ICU 75.1 75.1
|
Low-level C++ case mapping functions. More...
#include <casemap.h>
Static Public Member Functions | |
static int32_t | toLower (const char *locale, uint32_t options, const char16_t *src, int32_t srcLength, char16_t *dest, int32_t destCapacity, Edits *edits, UErrorCode &errorCode) |
Lowercases a UTF-16 string and optionally records edits. | |
static int32_t | toUpper (const char *locale, uint32_t options, const char16_t *src, int32_t srcLength, char16_t *dest, int32_t destCapacity, Edits *edits, UErrorCode &errorCode) |
Uppercases a UTF-16 string and optionally records edits. | |
static int32_t | toTitle (const char *locale, uint32_t options, BreakIterator *iter, const char16_t *src, int32_t srcLength, char16_t *dest, int32_t destCapacity, Edits *edits, UErrorCode &errorCode) |
Titlecases a UTF-16 string and optionally records edits. | |
static int32_t | fold (uint32_t options, const char16_t *src, int32_t srcLength, char16_t *dest, int32_t destCapacity, Edits *edits, UErrorCode &errorCode) |
Case-folds a UTF-16 string and optionally records edits. | |
static void | utf8ToLower (const char *locale, uint32_t options, StringPiece src, ByteSink &sink, Edits *edits, UErrorCode &errorCode) |
Lowercases a UTF-8 string and optionally records edits. | |
static void | utf8ToUpper (const char *locale, uint32_t options, StringPiece src, ByteSink &sink, Edits *edits, UErrorCode &errorCode) |
Uppercases a UTF-8 string and optionally records edits. | |
static void | utf8ToTitle (const char *locale, uint32_t options, BreakIterator *iter, StringPiece src, ByteSink &sink, Edits *edits, UErrorCode &errorCode) |
Titlecases a UTF-8 string and optionally records edits. | |
static void | utf8Fold (uint32_t options, StringPiece src, ByteSink &sink, Edits *edits, UErrorCode &errorCode) |
Case-folds a UTF-8 string and optionally records edits. | |
static int32_t | utf8ToLower (const char *locale, uint32_t options, const char *src, int32_t srcLength, char *dest, int32_t destCapacity, Edits *edits, UErrorCode &errorCode) |
Lowercases a UTF-8 string and optionally records edits. | |
static int32_t | utf8ToUpper (const char *locale, uint32_t options, const char *src, int32_t srcLength, char *dest, int32_t destCapacity, Edits *edits, UErrorCode &errorCode) |
Uppercases a UTF-8 string and optionally records edits. | |
static int32_t | utf8ToTitle (const char *locale, uint32_t options, BreakIterator *iter, const char *src, int32_t srcLength, char *dest, int32_t destCapacity, Edits *edits, UErrorCode &errorCode) |
Titlecases a UTF-8 string and optionally records edits. | |
static int32_t | utf8Fold (uint32_t options, const char *src, int32_t srcLength, char *dest, int32_t destCapacity, Edits *edits, UErrorCode &errorCode) |
Case-folds a UTF-8 string and optionally records edits. | |
|
static |
Case-folds a UTF-16 string and optionally records edits.
Case folding is locale-independent and not context-sensitive, but there is an option for whether to include or exclude mappings for dotted I and dotless i that are marked with 'T' in CaseFolding.txt.
The result may be longer or shorter than the original. The source string and the destination buffer must not overlap.
options | Options bit set, usually 0. See U_OMIT_UNCHANGED_TEXT, U_EDITS_NO_RESET, U_FOLD_CASE_DEFAULT, U_FOLD_CASE_EXCLUDE_SPECIAL_I. |
src | The original string. |
srcLength | The length of the original string. If -1, then src must be NUL-terminated. |
dest | A buffer for the result string. The result will be NUL-terminated if the buffer is large enough. The contents is undefined in case of failure. |
destCapacity | The size of the buffer (number of char16_ts). If it is 0, then dest may be nullptr and the function will only return the length of the result without writing any of the result string. |
edits | Records edits for index mapping, working with styled text, and getting only changes (if any). The Edits contents is undefined if any error occurs. This function calls edits->reset() first unless options includes U_EDITS_NO_RESET. edits can be nullptr. |
errorCode | Reference to an in/out error code value which must not indicate a failure before the function call. |
|
static |
Lowercases a UTF-16 string and optionally records edits.
