diff --git a/third_party/icu4c/libicu.ndk.json b/third_party/icu4c/libicu.ndk.json index 42149b48ee8e15acef291cf765991bbae222c418..358c158556a82b16edb1dc7ce7f57b21f152f921 100644 --- a/third_party/icu4c/libicu.ndk.json +++ b/third_party/icu4c/libicu.ndk.json @@ -2090,5 +2090,377 @@ { "first_introduced": "15", "name":"u_errorName" + }, + { + "first_introduced": "21", + "name":"u_charAge" + }, + { + "first_introduced": "21", + "name":"u_getUnicodeVersion" + }, + { + "first_introduced": "21", + "name":"uldn_open" + }, + { + "first_introduced": "21", + "name":"uldn_close" + }, + { + "first_introduced": "21", + "name":"uldn_getLocale" + }, + { + "first_introduced": "21", + "name":"uldn_getDialectHandling" + }, + { + "first_introduced": "21", + "name":"uldn_localeDisplayName" + }, + { + "first_introduced": "21", + "name":"uldn_languageDisplayName" + }, + { + "first_introduced": "21", + "name":"uldn_scriptDisplayName" + }, + { + "first_introduced": "21", + "name":"uldn_scriptCodeDisplayName" + }, + { + "first_introduced": "21", + "name":"uldn_regionDisplayName" + }, + { + "first_introduced": "21", + "name":"uldn_variantDisplayName" + }, + { + "first_introduced": "21", + "name":"uldn_keyDisplayName" + }, + { + "first_introduced": "21", + "name":"uldn_keyValueDisplayName" + }, + { + "first_introduced": "21", + "name":"uldn_openForContext" + }, + { + "first_introduced": "21", + "name":"uldn_getContext" + }, + { + "first_introduced": "21", + "name":"ulocdata_getCLDRVersion" + }, + { + "first_introduced": "21", + "name":"u_versionFromString" + }, + { + "first_introduced": "21", + "name":"u_versionFromUString" + }, + { + "first_introduced": "21", + "name":"u_versionToString" + }, + { + "first_introduced": "21", + "name":"u_getVersion" + }, + { + "first_introduced": "21", + "name":"UCPMapValueFilter" + }, + { + "first_introduced": "21", + "name":"U_IS_UNICODE_NONCHAR" + }, + { + "first_introduced": "21", + "name":"U_IS_UNICODE_CHAR" + }, + { + "first_introduced": "21", + "name":"U_IS_BMP" + }, + { + "first_introduced": "21", + "name":"U_IS_SUPPLEMENTARY" + }, + { + "first_introduced": "21", + "name":"U_IS_LEAD" + }, + { + "first_introduced": "21", + "name":"U_IS_TRAIL" + }, + { + "first_introduced": "21", + "name":"U_IS_SURROGATE" + }, + { + "first_introduced": "21", + "name":"U_IS_SURROGATE_LEAD" + }, + { + "first_introduced": "21", + "name":"U_IS_SURROGATE_TRAIL" + }, + { + "first_introduced": "21", + "name":"U16_IS_SINGLE" + }, + { + "first_introduced": "21", + "name":"U16_IS_LEAD" + }, + { + "first_introduced": "21", + "name":"U16_IS_TRAIL" + }, + { + "first_introduced": "21", + "name":"U16_IS_SURROGATE" + }, + { + "first_introduced": "21", + "name":"U16_IS_SURROGATE_LEAD" + }, + { + "first_introduced": "21", + "name":"U16_IS_SURROGATE_TRAIL" + }, + { + "first_introduced": "21", + "name":"U16_GET_SUPPLEMENTARY" + }, + { + "first_introduced": "21", + "name":"U16_LEAD" + }, + { + "first_introduced": "21", + "name":"U16_TRAIL" + }, + { + "first_introduced": "21", + "name":"U16_LENGTH" + }, + { + "first_introduced": "21", + "name":"U16_GET_UNSAFE" + }, + { + "first_introduced": "21", + "name":"U16_GET" + }, + { + "first_introduced": "21", + "name":"U16_GET_OR_FFFD" + }, + { + "first_introduced": "21", + "name":"U16_NEXT_UNSAFE" + }, + { + "first_introduced": "21", + "name":"U16_NEXT" + }, + { + "first_introduced": "21", + "name":"U16_NEXT_OR_FFFD" + }, + { + "first_introduced": "21", + "name":"U16_APPEND_UNSAFE" + }, + { + "first_introduced": "21", + "name":"U16_APPEND" + }, + { + "first_introduced": "21", + "name":"U16_FWD_1_UNSAFE" + }, + { + "first_introduced": "21", + "name":"U16_FWD_1" + }, + { + "first_introduced": "21", + "name":"U16_FWD_N_UNSAFE" + }, + { + "first_introduced": "21", + "name":"U16_FWD_N" + }, + { + "first_introduced": "21", + "name":"U16_SET_CP_START_UNSAFE" + }, + { + "first_introduced": "21", + "name":"U16_SET_CP_START" + }, + { + "first_introduced": "21", + "name":"U16_PREV_UNSAFE" + }, + { + "first_introduced": "21", + "name":"U16_PREV" + }, + { + "first_introduced": "21", + "name":"U16_PREV_OR_FFFD" + }, + { + "first_introduced": "21", + "name":"U16_BACK_1_UNSAFE" + }, + { + "first_introduced": "21", + "name":"U16_BACK_1" + }, + { + "first_introduced": "21", + "name":"U16_BACK_N_UNSAFE" + }, + { + "first_introduced": "21", + "name":"U16_BACK_N" + }, + { + "first_introduced": "21", + "name":"U16_SET_CP_LIMIT_UNSAFE" + }, + { + "first_introduced": "21", + "name":"U16_SET_CP_LIMIT" + }, + { + "first_introduced": "21", + "name":"U8_IS_SINGLE" + }, + { + "first_introduced": "21", + "name":"U8_IS_LEAD" + }, + { + "first_introduced": "21", + "name":"U8_IS_TRAIL" + }, + { + "first_introduced": "21", + "name":"U8_LENGTH" + }, + { + "first_introduced": "21", + "name":"U8_GET_UNSAFE" + }, + { + "first_introduced": "21", + "name":"U8_GET" + }, + { + "first_introduced": "21", + "name":"U8_GET_OR_FFFD" + }, + { + "first_introduced": "21", + "name":"U8_NEXT_UNSAFE" + }, + { + "first_introduced": "21", + "name":"U8_NEXT" + }, + { + "first_introduced": "21", + "name":"U8_NEXT_OR_FFFD" + }, + { + "first_introduced": "21", + "name":"U8_INTERNAL_NEXT_OR_SUB" + }, + { + "first_introduced": "21", + "name":"U8_APPEND_UNSAFE" + }, + { + "first_introduced": "21", + "name":"U8_APPEND" + }, + { + "first_introduced": "21", + "name":"U8_FWD_1_UNSAFE" + }, + { + "first_introduced": "21", + "name":"U8_FWD_1" + }, + { + "first_introduced": "21", + "name":"U8_FWD_N_UNSAFE" + }, + { + "first_introduced": "21", + "name":"U8_FWD_N" + }, + { + "first_introduced": "21", + "name":"U8_SET_CP_START_UNSAFE" + }, + { + "first_introduced": "21", + "name":"U8_SET_CP_START" + }, + { + "first_introduced": "21", + "name":"U8_TRUNCATE_IF_INCOMPLETE" + }, + { + "first_introduced": "21", + "name":"U8_PREV_UNSAFE" + }, + { + "first_introduced": "21", + "name":"U8_PREV" + }, + { + "first_introduced": "21", + "name":"U8_PREV_OR_FFFD" + }, + { + "first_introduced": "21", + "name":"U8_BACK_1_UNSAFE" + }, + { + "first_introduced": "21", + "name":"U8_BACK_1" + }, + { + "first_introduced": "21", + "name":"U8_BACK_N_UNSAFE" + }, + { + "first_introduced": "21", + "name":"U8_BACK_N" + }, + { + "first_introduced": "21", + "name":"U8_SET_CP_LIMIT_UNSAFE" + }, + { + "first_introduced": "21", + "name":"U8_SET_CP_LIMIT" } ] diff --git a/third_party/icu4c/ndk_headers/unicode/putil.h b/third_party/icu4c/ndk_headers/unicode/putil.h new file mode 100644 index 0000000000000000000000000000000000000000..c3b347d27b4a36efd55069945819fd7ba9b370a8 --- /dev/null +++ b/third_party/icu4c/ndk_headers/unicode/putil.h @@ -0,0 +1,67 @@ +// © 2016 and later: Unicode, Inc. and others. +// License & terms of use: http://www.unicode.org/copyright.html +/* +****************************************************************************** +* +* Copyright (C) 1997-2014, International Business Machines +* Corporation and others. All Rights Reserved. +* +****************************************************************************** +* +* FILE NAME : putil.h +* +* Date Name Description +* 05/14/98 nos Creation (content moved here from utypes.h). +* 06/17/99 erm Added IEEE_754 +* 07/22/98 stephen Added IEEEremainder, max, min, trunc +* 08/13/98 stephen Added isNegativeInfinity, isPositiveInfinity +* 08/24/98 stephen Added longBitsFromDouble +* 03/02/99 stephen Removed openFile(). Added AS400 support. +* 04/15/99 stephen Converted to C +* 11/15/99 helena Integrated S/390 changes for IEEE support. +* 01/11/00 helena Added u_getVersion. +****************************************************************************** +*/ + +#ifndef PUTIL_H +#define PUTIL_H + +#include "unicode/utypes.h" + /** + * \file + * \brief C API: Platform Utilities + */ + +/*==========================================================================*/ +/* Platform utilities */ +/*==========================================================================*/ + +/** + * Platform utilities isolates the platform dependencies of the + * library. For each platform which this code is ported to, these + * functions may have to be re-implemented. + */ + +/** + * @{ + * Filesystem file and path separator characters. + * Example: '/' and ':' on Unix, '\\' and ';' on Windows. + * @stable ICU 2.0 + */ +#if U_PLATFORM_USES_ONLY_WIN32_API +# define U_FILE_SEP_CHAR '\\' +# define U_FILE_ALT_SEP_CHAR '/' +# define U_PATH_SEP_CHAR ';' +# define U_FILE_SEP_STRING "\\" +# define U_FILE_ALT_SEP_STRING "/" +# define U_PATH_SEP_STRING ";" +#else +# define U_FILE_SEP_CHAR '/' +# define U_FILE_ALT_SEP_CHAR '/' +# define U_PATH_SEP_CHAR ':' +# define U_FILE_SEP_STRING "/" +# define U_FILE_ALT_SEP_STRING "/" +# define U_PATH_SEP_STRING ":" +#endif + +#endif diff --git a/third_party/icu4c/ndk_headers/unicode/uchar.h b/third_party/icu4c/ndk_headers/unicode/uchar.h index 1a529e4f8a050b51619a8ec0a5bbb77387a76fa5..a6605f52585abdd814fba094fe775e7217cc95d2 100644 --- a/third_party/icu4c/ndk_headers/unicode/uchar.h +++ b/third_party/icu4c/ndk_headers/unicode/uchar.h @@ -3866,6 +3866,37 @@ u_digit(UChar32 ch, int8_t radix); U_CAPI UChar32 U_EXPORT2 u_forDigit(int32_t digit, int8_t radix); +/** + * Get the "age" of the code point. + * The "age" is the Unicode version when the code point was first + * designated (as a non-character or for Private Use) + * or assigned a character. + * This can be useful to avoid emitting code points to receiving + * processes that do not accept newer characters. + * The data is from the UCD file DerivedAge.txt. + * + * @param c The code point. + * @param versionArray The Unicode version number array, to be filled in. + * + * @stable ICU 2.1 + */ +U_CAPI void U_EXPORT2 +u_charAge(UChar32 c, UVersionInfo versionArray); + +/** + * Gets the Unicode version information. + * The version array is filled in with the version information + * for the Unicode standard that is currently used by ICU. + * For example, Unicode version 3.1.1 is represented as an array with + * the values { 3, 1, 1, 0 }. + * + * @param versionArray an output array that will be filled in with + * the Unicode version number + * @stable ICU 2.0 + */ +U_CAPI void U_EXPORT2 +u_getUnicodeVersion(UVersionInfo versionArray); + #if !UCONFIG_NO_NORMALIZATION /** * Get the FC_NFKC_Closure property string for a character. diff --git a/third_party/icu4c/ndk_headers/unicode/ucpmap.h b/third_party/icu4c/ndk_headers/unicode/ucpmap.h new file mode 100644 index 0000000000000000000000000000000000000000..30f337ad130526b9231f5adade5258cc9e181422 --- /dev/null +++ b/third_party/icu4c/ndk_headers/unicode/ucpmap.h @@ -0,0 +1,104 @@ +// © 2018 and later: Unicode, Inc. and others. +// License & terms of use: http://www.unicode.org/copyright.html + +// ucpmap.h +// created: 2018sep03 Markus W. Scherer + +#ifndef __UCPMAP_H__ +#define __UCPMAP_H__ + +#include "unicode/utypes.h" + +U_CDECL_BEGIN + +/** + * \file + * \brief C API: This file defines an abstract map from Unicode code points to integer values. + * + * @see UCPMap + * @see UCPTrie + * @see UMutableCPTrie + */ + +/** + * Abstract map from Unicode code points (U+0000..U+10FFFF) to integer values. + * + * @see UCPTrie + * @see UMutableCPTrie + * @stable ICU 63 + */ +typedef struct UCPMap UCPMap; + +/** + * Selectors for how ucpmap_getRange() etc. should report value ranges overlapping with surrogates. + * Most users should use UCPMAP_RANGE_NORMAL. + * + * @see ucpmap_getRange + * @see ucptrie_getRange + * @see umutablecptrie_getRange + * @stable ICU 63 + */ +enum UCPMapRangeOption { + /** + * ucpmap_getRange() enumerates all same-value ranges as stored in the map. + * Most users should use this option. + * @stable ICU 63 + */ + UCPMAP_RANGE_NORMAL, + /** + * ucpmap_getRange() enumerates all same-value ranges as stored in the map, + * except that lead surrogates (U+D800..U+DBFF) are treated as having the + * surrogateValue, which is passed to getRange() as a separate parameter. + * The surrogateValue is not transformed via filter(). + * See U_IS_LEAD(c). + * + * Most users should use UCPMAP_RANGE_NORMAL instead. + * + * This option is useful for maps that map surrogate code *units* to + * special values optimized for UTF-16 string processing + * or for special error behavior for unpaired surrogates, + * but those values are not to be associated with the lead surrogate code *points*. + * @stable ICU 63 + */ + UCPMAP_RANGE_FIXED_LEAD_SURROGATES, + /** + * ucpmap_getRange() enumerates all same-value ranges as stored in the map, + * except that all surrogates (U+D800..U+DFFF) are treated as having the + * surrogateValue, which is passed to getRange() as a separate parameter. + * The surrogateValue is not transformed via filter(). + * See U_IS_SURROGATE(c). + * + * Most users should use UCPMAP_RANGE_NORMAL instead. + * + * This option is useful for maps that map surrogate code *units* to + * special values optimized for UTF-16 string processing + * or for special error behavior for unpaired surrogates, + * but those values are not to be associated with the lead surrogate code *points*. + * @stable ICU 63 + */ + UCPMAP_RANGE_FIXED_ALL_SURROGATES +}; +#ifndef U_IN_DOXYGEN +typedef enum UCPMapRangeOption UCPMapRangeOption; +#endif + +/** + * Callback function type: Modifies a map value. + * Optionally called by ucpmap_getRange()/ucptrie_getRange()/umutablecptrie_getRange(). + * The modified value will be returned by the getRange function. + * + * Can be used to ignore some of the value bits, + * make a filter for one of several values, + * return a value index computed from the map value, etc. + * + * @param context an opaque pointer, as passed into the getRange function + * @param value a value from the map + * @return the modified value + * @stable ICU 63 + */ +typedef uint32_t U_CALLCONV +UCPMapValueFilter(const void *context, uint32_t value); + +U_CDECL_END + +#endif diff --git a/third_party/icu4c/ndk_headers/unicode/uldnames.h b/third_party/icu4c/ndk_headers/unicode/uldnames.h new file mode 100644 index 0000000000000000000000000000000000000000..02e24f004972fd0d9284d7a5f228bf133a262b00 --- /dev/null +++ b/third_party/icu4c/ndk_headers/unicode/uldnames.h @@ -0,0 +1,284 @@ +// © 2016 and later: Unicode, Inc. and others. +// License & terms of use: http://www.unicode.org/copyright.html +/* +******************************************************************************* +* Copyright (C) 2010-2016, International Business Machines Corporation and +* others. All Rights Reserved. +******************************************************************************* +*/ + +#ifndef __ULDNAMES_H__ +#define __ULDNAMES_H__ + +/** + * \file + * \brief C API: Provides display names of Locale ids and their components. + */ + +#include "unicode/utypes.h" +#include "unicode/uscript.h" +#include "unicode/udisplaycontext.h" + +/** + * Enum used in LocaleDisplayNames::createInstance. + * @stable ICU 4.4 + */ +typedef enum { + /** + * Use standard names when generating a locale name, + * e.g. en_GB displays as 'English (United Kingdom)'. + * @stable ICU 4.4 + */ + ULDN_STANDARD_NAMES = 0, + /** + * Use dialect names, when generating a locale name, + * e.g. en_GB displays as 'British English'. + * @stable ICU 4.4 + */ + ULDN_DIALECT_NAMES +} UDialectHandling; + +/** + * Opaque C service object type for the locale display names API + * @stable ICU 4.4 + */ +struct ULocaleDisplayNames; + +/** + * C typedef for struct ULocaleDisplayNames. + * @stable ICU 4.4 + */ +typedef struct ULocaleDisplayNames ULocaleDisplayNames; + +#if !UCONFIG_NO_FORMATTING + +/** + * Returns an instance of LocaleDisplayNames that returns names + * formatted for the provided locale, using the provided + * dialectHandling. The usual value for dialectHandling is + * ULOC_STANDARD_NAMES. + * + * @param