• Home
  • Line#
  • Scopes#
  • Navigate#
  • Raw
  • Download
1 /*
2 **********************************************************************
3 *   Copyright (C) 1999-2004, International Business Machines
4 *   Corporation and others.  All Rights Reserved.
5 **********************************************************************
6  *  ucnv.h:
7  *  External APIs for the ICU's codeset conversion library
8  *  Bertrand A. Damiba
9  *
10  * Modification History:
11  *
12  *   Date        Name        Description
13  *   04/04/99    helena      Fixed internal header inclusion.
14  *   05/11/00    helena      Added setFallback and usesFallback APIs.
15  *   06/29/2000  helena      Major rewrite of the callback APIs.
16  *   12/07/2000  srl         Update of documentation
17  */
18 
19 /**
20  * \file
21  * \brief C API: Character conversion
22  *
23  * <h2>Character Conversion C API</h2>
24  *
25  * <p>This API is used to convert codepage or character encoded data to and
26  * from UTF-16. You can open a converter with {@link ucnv_open() }. With that
27  * converter, you can get its properties, set options, convert your data and
28  * close the converter.</p>
29  *
30  * <p>Since many software programs recogize different converter names for
31  * different types of converters, there are other functions in this API to
32  * iterate over the converter aliases. The functions {@link ucnv_getAvailableName() },
33  * {@link ucnv_getAlias() } and {@link ucnv_getStandardName() } are some of the
34  * more frequently used alias functions to get this information.</p>
35  *
36  * <p>When a converter encounters an illegal, irregular, invalid or unmappable character
37  * its default behavior is to use a substitution character to replace the
38  * bad byte sequence. This behavior can be changed by using {@link ucnv_getFromUCallBack() }
39  * or {@link ucnv_getToUCallBack() } on the converter. The header ucnv_err.h defines
40  * many other callback actions that can be used instead of a character substitution.</p>
41  *
42  * <p>More information about this API can be found in our
43  * <a href="http://oss.software.ibm.com/icu/userguide/conversion.html">User's
44  * Guide</a>.</p>
45  */
46 
47 #ifndef UCNV_H
48 #define UCNV_H
49 
50 #include "unicode/ucnv_err.h"
51 #include "unicode/uenum.h"
52 
53 #ifndef __USET_H__
54 
55 /**
56  * USet is the C API type for Unicode sets.
57  * It is forward-declared here to avoid including the header file if related
58  * conversion APIs are not used.
59  * See unicode/uset.h
60  *
61  * @see ucnv_getUnicodeSet
62  * @stable ICU 2.6
63  */
64 struct USet;
65 /** @stable ICU 2.6 */
66 typedef struct USet USet;
67 
68 #endif
69 
70 #if !UCONFIG_NO_CONVERSION
71 
72 U_CDECL_BEGIN
73 
74 /** Maximum length of a converter name including the terminating NULL @stable ICU 2.0 */
75 #define UCNV_MAX_CONVERTER_NAME_LENGTH 60
76 /** Maximum length of a converter name including path and terminating NULL @stable ICU 2.0 */
77 #define UCNV_MAX_FULL_FILE_NAME_LENGTH (600+UCNV_MAX_CONVERTER_NAME_LENGTH)
78 
79 /** Shift in for EBDCDIC_STATEFUL and iso2022 states @stable ICU 2.0 */
80 #define  UCNV_SI 0x0F
81 /** Shift out for EBDCDIC_STATEFUL and iso2022 states @stable ICU 2.0 */
82 #define  UCNV_SO 0x0E
83 
84 /**
85  * Enum for specifying basic types of converters
86  * @see ucnv_getType
87  * @stable ICU 2.0
88  */
89 typedef enum {
90     UCNV_UNSUPPORTED_CONVERTER = -1,
91     UCNV_SBCS = 0,
92     UCNV_DBCS = 1,
93     UCNV_MBCS = 2,
94     UCNV_LATIN_1 = 3,
95     UCNV_UTF8 = 4,
96     UCNV_UTF16_BigEndian = 5,
97     UCNV_UTF16_LittleEndian = 6,
98     UCNV_UTF32_BigEndian = 7,
99     UCNV_UTF32_LittleEndian = 8,
100     UCNV_EBCDIC_STATEFUL = 9,
101     UCNV_ISO_2022 = 10,
102 
103     UCNV_LMBCS_1 = 11,
104     UCNV_LMBCS_2,
105     UCNV_LMBCS_3,
106     UCNV_LMBCS_4,
107     UCNV_LMBCS_5,
108     UCNV_LMBCS_6,
109     UCNV_LMBCS_8,
110     UCNV_LMBCS_11,
111     UCNV_LMBCS_16,
112     UCNV_LMBCS_17,
113     UCNV_LMBCS_18,
114     UCNV_LMBCS_19,
115     UCNV_LMBCS_LAST = UCNV_LMBCS_19,
116     UCNV_HZ,
117     UCNV_SCSU,
118     UCNV_ISCII,
119     UCNV_US_ASCII,
120     UCNV_UTF7,
121     UCNV_BOCU1,
122     UCNV_UTF16,
123     UCNV_UTF32,
124     UCNV_CESU8,
125     UCNV_IMAP_MAILBOX,
126 
127     /* Number of converter types for which we have conversion routines. */
128     UCNV_NUMBER_OF_SUPPORTED_CONVERTER_TYPES
129 
130 } UConverterType;
131 
132 /**
133  * Enum for specifying which platform a converter ID refers to.
134  * The use of platform/CCSID is not recommended. See ucnv_openCCSID().
135  *
136  * @see ucnv_getPlatform
137  * @see ucnv_openCCSID
138  * @see ucnv_getCCSID
139  * @stable ICU 2.0
140  */
141 typedef enum {
142     UCNV_UNKNOWN = -1,
143     UCNV_IBM = 0
144 } UConverterPlatform;
145 
146 /**
147  * Function pointer for error callback in the codepage to unicode direction.
148  * Called when an error has occured in conversion to unicode, or on open/close of the callback (see reason).
149  * @param context Pointer to the callback's private data
150  * @param args Information about the conversion in progress
151  * @param codeUnits Points to 'length' bytes of the concerned codepage sequence
152  * @param length Size (in bytes) of the concerned codepage sequence
153  * @param reason Defines the reason the callback was invoked
154  * @see ucnv_setToUCallBack
155  * @see UConverterToUnicodeArgs
156  * @stable ICU 2.0
157  */
158 typedef void (U_EXPORT2 *UConverterToUCallback) (
159                   const void* context,
160                   UConverterToUnicodeArgs *args,
161                   const char *codeUnits,
162                   int32_t length,
163                   UConverterCallbackReason reason,
164                   UErrorCode *);
165 
166 /**
167  * Function pointer for error callback in the unicode to codepage direction.
168  * Called when an error has occured in conversion from unicode, or on open/close of the callback (see reason).
169  * @param context Pointer to the callback's private data
170  * @param args Information about the conversion in progress
171  * @param codeUnits Points to 'length' UChars of the concerned Unicode sequence
172  * @param length Size (in bytes) of the concerned codepage sequence
173  * @param codePoint Single UChar32 (UTF-32) containing the concerend Unicode codepoint.
174  * @param reason Defines the reason the callback was invoked
175  * @see ucnv_setFromUCallBack
176  * @stable ICU 2.0
177  */
178 typedef void (U_EXPORT2 *UConverterFromUCallback) (
179                     const void* context,
180                     UConverterFromUnicodeArgs *args,
181                     const UChar* codeUnits,
182                     int32_t length,
183                     UChar32 codePoint,
184                     UConverterCallbackReason reason,
185                     UErrorCode *);
186 
187 U_CDECL_END
188 
189 /**
190  * Character that separates converter names from options and options from each other.
191  * @see ucnv_open
192  * @stable ICU 2.0
193  */
194 #define UCNV_OPTION_SEP_CHAR ','
195 
196 /**
197  * String version of UCNV_OPTION_SEP_CHAR.
198  * @see ucnv_open
199  * @stable ICU 2.0
200  */
201 #define UCNV_OPTION_SEP_STRING ","
202 
203 /**
204  * Character that separates a converter option from its value.
205  * @see ucnv_open
206  * @stable ICU 2.0
207  */
208 #define UCNV_VALUE_SEP_CHAR '='
209 
210 /**
211  * String version of UCNV_VALUE_SEP_CHAR.
212  * @see ucnv_open
213  * @stable ICU 2.0
214  */
215 #define UCNV_VALUE_SEP_STRING "="
216 
217 /**
218  * Converter option for specifying a locale.
219  * For example, ucnv_open("SCSU,locale=ja", &errorCode);
220  * See convrtrs.txt.
221  *
222  * @see ucnv_open
223  * @stable ICU 2.0
224  */
225 #define UCNV_LOCALE_OPTION_STRING ",locale="
226 
227 /**
228  * Converter option for specifying a version selector (0..9) for some converters.
229  * For example, ucnv_open("UTF-7,version=1", &errorCode);
230  * See convrtrs.txt.
231  *
232  * @see ucnv_open
233  * @stable ICU 2.4
234  */
235 #define UCNV_VERSION_OPTION_STRING ",version="
236 
237 /**
238  * Converter option for EBCDIC SBCS or mixed-SBCS/DBCS (stateful) codepages.
239  * Swaps Unicode mappings for EBCDIC LF and NL codes, as used on
240  * S/390 (z/OS) Unix System Services (Open Edition).
241  * For example, ucnv_open("ibm-1047,swaplfnl", &errorCode);
242  * See convrtrs.txt.
243  *
244  * @see ucnv_open
245  * @stable ICU 2.4
246  */
247 #define UCNV_SWAP_LFNL_OPTION_STRING ",swaplfnl"
248 
249 /**
250  * Do a fuzzy compare of a two converter/alias names.  The comparison
251  * is case-insensitive.  It also ignores the characters '-', '_', and
252  * ' ' (dash, underscore, and space).  Thus the strings "UTF-8",
253  * "utf_8", and "Utf 8" are exactly equivalent.
254  *
255  * @param name1 a converter name or alias, zero-terminated
256  * @param name2 a converter name or alias, zero-terminated
257  * @return 0 if the names match, or a negative value if the name1
258  * lexically precedes name2, or a positive value if the name1
259  * lexically follows name2.
