2 **********************************************************************
3 * Copyright (C) 1999-2005, International Business Machines
4 * Corporation and others. All Rights Reserved.
5 **********************************************************************
7 * External APIs for the ICU's codeset conversion library
10 * Modification History:
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
21 * \brief C API: Character conversion
23 * <h2>Character Conversion C API</h2>
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>
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>
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_setFromUCallBack() }
39 * or {@link ucnv_setToUCallBack() } on the converter. The header ucnv_err.h defines
40 * many other callback actions that can be used instead of a character substitution.</p>
42 * <p>More information about this API can be found in our
43 * <a href="http://icu.sourceforge.net/userguide/conversion.html">User's
50 #include "unicode/ucnv_err.h"
51 #include "unicode/uenum.h"
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.
61 * @see ucnv_getUnicodeSet
65 /** @stable ICU 2.6 */
66 typedef struct USet USet;
70 #if !UCONFIG_NO_CONVERSION
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)
79 /** Shift in for EBDCDIC_STATEFUL and iso2022 states @stable ICU 2.0 */
81 /** Shift out for EBDCDIC_STATEFUL and iso2022 states @stable ICU 2.0 */
85 * Enum for specifying basic types of converters
90 UCNV_UNSUPPORTED_CONVERTER = -1,
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,
115 UCNV_LMBCS_LAST = UCNV_LMBCS_19,
127 /* Number of converter types for which we have conversion routines. */
128 UCNV_NUMBER_OF_SUPPORTED_CONVERTER_TYPES
133 * Enum for specifying which platform a converter ID refers to.
134 * The use of platform/CCSID is not recommended. See ucnv_openCCSID().
136 * @see ucnv_getPlatform
137 * @see ucnv_openCCSID
144 } UConverterPlatform;
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 * @param pErrorCode ICU error code in/out parameter.
155 * For converter callback functions, set to a conversion error
156 * before the call, and the callback may reset it to U_ZERO_ERROR.
157 * @see ucnv_setToUCallBack
158 * @see UConverterToUnicodeArgs
161 typedef void (U_EXPORT2 *UConverterToUCallback) (
163 UConverterToUnicodeArgs *args,
164 const char *codeUnits,
166 UConverterCallbackReason reason,
167 UErrorCode *pErrorCode);
170 * Function pointer for error callback in the unicode to codepage direction.
171 * Called when an error has occured in conversion from unicode, or on open/close of the callback (see reason).
172 * @param context Pointer to the callback's private data
173 * @param args Information about the conversion in progress
174 * @param codeUnits Points to 'length' UChars of the concerned Unicode sequence
175 * @param length Size (in bytes) of the concerned codepage sequence
176 * @param codePoint Single UChar32 (UTF-32) containing the concerend Unicode codepoint.
177 * @param reason Defines the reason the callback was invoked
178 * @param pErrorCode ICU error code in/out parameter.
179 * For converter callback functions, set to a conversion error
180 * before the call, and the callback may reset it to U_ZERO_ERROR.
181 * @see ucnv_setFromUCallBack
184 typedef void (U_EXPORT2 *UConverterFromUCallback) (
186 UConverterFromUnicodeArgs *args,
187 const UChar* codeUnits,
190 UConverterCallbackReason reason,
191 UErrorCode *pErrorCode);
196 * Character that separates converter names from options and options from each other.
200 #define UCNV_OPTION_SEP_CHAR ','
203 * String version of UCNV_OPTION_SEP_CHAR.
207 #define UCNV_OPTION_SEP_STRING ","
210 * Character that separates a converter option from its value.
214 #define UCNV_VALUE_SEP_CHAR '='
217 * String version of UCNV_VALUE_SEP_CHAR.
221 #define UCNV_VALUE_SEP_STRING "="
224 * Converter option for specifying a locale.
225 * For example, ucnv_open("SCSU,locale=ja", &errorCode);
231 #define UCNV_LOCALE_OPTION_STRING ",locale="
234 * Converter option for specifying a version selector (0..9) for some converters.
235 * For example, ucnv_open("UTF-7,version=1", &errorCode);
241 #define UCNV_VERSION_OPTION_STRING ",version="
244 * Converter option for EBCDIC SBCS or mixed-SBCS/DBCS (stateful) codepages.
245 * Swaps Unicode mappings for EBCDIC LF and NL codes, as used on
246 * S/390 (z/OS) Unix System Services (Open Edition).
247 * For example, ucnv_open("ibm-1047,swaplfnl", &errorCode);
253 #define UCNV_SWAP_LFNL_OPTION_STRING ",swaplfnl"
256 * Do a fuzzy compare of a two converter/alias names. The comparison
257 * is case-insensitive. It also ignores the characters '-', '_', and
258 * ' ' (dash, underscore, and space). Thus the strings "UTF-8",
259 * "utf_8", and "Utf 8" are exactly equivalent.
261 * @param name1 a converter name or alias, zero-terminated
262 * @param name2 a converter name or alias, zero-terminated
263 * @return 0 if the names match, or a negative value if the name1
264 * lexically precedes name2, or a positive value if the name1
265 * lexically follows name2.
268 U_STABLE int U_EXPORT2
269 ucnv_compareNames(const char *name1, const char *name2);
273 * Creates a UConverter object with the names specified as a C string.
274 * The actual name will be resolved with the alias file
275 * using a case-insensitive string comparison that ignores
276 * the delimiters '-', '_', and ' ' (dash, underscore, and space).
277 * E.g., the names "UTF8", "utf-8", and "Utf 8" are all equivalent.
278 * If <code>NULL</code> is passed for the converter name, it will create one with the
279 * getDefaultName return value.
281 * <p>A converter name for ICU 1.5 and above may contain options
282 * like a locale specification to control the specific behavior of
283 * the newly instantiated converter.
284 * The meaning of the options depends on the particular converter.
285 * If an option is not defined for or recognized by a given converter, then it is ignored.</p>
287 * <p>Options are appended to the converter name string, with a
288 * <code>UCNV_OPTION_SEP_CHAR</code> between the name and the first option and
289 * also between adjacent options.</p>
291 * <p>If the alias is ambiguous, then the preferred converter is used
292 * and the status is set to U_AMBIGUOUS_ALIAS_WARNING.</p>
294 * <p>The conversion behavior and names can vary between platforms. ICU may
295 * convert some characters differently from other platforms. Details on this topic
296 * are in the <a href="http://icu.sourceforge.net/userguide/conversion.html">User's
299 * @param converterName Name of the uconv table, may have options appended
300 * @param err outgoing error status <TT>U_MEMORY_ALLOCATION_ERROR, U_FILE_ACCESS_ERROR</TT>
301 * @return the created Unicode converter object, or <TT>NULL</TT> if an error occured
303 * @see ucnv_openCCSID
307 U_STABLE UConverter* U_EXPORT2
308 ucnv_open(const char *converterName, UErrorCode *err);
312 * Creates a Unicode converter with the names specified as unicode string.
313 * The name should be limited to the ASCII-7 alphanumerics range.
314 * The actual name will be resolved with the alias file
315 * using a case-insensitive string comparison that ignores
316 * the delimiters '-', '_', and ' ' (dash, underscore, and space).
317 * E.g., the names "UTF8", "utf-8", and "Utf 8" are all equivalent.
318 * If <TT>NULL</TT> is passed for the converter name, it will create
319 * one with the ucnv_getDefaultName() return value.
320 * If the alias is ambiguous, then the preferred converter is used
321 * and the status is set to U_AMBIGUOUS_ALIAS_WARNING.
322 * @param name : name of the uconv table in a zero terminated
324 * @param err outgoing error status <TT>U_MEMORY_ALLOCATION_ERROR,
325 * U_FILE_ACCESS_ERROR</TT>
326 * @return the created Unicode converter object, or <TT>NULL</TT> if an
329 * @see ucnv_openCCSID
331 * @see ucnv_getDefaultName
334 U_STABLE UConverter* U_EXPORT2
335 ucnv_openU(const UChar *name,
339 * Creates a UConverter object from a CCSID number and platform pair.
340 * Note that the usefulness of this function is limited to platforms with numeric
341 * encoding IDs. Only IBM and Microsoft platforms use numeric (16-bit) identifiers for
344 * In addition, IBM CCSIDs and Unicode conversion tables are not 1:1 related.
345 * For many IBM CCSIDs there are multiple (up to six) Unicode conversion tables, and
346 * for some Unicode conversion tables there are multiple CCSIDs.
347 * Some "alternate" Unicode conversion tables are provided by the
348 * IBM CDRA conversion table registry.
349 * The most prominent example of a systematic modification of conversion tables that is
350 * not provided in the form of conversion table files in the repository is
351 * that S/390 Unix System Services swaps the codes for Line Feed and New Line in all
352 * EBCDIC codepages, which requires such a swap in the Unicode conversion tables as well.