locale the display locale + * @param dialectHandling how to select names for locales + * @return a ULocaleDisplayNames instance + * @param pErrorCode the status code + * @stable ICU 4.4 + */ +U_CAPI ULocaleDisplayNames * U_EXPORT2 +uldn_open(const char * locale, + UDialectHandling dialectHandling, + UErrorCode *pErrorCode); + +/** + * Closes a ULocaleDisplayNames instance obtained from uldn_open(). + * @param ldn the ULocaleDisplayNames instance to be closed + * @stable ICU 4.4 + */ +U_CAPI void U_EXPORT2 +uldn_close(ULocaleDisplayNames *ldn); + +/* getters for state */ + +/** + * Returns the locale used to determine the display names. This is + * not necessarily the same locale passed to {@link #uldn_open}. + * @param ldn the LocaleDisplayNames instance + * @return the display locale + * @stable ICU 4.4 + */ +U_CAPI const char * U_EXPORT2 +uldn_getLocale(const ULocaleDisplayNames *ldn); + +/** + * Returns the dialect handling used in the display names. + * @param ldn the LocaleDisplayNames instance + * @return the dialect handling enum + * @stable ICU 4.4 + */ +U_CAPI UDialectHandling U_EXPORT2 +uldn_getDialectHandling(const ULocaleDisplayNames *ldn); + +/* names for entire locales */ + +/** + * Returns the display name of the provided locale. + * @param ldn the LocaleDisplayNames instance + * @param locale the locale whose display name to return + * @param result receives the display name + * @param maxResultSize the size of the result buffer + * @param pErrorCode the status code + * @return the actual buffer size needed for the display name. If it's + * greater than maxResultSize, the returned name will be truncated. + * @stable ICU 4.4 + */ +U_CAPI int32_t U_EXPORT2 +uldn_localeDisplayName(const ULocaleDisplayNames *ldn, + const char *locale, + UChar *result, + int32_t maxResultSize, + UErrorCode *pErrorCode); + +/* names for components of a locale */ + +/** + * Returns the display name of the provided language code. + * @param ldn the LocaleDisplayNames instance + * @param lang the language code whose display name to return + * @param result receives the display name + * @param maxResultSize the size of the result buffer + * @param pErrorCode the status code + * @return the actual buffer size needed for the display name. If it's + * greater than maxResultSize, the returned name will be truncated. + * @stable ICU 4.4 + */ +U_CAPI int32_t U_EXPORT2 +uldn_languageDisplayName(const ULocaleDisplayNames *ldn, + const char *lang, + UChar *result, + int32_t maxResultSize, + UErrorCode *pErrorCode); + +/** + * Returns the display name of the provided script. + * @param ldn the LocaleDisplayNames instance + * @param script the script whose display name to return + * @param result receives the display name + * @param maxResultSize the size of the result buffer + * @param pErrorCode the status code + * @return the actual buffer size needed for the display name. If it's + * greater than maxResultSize, the returned name will be truncated. + * @stable ICU 4.4 + */ +U_CAPI int32_t U_EXPORT2 +uldn_scriptDisplayName(const ULocaleDisplayNames *ldn, + const char *script, + UChar *result, + int32_t maxResultSize, + UErrorCode *pErrorCode); + +/** + * Returns the display name of the provided script code. + * @param ldn the LocaleDisplayNames instance + * @param scriptCode the script code whose display name to return + * @param result receives the display name + * @param maxResultSize the size of the result buffer + * @param pErrorCode the status code + * @return the actual buffer size needed for the display name. If it's + * greater than maxResultSize, the returned name will be truncated. + * @stable ICU 4.4 + */ +U_CAPI int32_t U_EXPORT2 +uldn_scriptCodeDisplayName(const ULocaleDisplayNames *ldn, + UScriptCode scriptCode, + UChar *result, + int32_t maxResultSize, + UErrorCode *pErrorCode); + +/** + * Returns the display name of the provided region code. + * @param ldn the LocaleDisplayNames instance + * @param region the region code whose display name to return + * @param result receives the display name + * @param maxResultSize the size of the result buffer + * @param pErrorCode the status code + * @return the actual buffer size needed for the display name. If it's + * greater than maxResultSize, the returned name will be truncated. + * @stable ICU 4.4 + */ +U_CAPI int32_t U_EXPORT2 +uldn_regionDisplayName(const ULocaleDisplayNames *ldn, + const char *region, + UChar *result, + int32_t maxResultSize, + UErrorCode *pErrorCode); + +/** + * Returns the display name of the provided variant + * @param ldn the LocaleDisplayNames instance + * @param variant the variant whose display name to return + * @param result receives the display name + * @param maxResultSize the size of the result buffer + * @param pErrorCode the status code + * @return the actual buffer size needed for the display name. If it's + * greater than maxResultSize, the returned name will be truncated. + * @stable ICU 4.4 + */ +U_CAPI int32_t U_EXPORT2 +uldn_variantDisplayName(const ULocaleDisplayNames *ldn, + const char *variant, + UChar *result, + int32_t maxResultSize, + UErrorCode *pErrorCode); + +/** + * Returns the display name of the provided locale key + * @param ldn