
michael@0: //========================================
michael@0: 
michael@0: //========================================
michael@0: // Privates
michael@0: //========================================
michael@0: 
michael@0: inline void
michael@0: UnicodeString::pinIndex(int32_t& start) const
michael@0: {
michael@0:   // pin index
michael@0:   if(start < 0) {
michael@0:     start = 0;
michael@0:   } else if(start > length()) {
michael@0:     start = length();
michael@0:   }
michael@0: }
michael@0: 
michael@0: inline void
michael@0: UnicodeString::pinIndices(int32_t& start,
michael@0:                           int32_t& _length) const
michael@0: {
michael@0:   // pin indices
michael@0:   int32_t len = length();
michael@0:   if(start < 0) {
michael@0:     start = 0;
michael@0:   } else if(start > len) {
michael@0:     start = len;
michael@0:   }
michael@0:   if(_length < 0) {
michael@0:     _length = 0;
michael@0:   } else if(_length > (len - start)) {
michael@0:     _length = (len - start);
michael@0:   }
michael@0: }
michael@0: 
michael@0: inline UChar*
michael@0: UnicodeString::getArrayStart()
michael@0: { return (fFlags&kUsingStackBuffer) ? fUnion.fStackBuffer : fUnion.fFields.fArray; }
michael@0: 
michael@0: inline const UChar*
michael@0: UnicodeString::getArrayStart() const
michael@0: { return (fFlags&kUsingStackBuffer) ? fUnion.fStackBuffer : fUnion.fFields.fArray; }
michael@0: 
michael@0: //========================================
michael@0: // Default constructor
michael@0: //========================================
michael@0: 
michael@0: inline
michael@0: UnicodeString::UnicodeString()
michael@0:   : fShortLength(0),
michael@0:     fFlags(kShortString)
michael@0: {}
michael@0: 
michael@0: //========================================
michael@0: // Read-only implementation methods
michael@0: //========================================
michael@0: inline int32_t
michael@0: UnicodeString::length() const
michael@0: { return fShortLength>=0 ? fShortLength : fUnion.fFields.fLength; }
michael@0: 
michael@0: inline int32_t
michael@0: UnicodeString::getCapacity() const
michael@0: { return (fFlags&kUsingStackBuffer) ? US_STACKBUF_SIZE : fUnion.fFields.fCapacity; }
michael@0: 
michael@0: inline int32_t
michael@0: UnicodeString::hashCode() const
michael@0: { return doHashCode(); }
michael@0: 
michael@0: inline UBool
michael@0: UnicodeString::isBogus() const
michael@0: { return (UBool)(fFlags & kIsBogus); }
michael@0: 
michael@0: inline UBool
michael@0: UnicodeString::isWritable() const
michael@0: { return (UBool)!(fFlags&(kOpenGetBuffer|kIsBogus)); }
michael@0: 
michael@0: inline UBool
michael@0: UnicodeString::isBufferWritable() const
michael@0: {
michael@0:   return (UBool)(
michael@0:       !(fFlags&(kOpenGetBuffer|kIsBogus|kBufferIsReadonly)) &&
michael@0:       (!(fFlags&kRefCounted) || refCount()==1));
michael@0: }
michael@0: 
michael@0: inline const UChar *
michael@0: UnicodeString::getBuffer() const {
michael@0:   if(fFlags&(kIsBogus|kOpenGetBuffer)) {
michael@0:     return 0;
michael@0:   } else if(fFlags&kUsingStackBuffer) {
michael@0:     return fUnion.fStackBuffer;
michael@0:   } else {
michael@0:     return fUnion.fFields.fArray;
michael@0:   }
michael@0: }
michael@0: 
michael@0: //========================================
michael@0: // Read-only alias methods
michael@0: //========================================
michael@0: inline int8_t
michael@0: UnicodeString::doCompare(int32_t start,
michael@0:               int32_t thisLength,
michael@0:               const UnicodeString& srcText,
michael@0:               int32_t srcStart,
michael@0:               int32_t srcLength) const
michael@0: {
michael@0:   if(srcText.isBogus()) {
michael@0:     return (int8_t)!isBogus(); // 0 if both are bogus, 1 otherwise
michael@0:   } else {
michael@0:     srcText.pinIndices(srcStart, srcLength);
michael@0:     return doCompare(start, thisLength, srcText.getArrayStart(), srcStart, srcLength);
michael@0:   }
michael@0: }
michael@0: 
michael@0: inline UBool
michael@0: UnicodeString::operator== (const UnicodeString& text) const
michael@0: {
michael@0:   if(isBogus()) {
michael@0:     return text.isBogus();
michael@0:   } else {
michael@0:     int32_t len = length(), textLength = text.length();
michael@0:     return !text.isBogus() && len == textLength && doEquals(text, len);
michael@0:   }
michael@0: }
michael@0: 
michael@0: inline UBool
michael@0: UnicodeString::operator!= (const UnicodeString& text) const
michael@0: { return (! operator==(text)); }
michael@0: 
michael@0: inline UBool