354 * Only IBM default conversion tables are accessible with ucnv_openCCSID().
355 * ucnv_getCCSID() will return the same CCSID for all conversion tables that are associated
358 * Currently, the only "platform" supported in the ICU converter API is UCNV_IBM.
360 * In summary, the use of CCSIDs and the associated API functions is not recommended.
362 * In order to open a converter with the default IBM CDRA Unicode conversion table,
363 * you can use this function or use the prefix "ibm-":
366 * sprintf(name, "ibm-%hu", ccsid);
367 * cnv=ucnv_open(name, &errorCode);
370 * In order to open a converter with the IBM S/390 Unix System Services variant
371 * of a Unicode/EBCDIC conversion table,
372 * you can use the prefix "ibm-" together with the option string UCNV_SWAP_LFNL_OPTION_STRING:
375 * sprintf(name, "ibm-%hu" UCNV_SWAP_LFNL_OPTION_STRING, ccsid);
376 * cnv=ucnv_open(name, &errorCode);
379 * In order to open a converter from a Microsoft codepage number, use the prefix "cp":
382 * sprintf(name, "cp%hu", codepageID);
383 * cnv=ucnv_open(name, &errorCode);
386 * If the alias is ambiguous, then the preferred converter is used
387 * and the status is set to U_AMBIGUOUS_ALIAS_WARNING.
389 * @param codepage codepage number to create
390 * @param platform the platform in which the codepage number exists
391 * @param err error status <TT>U_MEMORY_ALLOCATION_ERROR, U_FILE_ACCESS_ERROR</TT>
392 * @return the created Unicode converter object, or <TT>NULL</TT> if an error
398 * @see ucnv_getPlatform
399 * @see UConverterPlatform
402 U_STABLE UConverter* U_EXPORT2
403 ucnv_openCCSID(int32_t codepage,
404 UConverterPlatform platform,
408 * <p>Creates a UConverter object specified from a packageName and a converterName.</p>
410 * <p>The packageName and converterName must point to an ICU udata object, as defined by
411 * <code> udata_open( packageName, "cnv", converterName, err) </code> or equivalent.
412 * Typically, packageName will refer to a (.dat) file, or to a package registered with
413 * udata_setAppData(). Using a full file or directory pathname for packageName is deprecated.</p>
415 * <p>The name will NOT be looked up in the alias mechanism, nor will the converter be
416 * stored in the converter cache or the alias table. The only way to open further converters
417 * is call this function multiple times, or use the ucnv_safeClone() function to clone a
418 * 'master' converter.</p>
420 * <p>A future version of ICU may add alias table lookups and/or caching
421 * to this function.</p>
424 * <code>cnv = ucnv_openPackage("myapp", "myconverter", &err);</code>
427 * @param packageName name of the package (equivalent to 'path' in udata_open() call)
428 * @param converterName name of the data item to be used, without suffix.
429 * @param err outgoing error status <TT>U_MEMORY_ALLOCATION_ERROR, U_FILE_ACCESS_ERROR</TT>
430 * @return the created Unicode converter object, or <TT>NULL</TT> if an error occured
433 * @see ucnv_safeClone
437 U_STABLE UConverter* U_EXPORT2
438 ucnv_openPackage(const char *packageName, const char *converterName, UErrorCode *err);
441 * Thread safe converter cloning operation.
442 * For most efficient operation, pass in a stackBuffer (and a *pBufferSize)
443 * with at least U_CNV_SAFECLONE_BUFFERSIZE bytes of space.
444 * If the buffer size is sufficient, then the clone will use the stack buffer;
445 * otherwise, it will be allocated, and *pBufferSize will indicate
446 * the actual size. (This should not occur with U_CNV_SAFECLONE_BUFFERSIZE.)
448 * You must ucnv_close() the clone in any case.
450 * If *pBufferSize==0, (regardless of whether stackBuffer==NULL or not)
451 * then *pBufferSize will be changed to a sufficient size
452 * for cloning this converter,
453 * without actually cloning the converter ("pure pre-flighting").
455 * If *pBufferSize is greater than zero but not large enough for a stack-based
456 * clone, then the converter is cloned using newly allocated memory
457 * and *pBufferSize is changed to the necessary size.
459 * If the converter clone fits into the stack buffer but the stack buffer is not
460 * sufficiently aligned for the clone, then the clone will use an
461 * adjusted pointer and use an accordingly smaller buffer size.
463 * @param cnv converter to be cloned
464 * @param stackBuffer user allocated space for the new clone. If NULL new memory will be allocated.
465 * If buffer is not large enough, new memory will be allocated.
466 * Clients can use the U_CNV_SAFECLONE_BUFFERSIZE. This will probably be enough to avoid memory allocations.
467 * @param pBufferSize pointer to size of allocated space. pBufferSize must not be NULL.
468 * @param status to indicate whether the operation went on smoothly or there were errors
469 * An informational status value, U_SAFECLONE_ALLOCATED_WARNING,
470 * is used if any allocations were necessary.
471 * However, it is better to check if *pBufferSize grew for checking for
472 * allocations because warning codes can be overridden by subsequent
474 * @return pointer to the new clone
477 U_STABLE UConverter * U_EXPORT2
478 ucnv_safeClone(const UConverter *cnv,
480 int32_t *pBufferSize,
484 * \def U_CNV_SAFECLONE_BUFFERSIZE
485 * Definition of a buffer size that is designed to be large enough for
486 * converters to be cloned with ucnv_safeClone().
489 #define U_CNV_SAFECLONE_BUFFERSIZE 1024
492 * Deletes the unicode converter and releases resources associated
493 * with just this instance.
494 * Does not free up shared converter tables.
496 * @param converter the converter object to be deleted
499 * @see ucnv_openCCSID
502 U_STABLE void U_EXPORT2
503 ucnv_close(UConverter * converter);
506 * Fills in the output parameter, subChars, with the substitution characters
509 * @param converter the Unicode converter
510 * @param subChars the subsitution characters
511 * @param len on input the capacity of subChars, on output the number
512 * of bytes copied to it
513 * @param err the outgoing error status code.
514 * If the substitution character array is too small, an
515 * <TT>U_INDEX_OUTOFBOUNDS_ERROR</TT> will be returned.
516 * @see ucnv_setSubstChars
519 U_STABLE void U_EXPORT2
520 ucnv_getSubstChars(const UConverter *converter,
526 * Sets the substitution chars when converting from unicode to a codepage. The
527 * substitution is specified as a string of 1-4 bytes, and may contain
528 * <TT>NULL</TT> byte.
529 * @param converter the Unicode converter
530 * @param subChars the substitution character byte sequence we want set
531 * @param len the number of bytes in subChars
532 * @param err the error status code. <TT>U_INDEX_OUTOFBOUNDS_ERROR </TT> if
533 * len is bigger than the maximum number of bytes allowed in subchars
534 * @see ucnv_getSubstChars
537 U_STABLE void U_EXPORT2
538 ucnv_setSubstChars(UConverter *converter,
539 const char *subChars,
544 * Fills in the output parameter, errBytes, with the error characters from the
545 * last failing conversion.
547 * @param converter the Unicode converter
548 * @param errBytes the codepage bytes which were in error
549 * @param len on input the capacity of errBytes, on output the number of
550 * bytes which were copied to it
551 * @param err the error status code.
552 * If the substitution character array is too small, an
553 * <TT>U_INDEX_OUTOFBOUNDS_ERROR</TT> will be returned.
556 U_STABLE void U_EXPORT2
557 ucnv_getInvalidChars(const UConverter *converter,
563 * Fills in the output parameter, errChars, with the error characters from the
564 * last failing conversion.
566 * @param converter the Unicode converter
567 * @param errUChars the UChars which were in error
568 * @param len on input the capacity of errUChars, on output the number of
569 * UChars which were copied to it
570 * @param err the error status code.
571 * If the substitution character array is too small, an
572 * <TT>U_INDEX_OUTOFBOUNDS_ERROR</TT> will be returned.
575 U_STABLE void U_EXPORT2
576 ucnv_getInvalidUChars(const UConverter *converter,
582 * Resets the state of a converter to the default state. This is used
583 * in the case of an error, to restart a conversion from a known default state.
584 * It will also empty the internal output buffers.
585 * @param converter the Unicode converter
588 U_STABLE void U_EXPORT2
589 ucnv_reset(UConverter *converter);
592 * Resets the to-Unicode part of a converter state to the default state.
593 * This is used in the case of an error to restart a conversion to
594 * Unicode to a known default state. It will also empty the internal
595 * output buffers used for the conversion to Unicode codepoints.