Casing is locale-dependent and context-sensitive. The result may be longer or shorter than the original. The source string and the destination buffer must not overlap.
locale | The locale ID. ("" = root locale, nullptr = default locale.) |
options | Options bit set, usually 0. See U_OMIT_UNCHANGED_TEXT and U_EDITS_NO_RESET. |
src | The original string. |
srcLength | The length of the original string. If -1, then src must be NUL-terminated. |
dest | A buffer for the result string. The result will be NUL-terminated if the buffer is large enough. The contents is undefined in case of failure. |
destCapacity | The size of the buffer (number of char16_ts). If it is 0, then dest may be nullptr and the function will only return the length of the result without writing any of the result string. |
edits | Records edits for index mapping, working with styled text, and getting only changes (if any). The Edits contents is undefined if any error occurs. This function calls edits->reset() first unless options includes U_EDITS_NO_RESET. edits can be nullptr. |
errorCode | Reference to an in/out error code value which must not indicate a failure before the function call. |
|
static |
Titlecases a UTF-16 string and optionally records edits.
Casing is locale-dependent and context-sensitive. The result may be longer or shorter than the original. The source string and the destination buffer must not overlap.
Titlecasing uses a break iterator to find the first characters of words that are to be titlecased. It titlecases those characters and lowercases all others. (This can be modified with options bits.)
locale | The locale ID. ("" = root locale, nullptr = default locale.) |
options | Options bit set, usually 0. See U_OMIT_UNCHANGED_TEXT, U_EDITS_NO_RESET, U_TITLECASE_NO_LOWERCASE, U_TITLECASE_NO_BREAK_ADJUSTMENT, U_TITLECASE_ADJUST_TO_CASED, U_TITLECASE_WHOLE_STRING, U_TITLECASE_SENTENCES. |
iter | A break iterator to find the first characters of words that are to be titlecased. It is set to the source string (setText()) and used one or more times for iteration (first() and next()). If nullptr, then a word break iterator for the locale is used (or something equivalent). |
src | The original string. |
srcLength | The length of the original string. If -1, then src must be NUL-terminated. |
dest | A buffer for the result string. The result will be NUL-terminated if the buffer is large enough. The contents is undefined in case of failure. |
destCapacity | The size of the buffer (number of char16_ts). If it is 0, then dest may be nullptr and the function will only return the length of the result without writing any of the result string. |
edits | Records edits for index mapping, working with styled text, and getting only changes (if any). The Edits contents is undefined if any error occurs. This function calls edits->reset() first unless options includes U_EDITS_NO_RESET. edits can be nullptr. |
errorCode | Reference to an in/out error code value which must not indicate a failure before the function call. |
|
static |
Uppercases a UTF-16 string and optionally records edits.
Casing is locale-dependent and context-sensitive. The result may be longer or shorter than the original. The source string and the destination buffer must not overlap.
locale | The locale ID. ("" = root locale, nullptr = default locale.) |
options | Options bit set, usually 0. See U_OMIT_UNCHANGED_TEXT and U_EDITS_NO_RESET. |
src | The original string. |
srcLength | The length of the original string. If -1, then src must be NUL-terminated. |
dest | A buffer for the result string. The result will be NUL-terminated if the buffer is large enough. The contents is undefined in case of failure. |
destCapacity | The size of the buffer (number of char16_ts). If it is 0, then dest may be nullptr and the function will only return the length of the result without writing any of the result string. |
edits | Records edits for index mapping, working with styled text, and getting only changes (if any). The Edits contents is undefined if any error occurs. This function calls edits->reset() first unless options includes U_EDITS_NO_RESET. edits can be nullptr. |
errorCode | Reference to an in/out error code value which must not indicate a failure before the function call. |
|
static |
Case-folds a UTF-8 string and optionally records edits.
Case folding is locale-independent and not context-sensitive, but there is an option for whether to include or exclude mappings for dotted I and dotless i that are marked with 'T' in CaseFolding.txt.