260  * @stable ICU 2.0
261  */
262 U_STABLE int U_EXPORT2
263 ucnv_compareNames(const char *name1, const char *name2);
264 
265 
266 /**
267  * Creates a UConverter object with the names specified as a C string.
268  * The actual name will be resolved with the alias file
269  * using a case-insensitive string comparison that ignores
270  * the delimiters '-', '_', and ' ' (dash, underscore, and space).
271  * E.g., the names "UTF8", "utf-8", and "Utf 8" are all equivalent.
272  * If <code>NULL</code> is passed for the converter name, it will create one with the
273  * getDefaultName return value.
274  *
275  * <p>A converter name for ICU 1.5 and above may contain options
276  * like a locale specification to control the specific behavior of
277  * the newly instantiated converter.
278  * The meaning of the options depends on the particular converter.
279  * If an option is not defined for or recognized by a given converter, then it is ignored.</p>
280  *
281  * <p>Options are appended to the converter name string, with a
282  * <code>UCNV_OPTION_SEP_CHAR</code> between the name and the first option and
283  * also between adjacent options.</p>
284  *
285  * <p>If the alias is ambiguous, then the preferred converter is used
286  * and the status is set to U_AMBIGUOUS_ALIAS_WARNING.</p>
287  *
288  * <p>The conversion behavior and names can vary between platforms. ICU may
289  * convert some characters differently from other platforms. Details on this topic
290  * are in the <a href="http://oss.software.ibm.com/icu/userguide/conversion.html">User's
291  * Guide</a>.</p>
292  *
293  * @param converterName Name of the uconv table, may have options appended
294  * @param err outgoing error status <TT>U_MEMORY_ALLOCATION_ERROR, U_FILE_ACCESS_ERROR</TT>
295  * @return the created Unicode converter object, or <TT>NULL</TT> if an error occured
296  * @see ucnv_openU
297  * @see ucnv_openCCSID
298  * @see ucnv_close
299  * @stable ICU 2.0
300  */
301 U_STABLE UConverter* U_EXPORT2
302 ucnv_open(const char *converterName, UErrorCode *err);
303 
304 
305 /**
306  * Creates a Unicode converter with the names specified as unicode string.
307  * The name should be limited to the ASCII-7 alphanumerics range.
308  * The actual name will be resolved with the alias file
309  * using a case-insensitive string comparison that ignores
310  * the delimiters '-', '_', and ' ' (dash, underscore, and space).
311  * E.g., the names "UTF8", "utf-8", and "Utf 8" are all equivalent.
312  * If <TT>NULL</TT> is passed for the converter name, it will create
313  * one with the ucnv_getDefaultName() return value.
314  * If the alias is ambiguous, then the preferred converter is used
315  * and the status is set to U_AMBIGUOUS_ALIAS_WARNING.
316  * @param name : name of the uconv table in a zero terminated
317  *        Unicode string
318  * @param err outgoing error status <TT>U_MEMORY_ALLOCATION_ERROR,
319  *        U_FILE_ACCESS_ERROR</TT>
320  * @return the created Unicode converter object, or <TT>NULL</TT> if an
321  *        error occured
322  * @see ucnv_open
323  * @see ucnv_openCCSID
324  * @see ucnv_close
325  * @see ucnv_getDefaultName
326  * @stable ICU 2.0
327  */
328 U_STABLE UConverter* U_EXPORT2
329 ucnv_openU(const UChar *name,
330            UErrorCode *err);
331 
332 /**
333  * Creates a UConverter object from a CCSID number and platform pair.
334  * Note that the usefulness of this function is limited to platforms with numeric
335  * encoding IDs. Only IBM and Microsoft platforms use numeric (16-bit) identifiers for
336  * encodings.
337  *
338  * In addition, IBM CCSIDs and Unicode conversion tables are not 1:1 related.
339  * For many IBM CCSIDs there are multiple (up to six) Unicode conversion tables, and
340  * for some Unicode conversion tables there are multiple CCSIDs.
341  * Some "alternate" Unicode conversion tables are provided by the
342  * IBM CDRA conversion table registry.
343  * The most prominent example of a systematic modification of conversion tables that is
344  * not provided in the form of conversion table files in the repository is
345  * that S/390 Unix System Services swaps the codes for Line Feed and New Line in all
346  * EBCDIC codepages, which requires such a swap in the Unicode conversion tables as well.
347  *
348  * Only IBM default conversion tables are accessible with ucnv_openCCSID().
349  * ucnv_getCCSID() will return the same CCSID for all conversion tables that are associated
350  * with that CCSID.
351  *
352  * Currently, the only "platform" supported in the ICU converter API is UCNV_IBM.
353  *
354  * In summary, the use of CCSIDs and the associated API functions is not recommended.
355  *
356  * In order to open a converter with the default IBM CDRA Unicode conversion table,
357  * you can use this function or use the prefix "ibm-":
358  * \code
359  *     char name[20];
360  *     sprintf(name, "ibm-%hu", ccsid);
361  *     cnv=ucnv_open(name, &errorCode);
362  * \endcode
363  *
364  * In order to open a converter with the IBM S/390 Unix System Services variant
365  * of a Unicode/EBCDIC conversion table,
366  * you can use the prefix "ibm-" together with the option string UCNV_SWAP_LFNL_OPTION_STRING:
367  * \code
368  *     char name[20];
369  *     sprintf(name, "ibm-%hu" UCNV_SWAP_LFNL_OPTION_STRING, ccsid);
370  *     cnv=ucnv_open(name, &errorCode);
371  * \endcode
372  *
373  * In order to open a converter from a Microsoft codepage number, use the prefix "cp":
374  * \code
375  *     char name[20];
376  *     sprintf(name, "cp%hu", codepageID);
377  *     cnv=ucnv_open(name, &errorCode);
378  * \endcode
379  *
380  * If the alias is ambiguous, then the preferred converter is used
381  * and the status is set to U_AMBIGUOUS_ALIAS_WARNING.
382  *
383  * @param codepage codepage number to create
384  * @param platform the platform in which the codepage number exists
385  * @param err error status <TT>U_MEMORY_ALLOCATION_ERROR, U_FILE_ACCESS_ERROR</TT>
386  * @return the created Unicode converter object, or <TT>NULL</TT> if an error
387  *   occured.
388  * @see ucnv_open
389  * @see ucnv_openU
390  * @see ucnv_close
391  * @see ucnv_getCCSID
392  * @see ucnv_getPlatform
393  * @see UConverterPlatform
394  * @stable ICU 2.0
395  */
396 U_STABLE UConverter* U_EXPORT2
397 ucnv_openCCSID(int32_t codepage,
398                UConverterPlatform platform,
399                UErrorCode * err);
400 
401 /**
402  * <p>Creates a UConverter object specified from a packageName and a converterName.</p>
403  *
404  * <p>The packageName and converterName must point to an ICU udata object, as defined by
405  *   <code> udata_open( packageName, "cnv", converterName, err) </code> or equivalent.
406  * Typically, packageName will refer to a (.dat) file, or to a package registered with
407  * udata_setAppData().</p>
408  *
409  * <p>The name will NOT be looked up in the alias mechanism, nor will the converter be
410  * stored in the converter cache or the alias table. The only way to open further converters
411  * is call this function multiple times, or use the ucnv_safeClone() function to clone a
412  * 'master' converter.</p>
413  *
414  * <p>A future version of ICU may add alias table lookups and/or caching
415  * to this function.</p>
416  *
417  * <p>Example Use:
418  *      <code>cnv = ucnv_openPackage("myapp", "myconverter", &err);</code>
419  * </p>
420  *
421  * @param packageName name of the package (equivalent to 'path' in udata_open() call)
422  * @param converterName name of the data item to be used, without suffix.
423  * @param err outgoing error status <TT>U_MEMORY_ALLOCATION_ERROR, U_FILE_ACCESS_ERROR</TT>
424  * @return the created Unicode converter object, or <TT>NULL</TT> if an error occured
425  * @see udata_open
426  * @see ucnv_open
427  * @see ucnv_safeClone
428  * @see ucnv_close
429  * @stable ICU 2.2
430  */
431 U_STABLE UConverter* U_EXPORT2
432 ucnv_openPackage(const char *packageName, const char *converterName, UErrorCode *err);
433 
434 /**
435  * Thread safe cloning operation
436  * @param cnv converter to be cloned
437  * @param stackBuffer user allocated space for the new clone. If NULL new memory will be allocated.
438  *  If buffer is not large enough, new memory will be allocated.
439  *  Clients can use the U_CNV_SAFECLONE_BUFFERSIZE. This will probably be enough to avoid memory allocations.
440  * @param pBufferSize pointer to size of allocated space.
441  *  If *pBufferSize == 0, a sufficient size for use in cloning will
442  *  be returned ('pre-flighting')
443  *  If *pBufferSize is not enough for a stack-based safe clone,
444  *  new memory will be allocated.
445  * @param status to indicate whether the operation went on smoothly or there were errors
446  *  An informational status value, U_SAFECLONE_ALLOCATED_ERROR, is used if any allocations were necessary.
447  * @return pointer to the new clone
448  * @stable ICU 2.0
449  */
450 U_STABLE UConverter * U_EXPORT2
451 ucnv_safeClone(const UConverter *cnv,
452                void             *stackBuffer,
453                int32_t          *pBufferSize,
454                UErrorCode       *status);
455 
456 /**
457  * \def U_CNV_SAFECLONE_BUFFERSIZE
458  * Definition of a buffer size that is designed to be large enough for
459  * converters to be cloned with ucnv_safeClone().
460  * @stable ICU 2.0
461  */
462 #define U_CNV_SAFECLONE_BUFFERSIZE  1024
463 
464 /**
465  * Deletes the unicode converter and releases resources associated
466  * with just this instance.
467  * Does not free up shared converter tables.
468  *
469  * @param converter the converter object to be deleted
470  * @see ucnv_open
471  * @see ucnv_openU
472  * @see ucnv_openCCSID
473  * @stable ICU 2.0
474  */
475 U_STABLE void  U_EXPORT2
476 ucnv_close(UConverter * converter);
477 
478 /**
479  * Fills in the output parameter, subChars, with the substitution characters
480  * as multiple bytes.
481  *
482  * @param converter the Unicode converter
483  * @param subChars the subsitution characters
484  * @param len on input the capacity of subChars, on output the number
485  * of bytes copied to it
486  * @param  err the outgoing error status code.