596 * @param converter the Unicode converter
599 U_STABLE void U_EXPORT2
600 ucnv_resetToUnicode(UConverter *converter);
603 * Resets the from-Unicode part of a converter state to the default state.
604 * This is used in the case of an error to restart a conversion from
605 * Unicode to a known default state. It will also empty the internal output
606 * buffers used for the conversion from Unicode codepoints.
607 * @param converter the Unicode converter
610 U_STABLE void U_EXPORT2
611 ucnv_resetFromUnicode(UConverter *converter);
614 * Returns the maximum number of bytes that are output per UChar in conversion
615 * from Unicode using this converter.
616 * The returned number can be used with UCNV_GET_MAX_BYTES_FOR_STRING
617 * to calculate the size of a target buffer for conversion from Unicode.
619 * Note: Before ICU 2.8, this function did not return reliable numbers for
620 * some stateful converters (EBCDIC_STATEFUL, ISO-2022) and LMBCS.
622 * This number may not be the same as the maximum number of bytes per
623 * "conversion unit". In other words, it may not be the intuitively expected
624 * number of bytes per character that would be published for a charset,
625 * and may not fulfill any other purpose than the allocation of an output
626 * buffer of guaranteed sufficient size for a given input length and converter.
628 * Examples for special cases that are taken into account:
629 * - Supplementary code points may convert to more bytes than BMP code points.
630 * This function returns bytes per UChar (UTF-16 code unit), not per
631 * Unicode code point, for efficient buffer allocation.
632 * - State-shifting output (SI/SO, escapes, etc.) from stateful converters.
633 * - When m input UChars are converted to n output bytes, then the maximum m/n
634 * is taken into account.
636 * The number returned here does not take into account
637 * (see UCNV_GET_MAX_BYTES_FOR_STRING):
638 * - callbacks which output more than one charset character sequence per call,
639 * like escape callbacks
640 * - initial and final non-character bytes that are output by some converters
641 * (automatic BOMs, initial escape sequence, final SI, etc.)
643 * Examples for returned values:
646 * - UTF-16: 2 (2 per BMP, 4 per surrogate _pair_, BOM not counted)
647 * - UTF-8: 3 (3 per BMP, 4 per surrogate _pair_)
648 * - EBCDIC_STATEFUL (EBCDIC mixed SBCS/DBCS): 3 (SO + DBCS)
649 * - ISO-2022: 3 (always outputs UTF-8)
650 * - ISO-2022-JP: 6 (4-byte escape sequences + DBCS)
651 * - ISO-2022-CN: 8 (4-byte designator sequences + 2-byte SS2/SS3 + DBCS)
653 * @param converter The Unicode converter.
654 * @return The maximum number of bytes per UChar that are output by ucnv_fromUnicode(),
655 * to be used together with UCNV_GET_MAX_BYTES_FOR_STRING for buffer allocation.
657 * @see UCNV_GET_MAX_BYTES_FOR_STRING
658 * @see ucnv_getMinCharSize
661 U_STABLE int8_t U_EXPORT2
662 ucnv_getMaxCharSize(const UConverter *converter);
664 #ifndef U_HIDE_DRAFT_API
667 * Calculates the size of a buffer for conversion from Unicode to a charset.
668 * The calculated size is guaranteed to be sufficient for this conversion.
670 * It takes into account initial and final non-character bytes that are output
671 * by some converters.
672 * It does not take into account callbacks which output more than one charset
673 * character sequence per call, like escape callbacks.
674 * The default (substitution) callback only outputs one charset character sequence.
676 * @param length Number of UChars to be converted.
677 * @param maxCharSize Return value from ucnv_getMaxCharSize() for the converter
679 * @return Size of a buffer that will be large enough to hold the output bytes of
680 * converting length UChars with the converter that returned the maxCharSize.
682 * @see ucnv_getMaxCharSize
685 #define UCNV_GET_MAX_BYTES_FOR_STRING(length, maxCharSize) \
686 (((int32_t)(length)+10)*(int32_t)(maxCharSize))
688 #endif /*U_HIDE_DRAFT_API*/
691 * Returns the minimum byte length for characters in this codepage.
692 * This is usually either 1 or 2.
693 * @param converter the Unicode converter
694 * @return the minimum number of bytes allowed by this particular converter
695 * @see ucnv_getMaxCharSize
698 U_STABLE int8_t U_EXPORT2
699 ucnv_getMinCharSize(const UConverter *converter);
702 * Returns the display name of the converter passed in based on the Locale
703 * passed in. If the locale contains no display name, the internal ASCII
704 * name will be filled in.
706 * @param converter the Unicode converter.
707 * @param displayLocale is the specific Locale we want to localised for
708 * @param displayName user provided buffer to be filled in
709 * @param displayNameCapacity size of displayName Buffer
710 * @param err error status code
711 * @return displayNameLength number of UChar needed in displayName
715 U_STABLE int32_t U_EXPORT2
716 ucnv_getDisplayName(const UConverter *converter,
717 const char *displayLocale,
719 int32_t displayNameCapacity,
723 * Gets the internal, canonical name of the converter (zero-terminated).
724 * The lifetime of the returned string will be that of the converter
725 * passed to this function.
726 * @param converter the Unicode converter
727 * @param err UErrorCode status
728 * @return the internal name of the converter
729 * @see ucnv_getDisplayName
732 U_STABLE const char * U_EXPORT2
733 ucnv_getName(const UConverter *converter, UErrorCode *err);
736 * Gets a codepage number associated with the converter. This is not guaranteed
737 * to be the one used to create the converter. Some converters do not represent
738 * platform registered codepages and return zero for the codepage number.
739 * The error code fill-in parameter indicates if the codepage number
741 * Does not check if the converter is <TT>NULL</TT> or if converter's data
742 * table is <TT>NULL</TT>.
744 * Important: The use of CCSIDs is not recommended because it is limited
745 * to only two platforms in principle and only one (UCNV_IBM) in the current
747 * Also, CCSIDs are insufficient to identify IBM Unicode conversion tables precisely.
748 * For more details see ucnv_openCCSID().
750 * @param converter the Unicode converter
751 * @param err the error status code.
752 * @return If any error occurrs, -1 will be returned otherwise, the codepage number
754 * @see ucnv_openCCSID
755 * @see ucnv_getPlatform
758 U_STABLE int32_t U_EXPORT2
759 ucnv_getCCSID(const UConverter *converter,
763 * Gets a codepage platform associated with the converter. Currently,
764 * only <TT>UCNV_IBM</TT> will be returned.
765 * Does not test if the converter is <TT>NULL</TT> or if converter's data
766 * table is <TT>NULL</TT>.
767 * @param converter the Unicode converter
768 * @param err the error status code.
769 * @return The codepage platform
772 U_STABLE UConverterPlatform U_EXPORT2
773 ucnv_getPlatform(const UConverter *converter,
777 * Gets the type of the converter
778 * e.g. SBCS, MBCS, DBCS, UTF8, UTF16_BE, UTF16_LE, ISO_2022,
779 * EBCDIC_STATEFUL, LATIN_1
780 * @param converter a valid, opened converter
781 * @return the type of the converter
784 U_STABLE UConverterType U_EXPORT2
785 ucnv_getType(const UConverter * converter);
788 * Gets the "starter" (lead) bytes for converters of type MBCS.
789 * Will fill in an <TT>U_ILLEGAL_ARGUMENT_ERROR</TT> if converter passed in
790 * is not MBCS. Fills in an array of type UBool, with the value of the byte
791 * as offset to the array. For example, if (starters[0x20] == TRUE) at return,
792 * it means that the byte 0x20 is a starter byte in this converter.
793 * Context pointers are always owned by the caller.
795 * @param converter a valid, opened converter of type MBCS
796 * @param starters an array of size 256 to be filled in
797 * @param err error status, <TT>U_ILLEGAL_ARGUMENT_ERROR</TT> if the
798 * converter is not a type which can return starters.
802 U_STABLE void U_EXPORT2
803 ucnv_getStarters(const UConverter* converter,
809 * Selectors for Unicode sets that can be returned by ucnv_getUnicodeSet().
810 * @see ucnv_getUnicodeSet
813 typedef enum UConverterUnicodeSet {
814 /** Select the set of roundtrippable Unicode code points. @stable ICU 2.6 */
816 /** Number of UConverterUnicodeSet selectors. @stable ICU 2.6 */
818 } UConverterUnicodeSet;
822 * Returns the set of Unicode code points that can be converted by an ICU converter.
824 * The current implementation returns only one kind of set (UCNV_ROUNDTRIP_SET):
825 * The set of all Unicode code points that can be roundtrip-converted
826 * (converted without any data loss) with the converter.
827 * This set will not include code points that have fallback mappings
828 * or are only the result of reverse fallback mappings.