The result may be longer or shorter than the original. The source string and the destination buffer must not overlap.
options | Options bit set, usually 0. See U_OMIT_UNCHANGED_TEXT, U_EDITS_NO_RESET, U_FOLD_CASE_DEFAULT, U_FOLD_CASE_EXCLUDE_SPECIAL_I. |
src | The original string. |
srcLength | The length of the original string. If -1, then src must be NUL-terminated. |
dest | A buffer for the result string. The result will be NUL-terminated if the buffer is large enough. The contents is undefined in case of failure. |
destCapacity | The size of the buffer (number of bytes). If it is 0, then dest may be nullptr and the function will only return the length of the result without writing any of the result string. |
edits | Records edits for index mapping, working with styled text, and getting only changes (if any). The Edits contents is undefined if any error occurs. This function calls edits->reset() first unless options includes U_EDITS_NO_RESET. edits can be nullptr. |
errorCode | Reference to an in/out error code value which must not indicate a failure before the function call. |
|
static |
Case-folds a UTF-8 string and optionally records edits.
Case folding is locale-independent and not context-sensitive, but there is an option for whether to include or exclude mappings for dotted I and dotless i that are marked with 'T' in CaseFolding.txt.
The result may be longer or shorter than the original.
options | Options bit set, usually 0. See U_OMIT_UNCHANGED_TEXT and U_EDITS_NO_RESET. |
src | The original string. |
sink | A ByteSink to which the result string is written. sink.Flush() is called at the end. |
edits | Records edits for index mapping, working with styled text, and getting only changes (if any). The Edits contents is undefined if any error occurs. This function calls edits->reset() first unless options includes U_EDITS_NO_RESET. edits can be nullptr. |
errorCode | Reference to an in/out error code value which must not indicate a failure before the function call. |
|
static |
Lowercases a UTF-8 string and optionally records edits.
Casing is locale-dependent and context-sensitive. The result may be longer or shorter than the original. The source string and the destination buffer must not overlap.
locale | The locale ID. ("" = root locale, nullptr = default locale.) |
options | Options bit set, usually 0. See U_OMIT_UNCHANGED_TEXT and U_EDITS_NO_RESET. |
src | The original string. |
srcLength | The length of the original string. If -1, then src must be NUL-terminated. |
dest | A buffer for the result string. The result will be NUL-terminated if the buffer is large enough. The contents is undefined in case of failure. |
destCapacity | The size of the buffer (number of bytes). If it is 0, then dest may be nullptr and the function will only return the length of the result without writing any of the result string. |
edits | Records edits for index mapping, working with styled text, and getting only changes (if any). The Edits contents is undefined if any error occurs. This function calls edits->reset() first unless options includes U_EDITS_NO_RESET. edits can be nullptr. |
errorCode | Reference to an in/out error code value which must not indicate a failure before the function call. |
|
static |
Lowercases a UTF-8 string and optionally records edits.
Casing is locale-dependent and context-sensitive. The result may be longer or shorter than the original.
locale | The locale ID. ("" = root locale, nullptr = default locale.) |
options | Options bit set, usually 0. See U_OMIT_UNCHANGED_TEXT and U_EDITS_NO_RESET. |
src | The original string. |
sink | A ByteSink to which the result string is written. sink.Flush() is called at the end. |
edits | Records edits for index mapping, working with styled text, and getting only changes (if any). The Edits contents is undefined if any error occurs. This function calls edits->reset() first unless options includes U_EDITS_NO_RESET. edits can be nullptr. |
errorCode | Reference to an in/out error code value which must not indicate a failure before the function call. |
|
static |
Titlecases a UTF-8 string and optionally records edits.
Casing is locale-dependent and context-sensitive. The result may be longer or shorter than the original. The source string and the destination buffer must not overlap.