487  * If the substitution character array is too small, an
488  * <TT>U_INDEX_OUTOFBOUNDS_ERROR</TT> will be returned.
489  * @see ucnv_setSubstChars
490  * @stable ICU 2.0
491  */
492 U_STABLE void U_EXPORT2
493 ucnv_getSubstChars(const UConverter *converter,
494                    char *subChars,
495                    int8_t *len,
496                    UErrorCode *err);
497 
498 /**
499  * Sets the substitution chars when converting from unicode to a codepage. The
500  * substitution is specified as a string of 1-4 bytes, and may contain
501  *  <TT>NULL</TT> byte.
502  * @param converter the Unicode converter
503  * @param subChars the substitution character byte sequence we want set
504  * @param len the number of bytes in subChars
505  * @param err the error status code.  <TT>U_INDEX_OUTOFBOUNDS_ERROR </TT> if
506  * len is bigger than the maximum number of bytes allowed in subchars
507  * @see ucnv_getSubstChars
508  * @stable ICU 2.0
509  */
510 U_STABLE void U_EXPORT2
511 ucnv_setSubstChars(UConverter *converter,
512                    const char *subChars,
513                    int8_t len,
514                    UErrorCode *err);
515 
516 /**
517  * Fills in the output parameter, errBytes, with the error characters from the
518  * last failing conversion.
519  *
520  * @param converter the Unicode converter
521  * @param errBytes the codepage bytes which were in error
522  * @param len on input the capacity of errBytes, on output the number of
523  *  bytes which were copied to it
524  * @param err the error status code.
525  * If the substitution character array is too small, an
526  * <TT>U_INDEX_OUTOFBOUNDS_ERROR</TT> will be returned.
527  * @stable ICU 2.0
528  */
529 U_STABLE void U_EXPORT2
530 ucnv_getInvalidChars(const UConverter *converter,
531                      char *errBytes,
532                      int8_t *len,
533                      UErrorCode *err);
534 
535 /**
536  * Fills in the output parameter, errChars, with the error characters from the
537  * last failing conversion.
538  *
539  * @param converter the Unicode converter
540  * @param errUChars the UChars which were in error
541  * @param len on input the capacity of errUChars, on output the number of
542  *  UChars which were copied to it
543  * @param err the error status code.
544  * If the substitution character array is too small, an
545  * <TT>U_INDEX_OUTOFBOUNDS_ERROR</TT> will be returned.
546  * @stable ICU 2.0
547  */
548 U_STABLE void U_EXPORT2
549 ucnv_getInvalidUChars(const UConverter *converter,
550                       UChar *errUChars,
551                       int8_t *len,
552                       UErrorCode *err);
553 
554 /**
555  * Resets the state of a converter to the default state. This is used
556  * in the case of an error, to restart a conversion from a known default state.
557  * It will also empty the internal output buffers.
558  * @param converter the Unicode converter
559  * @stable ICU 2.0
560  */
561 U_STABLE void U_EXPORT2
562 ucnv_reset(UConverter *converter);
563 
564 /**
565  * Resets the to-Unicode part of a converter state to the default state.
566  * This is used in the case of an error to restart a conversion to
567  * Unicode to a known default state. It will also empty the internal
568  * output buffers used for the conversion to Unicode codepoints.
569  * @param converter the Unicode converter
570  * @stable ICU 2.0
571  */
572 U_STABLE void U_EXPORT2
573 ucnv_resetToUnicode(UConverter *converter);
574 
575 /**
576  * Resets the from-Unicode part of a converter state to the default state.
577  * This is used in the case of an error to restart a conversion from
578  * Unicode to a known default state. It will also empty the internal output
579  * buffers used for the conversion from Unicode codepoints.
580  * @param converter the Unicode converter
581  * @stable ICU 2.0
582  */
583 U_STABLE void U_EXPORT2
584 ucnv_resetFromUnicode(UConverter *converter);
585 
586 /**
587  * Returns the maximum number of bytes that are output per UChar in conversion
588  * from Unicode using this converter.
589  * The returned number can be used with UCNV_GET_MAX_BYTES_FOR_STRING
590  * to calculate the size of a target buffer for conversion from Unicode.
591  *
592  * Note: Before ICU 2.8, this function did not return reliable numbers for
593  * some stateful converters (EBCDIC_STATEFUL, ISO-2022) and LMBCS.
594  *
595  * This number may not be the same as the maximum number of bytes per
596  * "conversion unit". In other words, it may not be the intuitively expected
597  * number of bytes per character that would be published for a charset,
598  * and may not fulfill any other purpose than the allocation of an output
599  * buffer of guaranteed sufficient size for a given input length and converter.
600  *
601  * Examples for special cases that are taken into account:
602  * - Supplementary code points may convert to more bytes than BMP code points.
603  *   This function returns bytes per UChar (UTF-16 code unit), not per
604  *   Unicode code point, for efficient buffer allocation.
605  * - State-shifting output (SI/SO, escapes, etc.) from stateful converters.
606  * - When m input UChars are converted to n output bytes, then the maximum m/n
607  *   is taken into account.
608  *
609  * The number returned here does not take into account
610  * (see UCNV_GET_MAX_BYTES_FOR_STRING):
611  * - callbacks which output more than one charset character sequence per call,
612  *   like escape callbacks
613  * - initial and final non-character bytes that are output by some converters
614  *   (automatic BOMs, initial escape sequence, final SI, etc.)
615  *
616  * Examples for returned values:
617  * - SBCS charsets: 1
618  * - Shift-JIS: 2
619  * - UTF-16: 2 (2 per BMP, 4 per surrogate _pair_, BOM not counted)
620  * - UTF-8: 3 (3 per BMP, 4 per surrogate _pair_)
621  * - EBCDIC_STATEFUL (EBCDIC mixed SBCS/DBCS): 3 (SO + DBCS)
622  * - ISO-2022: 3 (always outputs UTF-8)
623  * - ISO-2022-JP: 6 (4-byte escape sequences + DBCS)
624  * - ISO-2022-CN: 8 (4-byte designator sequences + 2-byte SS2/SS3 + DBCS)
625  *
626  * @param converter The Unicode converter.
627  * @return The maximum number of bytes per UChar that are output by ucnv_fromUnicode(),
628  *         to be used together with UCNV_GET_MAX_BYTES_FOR_STRING for buffer allocation.
629  *
630  * @see UCNV_GET_MAX_BYTES_FOR_STRING
631  * @see ucnv_getMinCharSize
632  * @stable ICU 2.0
633  */
634 U_STABLE int8_t U_EXPORT2
635 ucnv_getMaxCharSize(const UConverter *converter);
636 
637 #ifndef U_HIDE_DRAFT_API
638 
639 /**
640  * Calculates the size of a buffer for conversion from Unicode to a charset.
641  * The calculated size is guaranteed to be sufficient for this conversion.
642  *
643  * It takes into account initial and final non-character bytes that are output
644  * by some converters.
645  * It does not take into account callbacks which output more than one charset
646  * character sequence per call, like escape callbacks.
647  * The default (substitution) callback only outputs one charset character sequence.
648  *
649  * @param length Number of UChars to be converted.
650  * @param maxCharSize Return value from ucnv_getMaxCharSize() for the converter
651  *                    that will be used.
652  * @return Size of a buffer that will be large enough to hold the output bytes of
653  *         converting length UChars with the converter that returned the maxCharSize.
654  *
655  * @see ucnv_getMaxCharSize
656  * @draft ICU 2.8
657  */
658 #define UCNV_GET_MAX_BYTES_FOR_STRING(length, maxCharSize) \
659      (((int32_t)(length)+10)*(int32_t)(maxCharSize))
660 
661 #endif /*U_HIDE_DRAFT_API*/
662 
663 /**
664  * Returns the minimum byte length for characters in this codepage.
665  * This is usually either 1 or 2.
666  * @param converter the Unicode converter
667  * @return the minimum number of bytes allowed by this particular converter
668  * @see ucnv_getMaxCharSize
669  * @stable ICU 2.0
670  */
671 U_STABLE int8_t U_EXPORT2
672 ucnv_getMinCharSize(const UConverter *converter);
673 
674 /**
675  * Returns the display name of the converter passed in based on the Locale
676  * passed in. If the locale contains no display name, the internal ASCII
677  * name will be filled in.
678  *
679  * @param converter the Unicode converter.
680  * @param displayLocale is the specific Locale we want to localised for
681  * @param displayName user provided buffer to be filled in
682  * @param displayNameCapacity size of displayName Buffer
683  * @param err error status code
684  * @return displayNameLength number of UChar needed in displayName
685  * @see ucnv_getName
686  * @stable ICU 2.0
687  */
688 U_STABLE int32_t U_EXPORT2
689 ucnv_getDisplayName(const UConverter *converter,
690                     const char *displayLocale,
691                     UChar *displayName,
692                     int32_t displayNameCapacity,
693                     UErrorCode *err);
694 
695 /**
696  * Gets the internal, canonical name of the converter (zero-terminated).
697  * The lifetime of the returned string will be that of the converter
698  * passed to this function.
699  * @param converter the Unicode converter
700  * @param err UErrorCode status
701  * @return the internal name of the converter
702  * @see ucnv_getDisplayName
703  * @stable ICU 2.0
704  */
705 U_STABLE const char * U_EXPORT2
706 ucnv_getName(const UConverter *converter, UErrorCode *err);
707 
708 /**
709  * Gets a codepage number associated with the converter. This is not guaranteed
710  * to be the one used to create the converter. Some converters do not represent
711  * platform registered codepages and return zero for the codepage number.
712  * The error code fill-in parameter indicates if the codepage number
713  * is available.
714  * Does not check if the converter is <TT>NULL</TT> or if converter's data
715  * table is <TT>NULL</TT>.
716  *
717  * Important: The use of CCSIDs is not recommended because it is limited
718  * to only two platforms in principle and only one (UCNV_IBM) in the current
719  * ICU converter API.
720  * Also, CCSIDs are insufficient to identify IBM Unicode conversion tables precisely.
721  * For more details see ucnv_openCCSID().
722  *
723  * @param converter the Unicode converter
724  * @param err the error status code.
725  * @return If any error occurrs, -1 will be returned otherwise, the codepage number
726  * will be returned
727  * @see ucnv_openCCSID
728  * @see ucnv_getPlatform
729  * @stable ICU 2.0
730  */
731 U_STABLE int32_t U_EXPORT2
732 ucnv_getCCSID(const UConverter *converter,
733               UErrorCode *err);
734 
735 /**
736  * Gets a codepage platform associated with the converter. Currently,
737  * only <TT>UCNV_IBM</TT> will be returned.