the LocaleDisplayNames instance + * @param key the locale key whose display name to return + * @param result receives the display name + * @param maxResultSize the size of the result buffer + * @param pErrorCode the status code + * @return the actual buffer size needed for the display name. If it's + * greater than maxResultSize, the returned name will be truncated. + * @stable ICU 4.4 + */ +U_CAPI int32_t U_EXPORT2 +uldn_keyDisplayName(const ULocaleDisplayNames *ldn, + const char *key, + UChar *result, + int32_t maxResultSize, + UErrorCode *pErrorCode); + +/** + * Returns the display name of the provided value (used with the provided key). + * @param ldn the LocaleDisplayNames instance + * @param key the locale key + * @param value the locale key's value + * @param result receives the display name + * @param maxResultSize the size of the result buffer + * @param pErrorCode the status code + * @return the actual buffer size needed for the display name. If it's + * greater than maxResultSize, the returned name will be truncated. + * @stable ICU 4.4 + */ +U_CAPI int32_t U_EXPORT2 +uldn_keyValueDisplayName(const ULocaleDisplayNames *ldn, + const char *key, + const char *value, + UChar *result, + int32_t maxResultSize, + UErrorCode *pErrorCode); + +/** +* Returns an instance of LocaleDisplayNames that returns names formatted +* for the provided locale, using the provided UDisplayContext settings. +* +* @param locale The display locale +* @param contexts List of one or more context settings (e.g. for dialect +* handling, capitalization, etc. +* @param length Number of items in the contexts list +* @param pErrorCode Pointer to UErrorCode input/output status. If at entry this indicates +* a failure status, the function will do nothing; otherwise this will be +* updated with any new status from the function. +* @return a ULocaleDisplayNames instance +* @stable ICU 51 +*/ +U_CAPI ULocaleDisplayNames * U_EXPORT2 +uldn_openForContext(const char * locale, UDisplayContext *contexts, + int32_t length, UErrorCode *pErrorCode); + +/** +* Returns the UDisplayContext value for the specified UDisplayContextType. +* @param ldn the ULocaleDisplayNames instance +* @param type the UDisplayContextType whose value to return +* @param pErrorCode Pointer to UErrorCode input/output status. If at entry this indicates +* a failure status, the function will do nothing; otherwise this will be +* updated with any new status from the function. +* @return the UDisplayContextValue for the specified type. +* @stable ICU 51 +*/ +U_CAPI UDisplayContext U_EXPORT2 +uldn_getContext(const ULocaleDisplayNames *ldn, UDisplayContextType type, + UErrorCode *pErrorCode); + +#endif /* !UCONFIG_NO_FORMATTING */ +#endif /* __ULDNAMES_H__ */ diff --git a/third_party/icu4c/ndk_headers/unicode/ulocdata.h b/third_party/icu4c/ndk_headers/unicode/ulocdata.h new file mode 100644 index 0000000000000000000000000000000000000000..c3dac41bfb61ee78171cd67f478d3475c98cd611 --- /dev/null +++ b/third_party/icu4c/ndk_headers/unicode/ulocdata.h @@ -0,0 +1,33 @@ +// © 2016 and later: Unicode, Inc. and others. +// License & terms of use: http://www.unicode.org/copyright.html +/* +****************************************************************************** +* * +* Copyright (C) 2003-2015, International Business Machines * +* Corporation and others. All Rights Reserved. * +* * +****************************************************************************** +* file name: ulocdata.h +* encoding: UTF-8 +* tab size: 8 (not used) +* indentation:4 +* +* created on: 2003Oct21 +* created by: Ram Viswanadha +*/ + +#ifndef __ULOCDATA_H__ +#define __ULOCDATA_H__ + +#include "unicode/uloc.h" + +/** + * Return the current CLDR version used by the library. + * @param versionArray fill-in that will receive the version number + * @param status error code - could be U_MISSING_RESOURCE_ERROR if the version was not found. + * @stable ICU 4.2 + */ +U_CAPI void U_EXPORT2 +ulocdata_getCLDRVersion(UVersionInfo versionArray, UErrorCode *status); + +#endif diff --git a/third_party/icu4c/ndk_headers/unicode/utf.h b/third_party/icu4c/ndk_headers/unicode/utf.h new file mode 100644 index 0000000000000000000000000000000000000000..00623e546ce257b6ad91b21ca9c8c19947cfd224 --- /dev/null +++ b/third_party/icu4c/ndk_headers/unicode/utf.h @@ -0,0 +1,222 @@ +// © 2016 and later: Unicode, Inc. and others. +// License & terms of use: http://www.unicode.org/copyright.html +/* +******************************************************************************* +* +* Copyright (C) 1999-2011, International Business Machines +* Corporation and others. All Rights Reserved. +* +******************************************************************************* +* file name: utf.h +* encoding: UTF-8 +* tab size: 8 (not used) +* indentation:4 +* +* created on: 1999sep09 +* created by: Markus W. Scherer +*/ + +/** + * \file + * \brief C API: Code point macros + * + * This file defines macros for checking whether a code point is + * a surrogate or a non-character etc. + * + * If U_NO_DEFAULT_INCLUDE_UTF_HEADERS is 0 then utf.h is included by utypes.h + * and itself includes utf8.h and utf16.h after some + * common definitions. + * If U_NO_DEFAULT_INCLUDE_UTF_HEADERS