Titlecasing uses a break iterator to find the first characters of words that are to be titlecased. It titlecases those characters and lowercases all others. (This can be modified with options bits.)
locale | The locale ID. ("" = root locale, nullptr = default locale.) |
options | Options bit set, usually 0. See U_OMIT_UNCHANGED_TEXT, U_EDITS_NO_RESET, U_TITLECASE_NO_LOWERCASE, U_TITLECASE_NO_BREAK_ADJUSTMENT, U_TITLECASE_ADJUST_TO_CASED, U_TITLECASE_WHOLE_STRING, U_TITLECASE_SENTENCES. |
iter | A break iterator to find the first characters of words that are to be titlecased. It is set to the source string (setUText()) and used one or more times for iteration (first() and next()). If nullptr, then a word break iterator for the locale is used (or something equivalent). |
src | The original string. |
srcLength | The length of the original string. If -1, then src must be NUL-terminated. |
dest | A buffer for the result string. The result will be NUL-terminated if the buffer is large enough. The contents is undefined in case of failure. |
destCapacity | The size of the buffer (number of bytes). If it is 0, then dest may be nullptr and the function will only return the length of the result without writing any of the result string. |
edits | Records edits for index mapping, working with styled text, and getting only changes (if any). The Edits contents is undefined if any error occurs. This function calls edits->reset() first unless options includes U_EDITS_NO_RESET. edits can be nullptr. |
errorCode | Reference to an in/out error code value which must not indicate a failure before the function call. |
|
static |
Titlecases a UTF-8 string and optionally records edits.
Casing is locale-dependent and context-sensitive. The result may be longer or shorter than the original.
Titlecasing uses a break iterator to find the first characters of words that are to be titlecased. It titlecases those characters and lowercases all others. (This can be modified with options bits.)
locale | The locale ID. ("" = root locale, nullptr = default locale.) |
options | Options bit set, usually 0. See U_OMIT_UNCHANGED_TEXT, U_EDITS_NO_RESET, U_TITLECASE_NO_LOWERCASE, U_TITLECASE_NO_BREAK_ADJUSTMENT, U_TITLECASE_ADJUST_TO_CASED, U_TITLECASE_WHOLE_STRING, U_TITLECASE_SENTENCES. |
iter | A break iterator to find the first characters of words that are to be titlecased. It is set to the source string (setUText()) and used one or more times for iteration (first() and next()). If nullptr, then a word break iterator for the locale is used (or something equivalent). |
src | The original string. |
sink | A ByteSink to which the result string is written. sink.Flush() is called at the end. |
edits | Records edits for index mapping, working with styled text, and getting only changes (if any). The Edits contents is undefined if any error occurs. This function calls edits->reset() first unless options includes U_EDITS_NO_RESET. edits can be nullptr. |
errorCode | Reference to an in/out error code value which must not indicate a failure before the function call. |
|
static |
Uppercases a UTF-8 string and optionally records edits.
Casing is locale-dependent and context-sensitive. The result may be longer or shorter than the original. The source string and the destination buffer must not overlap.
locale | The locale ID. ("" = root locale, nullptr = default locale.) |
options | Options bit set, usually 0. See U_OMIT_UNCHANGED_TEXT and U_EDITS_NO_RESET. |
src | The original string. |
srcLength | The length of the original string. If -1, then src must be NUL-terminated. |
dest | A buffer for the result string. The result will be NUL-terminated if the buffer is large enough. The contents is undefined in case of failure. |
destCapacity | The size of the buffer (number of bytes). If it is 0, then dest may be nullptr and the function will only return the length of the result without writing any of the result string. |
edits | Records edits for index mapping, working with styled text, and getting only changes (if any). The Edits contents is undefined if any error occurs. This function calls edits->reset() first unless options includes U_EDITS_NO_RESET. edits can be nullptr. |
errorCode | Reference to an in/out error code value which must not indicate a failure before the function call. |
|
static |
Uppercases a UTF-8 string and optionally records edits.
Casing is locale-dependent and context-sensitive. The result may be longer or shorter than the original.
locale | The locale ID. ("" = root locale, nullptr = default locale.) |
options | Options bit set, usually 0. See U_OMIT_UNCHANGED_TEXT and U_EDITS_NO_RESET. |
src | The original string. |
sink | A ByteSink to which the result string is written. sink.Flush() is called at the end. |
edits | Records edits for index mapping, working with styled text, and getting only changes (if any). The Edits contents is undefined if any error occurs. This function calls edits->reset() first unless options includes U_EDITS_NO_RESET. edits can be nullptr. |
errorCode | Reference to an in/out error code value which must not indicate a failure before the function call. |