738  * Does not test if the converter is <TT>NULL</TT> or if converter's data
739  * table is <TT>NULL</TT>.
740  * @param converter the Unicode converter
741  * @param err the error status code.
742  * @return The codepage platform
743  * @stable ICU 2.0
744  */
745 U_STABLE UConverterPlatform U_EXPORT2
746 ucnv_getPlatform(const UConverter *converter,
747                  UErrorCode *err);
748 
749 /**
750  * Gets the type of the converter
751  * e.g. SBCS, MBCS, DBCS, UTF8, UTF16_BE, UTF16_LE, ISO_2022,
752  * EBCDIC_STATEFUL, LATIN_1
753  * @param converter a valid, opened converter
754  * @return the type of the converter
755  * @stable ICU 2.0
756  */
757 U_STABLE UConverterType U_EXPORT2
758 ucnv_getType(const UConverter * converter);
759 
760 /**
761  * Gets the "starter" (lead) bytes for converters of type MBCS.
762  * Will fill in an <TT>U_ILLEGAL_ARGUMENT_ERROR</TT> if converter passed in
763  * is not MBCS. Fills in an array of type UBool, with the value of the byte
764  * as offset to the array. For example, if (starters[0x20] == TRUE) at return,
765  * it means that the byte 0x20 is a starter byte in this converter.
766  * Context pointers are always owned by the caller.
767  *
768  * @param converter a valid, opened converter of type MBCS
769  * @param starters an array of size 256 to be filled in
770  * @param err error status, <TT>U_ILLEGAL_ARGUMENT_ERROR</TT> if the
771  * converter is not a type which can return starters.
772  * @see ucnv_getType
773  * @stable ICU 2.0
774  */
775 U_STABLE void U_EXPORT2
776 ucnv_getStarters(const UConverter* converter,
777                  UBool starters[256],
778                  UErrorCode* err);
779 
780 
781 /**
782  * Selectors for Unicode sets that can be returned by ucnv_getUnicodeSet().
783  * @see ucnv_getUnicodeSet
784  * @stable ICU 2.6
785  */
786 typedef enum UConverterUnicodeSet {
787     /** Select the set of roundtrippable Unicode code points. @stable ICU 2.6 */
788     UCNV_ROUNDTRIP_SET,
789     /** Number of UConverterUnicodeSet selectors. @stable ICU 2.6 */
790     UCNV_SET_COUNT
791 } UConverterUnicodeSet;
792 
793 
794 /**
795  * Returns the set of Unicode code points that can be converted by an ICU converter.
796  *
797  * The current implementation returns only one kind of set (UCNV_ROUNDTRIP_SET):
798  * The set of all Unicode code points that can be roundtrip-converted
799  * (converted without any data loss) with the converter.
800  * This set will not include code points that have fallback mappings
801  * or are only the result of reverse fallback mappings.
802  * See UTR #22 "Character Mapping Markup Language"
803  * at http://www.unicode.org/reports/tr22/
804  *
805  * This is useful for example for
806  * - checking that a string or document can be roundtrip-converted with a converter,
807  *   without/before actually performing the conversion
808  * - testing if a converter can be used for text for typical text for a certain locale,
809  *   by comparing its roundtrip set with the set of ExemplarCharacters from
810  *   ICU's locale data or other sources
811  *
812  * In the future, there may be more UConverterUnicodeSet choices to select
813  * sets with different properties.
814  *
815  * @param cnv The converter for which a set is requested.
816  * @param setFillIn A valid USet *. It will be cleared by this function before
817  *            the converter's specific set is filled into the USet.
818  * @param whichSet A UConverterUnicodeSet selector;
819  *              currently UCNV_ROUNDTRIP_SET is the only supported value.
820  * @param pErrorCode ICU error code in/out parameter.
821  *                   Must fulfill U_SUCCESS before the function call.
822  *
823  * @see UConverterUnicodeSet
824  * @see uset_open
825  * @see uset_close
826  * @stable ICU 2.6
827  */
828 U_STABLE void U_EXPORT2
829 ucnv_getUnicodeSet(const UConverter *cnv,
830                    USet *setFillIn,
831                    UConverterUnicodeSet whichSet,
832                    UErrorCode *pErrorCode);
833 
834 /**
835  * Gets the current calback function used by the converter when an illegal
836  *  or invalid codepage sequence is found.
837  * Context pointers are always owned by the caller.
838  *
839  * @param converter the unicode converter
840  * @param action fillin: returns the callback function pointer
841  * @param context fillin: returns the callback's private void* context
842  * @see ucnv_setToUCallBack
843  * @stable ICU 2.0
844  */
845 U_STABLE void U_EXPORT2
846 ucnv_getToUCallBack (const UConverter * converter,
847                      UConverterToUCallback *action,
848                      const void **context);
849 
850 /**
851  * Gets the current callback function used by the converter when illegal
852  * or invalid Unicode sequence is found.
853  * Context pointers are always owned by the caller.
854  *
855  * @param converter the unicode converter
856  * @param action fillin: returns the callback function pointer
857  * @param context fillin: returns the callback's private void* context
858  * @see ucnv_setFromUCallBack
859  * @stable ICU 2.0
860  */
861 U_STABLE void U_EXPORT2
862 ucnv_getFromUCallBack (const UConverter * converter,
863                        UConverterFromUCallback *action,
864                        const void **context);
865 
866 /**
867  * Changes the callback function used by the converter when
868  * an illegal or invalid sequence is found.
869  * Context pointers are always owned by the caller.
870  * Predefined actions and contexts can be found in the ucnv_err.h header.
871  *
872  * @param converter the unicode converter
873  * @param newAction the new callback function
874  * @param newContext the new toUnicode callback context pointer. This can be NULL.
875  * @param oldAction fillin: returns the old callback function pointer. This can be NULL.
876  * @param oldContext fillin: returns the old callback's private void* context. This can be NULL.
877  * @param err The error code status
878  * @see ucnv_getToUCallBack
879  * @stable ICU 2.0
880  */
881 U_STABLE void U_EXPORT2
882 ucnv_setToUCallBack (UConverter * converter,
883                      UConverterToUCallback newAction,
884                      const void* newContext,
885                      UConverterToUCallback *oldAction,
886                      const void** oldContext,
887                      UErrorCode * err);
888 
889 /**
890  * Changes the current callback function used by the converter when
891  * an illegal or invalid sequence is found.
892  * Context pointers are always owned by the caller.
893  * Predefined actions and contexts can be found in the ucnv_err.h header.
894  *
895  * @param converter the unicode converter
896  * @param newAction the new callback function
897  * @param newContext the new fromUnicode callback context pointer. This can be NULL.
898  * @param oldAction fillin: returns the old callback function pointer. This can be NULL.
899  * @param oldContext fillin: returns the old callback's private void* context. This can be NULL.
900  * @param err The error code status
901  * @see ucnv_getFromUCallBack
902  * @stable ICU 2.0
903  */
904 U_STABLE void U_EXPORT2
905 ucnv_setFromUCallBack (UConverter * converter,
906                        UConverterFromUCallback newAction,
907                        const void *newContext,
908                        UConverterFromUCallback *oldAction,
909                        const void **oldContext,
910                        UErrorCode * err);
911 
912 /**
913  * Converts an array of unicode characters to an array of codepage
914  * characters. This function is optimized for converting a continuous
915  * stream of data in buffer-sized chunks, where the entire source and
916  * target does not fit in available buffers.
917  *
918  * The source pointer is an in/out parameter. It starts out pointing where the
919  * conversion is to begin, and ends up pointing after the last UChar consumed.
920  *
921  * Target similarly starts out pointer at the first available byte in the output
922  * buffer, and ends up pointing after the last byte written to the output.
923  *
924  * The converter always attempts to consume the entire source buffer, unless
925  * (1.) the target buffer is full, or (2.) a failing error is returned from the
926  * current callback function.  When a successful error status has been
927  * returned, it means that all of the source buffer has been
928  *  consumed. At that point, the caller should reset the source and
929  *  sourceLimit pointers to point to the next chunk.
930  *
931  * At the end of the stream (flush==TRUE), the input is completely consumed
932  * when *source==sourceLimit and no error code is set.
933  * The converter object is then automatically reset by this function.
934  * (This means that a converter need not be reset explicitly between data
935  * streams if it finishes the previous stream without errors.)
936  *
937  * This is a <I>stateful</I> conversion. Additionally, even when all source data has
938  * been consumed, some data may be in the converters' internal state.
939  * Call this function repeatedly, updating the target pointers with
940  * the next empty chunk of target in case of a
941  * <TT>U_BUFFER_OVERFLOW_ERROR</TT>, and updating the source  pointers
942  *  with the next chunk of source when a successful error status is
943  * returned, until there are no more chunks of source data.
944  * @param converter the Unicode converter
945  * @param target I/O parameter. Input : Points to the beginning of the buffer to copy
946  *  codepage characters to. Output : points to after the last codepage character copied
947  *  to <TT>target</TT>.
948  * @param targetLimit the pointer just after last of the <TT>target</TT> buffer
949  * @param source I/O parameter, pointer to pointer to the source Unicode character buffer.
950  * @param sourceLimit the pointer just after the last of the source buffer
951  * @param offsets if NULL is passed, nothing will happen to it, otherwise it needs to have the same number
952  * of allocated cells as <TT>target</TT>. Will fill in offsets from target to source pointer
953  * e.g: <TT>offsets[3]</TT> is equal to 6, it means that the <TT>target[3]</TT> was a result of transcoding <TT>source[6]</TT>
954  * For output data carried across calls, and other data without a specific source character
955  * (such as from escape sequences or callbacks)  -1 will be placed for offsets.
956  * @param flush set to <TT>TRUE</TT> if the current source buffer is the last available
957  * chunk of the source, <TT>FALSE</TT> otherwise. Note that if a failing status is returned,
958  * this function may have to be called multiple times with flush set to <TT>TRUE</TT> until
959  * the source buffer is consumed.
960  * @param err the error status.  <TT>U_ILLEGAL_ARGUMENT_ERROR</TT> will be set if the
961  * converter is <TT>NULL</TT>.