michael@0: UnicodeString::operator> (const UnicodeString& text) const
michael@0: { return doCompare(0, length(), text, 0, text.length()) == 1; }
michael@0: 
michael@0: inline UBool
michael@0: UnicodeString::operator< (const UnicodeString& text) const
michael@0: { return doCompare(0, length(), text, 0, text.length()) == -1; }
michael@0: 
michael@0: inline UBool
michael@0: UnicodeString::operator>= (const UnicodeString& text) const
michael@0: { return doCompare(0, length(), text, 0, text.length()) != -1; }
michael@0: 
michael@0: inline UBool
michael@0: UnicodeString::operator<= (const UnicodeString& text) const
michael@0: { return doCompare(0, length(), text, 0, text.length()) != 1; }
michael@0: 
michael@0: inline int8_t
michael@0: UnicodeString::compare(const UnicodeString& text) const
michael@0: { return doCompare(0, length(), text, 0, text.length()); }
michael@0: 
michael@0: inline int8_t
michael@0: UnicodeString::compare(int32_t start,
michael@0:                int32_t _length,
michael@0:                const UnicodeString& srcText) const
michael@0: { return doCompare(start, _length, srcText, 0, srcText.length()); }
michael@0: 
michael@0: inline int8_t
michael@0: UnicodeString::compare(const UChar *srcChars,
michael@0:                int32_t srcLength) const
michael@0: { return doCompare(0, length(), srcChars, 0, srcLength); }
michael@0: 
michael@0: inline int8_t
michael@0: UnicodeString::compare(int32_t start,
michael@0:                int32_t _length,
michael@0:                const UnicodeString& srcText,
michael@0:                int32_t srcStart,
michael@0:                int32_t srcLength) const
michael@0: { return doCompare(start, _length, srcText, srcStart, srcLength); }
michael@0: 
michael@0: inline int8_t
michael@0: UnicodeString::compare(int32_t start,
michael@0:                int32_t _length,
michael@0:                const UChar *srcChars) const
michael@0: { return doCompare(start, _length, srcChars, 0, _length); }
michael@0: 
michael@0: inline int8_t
michael@0: UnicodeString::compare(int32_t start,
michael@0:                int32_t _length,
michael@0:                const UChar *srcChars,
michael@0:                int32_t srcStart,
michael@0:                int32_t srcLength) const
michael@0: { return doCompare(start, _length, srcChars, srcStart, srcLength); }
michael@0: 
michael@0: inline int8_t
michael@0: UnicodeString::compareBetween(int32_t start,
michael@0:                   int32_t limit,
michael@0:                   const UnicodeString& srcText,
michael@0:                   int32_t srcStart,
michael@0:                   int32_t srcLimit) const
michael@0: { return doCompare(start, limit - start,
michael@0:            srcText, srcStart, srcLimit - srcStart); }
michael@0: 
michael@0: inline int8_t
michael@0: UnicodeString::doCompareCodePointOrder(int32_t start,
michael@0:                                        int32_t thisLength,
michael@0:                                        const UnicodeString& srcText,
michael@0:                                        int32_t srcStart,
michael@0:                                        int32_t srcLength) const
michael@0: {
michael@0:   if(srcText.isBogus()) {
michael@0:     return (int8_t)!isBogus(); // 0 if both are bogus, 1 otherwise
michael@0:   } else {
michael@0:     srcText.pinIndices(srcStart, srcLength);
michael@0:     return doCompareCodePointOrder(start, thisLength, srcText.getArrayStart(), srcStart, srcLength);
michael@0:   }
michael@0: }
michael@0: 
michael@0: inline int8_t
michael@0: UnicodeString::compareCodePointOrder(const UnicodeString& text) const
michael@0: { return doCompareCodePointOrder(0, length(), text, 0, text.length()); }
michael@0: 
michael@0: inline int8_t
michael@0: UnicodeString::compareCodePointOrder(int32_t start,
michael@0:                                      int32_t _length,
michael@0:                                      const UnicodeString& srcText) const
michael@0: { return doCompareCodePointOrder(start, _length, srcText, 0, srcText.length()); }
michael@0: 
michael@0: inline int8_t
michael@0: UnicodeString::compareCodePointOrder(const UChar *srcChars,
michael@0:                                      int32_t srcLength) const
michael@0: { return doCompareCodePointOrder(0, length(), srcChars, 0, srcLength); }
michael@0: 
michael@0: inline int8_t
michael@0: UnicodeString::compareCodePointOrder(int32_t start,
michael@0:                                      int32_t _length,
michael@0:                                      const UnicodeString& srcText,
michael@0:                                      int32_t srcStart,