is 1 then each of these headers must be + * included explicitly if their definitions are used. + * + * utf8.h and utf16.h define macros for efficiently getting code points + * in and out of UTF-8/16 strings. + * utf16.h macros have "U16_" prefixes. + * utf8.h defines similar macros with "U8_" prefixes for UTF-8 string handling. + * + * ICU mostly processes 16-bit Unicode strings. + * Most of the time, such strings are well-formed UTF-16. + * Single, unpaired surrogates must be handled as well, and are treated in ICU + * like regular code points where possible. + * (Pairs of surrogate code points are indistinguishable from supplementary + * code points encoded as pairs of supplementary code units.) + * + * In fact, almost all Unicode code points in normal text (>99%) + * are on the BMP (<=U+ffff) and even <=U+d7ff. + * ICU functions handle supplementary code points (U+10000..U+10ffff) + * but are optimized for the much more frequently occurring BMP code points. + * + * umachine.h defines UChar to be an unsigned 16-bit integer. + * Since ICU 59, ICU uses char16_t in C++, UChar only in C, + * and defines UChar=char16_t by default. See the UChar API docs for details. + * + * UChar32 is defined to be a signed 32-bit integer (int32_t), large enough for a 21-bit + * Unicode code point (Unicode scalar value, 0..0x10ffff) and U_SENTINEL (-1). + * Before ICU 2.4, the definition of UChar32 was similarly platform-dependent as + * the definition of UChar. For details see the documentation for UChar32 itself. + * + * utf.h defines a small number of C macros for single Unicode code points. + * These are simple checks for surrogates and non-characters. + * For actual Unicode character properties see uchar.h. + * + * By default, string operations must be done with error checking in case + * a string is not well-formed UTF-16 or UTF-8. + * + * The U16_ macros detect if a surrogate code unit is unpaired + * (lead unit without trail unit or vice versa) and just return the unit itself + * as the code point. + * + * The U8_ macros detect illegal byte sequences and return a negative value. + * Starting with ICU 60, the observable length of a single illegal byte sequence + * skipped by one of these macros follows the Unicode 6+ recommendation + * which is consistent with the W3C Encoding Standard. + * + * There are ..._OR_FFFD versions of both U16_ and U8_ macros + * that return U+FFFD for illegal code unit sequences. + * + * The regular "safe" macros require that the initial, passed-in string index + * is within bounds. They only check the index when they read more than one + * code unit. This is usually done with code similar to the following loop: + * 
while(i
+ *
+ * When it is safe to assume that text is well-formed UTF-16
+ * (does not contain single, unpaired surrogates), then one can use
+ * U16_..._UNSAFE macros.
+ * These do not check for proper code unit sequences or truncated text and may
+ * yield wrong results or even cause a crash if they are used with "malformed"
+ * text.
+ * In practice, U16_..._UNSAFE macros will produce slightly less code but
+ * should not be faster because the processing is only different when a
+ * surrogate code unit is detected, which will be rare.
+ *
+ * Similarly for UTF-8, there are "safe" macros without a suffix,
+ * and U8_..._UNSAFE versions.
+ * The performance differences are much larger here because UTF-8 provides so
+ * many opportunities for malformed sequences.
+ * The unsafe UTF-8 macros are entirely implemented inside the macro definitions
+ * and are fast, while the safe UTF-8 macros call functions for some complicated cases.
+ *
+ * Unlike with UTF-16, malformed sequences cannot be expressed with distinct
+ * code point values (0..U+10ffff). They are indicated with negative values instead.
+ *
+ * For more information see the ICU User Guide Strings chapter
+ * (https://unicode-org.github.io/icu/userguide/strings).
+ *
+ * Usage:
+ * ICU coding guidelines for if() statements should be followed when using these macros.
+ * Compound statements (curly braces {}) must be used  for if-else-while... 
+ * bodies and all macro statements should be terminated with semicolon.
+ *
+ * @stable ICU 2.4
+ */
+
+#ifndef __UTF_H__
+#define __UTF_H__
+
+#include "unicode/umachine.h"
+/* include the utfXX.h after the following definitions */
+
+/* single-code point definitions -------------------------------------------- */
+
+/**
+ * Is this code point a Unicode noncharacter?
+ * @param c 32-bit code point
+ * @return true or false
+ * @stable ICU 2.4
+ */
+#define U_IS_UNICODE_NONCHAR(c) \
+    ((c)>=0xfdd0 && \
+     ((c)<=0xfdef || ((c)&0xfffe)==0xfffe) && (c)<=0x10ffff)
+
+/**
+ * Is c a Unicode code point value (0..U+10ffff)
+ * that can be assigned a character?
+ *
+ * Code points that are not characters include:
+ * - single surrogate code points (U+d800..U+dfff, 2048 code points)
+ * - the last two code points on each plane (U+__fffe and U+__ffff, 34 code points)
+ * - U+fdd0..U+fdef (new with Unicode 3.1, 32 code points)
+ * - the highest Unicode code point value is U+10ffff
+ *
+ * This means that all code points below U+d800 are character code points,
+ * and that boundary is tested first for performance.