829 * See UTR #22 "Character Mapping Markup Language"
830 * at http://www.unicode.org/reports/tr22/
832 * This is useful for example for
833 * - checking that a string or document can be roundtrip-converted with a converter,
834 * without/before actually performing the conversion
835 * - testing if a converter can be used for text for typical text for a certain locale,
836 * by comparing its roundtrip set with the set of ExemplarCharacters from
837 * ICU's locale data or other sources
839 * In the future, there may be more UConverterUnicodeSet choices to select
840 * sets with different properties.
842 * @param cnv The converter for which a set is requested.
843 * @param setFillIn A valid USet *. It will be cleared by this function before
844 * the converter's specific set is filled into the USet.
845 * @param whichSet A UConverterUnicodeSet selector;
846 * currently UCNV_ROUNDTRIP_SET is the only supported value.
847 * @param pErrorCode ICU error code in/out parameter.
848 * Must fulfill U_SUCCESS before the function call.
850 * @see UConverterUnicodeSet
855 U_STABLE void U_EXPORT2
856 ucnv_getUnicodeSet(const UConverter *cnv,
858 UConverterUnicodeSet whichSet,
859 UErrorCode *pErrorCode);
862 * Gets the current calback function used by the converter when an illegal
863 * or invalid codepage sequence is found.
864 * Context pointers are always owned by the caller.
866 * @param converter the unicode converter
867 * @param action fillin: returns the callback function pointer
868 * @param context fillin: returns the callback's private void* context
869 * @see ucnv_setToUCallBack
872 U_STABLE void U_EXPORT2
873 ucnv_getToUCallBack (const UConverter * converter,
874 UConverterToUCallback *action,
875 const void **context);
878 * Gets the current callback function used by the converter when illegal
879 * or invalid Unicode sequence is found.
880 * Context pointers are always owned by the caller.
882 * @param converter the unicode converter
883 * @param action fillin: returns the callback function pointer
884 * @param context fillin: returns the callback's private void* context
885 * @see ucnv_setFromUCallBack
888 U_STABLE void U_EXPORT2
889 ucnv_getFromUCallBack (const UConverter * converter,
890 UConverterFromUCallback *action,
891 const void **context);
894 * Changes the callback function used by the converter when
895 * an illegal or invalid sequence is found.
896 * Context pointers are always owned by the caller.
897 * Predefined actions and contexts can be found in the ucnv_err.h header.
899 * @param converter the unicode converter
900 * @param newAction the new callback function
901 * @param newContext the new toUnicode callback context pointer. This can be NULL.
902 * @param oldAction fillin: returns the old callback function pointer. This can be NULL.
903 * @param oldContext fillin: returns the old callback's private void* context. This can be NULL.
904 * @param err The error code status
905 * @see ucnv_getToUCallBack
908 U_STABLE void U_EXPORT2
909 ucnv_setToUCallBack (UConverter * converter,
910 UConverterToUCallback newAction,
911 const void* newContext,
912 UConverterToUCallback *oldAction,
913 const void** oldContext,
917 * Changes the current callback function used by the converter when
918 * an illegal or invalid sequence is found.
919 * Context pointers are always owned by the caller.
920 * Predefined actions and contexts can be found in the ucnv_err.h header.
922 * @param converter the unicode converter
923 * @param newAction the new callback function
924 * @param newContext the new fromUnicode callback context pointer. This can be NULL.
925 * @param oldAction fillin: returns the old callback function pointer. This can be NULL.
926 * @param oldContext fillin: returns the old callback's private void* context. This can be NULL.
927 * @param err The error code status
928 * @see ucnv_getFromUCallBack
931 U_STABLE void U_EXPORT2
932 ucnv_setFromUCallBack (UConverter * converter,
933 UConverterFromUCallback newAction,
934 const void *newContext,
935 UConverterFromUCallback *oldAction,
936 const void **oldContext,
940 * Converts an array of unicode characters to an array of codepage
941 * characters. This function is optimized for converting a continuous
942 * stream of data in buffer-sized chunks, where the entire source and
943 * target does not fit in available buffers.
945 * The source pointer is an in/out parameter. It starts out pointing where the
946 * conversion is to begin, and ends up pointing after the last UChar consumed.
948 * Target similarly starts out pointer at the first available byte in the output
949 * buffer, and ends up pointing after the last byte written to the output.
951 * The converter always attempts to consume the entire source buffer, unless
952 * (1.) the target buffer is full, or (2.) a failing error is returned from the
953 * current callback function. When a successful error status has been
954 * returned, it means that all of the source buffer has been
955 * consumed. At that point, the caller should reset the source and
956 * sourceLimit pointers to point to the next chunk.
958 * At the end of the stream (flush==TRUE), the input is completely consumed
959 * when *source==sourceLimit and no error code is set.
960 * The converter object is then automatically reset by this function.
961 * (This means that a converter need not be reset explicitly between data
962 * streams if it finishes the previous stream without errors.)
964 * This is a <I>stateful</I> conversion. Additionally, even when all source data has
965 * been consumed, some data may be in the converters' internal state.
966 * Call this function repeatedly, updating the target pointers with
967 * the next empty chunk of target in case of a
968 * <TT>U_BUFFER_OVERFLOW_ERROR</TT>, and updating the source pointers
969 * with the next chunk of source when a successful error status is
970 * returned, until there are no more chunks of source data.
971 * @param converter the Unicode converter
972 * @param target I/O parameter. Input : Points to the beginning of the buffer to copy
973 * codepage characters to. Output : points to after the last codepage character copied
974 * to <TT>target</TT>.
975 * @param targetLimit the pointer just after last of the <TT>target</TT> buffer
976 * @param source I/O parameter, pointer to pointer to the source Unicode character buffer.
977 * @param sourceLimit the pointer just after the last of the source buffer
978 * @param offsets if NULL is passed, nothing will happen to it, otherwise it needs to have the same number
979 * of allocated cells as <TT>target</TT>. Will fill in offsets from target to source pointer
980 * 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>
981 * For output data carried across calls, and other data without a specific source character
982 * (such as from escape sequences or callbacks) -1 will be placed for offsets.
983 * @param flush set to <TT>TRUE</TT> if the current source buffer is the last available
984 * chunk of the source, <TT>FALSE</TT> otherwise. Note that if a failing status is returned,
985 * this function may have to be called multiple times with flush set to <TT>TRUE</TT> until
986 * the source buffer is consumed.
987 * @param err the error status. <TT>U_ILLEGAL_ARGUMENT_ERROR</TT> will be set if the
988 * converter is <TT>NULL</TT>.
989 * <code>U_BUFFER_OVERFLOW_ERROR</code> will be set if the target is full and there is
990 * still data to be written to the target.
991 * @see ucnv_fromUChars
993 * @see ucnv_getMinCharSize
994 * @see ucnv_setToUCallBack
997 U_STABLE void U_EXPORT2
998 ucnv_fromUnicode (UConverter * converter,
1000 const char *targetLimit,
1001 const UChar ** source,
1002 const UChar * sourceLimit,
1008 * Converts a buffer of codepage bytes into an array of unicode UChars
1009 * characters. This function is optimized for converting a continuous
1010 * stream of data in buffer-sized chunks, where the entire source and
1011 * target does not fit in available buffers.
1013 * The source pointer is an in/out parameter. It starts out pointing where the
1014 * conversion is to begin, and ends up pointing after the last byte of source consumed.
1016 * Target similarly starts out pointer at the first available UChar in the output
1017 * buffer, and ends up pointing after the last UChar written to the output.
1018 * It does NOT necessarily keep UChar sequences together.
1020 * The converter always attempts to consume the entire source buffer, unless
1021 * (1.) the target buffer is full, or (2.) a failing error is returned from the
1022 * current callback function. When a successful error status has been
1023 * returned, it means that all of the source buffer has been
1024 * consumed. At that point, the caller should reset the source and
1025 * sourceLimit pointers to point to the next chunk.
1027 * At the end of the stream (flush==TRUE), the input is completely consumed
1028 * when *source==sourceLimit and no error code is set
1029 * The converter object is then automatically reset by this function.
1030 * (This means that a converter need not be reset explicitly between data
1031 * streams if it finishes the previous stream without errors.)
1033 * This is a <I>stateful</I> conversion. Additionally, even when all source data has
1034 * been consumed, some data may be in the converters' internal state.
1035 * Call this function repeatedly, updating the target pointers with
1036 * the next empty chunk of target in case of a
1037 * <TT>U_BUFFER_OVERFLOW_ERROR</TT>, and updating the source pointers
1038 * with the next chunk of source when a successful error status is
1039 * returned, until there are no more chunks of source data.
1040 * @param converter the Unicode converter
1041 * @param target I/O parameter. Input : Points to the beginning of the buffer to copy
1042 * UChars into. Output : points to after the last UChar copied.