michael@0:                                      int32_t srcLength) const
michael@0: { return doCompareCodePointOrder(start, _length, srcText, srcStart, srcLength); }
michael@0: 
michael@0: inline int8_t
michael@0: UnicodeString::compareCodePointOrder(int32_t start,
michael@0:                                      int32_t _length,
michael@0:                                      const UChar *srcChars) const
michael@0: { return doCompareCodePointOrder(start, _length, srcChars, 0, _length); }
michael@0: 
michael@0: inline int8_t
michael@0: UnicodeString::compareCodePointOrder(int32_t start,
michael@0:                                      int32_t _length,
michael@0:                                      const UChar *srcChars,
michael@0:                                      int32_t srcStart,
michael@0:                                      int32_t srcLength) const
michael@0: { return doCompareCodePointOrder(start, _length, srcChars, srcStart, srcLength); }
michael@0: 
michael@0: inline int8_t
michael@0: UnicodeString::compareCodePointOrderBetween(int32_t start,
michael@0:                                             int32_t limit,
michael@0:                                             const UnicodeString& srcText,
michael@0:                                             int32_t srcStart,
michael@0:                                             int32_t srcLimit) const
michael@0: { return doCompareCodePointOrder(start, limit - start,
michael@0:            srcText, srcStart, srcLimit - srcStart); }
michael@0: 
michael@0: inline int8_t
michael@0: UnicodeString::doCaseCompare(int32_t start,
michael@0:                              int32_t thisLength,
michael@0:                              const UnicodeString &srcText,
michael@0:                              int32_t srcStart,
michael@0:                              int32_t srcLength,
michael@0:                              uint32_t options) const
michael@0: {
michael@0:   if(srcText.isBogus()) {
michael@0:     return (int8_t)!isBogus(); // 0 if both are bogus, 1 otherwise
michael@0:   } else {
michael@0:     srcText.pinIndices(srcStart, srcLength);
michael@0:     return doCaseCompare(start, thisLength, srcText.getArrayStart(), srcStart, srcLength, options);
michael@0:   }
michael@0: }
michael@0: 
michael@0: inline int8_t
michael@0: UnicodeString::caseCompare(const UnicodeString &text, uint32_t options) const {
michael@0:   return doCaseCompare(0, length(), text, 0, text.length(), options);
michael@0: }
michael@0: 
michael@0: inline int8_t
michael@0: UnicodeString::caseCompare(int32_t start,
michael@0:                            int32_t _length,
michael@0:                            const UnicodeString &srcText,
michael@0:                            uint32_t options) const {
michael@0:   return doCaseCompare(start, _length, srcText, 0, srcText.length(), options);
michael@0: }
michael@0: 
michael@0: inline int8_t
michael@0: UnicodeString::caseCompare(const UChar *srcChars,
michael@0:                            int32_t srcLength,
michael@0:                            uint32_t options) const {
michael@0:   return doCaseCompare(0, length(), srcChars, 0, srcLength, options);
michael@0: }
michael@0: 
michael@0: inline int8_t
michael@0: UnicodeString::caseCompare(int32_t start,
michael@0:                            int32_t _length,
michael@0:                            const UnicodeString &srcText,
michael@0:                            int32_t srcStart,
michael@0:                            int32_t srcLength,
michael@0:                            uint32_t options) const {
michael@0:   return doCaseCompare(start, _length, srcText, srcStart, srcLength, options);
michael@0: }
michael@0: 
michael@0: inline int8_t
michael@0: UnicodeString::caseCompare(int32_t start,
michael@0:                            int32_t _length,
michael@0:                            const UChar *srcChars,
michael@0:                            uint32_t options) const {
michael@0:   return doCaseCompare(start, _length, srcChars, 0, _length, options);
michael@0: }
michael@0: 
michael@0: inline int8_t
michael@0: UnicodeString::caseCompare(int32_t start,
michael@0:                            int32_t _length,
michael@0:                            const UChar *srcChars,
michael@0:                            int32_t srcStart,
michael@0:                            int32_t srcLength,
michael@0:                            uint32_t options) const {
michael@0:   return doCaseCompare(start, _length, srcChars, srcStart, srcLength, options);
michael@0: }
michael@0: 