+ *
+ * @param c 32-bit code point
+ * @return true or false
+ * @stable ICU 2.4
+ */
+#define U_IS_UNICODE_CHAR(c) \
+    ((uint32_t)(c)<0xd800 || \
+        (0xdfff<(c) && (c)<=0x10ffff && !U_IS_UNICODE_NONCHAR(c)))
+
+/**
+ * Is this code point a BMP code point (U+0000..U+ffff)?
+ * @param c 32-bit code point
+ * @return true or false
+ * @stable ICU 2.8
+ */
+#define U_IS_BMP(c) ((uint32_t)(c)<=0xffff)
+
+/**
+ * Is this code point a supplementary code point (U+10000..U+10ffff)?
+ * @param c 32-bit code point
+ * @return true or false
+ * @stable ICU 2.8
+ */
+#define U_IS_SUPPLEMENTARY(c) ((uint32_t)((c)-0x10000)<=0xfffff)
+ 
+/**
+ * Is this code point a lead surrogate (U+d800..U+dbff)?
+ * @param c 32-bit code point
+ * @return true or false
+ * @stable ICU 2.4
+ */
+#define U_IS_LEAD(c) (((c)&0xfffffc00)==0xd800)
+
+/**
+ * Is this code point a trail surrogate (U+dc00..U+dfff)?
+ * @param c 32-bit code point
+ * @return true or false
+ * @stable ICU 2.4
+ */
+#define U_IS_TRAIL(c) (((c)&0xfffffc00)==0xdc00)
+
+/**
+ * Is this code point a surrogate (U+d800..U+dfff)?
+ * @param c 32-bit code point
+ * @return true or false
+ * @stable ICU 2.4
+ */
+#define U_IS_SURROGATE(c) (((c)&0xfffff800)==0xd800)
+
+/**
+ * Assuming c is a surrogate code point (U_IS_SURROGATE(c)),
+ * is it a lead surrogate?
+ * @param c 32-bit code point
+ * @return true or false
+ * @stable ICU 2.4
+ */
+#define U_IS_SURROGATE_LEAD(c) (((c)&0x400)==0)
+
+/**
+ * Assuming c is a surrogate code point (U_IS_SURROGATE(c)),
+ * is it a trail surrogate?
+ * @param c 32-bit code point
+ * @return true or false
+ * @stable ICU 4.2
+ */
+#define U_IS_SURROGATE_TRAIL(c) (((c)&0x400)!=0)
+
+/* include the utfXX.h ------------------------------------------------------ */
+
+#if !U_NO_DEFAULT_INCLUDE_UTF_HEADERS
+
+#include "unicode/utf8.h"
+#include "unicode/utf16.h"
+
+#endif  /* !U_NO_DEFAULT_INCLUDE_UTF_HEADERS */
+
+#endif  /* __UTF_H__ */
diff --git a/third_party/icu4c/ndk_headers/unicode/utf16.h b/third_party/icu4c/ndk_headers/unicode/utf16.h
new file mode 100644
index 0000000000000000000000000000000000000000..93c14e124897faf96a0c6f7d9566e85f94f4c829
--- /dev/null
+++ b/third_party/icu4c/ndk_headers/unicode/utf16.h
@@ -0,0 +1,728 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 1999-2012, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  utf16.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 1999sep09
+*   created by: Markus W. Scherer
+*/
+
+/**
+ * \file
+ * \brief C API: 16-bit Unicode handling macros
+ * 
+ * This file defines macros to deal with 16-bit Unicode (UTF-16) code units and strings.
+ *
+ * For more information see utf.h and the ICU User Guide Strings chapter
+ * (https://unicode-org.github.io/icu/userguide/strings).
+ *
+ * Usage:
+ * ICU coding guidelines for if() statements should be followed when using these macros.
+ * Compound statements (curly braces {}) must be used  for if-else-while... 
+ * bodies and all macro statements should be terminated with semicolon.
+ */
+
+#ifndef __UTF16_H__
+#define __UTF16_H__
+
+#include 
+#include "unicode/umachine.h"
+#ifndef __UTF_H__
+#   include "unicode/utf.h"
+#endif
+
+/* single-code point definitions -------------------------------------------- */
+
+/**
+ * Does this code unit alone encode a code point (BMP, not a surrogate)?
+ * @param c 16-bit code unit
+ * @return true or false
+ * @stable ICU 2.4
+ */
+#define U16_IS_SINGLE(c) !U_IS_SURROGATE(c)
+
+/**
+ * Is this code unit a lead surrogate (U+d800..U+dbff)?
+ * @param c 16-bit code unit
+ * @return true or false
+ * @stable ICU 2.4
+ */
+#define U16_IS_LEAD(c) (((c)&0xfffffc00)==0xd800)
+
+/**
+ * Is this code unit a trail surrogate (U+dc00..U+dfff)?
+ * @param c 16-bit code unit
+ * @return true or false
+ * @stable ICU 2.4
+ */
+#define U16_IS_TRAIL(c) (((c)&0xfffffc00)==0xdc00)
+
+/**
+ * Is this code unit a surrogate (U+d800..U+dfff)?
+ * @param c 16-bit code unit
+ * @return true or false
+ * @stable ICU 2.4
+ */
+#define U16_IS_SURROGATE(c) U_IS_SURROGATE(c)
+
+/**
+ * Assuming c is a surrogate code point (U16_IS_SURROGATE(c)),
+ * is it a lead surrogate?
+ * @param c 16-bit code unit
+ * @return true or false
+ * @stable ICU 2.4
+ */
+#define U16_IS_SURROGATE_LEAD(c) (((c)&0x400)==0)
+
+/**
+ * Assuming c is a surrogate code point (U16_IS_SURROGATE(c)),
+ * is it a trail surrogate?