1043 * @param targetLimit the pointer just after the end of the <TT>target</TT> buffer
1044 * @param source I/O parameter, pointer to pointer to the source codepage buffer.
1045 * @param sourceLimit the pointer to the byte after the end of the source buffer
1046 * @param offsets if NULL is passed, nothing will happen to it, otherwise it needs to have the same number
1047 * of allocated cells as <TT>target</TT>. Will fill in offsets from target to source pointer
1048 * 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>
1049 * For output data carried across calls, and other data without a specific source character
1050 * (such as from escape sequences or callbacks) -1 will be placed for offsets.
1051 * @param flush set to <TT>TRUE</TT> if the current source buffer is the last available
1052 * chunk of the source, <TT>FALSE</TT> otherwise. Note that if a failing status is returned,
1053 * this function may have to be called multiple times with flush set to <TT>TRUE</TT> until
1054 * the source buffer is consumed.
1055 * @param err the error status. <TT>U_ILLEGAL_ARGUMENT_ERROR</TT> will be set if the
1056 * converter is <TT>NULL</TT>.
1057 * <code>U_BUFFER_OVERFLOW_ERROR</code> will be set if the target is full and there is
1058 * still data to be written to the target.
1059 * @see ucnv_fromUChars
1061 * @see ucnv_getMinCharSize
1062 * @see ucnv_setFromUCallBack
1063 * @see ucnv_getNextUChar
1066 U_STABLE void U_EXPORT2
1067 ucnv_toUnicode(UConverter *converter,
1069 const UChar *targetLimit,
1070 const char **source,
1071 const char *sourceLimit,
1077 * Convert the Unicode string into a codepage string using an existing UConverter.
1078 * The output string is NUL-terminated if possible.
1080 * This function is a more convenient but less powerful version of ucnv_fromUnicode().
1081 * It is only useful for whole strings, not for streaming conversion.
1083 * The maximum output buffer capacity required (barring output from callbacks) will be
1084 * UCNV_GET_MAX_BYTES_FOR_STRING(srcLength, ucnv_getMaxCharSize(cnv)).
1086 * @param cnv the converter object to be used (ucnv_resetFromUnicode() will be called)
1087 * @param src the input Unicode string
1088 * @param srcLength the input string length, or -1 if NUL-terminated
1089 * @param dest destination string buffer, can be NULL if destCapacity==0
1090 * @param destCapacity the number of chars available at dest
1091 * @param pErrorCode normal ICU error code;
1092 * common error codes that may be set by this function include
1093 * U_BUFFER_OVERFLOW_ERROR, U_STRING_NOT_TERMINATED_WARNING,
1094 * U_ILLEGAL_ARGUMENT_ERROR, and conversion errors
1095 * @return the length of the output string, not counting the terminating NUL;
1096 * if the length is greater than destCapacity, then the string will not fit
1097 * and a buffer of the indicated length would need to be passed in
1098 * @see ucnv_fromUnicode
1100 * @see UCNV_GET_MAX_BYTES_FOR_STRING
1103 U_STABLE int32_t U_EXPORT2
1104 ucnv_fromUChars(UConverter *cnv,
1105 char *dest, int32_t destCapacity,
1106 const UChar *src, int32_t srcLength,
1107 UErrorCode *pErrorCode);
1110 * Convert the codepage string into a Unicode string using an existing UConverter.
1111 * The output string is NUL-terminated if possible.
1113 * This function is a more convenient but less powerful version of ucnv_toUnicode().
1114 * It is only useful for whole strings, not for streaming conversion.
1116 * The maximum output buffer capacity required (barring output from callbacks) will be
1117 * 2*srcLength (each char may be converted into a surrogate pair).
1119 * @param cnv the converter object to be used (ucnv_resetToUnicode() will be called)
1120 * @param src the input codepage string
1121 * @param srcLength the input string length, or -1 if NUL-terminated
1122 * @param dest destination string buffer, can be NULL if destCapacity==0
1123 * @param destCapacity the number of UChars available at dest
1124 * @param pErrorCode normal ICU error code;
1125 * common error codes that may be set by this function include
1126 * U_BUFFER_OVERFLOW_ERROR, U_STRING_NOT_TERMINATED_WARNING,
1127 * U_ILLEGAL_ARGUMENT_ERROR, and conversion errors
1128 * @return the length of the output string, not counting the terminating NUL;
1129 * if the length is greater than destCapacity, then the string will not fit
1130 * and a buffer of the indicated length would need to be passed in
1131 * @see ucnv_toUnicode
1135 U_STABLE int32_t U_EXPORT2
1136 ucnv_toUChars(UConverter *cnv,
1137 UChar *dest, int32_t destCapacity,
1138 const char *src, int32_t srcLength,
1139 UErrorCode *pErrorCode);
1142 * Convert a codepage buffer into Unicode one character at a time.
1143 * The input is completely consumed when the U_INDEX_OUTOFBOUNDS_ERROR is set.
1145 * Advantage compared to ucnv_toUnicode() or ucnv_toUChars():
1146 * - Faster for small amounts of data, for most converters, e.g.,
1147 * US-ASCII, ISO-8859-1, UTF-8/16/32, and most "normal" charsets.
1148 * (For complex converters, e.g., SCSU, UTF-7 and ISO 2022 variants,
1149 * it uses ucnv_toUnicode() internally.)
1152 * Limitations compared to ucnv_toUnicode():
1153 * - Always assumes flush=TRUE.
1154 * This makes ucnv_getNextUChar() unsuitable for "streaming" conversion,
1155 * that is, for where the input is supplied in multiple buffers,
1156 * because ucnv_getNextUChar() will assume the end of the input at the end
1157 * of the first buffer.
1158 * - Does not provide offset output.
1160 * It is possible to "mix" ucnv_getNextUChar() and ucnv_toUnicode() because
1161 * ucnv_getNextUChar() uses the current state of the converter
1162 * (unlike ucnv_toUChars() which always resets first).
1163 * However, if ucnv_getNextUChar() is called after ucnv_toUnicode()
1164 * stopped in the middle of a character sequence (with flush=FALSE),
1165 * then ucnv_getNextUChar() will always use the slower ucnv_toUnicode()
1166 * internally until the next character boundary.
1167 * (This is new in ICU 2.6. In earlier releases, ucnv_getNextUChar() had to
1168 * start at a character boundary.)
1170 * Instead of using ucnv_getNextUChar(), it is recommended
1171 * to convert using ucnv_toUnicode() or ucnv_toUChars()
1172 * and then iterate over the text using U16_NEXT() or a UCharIterator (uiter.h)
1173 * or a C++ CharacterIterator or similar.
1174 * This allows streaming conversion and offset output, for example.
1176 * <p>Handling of surrogate pairs and supplementary-plane code points:<br>
1177 * There are two different kinds of codepages that provide mappings for surrogate characters:
1179 * <li>Codepages like UTF-8, UTF-32, and GB 18030 provide direct representations for Unicode
1180 * code points U+10000-U+10ffff as well as for single surrogates U+d800-U+dfff.
1181 * Each valid sequence will result in exactly one returned code point.
1182 * If a sequence results in a single surrogate, then that will be returned
1183 * by itself, even if a neighboring sequence encodes the matching surrogate.</li>
1184 * <li>Codepages like SCSU and LMBCS (and UTF-16) provide direct representations only for BMP code points
1185 * including surrogates. Code points in supplementary planes are represented with
1186 * two sequences, each encoding a surrogate.
1187 * For these codepages, matching pairs of surrogates will be combined into single
1188 * code points for returning from this function.
1189 * (Note that SCSU is actually a mix of these codepage types.)</li>
1192 * @param converter an open UConverter
1193 * @param source the address of a pointer to the codepage buffer, will be
1194 * updated to point after the bytes consumed in the conversion call.
1195 * @param sourceLimit points to the end of the input buffer
1196 * @param err fills in error status (see ucnv_toUnicode)
1197 * <code>U_INDEX_OUTOFBOUNDS_ERROR</code> will be set if the input
1198 * is empty or does not convert to any output (e.g.: pure state-change
1199 * codes SI/SO, escape sequences for ISO 2022,
1200 * or if the callback did not output anything, ...).
1201 * This function will not set a <code>U_BUFFER_OVERFLOW_ERROR</code> because
1202 * the "buffer" is the return code. However, there might be subsequent output
1203 * stored in the converter object
1204 * that will be returned in following calls to this function.
1205 * @return a UChar32 resulting from the partial conversion of source
1206 * @see ucnv_toUnicode
1207 * @see ucnv_toUChars
1211 U_STABLE UChar32 U_EXPORT2
1212 ucnv_getNextUChar(UConverter * converter,
1213 const char **source,
1214 const char * sourceLimit,
1218 * Convert from one external charset to another using two existing UConverters.