michael@0: inline int8_t
michael@0: UnicodeString::caseCompareBetween(int32_t start,
michael@0:                                   int32_t limit,
michael@0:                                   const UnicodeString &srcText,
michael@0:                                   int32_t srcStart,
michael@0:                                   int32_t srcLimit,
michael@0:                                   uint32_t options) const {
michael@0:   return doCaseCompare(start, limit - start, srcText, srcStart, srcLimit - srcStart, options);
michael@0: }
michael@0: 
michael@0: inline int32_t
michael@0: UnicodeString::indexOf(const UnicodeString& srcText,
michael@0:                int32_t srcStart,
michael@0:                int32_t srcLength,
michael@0:                int32_t start,
michael@0:                int32_t _length) const
michael@0: {
michael@0:   if(!srcText.isBogus()) {
michael@0:     srcText.pinIndices(srcStart, srcLength);
michael@0:     if(srcLength > 0) {
michael@0:       return indexOf(srcText.getArrayStart(), srcStart, srcLength, start, _length);
michael@0:     }
michael@0:   }
michael@0:   return -1;
michael@0: }
michael@0: 
michael@0: inline int32_t
michael@0: UnicodeString::indexOf(const UnicodeString& text) const
michael@0: { return indexOf(text, 0, text.length(), 0, length()); }
michael@0: 
michael@0: inline int32_t
michael@0: UnicodeString::indexOf(const UnicodeString& text,
michael@0:                int32_t start) const {
michael@0:   pinIndex(start);
michael@0:   return indexOf(text, 0, text.length(), start, length() - start);
michael@0: }
michael@0: 
michael@0: inline int32_t
michael@0: UnicodeString::indexOf(const UnicodeString& text,
michael@0:                int32_t start,
michael@0:                int32_t _length) const
michael@0: { return indexOf(text, 0, text.length(), start, _length); }
michael@0: 
michael@0: inline int32_t
michael@0: UnicodeString::indexOf(const UChar *srcChars,
michael@0:                int32_t srcLength,
michael@0:                int32_t start) const {
michael@0:   pinIndex(start);
michael@0:   return indexOf(srcChars, 0, srcLength, start, length() - start);
michael@0: }
michael@0: 
michael@0: inline int32_t
michael@0: UnicodeString::indexOf(const UChar *srcChars,
michael@0:                int32_t srcLength,
michael@0:                int32_t start,
michael@0:                int32_t _length) const
michael@0: { return indexOf(srcChars, 0, srcLength, start, _length); }
michael@0: 
michael@0: inline int32_t
michael@0: UnicodeString::indexOf(UChar c,
michael@0:                int32_t start,
michael@0:                int32_t _length) const
michael@0: { return doIndexOf(c, start, _length); }
michael@0: 
michael@0: inline int32_t
michael@0: UnicodeString::indexOf(UChar32 c,
michael@0:                int32_t start,
michael@0:                int32_t _length) const
michael@0: { return doIndexOf(c, start, _length); }
michael@0: 
michael@0: inline int32_t
michael@0: UnicodeString::indexOf(UChar c) const
michael@0: { return doIndexOf(c, 0, length()); }
michael@0: 
michael@0: inline int32_t
michael@0: UnicodeString::indexOf(UChar32 c) const
michael@0: { return indexOf(c, 0, length()); }
michael@0: 
michael@0: inline int32_t
michael@0: UnicodeString::indexOf(UChar c,
michael@0:                int32_t start) const {
michael@0:   pinIndex(start);
michael@0:   return doIndexOf(c, start, length() - start);
michael@0: }
michael@0: 
michael@0: inline int32_t
michael@0: UnicodeString::indexOf(UChar32 c,
michael@0:                int32_t start) const {
michael@0:   pinIndex(start);
michael@0:   return indexOf(c, start, length() - start);
michael@0: }
michael@0: 
michael@0: inline int32_t
michael@0: UnicodeString::lastIndexOf(const UChar *srcChars,
michael@0:                int32_t srcLength,
michael@0:                int32_t start,
michael@0:                int32_t _length) const
michael@0: { return lastIndexOf(srcChars, 0, srcLength, start, _length); }
michael@0: 
michael@0: inline int32_t
michael@0: UnicodeString::lastIndexOf(const UChar *srcChars,
michael@0:                int32_t srcLength,
michael@0:                int32_t start) const {
michael@0:   pinIndex(start);
michael@0:   return lastIndexOf(srcChars, 0, srcLength, start, length() - start);
michael@0: }
michael@0: 
michael@0: inline int32_t
michael@0: UnicodeString::lastIndexOf(const UnicodeString& srcText,
michael@0:                int32_t srcStart,
michael@0:                int32_t srcLength,
michael@0:                int32_t start,
michael@0:                int32_t _length) const
michael@0: {
michael@0:   if(!srcText.isBogus()) {
