icuSources/common/unicode/ucnvsel.h

   1 /*
   2 *******************************************************************************
   3 *
   4 *   Copyright (C) 2008-2010, International Business Machines
   5 *   Corporation, Google and others.  All Rights Reserved.
   6 *
   7 *******************************************************************************
   8 */
   9 /*
  10  * Author : eldawy@google.com (Mohamed Eldawy)
  11  * ucnvsel.h
  12  *
  13  * Purpose: To generate a list of encodings capable of handling
  14  * a given Unicode text
  15  *
  16  * Started 09-April-2008
  17  */
  18
  19 #ifndef __ICU_UCNV_SEL_H__
  20 #define __ICU_UCNV_SEL_H__
  21
  22 #include "unicode/uset.h"
  23 #include "unicode/utypes.h"
  24 #include "unicode/utf16.h"
  25 #include "unicode/uenum.h"
  26 #include "unicode/ucnv.h"
  27 #include "unicode/localpointer.h"
  28
  29 /**
  30  * \file
  31  *
  32  * A converter selector is built with a set of encoding/charset names
  33  * and given an input string returns the set of names of the
  34  * corresponding converters which can convert the string.
  35  *
  36  * A converter selector can be serialized into a buffer and reopened
  37  * from the serialized form.
  38  */
  39
  40 /**
  41  * @{
  42  * The selector data structure
  43  */
  44 struct UConverterSelector;
  45 typedef struct UConverterSelector UConverterSelector;
  46 /** @} */
  47
  48 /**
  49  * Open a selector.
  50  * If converterListSize is 0, build for all available converters.
  51  * If excludedCodePoints is NULL, don't exclude any code points.
  52  *
  53  * @param converterList a pointer to encoding names needed to be involved.
  54  *                      Can be NULL if converterListSize==0.
  55  *                      The list and the names will be cloned, and the caller
  56  *                      retains ownership of the original.
  57  * @param converterListSize number of encodings in above list.
  58  *                          If 0, builds a selector for all available converters.
  59  * @param excludedCodePoints a set of code points to be excluded from consideration.
  60  *                           That is, excluded code points in a string do not change
  61  *                           the selection result. (They might be handled by a callback.)
  62  *                           Use NULL to exclude nothing.
  63  * @param whichSet what converter set to use? Use this to determine whether
  64  *                 to consider only roundtrip mappings or also fallbacks.
  65  * @param status an in/out ICU UErrorCode
  66  * @return the new selector
  67  *
  68  * @stable ICU 4.2
  69  */
  70 U_STABLE UConverterSelector* U_EXPORT2
  71 ucnvsel_open(const char* const*  converterList, int32_t converterListSize,
  72              const USet* excludedCodePoints,
  73              const UConverterUnicodeSet whichSet, UErrorCode* status);
  74
  75 /**
  76  * Closes a selector.
  77  * If any Enumerations were returned by ucnv_select*, they become invalid.
  78  * They can be closed before or after calling ucnv_closeSelector,
  79  * but should never be used after the selector is closed.
  80  *
  81  * @see ucnv_selectForString
  82  * @see ucnv_selectForUTF8
  83  *
  84  * @param sel selector to close
  85  *
  86  * @stable ICU 4.2
  87  */
  88 U_STABLE void U_EXPORT2
  89 ucnvsel_close(UConverterSelector *sel);
  90
  91 #if U_SHOW_CPLUSPLUS_API
  92
  93 U_NAMESPACE_BEGIN
  94
  95 /**
  96  * \class LocalUConverterSelectorPointer
  97  * "Smart pointer" class, closes a UConverterSelector via ucnvsel_close().
  98  * For most methods see the LocalPointerBase base class.
  99  *
 100  * @see LocalPointerBase
 101  * @see LocalPointer
 102  * @stable ICU 4.4
 103  */
 104 U_DEFINE_LOCAL_OPEN_POINTER(LocalUConverterSelectorPointer, UConverterSelector, ucnvsel_close);
 105
 106 U_NAMESPACE_END
 107
 108 #endif
 109
 110 /**
 111  * Open a selector from its serialized form.
 112  * The buffer must remain valid and unchanged for the lifetime of the selector.
 113  * This is much faster than creating a selector from scratch.
 114  * Using a serialized form from a different machine (endianness/charset) is supported.
 115  *
 116  * @param buffer pointer to the serialized form of a converter selector;
 117  *               must be 32-bit-aligned
 118  * @param length the capacity of this buffer (can be equal to or larger than
 119  *               the actual data length)
 120  * @param status an in/out ICU UErrorCode
 121  * @return the new selector
 122  *
 123  * @stable ICU 4.2
 124  */
 125 U_STABLE UConverterSelector* U_EXPORT2
 126 ucnvsel_openFromSerialized(const void* buffer, int32_t length, UErrorCode* status);
 127
 128 /**
 129  * Serialize a selector into a linear buffer.
 130  * The serialized form is portable to different machines.
 131  *
 132  * @param sel selector to consider
 133  * @param buffer pointer to 32-bit-aligned memory to be filled with the
 134  *               serialized form of this converter selector
 135  * @param bufferCapacity the capacity of this buffer
 136  * @param status an in/out ICU UErrorCode
 137  * @return the required buffer capacity to hold serialize data (even if the call fails
 138  *         with a U_BUFFER_OVERFLOW_ERROR, it will return the required capacity)
 139  *
 140  * @stable ICU 4.2
 141  */
 142 U_STABLE int32_t U_EXPORT2
 143 ucnvsel_serialize(const UConverterSelector* sel,
 144                   void* buffer, int32_t bufferCapacity, UErrorCode* status);
 145
 146 /**
 147  * Select converters that can map all characters in a UTF-16 string,
 148  * ignoring the excluded code points.
 149  *
 150  * @param sel a selector
 151  * @param s UTF-16 string
 152  * @param length length of the string, or -1 if NUL-terminated
 153  * @param status an in/out ICU UErrorCode
 154  * @return an enumeration containing encoding names.
 155  *         The returned encoding names and their order will be the same as
 156  *         supplied when building the selector.
 157  *
 158  * @stable ICU 4.2
 159  */
 160 U_STABLE UEnumeration * U_EXPORT2
 161 ucnvsel_selectForString(const UConverterSelector* sel,
 162                         const UChar *s, int32_t length, UErrorCode *status);
 163
 164 /**
 165  * Select converters that can map all characters in a UTF-8 string,
 166  * ignoring the excluded code points.
 167  *
 168  * @param sel a selector
 169  * @param s UTF-8 string
 170  * @param length length of the string, or -1 if NUL-terminated
 171  * @param status an in/out ICU UErrorCode
 172  * @return an enumeration containing encoding names.
 173  *         The returned encoding names and their order will be the same as
 174  *         supplied when building the selector.
 175  *
 176  * @stable ICU 4.2
 177  */
 178 U_STABLE UEnumeration * U_EXPORT2
 179 ucnvsel_selectForUTF8(const UConverterSelector* sel,
 180                       const char *s, int32_t length, UErrorCode *status);
 181
 182 #endif  /* __ICU_UCNV_SEL_H__ */