1219 * Internally, two conversions - ucnv_toUnicode() and ucnv_fromUnicode() -
1220 * are used, "pivoting" through 16-bit Unicode.
1222 * There is a similar function, ucnv_convert(),
1223 * which has the following limitations:
1224 * - it takes charset names, not converter objects, so that
1225 * - two converters are opened for each call
1226 * - only single-string conversion is possible, not streaming operation
1227 * - it does not provide enough information to find out,
1228 * in case of failure, whether the toUnicode or
1229 * the fromUnicode conversion failed
1231 * By contrast, ucnv_convertEx()
1232 * - takes UConverter parameters instead of charset names
1233 * - fully exposes the pivot buffer for complete error handling
1235 * ucnv_convertEx() also provides further convenience:
1236 * - an option to reset the converters at the beginning
1237 * (if reset==TRUE, see parameters;
1238 * also sets *pivotTarget=*pivotSource=pivotStart)
1239 * - allow NUL-terminated input
1240 * (only a single NUL byte, will not work for charsets with multi-byte NULs)
1241 * (if sourceLimit==NULL, see parameters)
1242 * - terminate with a NUL on output
1243 * (only a single NUL byte, not useful for charsets with multi-byte NULs),
1244 * or set U_STRING_NOT_TERMINATED_WARNING if the output exactly fills
1246 * - the pivot buffer can be provided internally;
1247 * in this case, the caller will not be able to get details about where an
1249 * (if pivotStart==NULL, see below)
1251 * The function returns when one of the following is true:
1252 * - the entire source text has been converted successfully to the target buffer
1253 * - a target buffer overflow occurred (U_BUFFER_OVERFLOW_ERROR)
1254 * - a conversion error occurred
1255 * (other U_FAILURE(), see description of pErrorCode)
1257 * Limitation compared to the direct use of
1258 * ucnv_fromUnicode() and ucnv_toUnicode():
1259 * ucnv_convertEx() does not provide offset information.
1261 * Limitation compared to ucnv_fromUChars() and ucnv_toUChars():
1262 * ucnv_convertEx() does not support preflighting directly.
1264 * Sample code for converting a single string from
1265 * one external charset to UTF-8, ignoring the location of errors:
1269 * myToUTF8(UConverter *cnv,
1270 * const char *s, int32_t length,
1271 * char *u8, int32_t capacity,
1272 * UErrorCode *pErrorCode) {
1273 * UConverter *utf8Cnv;
1276 * if(U_FAILURE(*pErrorCode)) {
1280 * utf8Cnv=myGetCachedUTF8Converter(pErrorCode);
1281 * if(U_FAILURE(*pErrorCode)) {
1286 * ucnv_convertEx(cnv, utf8Cnv,
1287 * &target, u8+capacity,
1288 * &s, length>=0 ? s+length : NULL,
1289 * NULL, NULL, NULL, NULL,
1293 * myReleaseCachedUTF8Converter(utf8Cnv);
1295 * // return the output string length, but without preflighting
1296 * return (int32_t)(target-u8);
1300 * @param targetCnv Output converter, used to convert from the UTF-16 pivot
1301 * to the target using ucnv_fromUnicode().
1302 * @param sourceCnv Input converter, used to convert from the source to
1303 * the UTF-16 pivot using ucnv_toUnicode().
1304 * @param target I/O parameter, same as for ucnv_fromUChars().
1305 * Input: *target points to the beginning of the target buffer.
1306 * Output: *target points to the first unit after the last char written.
1307 * @param targetLimit Pointer to the first unit after the target buffer.
1308 * @param source I/O parameter, same as for ucnv_toUChars().
1309 * Input: *source points to the beginning of the source buffer.
1310 * Output: *source points to the first unit after the last char read.
1311 * @param sourceLimit Pointer to the first unit after the source buffer.
1312 * @param pivotStart Pointer to the UTF-16 pivot buffer. If pivotStart==NULL,
1313 * then an internal buffer is used and the other pivot
1314 * arguments are ignored and can be NULL as well.
1315 * @param pivotSource I/O parameter, same as source in ucnv_fromUChars() for
1316 * conversion from the pivot buffer to the target buffer.
1317 * @param pivotTarget I/O parameter, same as target in ucnv_toUChars() for
1318 * conversion from the source buffer to the pivot buffer.
1319 * It must be pivotStart<=*pivotSource<=*pivotTarget<=pivotLimit
1320 * and pivotStart<pivotLimit (unless pivotStart==NULL).
1321 * @param pivotLimit Pointer to the first unit after the pivot buffer.
1322 * @param reset If TRUE, then ucnv_resetToUnicode(sourceCnv) and
1323 * ucnv_resetFromUnicode(targetCnv) are called, and the
1324 * pivot pointers are reset (*pivotTarget=*pivotSource=pivotStart).
1325 * @param flush If true, indicates the end of the input.
1326 * Passed directly to ucnv_toUnicode(), and carried over to
1327 * ucnv_fromUnicode() when the source is empty as well.
1328 * @param pErrorCode ICU error code in/out parameter.
1329 * Must fulfill U_SUCCESS before the function call.
1330 * U_BUFFER_OVERFLOW_ERROR always refers to the target buffer
1331 * because overflows into the pivot buffer are handled internally.
1332 * Other conversion errors are from the source-to-pivot
1333 * conversion if *pivotSource==pivotStart, otherwise from
1334 * the pivot-to-target conversion.
1337 * @see ucnv_fromAlgorithmic
1338 * @see ucnv_toAlgorithmic
1339 * @see ucnv_fromUnicode
1340 * @see ucnv_toUnicode
1341 * @see ucnv_fromUChars
1342 * @see ucnv_toUChars
1345 U_STABLE void U_EXPORT2
1346 ucnv_convertEx(UConverter *targetCnv, UConverter *sourceCnv,
1347 char **target, const char *targetLimit,
1348 const char **source, const char *sourceLimit,
1349 UChar *pivotStart, UChar **pivotSource,
1350 UChar **pivotTarget, const UChar *pivotLimit,
1351 UBool reset, UBool flush,
1352 UErrorCode *pErrorCode);
1355 * Convert from one external charset to another.
1356 * Internally, two converters are opened according to the name arguments,
1357 * then the text is converted to and from the 16-bit Unicode "pivot"
1358 * using ucnv_convertEx(), then the converters are closed again.
1360 * This is a convenience function, not an efficient way to convert a lot of text:
1362 * - takes charset names, not converter objects, so that
1363 * - two converters are opened for each call
1364 * - only single-string conversion is possible, not streaming operation
1365 * - does not provide enough information to find out,
1366 * in case of failure, whether the toUnicode or
1367 * the fromUnicode conversion failed
1368 * - allows NUL-terminated input
1369 * (only a single NUL byte, will not work for charsets with multi-byte NULs)
1370 * (if sourceLength==-1, see parameters)
1371 * - terminate with a NUL on output
1372 * (only a single NUL byte, not useful for charsets with multi-byte NULs),
1373 * or set U_STRING_NOT_TERMINATED_WARNING if the output exactly fills
1375 * - a pivot buffer is provided internally
1377 * The function returns when one of the following is true:
1378 * - the entire source text has been converted successfully to the target buffer
1379 * and either the target buffer is terminated with a single NUL byte
1380 * or the error code is set to U_STRING_NOT_TERMINATED_WARNING
1381 * - a target buffer overflow occurred (U_BUFFER_OVERFLOW_ERROR)
1382 * and the full output string length is returned ("preflighting")
1383 * - a conversion error occurred
1384 * (other U_FAILURE(), see description of pErrorCode)
1386 * @param toConverterName The name of the converter that is used to convert
1387 * from the UTF-16 pivot buffer to the target.
1388 * @param fromConverterName The name of the converter that is used to convert
1389 * from the source to the UTF-16 pivot buffer.
1390 * @param target Pointer to the output buffer.
1391 * @param targetCapacity Capacity of the target, in bytes.
1392 * @param source Pointer to the input buffer.
1393 * @param sourceLength Length of the input text, in bytes, or -1 for NUL-terminated input.
1394 * @param pErrorCode ICU error code in/out parameter.
1395 * Must fulfill U_SUCCESS before the function call.
1396 * @return Length of the complete output text in bytes, even if it exceeds the targetCapacity
1397 * and a U_BUFFER_OVERFLOW_ERROR is set.
1399 * @see ucnv_convertEx
1400 * @see ucnv_fromAlgorithmic
1401 * @see ucnv_toAlgorithmic
1402 * @see ucnv_fromUnicode
1403 * @see ucnv_toUnicode
1404 * @see ucnv_fromUChars
1405 * @see ucnv_toUChars
1406 * @see ucnv_getNextUChar
1409 U_STABLE int32_t U_EXPORT2
1410 ucnv_convert(const char *toConverterName,
1411 const char *fromConverterName,
1413 int32_t targetCapacity,
1415 int32_t sourceLength,
1416 UErrorCode *pErrorCode);
1419 * Convert from one external charset to another.