962  * <code>U_BUFFER_OVERFLOW_ERROR</code> will be set if the target is full and there is
963  * still data to be written to the target.
964  * @see ucnv_fromUChars
965  * @see ucnv_convert
966  * @see ucnv_getMinCharSize
967  * @see ucnv_setToUCallBack
968  * @stable ICU 2.0
969  */
970 U_STABLE void U_EXPORT2
971 ucnv_fromUnicode (UConverter * converter,
972                   char **target,
973                   const char *targetLimit,
974                   const UChar ** source,
975                   const UChar * sourceLimit,
976                   int32_t* offsets,
977                   UBool flush,
978                   UErrorCode * err);
979 
980 /**
981  * Converts a buffer of codepage bytes into an array of unicode UChars
982  * characters. This function is optimized for converting a continuous
983  * stream of data in buffer-sized chunks, where the entire source and
984  * target does not fit in available buffers.
985  *
986  * The source pointer is an in/out parameter. It starts out pointing where the
987  * conversion is to begin, and ends up pointing after the last byte of source consumed.
988  *
989  * Target similarly starts out pointer at the first available UChar in the output
990  * buffer, and ends up pointing after the last UChar written to the output.
991  * It does NOT necessarily keep UChar sequences together.
992  *
993  * The converter always attempts to consume the entire source buffer, unless
994  * (1.) the target buffer is full, or (2.) a failing error is returned from the
995  * current callback function.  When a successful error status has been
996  * returned, it means that all of the source buffer has been
997  *  consumed. At that point, the caller should reset the source and
998  *  sourceLimit pointers to point to the next chunk.
999  *
1000  * At the end of the stream (flush==TRUE), the input is completely consumed
1001  * when *source==sourceLimit and no error code is set
1002  * The converter object is then automatically reset by this function.
1003  * (This means that a converter need not be reset explicitly between data
1004  * streams if it finishes the previous stream without errors.)
1005  *
1006  * This is a <I>stateful</I> conversion. Additionally, even when all source data has
1007  * been consumed, some data may be in the converters' internal state.
1008  * Call this function repeatedly, updating the target pointers with
1009  * the next empty chunk of target in case of a
1010  * <TT>U_BUFFER_OVERFLOW_ERROR</TT>, and updating the source  pointers
1011  *  with the next chunk of source when a successful error status is
1012  * returned, until there are no more chunks of source data.
1013  * @param converter the Unicode converter
1014  * @param target I/O parameter. Input : Points to the beginning of the buffer to copy
1015  *  UChars into. Output : points to after the last UChar copied.
1016  * @param targetLimit the pointer just after the end of the <TT>target</TT> buffer
1017  * @param source I/O parameter, pointer to pointer to the source codepage buffer.
1018  * @param sourceLimit the pointer to the byte after the end of the source buffer
1019  * @param offsets if NULL is passed, nothing will happen to it, otherwise it needs to have the same number
1020  * of allocated cells as <TT>target</TT>. Will fill in offsets from target to source pointer
1021  * e.g: <TT>offsets[3]</TT> is equal to 6, it means that the <TT>target[3]</TT> was a result of transcoding <TT>source[6]</TT>
1022  * For output data carried across calls, and other data without a specific source character
1023  * (such as from escape sequences or callbacks)  -1 will be placed for offsets.
1024  * @param flush set to <TT>TRUE</TT> if the current source buffer is the last available
1025  * chunk of the source, <TT>FALSE</TT> otherwise. Note that if a failing status is returned,
1026  * this function may have to be called multiple times with flush set to <TT>TRUE</TT> until
1027  * the source buffer is consumed.
1028  * @param err the error status.  <TT>U_ILLEGAL_ARGUMENT_ERROR</TT> will be set if the
1029  * converter is <TT>NULL</TT>.
1030  * <code>U_BUFFER_OVERFLOW_ERROR</code> will be set if the target is full and there is
1031  * still data to be written to the target.
1032  * @see ucnv_fromUChars
1033  * @see ucnv_convert
1034  * @see ucnv_getMinCharSize
1035  * @see ucnv_setFromUCallBack
1036  * @see ucnv_getNextUChar
1037  * @stable ICU 2.0
1038  */
1039 U_STABLE void U_EXPORT2
1040 ucnv_toUnicode(UConverter *converter,
1041                UChar **target,
1042                const UChar *targetLimit,
1043                const char **source,
1044                const char *sourceLimit,
1045                int32_t *offsets,
1046                UBool flush,
1047                UErrorCode *err);
1048 
1049 /**
1050  * Convert the Unicode string into a codepage string using an existing UConverter.
1051  * The output string is NUL-terminated if possible.
1052  *
1053  * This function is a more convenient but less powerful version of ucnv_fromUnicode().
1054  * It is only useful for whole strings, not for streaming conversion.
1055  *
1056  * The maximum output buffer capacity required (barring output from callbacks) will be
1057  * UCNV_GET_MAX_BYTES_FOR_STRING(srcLength, ucnv_getMaxCharSize(cnv)).
1058  *
1059  * @param cnv the converter object to be used (ucnv_resetFromUnicode() will be called)
1060  * @param src the input Unicode string
1061  * @param srcLength the input string length, or -1 if NUL-terminated
1062  * @param dest destination string buffer, can be NULL if destCapacity==0
1063  * @param destCapacity the number of chars available at dest
1064  * @param pErrorCode normal ICU error code;
1065  *                  common error codes that may be set by this function include
1066  *                  U_BUFFER_OVERFLOW_ERROR, U_STRING_NOT_TERMINATED_WARNING,
1067  *                  U_ILLEGAL_ARGUMENT_ERROR, and conversion errors
1068  * @return the length of the output string, not counting the terminating NUL;
1069  *         if the length is greater than destCapacity, then the string will not fit
1070  *         and a buffer of the indicated length would need to be passed in
1071  * @see ucnv_fromUnicode
1072  * @see ucnv_convert
1073  * @see UCNV_GET_MAX_BYTES_FOR_STRING
1074  * @stable ICU 2.0
1075  */
1076 U_STABLE int32_t U_EXPORT2
1077 ucnv_fromUChars(UConverter *cnv,
1078                 char *dest, int32_t destCapacity,
1079                 const UChar *src, int32_t srcLength,
1080                 UErrorCode *pErrorCode);
1081 
1082 /**
1083  * Convert the codepage string into a Unicode string using an existing UConverter.
1084  * The output string is NUL-terminated if possible.
1085  *
1086  * This function is a more convenient but less powerful version of ucnv_toUnicode().
1087  * It is only useful for whole strings, not for streaming conversion.
1088  *
1089  * The maximum output buffer capacity required (barring output from callbacks) will be
1090  * 2*srcLength (each char may be converted into a surrogate pair).
1091  *
1092  * @param cnv the converter object to be used (ucnv_resetToUnicode() will be called)
1093  * @param src the input codepage string
1094  * @param srcLength the input string length, or -1 if NUL-terminated
1095  * @param dest destination string buffer, can be NULL if destCapacity==0
1096  * @param destCapacity the number of UChars available at dest
1097  * @param pErrorCode normal ICU error code;
1098  *                  common error codes that may be set by this function include
1099  *                  U_BUFFER_OVERFLOW_ERROR, U_STRING_NOT_TERMINATED_WARNING,
1100  *                  U_ILLEGAL_ARGUMENT_ERROR, and conversion errors
1101  * @return the length of the output string, not counting the terminating NUL;
1102  *         if the length is greater than destCapacity, then the string will not fit
1103  *         and a buffer of the indicated length would need to be passed in
1104  * @see ucnv_toUnicode
1105  * @see ucnv_convert
1106  * @stable ICU 2.0
1107  */
1108 U_STABLE int32_t U_EXPORT2
1109 ucnv_toUChars(UConverter *cnv,
1110               UChar *dest, int32_t destCapacity,
1111               const char *src, int32_t srcLength,
1112               UErrorCode *pErrorCode);
1113 
1114 /**
1115  * Convert a codepage buffer into Unicode one character at a time.
1116  * The input is completely consumed when the U_INDEX_OUTOFBOUNDS_ERROR is set.
1117  *
1118  * Advantage compared to ucnv_toUnicode() or ucnv_toUChars():
1119  * - Faster for small amounts of data, for most converters, e.g.,
1120  *   US-ASCII, ISO-8859-1, UTF-8/16/32, and most "normal" charsets.
1121  *   (For complex converters, e.g., SCSU, UTF-7 and ISO 2022 variants,
1122  *    it uses ucnv_toUnicode() internally.)
1123  * - Convenient.
1124  *
1125  * Limitations compared to ucnv_toUnicode():
1126  * - Always assumes flush=TRUE.
1127  *   This makes ucnv_getNextUChar() unsuitable for "streaming" conversion,
1128  *   that is, for where the input is supplied in multiple buffers,
1129  *   because ucnv_getNextUChar() will assume the end of the input at the end
1130  *   of the first buffer.
1131  * - Does not provide offset output.
1132  *
1133  * It is possible to "mix" ucnv_getNextUChar() and ucnv_toUnicode() because
1134  * ucnv_getNextUChar() uses the current state of the converter
1135  * (unlike ucnv_toUChars() which always resets first).
1136  * However, if ucnv_getNextUChar() is called after ucnv_toUnicode()
1137  * stopped in the middle of a character sequence (with flush=FALSE),
1138  * then ucnv_getNextUChar() will always use the slower ucnv_toUnicode()
1139  * internally until the next character boundary.
1140  * (This is new in ICU 2.6. In earlier releases, ucnv_getNextUChar() had to
1141  * start at a character boundary.)
1142  *
1143  * Instead of using ucnv_getNextUChar(), it is recommended
1144  * to convert using ucnv_toUnicode() or ucnv_toUChars()
1145  * and then iterate over the text using U16_NEXT() or a UCharIterator (uiter.h)
1146  * or a C++ CharacterIterator or similar.
1147  * This allows streaming conversion and offset output, for example.
1148  *
1149  * <p>Handling of surrogate pairs and supplementary-plane code points:<br>
1150  * There are two different kinds of codepages that provide mappings for surrogate characters:
1151  * <ul>
1152  *   <li>Codepages like UTF-8, UTF-32, and GB 18030 provide direct representations for Unicode
1153  *       code points U+10000-U+10ffff as well as for single surrogates U+d800-U+dfff.
1154  *       Each valid sequence will result in exactly one returned code point.