michael@0:     srcText.pinIndices(srcStart, srcLength);
michael@0:     if(srcLength > 0) {
michael@0:       return lastIndexOf(srcText.getArrayStart(), srcStart, srcLength, start, _length);
michael@0:     }
michael@0:   }
michael@0:   return -1;
michael@0: }
michael@0: 
michael@0: inline int32_t
michael@0: UnicodeString::lastIndexOf(const UnicodeString& text,
michael@0:                int32_t start,
michael@0:                int32_t _length) const
michael@0: { return lastIndexOf(text, 0, text.length(), start, _length); }
michael@0: 
michael@0: inline int32_t
michael@0: UnicodeString::lastIndexOf(const UnicodeString& text,
michael@0:                int32_t start) const {
michael@0:   pinIndex(start);
michael@0:   return lastIndexOf(text, 0, text.length(), start, length() - start);
michael@0: }
michael@0: 
michael@0: inline int32_t
michael@0: UnicodeString::lastIndexOf(const UnicodeString& text) const
michael@0: { return lastIndexOf(text, 0, text.length(), 0, length()); }
michael@0: 
michael@0: inline int32_t
michael@0: UnicodeString::lastIndexOf(UChar c,
michael@0:                int32_t start,
michael@0:                int32_t _length) const
michael@0: { return doLastIndexOf(c, start, _length); }
michael@0: 
michael@0: inline int32_t
michael@0: UnicodeString::lastIndexOf(UChar32 c,
michael@0:                int32_t start,
michael@0:                int32_t _length) const {
michael@0:   return doLastIndexOf(c, start, _length);
michael@0: }
michael@0: 
michael@0: inline int32_t
michael@0: UnicodeString::lastIndexOf(UChar c) const
michael@0: { return doLastIndexOf(c, 0, length()); }
michael@0: 
michael@0: inline int32_t
michael@0: UnicodeString::lastIndexOf(UChar32 c) const {
michael@0:   return lastIndexOf(c, 0, length());
michael@0: }
michael@0: 
michael@0: inline int32_t
michael@0: UnicodeString::lastIndexOf(UChar c,
michael@0:                int32_t start) const {
michael@0:   pinIndex(start);
michael@0:   return doLastIndexOf(c, start, length() - start);
michael@0: }
michael@0: 
michael@0: inline int32_t
michael@0: UnicodeString::lastIndexOf(UChar32 c,
michael@0:                int32_t start) const {
michael@0:   pinIndex(start);
michael@0:   return lastIndexOf(c, start, length() - start);
michael@0: }
michael@0: 
michael@0: inline UBool
michael@0: UnicodeString::startsWith(const UnicodeString& text) const
michael@0: { return compare(0, text.length(), text, 0, text.length()) == 0; }
michael@0: 
michael@0: inline UBool
michael@0: UnicodeString::startsWith(const UnicodeString& srcText,
michael@0:               int32_t srcStart,
michael@0:               int32_t srcLength) const
michael@0: { return doCompare(0, srcLength, srcText, srcStart, srcLength) == 0; }
michael@0: 
michael@0: inline UBool
michael@0: UnicodeString::startsWith(const UChar *srcChars, int32_t srcLength) const {
michael@0:   if(srcLength < 0) {
michael@0:     srcLength = u_strlen(srcChars);
michael@0:   }
michael@0:   return doCompare(0, srcLength, srcChars, 0, srcLength) == 0;
michael@0: }
michael@0: 
michael@0: inline UBool
michael@0: UnicodeString::startsWith(const UChar *srcChars, int32_t srcStart, int32_t srcLength) const {
michael@0:   if(srcLength < 0) {
michael@0:     srcLength = u_strlen(srcChars);
michael@0:   }
michael@0:   return doCompare(0, srcLength, srcChars, srcStart, srcLength) == 0;
michael@0: }
michael@0: 
michael@0: inline UBool
michael@0: UnicodeString::endsWith(const UnicodeString& text) const
michael@0: { return doCompare(length() - text.length(), text.length(),
michael@0:            text, 0, text.length()) == 0; }
michael@0: 
michael@0: inline UBool
michael@0: UnicodeString::endsWith(const UnicodeString& srcText,
michael@0:             int32_t srcStart,
michael@0:             int32_t srcLength) const {
michael@0:   srcText.pinIndices(srcStart, srcLength);
michael@0:   return doCompare(length() - srcLength, srcLength,
michael@0:                    srcText, srcStart, srcLength) == 0;
michael@0: }
michael@0: 
michael@0: inline UBool
michael@0: UnicodeString::endsWith(const UChar *srcChars,
michael@0:             int32_t srcLength) const {
michael@0:   if(srcLength < 0) {
michael@0:     srcLength = u_strlen(srcChars);
michael@0:   }
michael@0:   return doCompare(length() - srcLength, srcLength,
michael@0:                    srcChars, 0, srcLength) == 0;
michael@0: }
michael@0: 
michael@0: inline UBool
michael@0: UnicodeString::endsWith(const UChar *srcChars,
michael@0:             int32_t srcStart,
michael@0:             int32_t srcLength) const {
michael@0:   if(srcLength < 0) {