1420 * Internally, the text is converted to and from the 16-bit Unicode "pivot"
1421 * using ucnv_convertEx(). ucnv_toAlgorithmic() works exactly like ucnv_convert()
1422 * except that the two converters need not be looked up and opened completely.
1424 * The source-to-pivot conversion uses the cnv converter parameter.
1425 * The pivot-to-target conversion uses a purely algorithmic converter
1426 * according to the specified type, e.g., UCNV_UTF8 for a UTF-8 converter.
1428 * Internally, the algorithmic converter is opened and closed for each
1429 * function call, which is more efficient than using the public ucnv_open()
1430 * but somewhat less efficient than only resetting an existing converter
1431 * and using ucnv_convertEx().
1433 * This function is more convenient than ucnv_convertEx() for single-string
1434 * conversions, especially when "preflighting" is desired (returning the length
1435 * of the complete output even if it does not fit into the target buffer;
1436 * see the User Guide Strings chapter). See ucnv_convert() for details.
1438 * @param algorithmicType UConverterType constant identifying the desired target
1439 * charset as a purely algorithmic converter.
1440 * Those are converters for Unicode charsets like
1441 * UTF-8, BOCU-1, SCSU, UTF-7, IMAP-mailbox-name, etc.,
1442 * as well as US-ASCII and ISO-8859-1.
1443 * @param cnv The converter that is used to convert
1444 * from the source to the UTF-16 pivot buffer.
1445 * @param target Pointer to the output buffer.
1446 * @param targetCapacity Capacity of the target, in bytes.
1447 * @param source Pointer to the input buffer.
1448 * @param sourceLength Length of the input text, in bytes
1449 * @param pErrorCode ICU error code in/out parameter.
1450 * Must fulfill U_SUCCESS before the function call.
1451 * @return Length of the complete output text in bytes, even if it exceeds the targetCapacity
1452 * and a U_BUFFER_OVERFLOW_ERROR is set.
1454 * @see ucnv_fromAlgorithmic
1456 * @see ucnv_convertEx
1457 * @see ucnv_fromUnicode
1458 * @see ucnv_toUnicode
1459 * @see ucnv_fromUChars
1460 * @see ucnv_toUChars
1463 U_STABLE int32_t U_EXPORT2
1464 ucnv_toAlgorithmic(UConverterType algorithmicType,
1466 char *target, int32_t targetCapacity,
1467 const char *source, int32_t sourceLength,
1468 UErrorCode *pErrorCode);
1471 * Convert from one external charset to another.
1472 * Internally, the text is converted to and from the 16-bit Unicode "pivot"
1473 * using ucnv_convertEx(). ucnv_fromAlgorithmic() works exactly like ucnv_convert()
1474 * except that the two converters need not be looked up and opened completely.
1476 * The source-to-pivot conversion uses a purely algorithmic converter
1477 * according to the specified type, e.g., UCNV_UTF8 for a UTF-8 converter.
1478 * The pivot-to-target conversion uses the cnv converter parameter.
1480 * Internally, the algorithmic converter is opened and closed for each
1481 * function call, which is more efficient than using the public ucnv_open()
1482 * but somewhat less efficient than only resetting an existing converter
1483 * and using ucnv_convertEx().
1485 * This function is more convenient than ucnv_convertEx() for single-string
1486 * conversions, especially when "preflighting" is desired (returning the length
1487 * of the complete output even if it does not fit into the target buffer;
1488 * see the User Guide Strings chapter). See ucnv_convert() for details.
1490 * @param cnv The converter that is used to convert
1491 * from the UTF-16 pivot buffer to the target.
1492 * @param algorithmicType UConverterType constant identifying the desired source
1493 * charset as a purely algorithmic converter.
1494 * Those are converters for Unicode charsets like
1495 * UTF-8, BOCU-1, SCSU, UTF-7, IMAP-mailbox-name, etc.,
1496 * as well as US-ASCII and ISO-8859-1.
1497 * @param target Pointer to the output buffer.
1498 * @param targetCapacity Capacity of the target, in bytes.
1499 * @param source Pointer to the input buffer.
1500 * @param sourceLength Length of the input text, in bytes
1501 * @param pErrorCode ICU error code in/out parameter.
1502 * Must fulfill U_SUCCESS before the function call.
1503 * @return Length of the complete output text in bytes, even if it exceeds the targetCapacity
1504 * and a U_BUFFER_OVERFLOW_ERROR is set.
1506 * @see ucnv_fromAlgorithmic
1508 * @see ucnv_convertEx
1509 * @see ucnv_fromUnicode
1510 * @see ucnv_toUnicode
1511 * @see ucnv_fromUChars
1512 * @see ucnv_toUChars
1515 U_STABLE int32_t U_EXPORT2
1516 ucnv_fromAlgorithmic(UConverter *cnv,
1517 UConverterType algorithmicType,
1518 char *target, int32_t targetCapacity,
1519 const char *source, int32_t sourceLength,
1520 UErrorCode *pErrorCode);
1523 * Frees up memory occupied by unused, cached converter shared data.
1525 * @return the number of cached converters successfully deleted
1529 U_STABLE int32_t U_EXPORT2
1530 ucnv_flushCache(void);
1533 * Returns the number of available converters, as per the alias file.
1535 * @return the number of available converters
1536 * @see ucnv_getAvailableName
1539 U_STABLE int32_t U_EXPORT2
1540 ucnv_countAvailable(void);
1543 * Gets the canonical converter name of the specified converter from a list of
1544 * all available converters contaied in the alias file. All converters
1545 * in this list can be opened.
1547 * @param n the index to a converter available on the system (in the range <TT>[0..ucnv_countAvaiable()]</TT>)
1548 * @return a pointer a string (library owned), or <TT>NULL</TT> if the index is out of bounds.
1549 * @see ucnv_countAvailable
1552 U_STABLE const char* U_EXPORT2
1553 ucnv_getAvailableName(int32_t n);
1556 * Returns a UEnumeration to enumerate all of the canonical converter
1557 * names, as per the alias file, regardless of the ability to open each
1560 * @return A UEnumeration object for getting all the recognized canonical
1562 * @see ucnv_getAvailableName
1567 U_STABLE UEnumeration * U_EXPORT2
1568 ucnv_openAllNames(UErrorCode *pErrorCode);
1571 * Gives the number of aliases for a given converter or alias name.
1572 * If the alias is ambiguous, then the preferred converter is used
1573 * and the status is set to U_AMBIGUOUS_ALIAS_WARNING.
1574 * This method only enumerates the listed entries in the alias file.
1575 * @param alias alias name
1576 * @param pErrorCode error status
1577 * @return number of names on alias list for given alias
1580 U_STABLE uint16_t U_EXPORT2
1581 ucnv_countAliases(const char *alias, UErrorCode *pErrorCode);
1584 * Gives the name of the alias at given index of alias list.
1585 * This method only enumerates the listed entries in the alias file.
1586 * If the alias is ambiguous, then the preferred converter is used
1587 * and the status is set to U_AMBIGUOUS_ALIAS_WARNING.
1588 * @param alias alias name
1589 * @param n index in alias list
1590 * @param pErrorCode result of operation
1591 * @return returns the name of the alias at given index
1592 * @see ucnv_countAliases
1595 U_STABLE const char * U_EXPORT2
1596 ucnv_getAlias(const char *alias, uint16_t n, UErrorCode *pErrorCode);
1599 * Fill-up the list of alias names for the given alias.
1600 * This method only enumerates the listed entries in the alias file.
1601 * If the alias is ambiguous, then the preferred converter is used
1602 * and the status is set to U_AMBIGUOUS_ALIAS_WARNING.
1603 * @param alias alias name
1604 * @param aliases fill-in list, aliases is a pointer to an array of
1605 * <code>ucnv_countAliases()</code> string-pointers
1606 * (<code>const char *</code>) that will be filled in.
1607 * The strings themselves are owned by the library.
1608 * @param pErrorCode result of operation
1611 U_STABLE void U_EXPORT2
1612 ucnv_getAliases(const char *alias, const char **aliases, UErrorCode *pErrorCode);
1615 * Return a new UEnumeration object for enumerating all the
1616 * alias names for a given converter that are recognized by a standard.
1617 * This method only enumerates the listed entries in the alias file.
1618 * The convrtrs.txt file can be modified to change the results of
1620 * The first result in this list is the same result given by
1621 * <code>ucnv_getStandardName</code>, which is the default alias for
1622 * the specified standard name. The returned object must be closed with
1623 * <code>uenum_close</code> when you are done with the object.