1155  *       If a sequence results in a single surrogate, then that will be returned
1156  *       by itself, even if a neighboring sequence encodes the matching surrogate.</li>
1157  *   <li>Codepages like SCSU and LMBCS (and UTF-16) provide direct representations only for BMP code points
1158  *       including surrogates. Code points in supplementary planes are represented with
1159  *       two sequences, each encoding a surrogate.
1160  *       For these codepages, matching pairs of surrogates will be combined into single
1161  *       code points for returning from this function.
1162  *       (Note that SCSU is actually a mix of these codepage types.)</li>
1163  * </ul></p>
1164  *
1165  * @param converter an open UConverter
1166  * @param source the address of a pointer to the codepage buffer, will be
1167  *  updated to point after the bytes consumed in the conversion call.
1168  * @param sourceLimit points to the end of the input buffer
1169  * @param err fills in error status (see ucnv_toUnicode)
1170  * <code>U_INDEX_OUTOFBOUNDS_ERROR</code> will be set if the input
1171  * is empty or does not convert to any output (e.g.: pure state-change
1172  * codes SI/SO, escape sequences for ISO 2022,
1173  * or if the callback did not output anything, ...).
1174  * This function will not set a <code>U_BUFFER_OVERFLOW_ERROR</code> because
1175  *  the "buffer" is the return code. However, there might be subsequent output
1176  *  stored in the converter object
1177  * that will be returned in following calls to this function.
1178  * @return a UChar32 resulting from the partial conversion of source
1179  * @see ucnv_toUnicode
1180  * @see ucnv_toUChars
1181  * @see ucnv_convert
1182  * @stable ICU 2.0
1183  */
1184 U_STABLE UChar32 U_EXPORT2
1185 ucnv_getNextUChar(UConverter * converter,
1186                   const char **source,
1187                   const char * sourceLimit,
1188                   UErrorCode * err);
1189 
1190 /**
1191  * Convert from one external charset to another using two existing UConverters.
1192  * Internally, two conversions - ucnv_toUnicode() and ucnv_fromUnicode() -
1193  * are used, "pivoting" through 16-bit Unicode.
1194  *
1195  * There is a similar function, ucnv_convert(),
1196  * which has the following limitations:
1197  * - it takes charset names, not converter objects, so that
1198  *   - two converters are opened for each call
1199  *   - only single-string conversion is possible, not streaming operation
1200  * - it does not provide enough information to find out,
1201  *   in case of failure, whether the toUnicode or
1202  *   the fromUnicode conversion failed
1203  *
1204  * By contrast, ucnv_convertEx()
1205  * - takes UConverter parameters instead of charset names
1206  * - fully exposes the pivot buffer for complete error handling
1207  *
1208  * ucnv_convertEx() also provides further convenience:
1209  * - an option to reset the converters at the beginning
1210  *   (if reset==TRUE, see parameters;
1211  *    also sets *pivotTarget=*pivotSource=pivotStart)
1212  * - allow NUL-terminated input
1213  *   (only a single NUL byte, will not work for charsets with multi-byte NULs)
1214  *   (if sourceLimit==NULL, see parameters)
1215  * - terminate with a NUL on output
1216  *   (only a single NUL byte, not useful for charsets with multi-byte NULs),
1217  *   or set U_STRING_NOT_TERMINATED_WARNING if the output exactly fills
1218  *   the target buffer
1219  * - the pivot buffer can be provided internally;
1220  *   in this case, the caller will not be able to get details about where an
1221  *   error occurred
1222  *   (if pivotStart==NULL, see below)
1223  *
1224  * The function returns when one of the following is true:
1225  * - the entire source text has been converted successfully to the target buffer
1226  * - a target buffer overflow occurred (U_BUFFER_OVERFLOW_ERROR)
1227  * - a conversion error occurred
1228  *   (other U_FAILURE(), see description of pErrorCode)
1229  *
1230  * Limitation compared to the direct use of
1231  * ucnv_fromUnicode() and ucnv_toUnicode():
1232  * ucnv_convertEx() does not provide offset information.
1233  *
1234  * Limitation compared to ucnv_fromUChars() and ucnv_toUChars():
1235  * ucnv_convertEx() does not support preflighting directly.
1236  *
1237  * Sample code for converting a single string from
1238  * one external charset to UTF-8, ignoring the location of errors:
1239  *
1240  * \code
1241  * int32_t
1242  * myToUTF8(UConverter *cnv,
1243  *          const char *s, int32_t length,
1244  *          char *u8, int32_t capacity,
1245  *          UErrorCode *pErrorCode) {
1246  *     UConverter *utf8Cnv;
1247  *     char *target;
1248  *
1249  *     if(U_FAILURE(*pErrorCode)) {
1250  *         return 0;
1251  *     }
1252  *
1253  *     utf8Cnv=myGetCachedUTF8Converter(pErrorCode);
1254  *     if(U_FAILURE(*pErrorCode)) {
1255  *         return 0;
1256  *     }
1257  *
1258  *     target=u8;
1259  *     ucnv_convertEx(cnv, utf8Cnv,
1260  *                    &target, u8+capacity,
1261  *                    &s, length>=0 ? s+length : NULL,
1262  *                    NULL, NULL, NULL, NULL,
1263  *                    TRUE, TRUE,
1264  *                    pErrorCode);
1265  *
1266  *     myReleaseCachedUTF8Converter(utf8Cnv);
1267  *
1268  *     // return the output string length, but without preflighting
1269  *     return (int32_t)(target-u8);
1270  * }
1271  * \endcode
1272  *
1273  * @param targetCnv     Output converter, used to convert from the UTF-16 pivot
1274  *                      to the target using ucnv_fromUnicode().
1275  * @param sourceCnv     Input converter, used to convert from the source to
1276  *                      the UTF-16 pivot using ucnv_toUnicode().
1277  * @param target        I/O parameter, same as for ucnv_fromUChars().
1278  *                      Input: *target points to the beginning of the target buffer.
1279  *                      Output: *target points to the first unit after the last char written.
1280  * @param targetLimit   Pointer to the first unit after the target buffer.
1281  * @param source        I/O parameter, same as for ucnv_toUChars().
1282  *                      Input: *source points to the beginning of the source buffer.
1283  *                      Output: *source points to the first unit after the last char read.
1284  * @param sourceLimit   Pointer to the first unit after the source buffer.
1285  * @param pivotStart    Pointer to the UTF-16 pivot buffer. If pivotStart==NULL,
1286  *                      then an internal buffer is used and the other pivot
1287  *                      arguments are ignored and can be NULL as well.
1288  * @param pivotSource   I/O parameter, same as source in ucnv_fromUChars() for
1289  *                      conversion from the pivot buffer to the target buffer.
1290  * @param pivotTarget   I/O parameter, same as target in ucnv_toUChars() for
1291  *                      conversion from the source buffer to the pivot buffer.
1292  *                      It must be pivotStart<=*pivotSource<=*pivotTarget<=pivotLimit
1293  *                      and pivotStart<pivotLimit (unless pivotStart==NULL).
1294  * @param pivotLimit    Pointer to the first unit after the pivot buffer.
1295  * @param reset         If TRUE, then ucnv_resetToUnicode(sourceCnv) and
1296  *                      ucnv_resetFromUnicode(targetCnv) are called, and the
1297  *                      pivot pointers are reset (*pivotTarget=*pivotSource=pivotStart).
1298  * @param flush         If true, indicates the end of the input.
1299  *                      Passed directly to ucnv_toUnicode(), and carried over to
1300  *                      ucnv_fromUnicode() when the source is empty as well.
1301  * @param pErrorCode    ICU error code in/out parameter.
1302  *                      Must fulfill U_SUCCESS before the function call.
1303  *                      U_BUFFER_OVERFLOW_ERROR always refers to the target buffer
1304  *                      because overflows into the pivot buffer are handled internally.
1305  *                      Other conversion errors are from the source-to-pivot
1306  *                      conversion if *pivotSource==pivotStart, otherwise from
1307  *                      the pivot-to-target conversion.
1308  *
1309  * @see ucnv_convert
1310  * @see ucnv_fromAlgorithmic
1311  * @see ucnv_toAlgorithmic
1312  * @see ucnv_fromUnicode
1313  * @see ucnv_toUnicode
1314  * @see ucnv_fromUChars
1315  * @see ucnv_toUChars
1316  * @stable ICU 2.6
1317  */
1318 U_STABLE void U_EXPORT2
1319 ucnv_convertEx(UConverter *targetCnv, UConverter *sourceCnv,
1320                char **target, const char *targetLimit,
1321                const char **source, const char *sourceLimit,
1322                UChar *pivotStart, UChar **pivotSource,
1323                UChar **pivotTarget, const UChar *pivotLimit,
1324                UBool reset, UBool flush,
1325                UErrorCode *pErrorCode);
1326 
1327 /**
1328  * Convert from one external charset to another.
1329  * Internally, two converters are opened according to the name arguments,
1330  * then the text is converted to and from the 16-bit Unicode "pivot"
1331  * using ucnv_convertEx(), then the converters are closed again.
1332  *
1333  * This is a convenience function, not an efficient way to convert a lot of text:
1334  * ucnv_convert()
1335  * - takes charset names, not converter objects, so that
1336  *   - two converters are opened for each call
1337  *   - only single-string conversion is possible, not streaming operation
1338  * - does not provide enough information to find out,
1339  *   in case of failure, whether the toUnicode or
1340  *   the fromUnicode conversion failed
1341  * - allows NUL-terminated input
1342  *   (only a single NUL byte, will not work for charsets with multi-byte NULs)
1343  *   (if sourceLength==-1, see parameters)
1344  * - terminate with a NUL on output
1345  *   (only a single NUL byte, not useful for charsets with multi-byte NULs),
1346  *   or set U_STRING_NOT_TERMINATED_WARNING if the output exactly fills
1347  *   the target buffer
1348  * - a pivot buffer is provided internally
1349  *
1350  * The function returns when one of the following is true:
1351  * - the entire source text has been converted successfully to the target buffer
1352  *   and either the target buffer is terminated with a single NUL byte
1353  *   or the error code is set to U_STRING_NOT_TERMINATED_WARNING
1354  * - a target buffer overflow occurred (U_BUFFER_OVERFLOW_ERROR)
1355  *   and the full output string length is returned ("preflighting")
1356  * - a conversion error occurred
1357  *   (other U_FAILURE(), see description of pErrorCode)
1358  *
1359  * @param toConverterName   The name of the converter that is used to convert
1360  *                          from the UTF-16 pivot buffer to the target.