michael@0:     srcLength = u_strlen(srcChars + srcStart);
michael@0:   }
michael@0:   return doCompare(length() - srcLength, srcLength,
michael@0:                    srcChars, srcStart, srcLength) == 0;
michael@0: }
michael@0: 
michael@0: //========================================
michael@0: // replace
michael@0: //========================================
michael@0: inline UnicodeString&
michael@0: UnicodeString::replace(int32_t start,
michael@0:                int32_t _length,
michael@0:                const UnicodeString& srcText)
michael@0: { return doReplace(start, _length, srcText, 0, srcText.length()); }
michael@0: 
michael@0: inline UnicodeString&
michael@0: UnicodeString::replace(int32_t start,
michael@0:                int32_t _length,
michael@0:                const UnicodeString& srcText,
michael@0:                int32_t srcStart,
michael@0:                int32_t srcLength)
michael@0: { return doReplace(start, _length, srcText, srcStart, srcLength); }
michael@0: 
michael@0: inline UnicodeString&
michael@0: UnicodeString::replace(int32_t start,
michael@0:                int32_t _length,
michael@0:                const UChar *srcChars,
michael@0:                int32_t srcLength)
michael@0: { return doReplace(start, _length, srcChars, 0, srcLength); }
michael@0: 
michael@0: inline UnicodeString&
michael@0: UnicodeString::replace(int32_t start,
michael@0:                int32_t _length,
michael@0:                const UChar *srcChars,
michael@0:                int32_t srcStart,
michael@0:                int32_t srcLength)
michael@0: { return doReplace(start, _length, srcChars, srcStart, srcLength); }
michael@0: 
michael@0: inline UnicodeString&
michael@0: UnicodeString::replace(int32_t start,
michael@0:                int32_t _length,
michael@0:                UChar srcChar)
michael@0: { return doReplace(start, _length, &srcChar, 0, 1); }
michael@0: 
michael@0: inline UnicodeString&
michael@0: UnicodeString::replaceBetween(int32_t start,
michael@0:                   int32_t limit,
michael@0:                   const UnicodeString& srcText)
michael@0: { return doReplace(start, limit - start, srcText, 0, srcText.length()); }
michael@0: 
michael@0: inline UnicodeString&
michael@0: UnicodeString::replaceBetween(int32_t start,
michael@0:                   int32_t limit,
michael@0:                   const UnicodeString& srcText,
michael@0:                   int32_t srcStart,
michael@0:                   int32_t srcLimit)
michael@0: { return doReplace(start, limit - start, srcText, srcStart, srcLimit - srcStart); }
michael@0: 
michael@0: inline UnicodeString&
michael@0: UnicodeString::findAndReplace(const UnicodeString& oldText,
michael@0:                   const UnicodeString& newText)
michael@0: { return findAndReplace(0, length(), oldText, 0, oldText.length(),
michael@0:             newText, 0, newText.length()); }
michael@0: 
michael@0: inline UnicodeString&
michael@0: UnicodeString::findAndReplace(int32_t start,
michael@0:                   int32_t _length,
michael@0:                   const UnicodeString& oldText,
michael@0:                   const UnicodeString& newText)
michael@0: { return findAndReplace(start, _length, oldText, 0, oldText.length(),
michael@0:             newText, 0, newText.length()); }
michael@0: 
michael@0: // ============================
michael@0: // extract
michael@0: // ============================
michael@0: inline void
michael@0: UnicodeString::doExtract(int32_t start,
michael@0:              int32_t _length,
michael@0:              UnicodeString& target) const
michael@0: { target.replace(0, target.length(), *this, start, _length); }
michael@0: 
michael@0: inline void
michael@0: UnicodeString::extract(int32_t start,
michael@0:                int32_t _length,
michael@0:                UChar *target,
michael@0:                int32_t targetStart) const
michael@0: { doExtract(start, _length, target, targetStart); }
michael@0: 
michael@0: inline void
michael@0: UnicodeString::extract(int32_t start,
michael@0:                int32_t _length,
michael@0:                UnicodeString& target) const
michael@0: { doExtract(start, _length, target); }
michael@0: 
michael@0: #if !UCONFIG_NO_CONVERSION
michael@0: 
michael@0: inline int32_t
michael@0: UnicodeString::extract(int32_t start,
michael@0:                int32_t _length,
michael@0:                char *dst,
michael@0:                const char *codepage) const
michael@0: 
michael@0: {
michael@0:   // This dstSize value will be checked explicitly