1625 * @param convName original converter name
1626 * @param standard name of the standard governing the names; MIME and IANA
1627 * are such standards
1628 * @param pErrorCode The error code
1629 * @return A UEnumeration object for getting all aliases that are recognized
1630 * by a standard. If any of the parameters are invalid, NULL
1632 * @see ucnv_getStandardName
1637 U_STABLE UEnumeration * U_EXPORT2
1638 ucnv_openStandardNames(const char *convName,
1639 const char *standard,
1640 UErrorCode *pErrorCode);
1643 * Gives the number of standards associated to converter names.
1644 * @return number of standards
1647 U_STABLE uint16_t U_EXPORT2
1648 ucnv_countStandards(void);
1651 * Gives the name of the standard at given index of standard list.
1652 * @param n index in standard list
1653 * @param pErrorCode result of operation
1654 * @return returns the name of the standard at given index. Owned by the library.
1657 U_STABLE const char * U_EXPORT2
1658 ucnv_getStandard(uint16_t n, UErrorCode *pErrorCode);
1661 * Returns a standard name for a given converter name.
1663 * Example alias table:<br>
1664 * conv alias1 { STANDARD1 } alias2 { STANDARD1* }
1666 * Result of ucnv_getStandardName("conv", "STANDARD1") from example
1670 * @param name original converter name
1671 * @param standard name of the standard governing the names; MIME and IANA
1672 * are such standards
1673 * @param pErrorCode result of operation
1674 * @return returns the standard converter name;
1675 * if a standard converter name cannot be determined,
1676 * then <code>NULL</code> is returned. Owned by the library.
1679 U_STABLE const char * U_EXPORT2
1680 ucnv_getStandardName(const char *name, const char *standard, UErrorCode *pErrorCode);
1683 * This function will return the internal canonical converter name of the
1684 * tagged alias. This is the opposite of ucnv_openStandardNames, which
1685 * returns the tagged alias given the canonical name.
1687 * Example alias table:<br>
1688 * conv alias1 { STANDARD1 } alias2 { STANDARD1* }
1690 * Result of ucnv_getStandardName("alias1", "STANDARD1") from example
1694 * @return returns the canonical converter name;
1695 * if a standard or alias name cannot be determined,
1696 * then <code>NULL</code> is returned. The returned string is
1697 * owned by the library.
1698 * @see ucnv_getStandardName
1701 U_STABLE const char * U_EXPORT2
1702 ucnv_getCanonicalName(const char *alias, const char *standard, UErrorCode *pErrorCode);
1705 * returns the current default converter name.
1707 * @return returns the current default converter name;
1708 * if a default converter name cannot be determined,
1709 * then <code>NULL</code> is returned.
1710 * Storage owned by the library
1711 * @see ucnv_setDefaultName
1714 U_STABLE const char * U_EXPORT2
1715 ucnv_getDefaultName(void);
1718 * sets the current default converter name. Caller must own the storage for 'name'
1719 * and preserve it indefinitely.
1720 * @param name the converter name to be the default (must exist).
1721 * @see ucnv_getDefaultName
1722 * @system SYSTEM API
1725 U_STABLE void U_EXPORT2
1726 ucnv_setDefaultName(const char *name);
1729 * Fixes the backslash character mismapping. For example, in SJIS, the backslash
1730 * character in the ASCII portion is also used to represent the yen currency sign.
1731 * When mapping from Unicode character 0x005C, it's unclear whether to map the
1732 * character back to yen or backslash in SJIS. This function will take the input
1733 * buffer and replace all the yen sign characters with backslash. This is necessary
1734 * when the user tries to open a file with the input buffer on Windows.
1735 * This function will test the converter to see whether such mapping is
1736 * required. You can sometimes avoid using this function by using the correct version
1739 * @param cnv The converter representing the target codepage.
1740 * @param source the input buffer to be fixed
1741 * @param sourceLen the length of the input buffer
1742 * @see ucnv_isAmbiguous
1745 U_STABLE void U_EXPORT2
1746 ucnv_fixFileSeparator(const UConverter *cnv, UChar *source, int32_t sourceLen);
1749 * Determines if the converter contains ambiguous mappings of the same
1751 * @param cnv the converter to be tested
1752 * @return TRUE if the converter contains ambiguous mapping of the same
1753 * character, FALSE otherwise.
1756 U_STABLE UBool U_EXPORT2
1757 ucnv_isAmbiguous(const UConverter *cnv);
1760 * Sets the converter to use fallback mapping or not.
1761 * @param cnv The converter to set the fallback mapping usage on.
1762 * @param usesFallback TRUE if the user wants the converter to take advantage of the fallback
1763 * mapping, FALSE otherwise.
1766 U_STABLE void U_EXPORT2
1767 ucnv_setFallback(UConverter *cnv, UBool usesFallback);
1770 * Determines if the converter uses fallback mappings or not.
1771 * @param cnv The converter to be tested
1772 * @return TRUE if the converter uses fallback, FALSE otherwise.
1775 U_STABLE UBool U_EXPORT2
1776 ucnv_usesFallback(const UConverter *cnv);
1779 * Detects Unicode signature byte sequences at the start of the byte stream
1780 * and returns the charset name of the indicated Unicode charset.
1781 * NULL is returned when no Unicode signature is recognized.
1782 * The number of bytes in the signature is output as well.
1784 * The caller can ucnv_open() a converter using the charset name.
1785 * The first code unit (UChar) from the start of the stream will be U+FEFF
1786 * (the Unicode BOM/signature character) and can usually be ignored.
1788 * For most Unicode charsets it is also possible to ignore the indicated
1789 * number of initial stream bytes and start converting after them.
1790 * However, there are stateful Unicode charsets (UTF-7 and BOCU-1) for which
1791 * this will not work. Therefore, it is best to ignore the first output UChar
1792 * instead of the input signature bytes.
1796 * UErrorCode err = U_ZERO_ERROR;
1797 * char input[] = { '\xEF','\xBB', '\xBF','\x41','\x42','\x43' };
1798 * int32_t signatureLength = 0;
1799 * char *encoding = ucnv_detectUnicodeSignatures(input,sizeof(input),&signatureLength,&err);
1800 * UConverter *conv = NULL;
1801 * UChar output[100];
1802 * UChar *target = output, *out;
1803 * char *source = input;
1804 * if(encoding!=NULL && U_SUCCESS(err)){
1805 * // should signature be discarded ?
1806 * conv = ucnv_open(encoding, &err);
1807 * // do the conversion
1808 * ucnv_toUnicode(conv,
1809 * target, output + sizeof(output)/U_SIZEOF_UCHAR,
1810 * source, input + sizeof(input),
1811 * NULL, TRUE, &err);
1813 * if (discardSignature){
1814 * ++out; // ignore initial U+FEFF
1816 * while(out != target) {
1817 * printf("%04x ", *out++);
1824 * @param source The source string in which the signature should be detected.
1825 * @param sourceLength Length of the input string, or -1 if terminated with a NUL byte.
1826 * @param signatureLength A pointer to int32_t to receive the number of bytes that make up the signature
1827 * of the detected UTF. 0 if not detected.
1828 * Can be a NULL pointer.
1829 * @param pErrorCode ICU error code in/out parameter.
1830 * Must fulfill U_SUCCESS before the function call.
1831 * @return The name of the encoding detected. NULL if encoding is not detected.
1834 U_STABLE const char* U_EXPORT2
1835 ucnv_detectUnicodeSignature(const char* source,
1836 int32_t sourceLength,
1837 int32_t *signatureLength,
1838 UErrorCode *pErrorCode);
1841 * Returns the number of UChars held in the converter's internal state
1842 * because more input is needed for completing the conversion. This function is
1843 * useful for mapping semantics of ICU's converter interface to those of iconv,
1844 * and this information is not needed for normal conversion.
1845 * @param cnv The converter in which the input is held
1846 * @param status ICU error code in/out parameter.
1847 * Must fulfill U_SUCCESS before the function call.
1848 * @return The number of UChars in the state. -1 if an error is encountered.
1851 U_DRAFT int32_t U_EXPORT2
1852 ucnv_fromUCountPending(const UConverter* cnv, UErrorCode* status);
1855 * Returns the number of chars held in the converter's internal state
1856 * because more input is needed for completing the conversion. This function is
1857 * useful for mapping semantics of ICU's converter interface to those of iconv,
1858 * and this information is not needed for normal conversion.
1859 * @param cnv The converter in which the input is held as internal state
1860 * @param status ICU error code in/out parameter.
1861 * Must fulfill U_SUCCESS before the function call.
1862 * @return The number of chars in the state. -1 if an error is encountered.
1865 U_DRAFT int32_t U_EXPORT2
1866 ucnv_toUCountPending(const UConverter* cnv, UErrorCode* status);