1361  * @param fromConverterName The name of the converter that is used to convert
1362  *                          from the source to the UTF-16 pivot buffer.
1363  * @param target            Pointer to the output buffer.
1364  * @param targetCapacity    Capacity of the target, in bytes.
1365  * @param source            Pointer to the input buffer.
1366  * @param sourceLength      Length of the input text, in bytes, or -1 for NUL-terminated input.
1367  * @param pErrorCode        ICU error code in/out parameter.
1368  *                          Must fulfill U_SUCCESS before the function call.
1369  * @return Length of the complete output text in bytes, even if it exceeds the targetCapacity
1370  *         and a U_BUFFER_OVERFLOW_ERROR is set.
1371  *
1372  * @see ucnv_convertEx
1373  * @see ucnv_fromAlgorithmic
1374  * @see ucnv_toAlgorithmic
1375  * @see ucnv_fromUnicode
1376  * @see ucnv_toUnicode
1377  * @see ucnv_fromUChars
1378  * @see ucnv_toUChars
1379  * @see ucnv_getNextUChar
1380  * @stable ICU 2.0
1381  */
1382 U_STABLE int32_t U_EXPORT2
1383 ucnv_convert(const char *toConverterName,
1384              const char *fromConverterName,
1385              char *target,
1386              int32_t targetCapacity,
1387              const char *source,
1388              int32_t sourceLength,
1389              UErrorCode *pErrorCode);
1390 
1391 /**
1392  * Convert from one external charset to another.
1393  * Internally, the text is converted to and from the 16-bit Unicode "pivot"
1394  * using ucnv_convertEx(). ucnv_toAlgorithmic() works exactly like ucnv_convert()
1395  * except that the two converters need not be looked up and opened completely.
1396  *
1397  * The source-to-pivot conversion uses the cnv converter parameter.
1398  * The pivot-to-target conversion uses a purely algorithmic converter
1399  * according to the specified type, e.g., UCNV_UTF8 for a UTF-8 converter.
1400  *
1401  * Internally, the algorithmic converter is opened and closed for each
1402  * function call, which is more efficient than using the public ucnv_open()
1403  * but somewhat less efficient than only resetting an existing converter
1404  * and using ucnv_convertEx().
1405  *
1406  * This function is more convenient than ucnv_convertEx() for single-string
1407  * conversions, especially when "preflighting" is desired (returning the length
1408  * of the complete output even if it does not fit into the target buffer;
1409  * see the User Guide Strings chapter). See ucnv_convert() for details.
1410  *
1411  * @param algorithmicType   UConverterType constant identifying the desired target
1412  *                          charset as a purely algorithmic converter.
1413  *                          Those are converters for Unicode charsets like
1414  *                          UTF-8, BOCU-1, SCSU, UTF-7, IMAP-mailbox-name, etc.,
1415  *                          as well as US-ASCII and ISO-8859-1.
1416  * @param cnv               The converter that is used to convert
1417  *                          from the source to the UTF-16 pivot buffer.
1418  * @param target            Pointer to the output buffer.
1419  * @param targetCapacity    Capacity of the target, in bytes.
1420  * @param source            Pointer to the input buffer.
1421  * @param sourceLength      Length of the input text, in bytes
1422  * @param pErrorCode        ICU error code in/out parameter.
1423  *                          Must fulfill U_SUCCESS before the function call.
1424  * @return Length of the complete output text in bytes, even if it exceeds the targetCapacity
1425  *         and a U_BUFFER_OVERFLOW_ERROR is set.
1426  *
1427  * @see ucnv_fromAlgorithmic
1428  * @see ucnv_convert
1429  * @see ucnv_convertEx
1430  * @see ucnv_fromUnicode
1431  * @see ucnv_toUnicode
1432  * @see ucnv_fromUChars
1433  * @see ucnv_toUChars
1434  * @stable ICU 2.6
1435  */
1436 U_STABLE int32_t U_EXPORT2
1437 ucnv_toAlgorithmic(UConverterType algorithmicType,
1438                    UConverter *cnv,
1439                    char *target, int32_t targetCapacity,
1440                    const char *source, int32_t sourceLength,
1441                    UErrorCode *pErrorCode);
1442 
1443 /**
1444  * Convert from one external charset to another.
1445  * Internally, the text is converted to and from the 16-bit Unicode "pivot"
1446  * using ucnv_convertEx(). ucnv_fromAlgorithmic() works exactly like ucnv_convert()
1447  * except that the two converters need not be looked up and opened completely.
1448  *
1449  * The source-to-pivot conversion uses a purely algorithmic converter
1450  * according to the specified type, e.g., UCNV_UTF8 for a UTF-8 converter.
1451  * The pivot-to-target conversion uses the cnv converter parameter.
1452  *
1453  * Internally, the algorithmic converter is opened and closed for each
1454  * function call, which is more efficient than using the public ucnv_open()
1455  * but somewhat less efficient than only resetting an existing converter
1456  * and using ucnv_convertEx().
1457  *
1458  * This function is more convenient than ucnv_convertEx() for single-string
1459  * conversions, especially when "preflighting" is desired (returning the length
1460  * of the complete output even if it does not fit into the target buffer;
1461  * see the User Guide Strings chapter). See ucnv_convert() for details.
1462  *
1463  * @param cnv               The converter that is used to convert
1464  *                          from the UTF-16 pivot buffer to the target.
1465  * @param algorithmicType   UConverterType constant identifying the desired source
1466  *                          charset as a purely algorithmic converter.
1467  *                          Those are converters for Unicode charsets like
1468  *                          UTF-8, BOCU-1, SCSU, UTF-7, IMAP-mailbox-name, etc.,
1469  *                          as well as US-ASCII and ISO-8859-1.
1470  * @param target            Pointer to the output buffer.
1471  * @param targetCapacity    Capacity of the target, in bytes.
1472  * @param source            Pointer to the input buffer.
1473  * @param sourceLength      Length of the input text, in bytes
1474  * @param pErrorCode        ICU error code in/out parameter.
1475  *                          Must fulfill U_SUCCESS before the function call.
1476  * @return Length of the complete output text in bytes, even if it exceeds the targetCapacity
1477  *         and a U_BUFFER_OVERFLOW_ERROR is set.
1478  *
1479  * @see ucnv_fromAlgorithmic
1480  * @see ucnv_convert
1481  * @see ucnv_convertEx
1482  * @see ucnv_fromUnicode
1483  * @see ucnv_toUnicode
1484  * @see ucnv_fromUChars
1485  * @see ucnv_toUChars
1486  * @stable ICU 2.6
1487  */
1488 U_STABLE int32_t U_EXPORT2
1489 ucnv_fromAlgorithmic(UConverter *cnv,
1490                      UConverterType algorithmicType,
1491                      char *target, int32_t targetCapacity,
1492                      const char *source, int32_t sourceLength,
1493                      UErrorCode *pErrorCode);
1494 
1495 /**
1496  * Frees up memory occupied by unused, cached converter shared data.
1497  *
1498  * @return the number of cached converters successfully deleted
1499  * @see ucnv_close
1500  * @stable ICU 2.0
1501  */
1502 U_STABLE int32_t U_EXPORT2
1503 ucnv_flushCache(void);
1504 
1505 /**
1506  * Returns the number of available converters, as per the alias file.
1507  *
1508  * @return the number of available converters
1509  * @see ucnv_getAvailableName
1510  * @stable ICU 2.0
1511  */
1512 U_STABLE int32_t U_EXPORT2
1513 ucnv_countAvailable(void);
1514 
1515 /**
1516  * Gets the canonical converter name of the specified converter from a list of
1517  * all available converters contaied in the alias file. All converters
1518  * in this list can be opened.
1519  *
1520  * @param n the index to a converter available on the system (in the range <TT>[0..ucnv_countAvaiable()]</TT>)
1521  * @return a pointer a string (library owned), or <TT>NULL</TT> if the index is out of bounds.
1522  * @see ucnv_countAvailable
1523  * @stable ICU 2.0
1524  */
1525 U_STABLE const char* U_EXPORT2
1526 ucnv_getAvailableName(int32_t n);
1527 
1528 /**
1529  * Returns a UEnumeration to enumerate all of the canonical converter
1530  * names, as per the alias file, regardless of the ability to open each
1531  * converter.
1532  *
1533  * @return A UEnumeration object for getting all the recognized canonical
1534  *   converter names.
1535  * @see ucnv_getAvailableName
1536  * @see uenum_close
1537  * @see uenum_next
1538  * @stable ICU 2.4
1539  */
1540 U_STABLE UEnumeration * U_EXPORT2
1541 ucnv_openAllNames(UErrorCode *pErrorCode);
1542 
1543 /**
1544  * Gives the number of aliases for a given converter or alias name.
1545  * If the alias is ambiguous, then the preferred converter is used
1546  * and the status is set to U_AMBIGUOUS_ALIAS_WARNING.
1547  * This method only enumerates the listed entries in the alias file.
1548  * @param alias alias name
1549  * @param pErrorCode error status
1550  * @return number of names on alias list for given alias
1551  * @stable ICU 2.0
1552  */
1553 U_STABLE uint16_t U_EXPORT2
1554 ucnv_countAliases(const char *alias, UErrorCode *pErrorCode);
1555 
1556 /**
1557  * Gives the name of the alias at given index of alias list.
1558  * This method only enumerates the listed entries in the alias file.
1559  * If the alias is ambiguous, then the preferred converter is used
1560  * and the status is set to U_AMBIGUOUS_ALIAS_WARNING.
1561  * @param alias alias name
1562  * @param n index in alias list
1563  * @param pErrorCode result of operation
1564  * @return returns the name of the alias at given index
1565  * @see ucnv_countAliases
1566  * @stable ICU 2.0
1567  */
1568 U_STABLE const char * U_EXPORT2
1569 ucnv_getAlias(const char *alias, uint16_t n, UErrorCode *pErrorCode);
1570 
1571 /**
1572  * Fill-up the list of alias names for the given alias.
1573  * This method only enumerates the listed entries in the alias file.
1574  * If the alias is ambiguous, then the preferred converter is used
1575  * and the status is set to U_AMBIGUOUS_ALIAS_WARNING.