michael@0:   return extract(start, _length, dst, dst!=0 ? 0xffffffff : 0, codepage);
michael@0: }
michael@0: 
michael@0: #endif
michael@0: 
michael@0: inline void
michael@0: UnicodeString::extractBetween(int32_t start,
michael@0:                   int32_t limit,
michael@0:                   UChar *dst,
michael@0:                   int32_t dstStart) const {
michael@0:   pinIndex(start);
michael@0:   pinIndex(limit);
michael@0:   doExtract(start, limit - start, dst, dstStart);
michael@0: }
michael@0: 
michael@0: inline UnicodeString
michael@0: UnicodeString::tempSubStringBetween(int32_t start, int32_t limit) const {
michael@0:     return tempSubString(start, limit - start);
michael@0: }
michael@0: 
michael@0: inline UChar
michael@0: UnicodeString::doCharAt(int32_t offset) const
michael@0: {
michael@0:   if((uint32_t)offset < (uint32_t)length()) {
michael@0:     return getArrayStart()[offset];
michael@0:   } else {
michael@0:     return kInvalidUChar;
michael@0:   }
michael@0: }
michael@0: 
michael@0: inline UChar
michael@0: UnicodeString::charAt(int32_t offset) const
michael@0: { return doCharAt(offset); }
michael@0: 
michael@0: inline UChar
michael@0: UnicodeString::operator[] (int32_t offset) const
michael@0: { return doCharAt(offset); }
michael@0: 
michael@0: inline UBool
michael@0: UnicodeString::isEmpty() const {
michael@0:   return fShortLength == 0;
michael@0: }
michael@0: 
michael@0: //========================================
michael@0: // Write implementation methods
michael@0: //========================================
michael@0: inline void
michael@0: UnicodeString::setLength(int32_t len) {
michael@0:   if(len <= 127) {
michael@0:     fShortLength = (int8_t)len;
michael@0:   } else {
michael@0:     fShortLength = (int8_t)-1;
michael@0:     fUnion.fFields.fLength = len;
michael@0:   }
michael@0: }
michael@0: 
michael@0: inline void
michael@0: UnicodeString::setToEmpty() {
michael@0:   fShortLength = 0;
michael@0:   fFlags = kShortString;
michael@0: }
michael@0: 
michael@0: inline void
michael@0: UnicodeString::setArray(UChar *array, int32_t len, int32_t capacity) {
michael@0:   setLength(len);
michael@0:   fUnion.fFields.fArray = array;
michael@0:   fUnion.fFields.fCapacity = capacity;
michael@0: }
michael@0: 
michael@0: inline UnicodeString&
michael@0: UnicodeString::operator= (UChar ch)
michael@0: { return doReplace(0, length(), &ch, 0, 1); }
michael@0: 
michael@0: inline UnicodeString&
michael@0: UnicodeString::operator= (UChar32 ch)
michael@0: { return replace(0, length(), ch); }
michael@0: 
michael@0: inline UnicodeString&
michael@0: UnicodeString::setTo(const UnicodeString& srcText,
michael@0:              int32_t srcStart,
michael@0:              int32_t srcLength)
michael@0: {
michael@0:   unBogus();
michael@0:   return doReplace(0, length(), srcText, srcStart, srcLength);
michael@0: }
michael@0: 
michael@0: inline UnicodeString&
michael@0: UnicodeString::setTo(const UnicodeString& srcText,
michael@0:              int32_t srcStart)
michael@0: {
michael@0:   unBogus();
michael@0:   srcText.pinIndex(srcStart);
michael@0:   return doReplace(0, length(), srcText, srcStart, srcText.length() - srcStart);
michael@0: }
michael@0: 
michael@0: inline UnicodeString&
michael@0: UnicodeString::setTo(const UnicodeString& srcText)
michael@0: {
michael@0:   return copyFrom(srcText);
michael@0: }
michael@0: 
michael@0: inline UnicodeString&
michael@0: UnicodeString::setTo(const UChar *srcChars,
michael@0:              int32_t srcLength)
michael@0: {
michael@0:   unBogus();
michael@0:   return doReplace(0, length(), srcChars, 0, srcLength);
michael@0: }
michael@0: 
michael@0: inline UnicodeString&
michael@0: UnicodeString::setTo(UChar srcChar)
michael@0: {
michael@0:   unBogus();
michael@0:   return doReplace(0, length(), &srcChar, 0, 1);
michael@0: }
michael@0: 
michael@0: inline UnicodeString&
michael@0: UnicodeString::setTo(UChar32 srcChar)
michael@0: {
michael@0:   unBogus();
michael@0:   return replace(0, length(), srcChar);
michael@0: }
michael@0: 
michael@0: inline UnicodeString&
michael@0: UnicodeString::append(const UnicodeString& srcText,
michael@0:               int32_t srcStart,
michael@0:               int32_t srcLength)
michael@0: { return doReplace(length(), 0, srcText, srcStart, srcLength); }
michael@0: 
michael@0: inline UnicodeString&
michael@0: UnicodeString::append(const UnicodeString& srcText)
michael@0: { return doReplace(length(), 0, srcText, 0, srcText.length()); }
michael@0: 
michael@0: inline UnicodeString&
michael@0: UnicodeString::append(const UChar *srcChars,
michael@0:               int32_t srcStart,