1576  * @param alias alias name
1577  * @param aliases fill-in list, aliases is a pointer to an array of
1578  *        <code>ucnv_countAliases()</code> string-pointers
1579  *        (<code>const char *</code>) that will be filled in.
1580  *        The strings themselves are owned by the library.
1581  * @param pErrorCode result of operation
1582  * @stable ICU 2.0
1583  */
1584 U_STABLE void U_EXPORT2
1585 ucnv_getAliases(const char *alias, const char **aliases, UErrorCode *pErrorCode);
1586 
1587 /**
1588  * Return a new UEnumeration object for enumerating all the
1589  * alias names for a given converter that are recognized by a standard.
1590  * This method only enumerates the listed entries in the alias file.
1591  * The convrtrs.txt file can be modified to change the results of
1592  * this function.
1593  * The first result in this list is the same result given by
1594  * <code>ucnv_getStandardName</code>, which is the default alias for
1595  * the specified standard name. The returned object must be closed with
1596  * <code>uenum_close</code> when you are done with the object.
1597  *
1598  * @param convName original converter name
1599  * @param standard name of the standard governing the names; MIME and IANA
1600  *      are such standards
1601  * @param pErrorCode The error code
1602  * @return A UEnumeration object for getting all aliases that are recognized
1603  *      by a standard. If any of the parameters are invalid, NULL
1604  *      is returned.
1605  * @see ucnv_getStandardName
1606  * @see uenum_close
1607  * @see uenum_next
1608  * @stable ICU 2.2
1609  */
1610 U_STABLE UEnumeration * U_EXPORT2
1611 ucnv_openStandardNames(const char *convName,
1612                        const char *standard,
1613                        UErrorCode *pErrorCode);
1614 
1615 /**
1616  * Gives the number of standards associated to converter names.
1617  * @return number of standards
1618  * @stable ICU 2.0
1619  */
1620 U_STABLE uint16_t U_EXPORT2
1621 ucnv_countStandards(void);
1622 
1623 /**
1624  * Gives the name of the standard at given index of standard list.
1625  * @param n index in standard list
1626  * @param pErrorCode result of operation
1627  * @return returns the name of the standard at given index. Owned by the library.
1628  * @stable ICU 2.0
1629  */
1630 U_STABLE const char * U_EXPORT2
1631 ucnv_getStandard(uint16_t n, UErrorCode *pErrorCode);
1632 
1633 /**
1634  * Returns a standard name for a given converter name.
1635  * <p>
1636  * Example alias table:<br>
1637  * conv alias1 { STANDARD1 } alias2 { STANDARD1* }
1638  * <p>
1639  * Result of ucnv_getStandardName("conv", "STANDARD1") from example
1640  * alias table:<br>
1641  * <b>"alias2"</b>
1642  *
1643  * @param name original converter name
1644  * @param standard name of the standard governing the names; MIME and IANA
1645  *        are such standards
1646  * @param pErrorCode result of operation
1647  * @return returns the standard converter name;
1648  *         if a standard converter name cannot be determined,
1649  *         then <code>NULL</code> is returned. Owned by the library.
1650  * @stable ICU 2.0
1651  */
1652 U_STABLE const char * U_EXPORT2
1653 ucnv_getStandardName(const char *name, const char *standard, UErrorCode *pErrorCode);
1654 
1655 /**
1656  * This function will return the internal canonical converter name of the
1657  * tagged alias. This is the opposite of ucnv_openStandardNames, which
1658  * returns the tagged alias given the canonical name.
1659  * <p>
1660  * Example alias table:<br>
1661  * conv alias1 { STANDARD1 } alias2 { STANDARD1* }
1662  * <p>
1663  * Result of ucnv_getStandardName("alias1", "STANDARD1") from example
1664  * alias table:<br>
1665  * <b>"conv"</b>
1666  *
1667  * @return returns the canonical converter name;
1668  *         if a standard or alias name cannot be determined,
1669  *         then <code>NULL</code> is returned. The returned string is
1670  *         owned by the library.
1671  * @see ucnv_getStandardName
1672  * @stable ICU 2.4
1673  */
1674 U_STABLE const char * U_EXPORT2
1675 ucnv_getCanonicalName(const char *alias, const char *standard, UErrorCode *pErrorCode);
1676 
1677 /**
1678  * returns the current default converter name.
1679  *
1680  * @return returns the current default converter name;
1681  *         if a default converter name cannot be determined,
1682  *         then <code>NULL</code> is returned.
1683  *         Storage owned by the library
1684  * @see ucnv_setDefaultName
1685  * @stable ICU 2.0
1686  */
1687 U_STABLE const char * U_EXPORT2
1688 ucnv_getDefaultName(void);
1689 
1690 /**
1691  * sets the current default converter name. Caller must own the storage for 'name'
1692  * and preserve it indefinitely.
1693  * @param name the converter name to be the default (must exist).
1694  * @see ucnv_getDefaultName
1695  * @system SYSTEM API
1696  * @stable ICU 2.0
1697  */
1698 U_STABLE void U_EXPORT2
1699 ucnv_setDefaultName(const char *name);
1700 
1701 /**
1702  * Fixes the backslash character mismapping.  For example, in SJIS, the backslash
1703  * character in the ASCII portion is also used to represent the yen currency sign.
1704  * When mapping from Unicode character 0x005C, it's unclear whether to map the
1705  * character back to yen or backslash in SJIS.  This function will take the input
1706  * buffer and replace all the yen sign characters with backslash.  This is necessary
1707  * when the user tries to open a file with the input buffer on Windows.
1708  * This function will test the converter to see whether such mapping is
1709  * required.  You can sometimes avoid using this function by using the correct version
1710  * of Shift-JIS.
1711  *
1712  * @param cnv The converter representing the target codepage.
1713  * @param source the input buffer to be fixed
1714  * @param sourceLen the length of the input buffer
1715  * @see ucnv_isAmbiguous
1716  * @stable ICU 2.0
1717  */
1718 U_STABLE void U_EXPORT2
1719 ucnv_fixFileSeparator(const UConverter *cnv, UChar *source, int32_t sourceLen);
1720 
1721 /**
1722  * Determines if the converter contains ambiguous mappings of the same
1723  * character or not.
1724  * @param cnv the converter to be tested
1725  * @return TRUE if the converter contains ambiguous mapping of the same
1726  * character, FALSE otherwise.
1727  * @stable ICU 2.0
1728  */
1729 U_STABLE UBool U_EXPORT2
1730 ucnv_isAmbiguous(const UConverter *cnv);
1731 
1732 /**
1733  * Sets the converter to use fallback mapping or not.
1734  * @param cnv The converter to set the fallback mapping usage on.
1735  * @param usesFallback TRUE if the user wants the converter to take advantage of the fallback
1736  * mapping, FALSE otherwise.
1737  * @stable ICU 2.0
1738  */
1739 U_STABLE void U_EXPORT2
1740 ucnv_setFallback(UConverter *cnv, UBool usesFallback);
1741 
1742 /**
1743  * Determines if the converter uses fallback mappings or not.
1744  * @param cnv The converter to be tested
1745  * @return TRUE if the converter uses fallback, FALSE otherwise.
1746  * @stable ICU 2.0
1747  */
1748 U_STABLE UBool U_EXPORT2
1749 ucnv_usesFallback(const UConverter *cnv);
1750 
1751 /**
1752  * Detects Unicode signature byte sequences at the start of the byte stream
1753  * and returns the charset name of the indicated Unicode charset.
1754  * NULL is returned when no Unicode signature is recognized.
1755  * The number of bytes in the signature is output as well.
1756  *
1757  * The caller can ucnv_open() a converter using the charset name.
1758  * The first code unit (UChar) from the start of the stream will be U+FEFF
1759  * (the Unicode BOM/signature character) and can usually be ignored.
1760  *
1761  * For most Unicode charsets it is also possible to ignore the indicated
1762  * number of initial stream bytes and start converting after them.
1763  * However, there are stateful Unicode charsets (UTF-7 and BOCU-1) for which
1764  * this will not work. Therefore, it is best to ignore the first output UChar
1765  * instead of the input signature bytes.
1766  * <p>
1767  * Usage:
1768  * @code
1769  *      UErrorCode err = U_ZERO_ERROR;
1770  *      char input[] = { '\xEF','\xBB', '\xBF','\x41','\x42','\x43' };
1771  *      int32_t signatureLength = 0;
1772  *      char *encoding = ucnv_detectUnicodeSignatures(input,sizeof(input),&signatureLength,&err);
1773  *      UConverter *conv = NULL;
1774  *      UChar output[100];
1775  *      UChar *target = output, *out;
1776  *      char *source = input;
1777  *      if(encoding!=NULL && U_SUCCESS(err)){
1778  *          // should signature be discarded ?
1779  *          conv = ucnv_open(encoding, &err);
1780  *          // do the conversion
1781  *          ucnv_toUnicode(conv,
1782  *                         target, output + sizeof(output)/U_SIZEOF_UCHAR,
1783  *                         source, input + sizeof(input),
1784  *                         NULL, TRUE, &err);
1785  *          out = output;
1786  *          if (discardSignature){
1787  *              ++out; // ignore initial U+FEFF
1788  *          }
1789  *          while(out != target) {
1790  *              printf("%04x ", *out++);
1791  *          }
1792  *          puts("");
1793  *      }
1794  *
1795  * @endcode
1796  *
1797  * @param source            The source string in which the signature should be detected.
1798  * @param sourceLength      Length of the input string, or -1 if terminated with a NUL byte.
1799  * @param signatureLength   A pointer to int32_t to receive the number of bytes that make up the signature
1800  *                          of the detected UTF. 0 if not detected.
1801  *                          Can be a NULL pointer.
1802  * @param pErrorCode        A pointer to receive information about any errors that may occur during detection.
1803  *                          Must be a valid pointer to an error code value, which must not indicate a failure
1804  *                          before the function call.
1805  * @return The name of the encoding detected. NULL if encoding is not detected.
1806  * @stable ICU 2.4
1807  */
1808 U_STABLE const char* U_EXPORT2
1809 ucnv_detectUnicodeSignature(const char* source,
1810                             int32_t sourceLength,
1811                             int32_t *signatureLength,
1812                             UErrorCode *pErrorCode);
1813 
1814 #endif
1815 
1816 #endif
1817 /*_UCNV*/
1818