michael@0:               int32_t srcLength)
michael@0: { return doReplace(length(), 0, srcChars, srcStart, srcLength); }
michael@0: 
michael@0: inline UnicodeString&
michael@0: UnicodeString::append(const UChar *srcChars,
michael@0:               int32_t srcLength)
michael@0: { return doReplace(length(), 0, srcChars, 0, srcLength); }
michael@0: 
michael@0: inline UnicodeString&
michael@0: UnicodeString::append(UChar srcChar)
michael@0: { return doReplace(length(), 0, &srcChar, 0, 1); }
michael@0: 
michael@0: inline UnicodeString&
michael@0: UnicodeString::operator+= (UChar ch)
michael@0: { return doReplace(length(), 0, &ch, 0, 1); }
michael@0: 
michael@0: inline UnicodeString&
michael@0: UnicodeString::operator+= (UChar32 ch) {
michael@0:   return append(ch);
michael@0: }
michael@0: 
michael@0: inline UnicodeString&
michael@0: UnicodeString::operator+= (const UnicodeString& srcText)
michael@0: { return doReplace(length(), 0, srcText, 0, srcText.length()); }
michael@0: 
michael@0: inline UnicodeString&
michael@0: UnicodeString::insert(int32_t start,
michael@0:               const UnicodeString& srcText,
michael@0:               int32_t srcStart,
michael@0:               int32_t srcLength)
michael@0: { return doReplace(start, 0, srcText, srcStart, srcLength); }
michael@0: 
michael@0: inline UnicodeString&
michael@0: UnicodeString::insert(int32_t start,
michael@0:               const UnicodeString& srcText)
michael@0: { return doReplace(start, 0, srcText, 0, srcText.length()); }
michael@0: 
michael@0: inline UnicodeString&
michael@0: UnicodeString::insert(int32_t start,
michael@0:               const UChar *srcChars,
michael@0:               int32_t srcStart,
michael@0:               int32_t srcLength)
michael@0: { return doReplace(start, 0, srcChars, srcStart, srcLength); }
michael@0: 
michael@0: inline UnicodeString&
michael@0: UnicodeString::insert(int32_t start,
michael@0:               const UChar *srcChars,
michael@0:               int32_t srcLength)
michael@0: { return doReplace(start, 0, srcChars, 0, srcLength); }
michael@0: 
michael@0: inline UnicodeString&
michael@0: UnicodeString::insert(int32_t start,
michael@0:               UChar srcChar)
michael@0: { return doReplace(start, 0, &srcChar, 0, 1); }
michael@0: 
michael@0: inline UnicodeString&
michael@0: UnicodeString::insert(int32_t start,
michael@0:               UChar32 srcChar)
michael@0: { return replace(start, 0, srcChar); }
michael@0: 
michael@0: 
michael@0: inline UnicodeString&
michael@0: UnicodeString::remove()
michael@0: {
michael@0:   // remove() of a bogus string makes the string empty and non-bogus
michael@0:   if(isBogus()) {
michael@0:     setToEmpty();
michael@0:   } else {
michael@0:     fShortLength = 0;
michael@0:   }
michael@0:   return *this;
michael@0: }
michael@0: 
michael@0: inline UnicodeString&
michael@0: UnicodeString::remove(int32_t start,
michael@0:              int32_t _length)
michael@0: {
michael@0:     if(start <= 0 && _length == INT32_MAX) {
michael@0:         // remove(guaranteed everything) of a bogus string makes the string empty and non-bogus
michael@0:         return remove();
michael@0:     }
michael@0:     return doReplace(start, _length, NULL, 0, 0);
michael@0: }
michael@0: 
michael@0: inline UnicodeString&
michael@0: UnicodeString::removeBetween(int32_t start,
michael@0:                 int32_t limit)
michael@0: { return doReplace(start, limit - start, NULL, 0, 0); }
michael@0: 
michael@0: inline UnicodeString &
michael@0: UnicodeString::retainBetween(int32_t start, int32_t limit) {
michael@0:   truncate(limit);
michael@0:   return doReplace(0, start, NULL, 0, 0);
michael@0: }
michael@0: 
michael@0: inline UBool
michael@0: UnicodeString::truncate(int32_t targetLength)
michael@0: {
michael@0:   if(isBogus() && targetLength == 0) {
michael@0:     // truncate(0) of a bogus string makes the string empty and non-bogus
michael@0:     unBogus();
michael@0:     return FALSE;
michael@0:   } else if((uint32_t)targetLength < (uint32_t)length()) {
michael@0:     setLength(targetLength);
michael@0:     return TRUE;
michael@0:   } else {
michael@0:     return FALSE;
michael@0:   }
michael@0: }
michael@0: 
michael@0: inline UnicodeString&
michael@0: UnicodeString::reverse()
michael@0: { return doReverse(0, length()); }
michael@0: 
michael@0: inline UnicodeString&
michael@0: UnicodeString::reverse(int32_t start,
michael@0:                int32_t _length)
michael@0: { return doReverse(start, _length); }
michael@0: 
michael@0: U_NAMESPACE_END
michael@0: 
michael@0: #endif