+ * @param c 16-bit code unit
+ * @return true or false
+ * @stable ICU 4.2
+ */
+#define U16_IS_SURROGATE_TRAIL(c) (((c)&0x400)!=0)
+
+/**
+ * Get a supplementary code point value (U+10000..U+10ffff)
+ * from its lead and trail surrogates.
+ * The result is undefined if the input values are not
+ * lead and trail surrogates.
+ *
+ * @param lead lead surrogate (U+d800..U+dbff)
+ * @param trail trail surrogate (U+dc00..U+dfff)
+ * @return supplementary code point (U+10000..U+10ffff)
+ * @stable ICU 2.4
+ */
+#define U16_GET_SUPPLEMENTARY(lead, trail) \
+    (((UChar32)(lead)<<10UL)+(UChar32)(trail)-U16_SURROGATE_OFFSET)
+
+
+/**
+ * Get the lead surrogate (0xd800..0xdbff) for a
+ * supplementary code point (0x10000..0x10ffff).
+ * @param supplementary 32-bit code point (U+10000..U+10ffff)
+ * @return lead surrogate (U+d800..U+dbff) for supplementary
+ * @stable ICU 2.4
+ */
+#define U16_LEAD(supplementary) (UChar)(((supplementary)>>10)+0xd7c0)
+
+/**
+ * Get the trail surrogate (0xdc00..0xdfff) for a
+ * supplementary code point (0x10000..0x10ffff).
+ * @param supplementary 32-bit code point (U+10000..U+10ffff)
+ * @return trail surrogate (U+dc00..U+dfff) for supplementary
+ * @stable ICU 2.4
+ */
+#define U16_TRAIL(supplementary) (UChar)(((supplementary)&0x3ff)|0xdc00)
+
+/**
+ * How many 16-bit code units are used to encode this Unicode code point? (1 or 2)
+ * The result is not defined if c is not a Unicode code point (U+0000..U+10ffff).
+ * @param c 32-bit code point
+ * @return 1 or 2
+ * @stable ICU 2.4
+ */
+#define U16_LENGTH(c) ((uint32_t)(c)<=0xffff ? 1 : 2)
+
+/**
+ * The maximum number of 16-bit code units per Unicode code point (U+0000..U+10ffff).
+ * @return 2
+ * @stable ICU 2.4
+ */
+#define U16_MAX_LENGTH 2
+
+/**
+ * Get a code point from a string at a random-access offset,
+ * without changing the offset.
+ * "Unsafe" macro, assumes well-formed UTF-16.
+ *
+ * The offset may point to either the lead or trail surrogate unit
+ * for a supplementary code point, in which case the macro will read
+ * the adjacent matching surrogate as well.
+ * The result is undefined if the offset points to a single, unpaired surrogate.
+ * Iteration through a string is more efficient with U16_NEXT_UNSAFE or U16_NEXT.
+ *
+ * @param s const UChar * string
+ * @param i string offset
+ * @param c output UChar32 variable
+ * @see U16_GET
+ * @stable ICU 2.4
+ */
+#define U16_GET_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \
+    (c)=(s)[i]; \
+    if(U16_IS_SURROGATE(c)) { \
+        if(U16_IS_SURROGATE_LEAD(c)) { \
+            (c)=U16_GET_SUPPLEMENTARY((c), (s)[(i)+1]); \
+        } else { \
+            (c)=U16_GET_SUPPLEMENTARY((s)[(i)-1], (c)); \
+        } \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Get a code point from a string at a random-access offset,
+ * without changing the offset.
+ * "Safe" macro, handles unpaired surrogates and checks for string boundaries.
+ *
+ * The offset may point to either the lead or trail surrogate unit
+ * for a supplementary code point, in which case the macro will read
+ * the adjacent matching surrogate as well.
+ *
+ * The length can be negative for a NUL-terminated string.
+ *
+ * If the offset points to a single, unpaired surrogate, then
+ * c is set to that unpaired surrogate.
+ * Iteration through a string is more efficient with U16_NEXT_UNSAFE or U16_NEXT.
+ *
+ * @param s const UChar * string
+ * @param start starting string offset (usually 0)
+ * @param i string offset, must be start<=i(start) && U16_IS_LEAD(__c2=(s)[(i)-1])) { \
+                (c)=U16_GET_SUPPLEMENTARY(__c2, (c)); \
+            } \
+        } \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Get a code point from a string at a random-access offset,
+ * without changing the offset.
+ * "Safe" macro, handles unpaired surrogates and checks for string boundaries.
+ *
+ * The offset may point to either the lead or trail surrogate unit
+ * for a supplementary code point, in which case the macro will read
+ * the adjacent matching surrogate as well.
+ *
+ * The length can be negative for a NUL-terminated string.
+ *
+ * If the offset points to a single, unpaired surrogate, then
+ * c is set to U+FFFD.
+ * Iteration through a string is more efficient with U16_NEXT_UNSAFE or U16_NEXT_OR_FFFD.
+ *
+ * @param s const UChar * string
+ * @param start starting string offset (usually 0)
+ * @param i string offset, must be start<=i(start) && U16_IS_LEAD(__c2=(s)[(i)-1])) { \
+                (c)=U16_GET_SUPPLEMENTARY(__c2, (c)); \
+            } else { \
+                (c)=0xfffd; \
+            } \
+        } \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/* definitions with forward iteration --------------------------------------- */
+
+/**
+ * Get a code point from a string at a code point boundary offset,
+ * and advance the offset to the next code point boundary.
+ * (Post-incrementing forward iteration.)
+ * "Unsafe" macro, assumes well-formed UTF-16.
+ *
+ * The offset may point to the lead surrogate unit
+ * for a supplementary code point, in which case the macro will read
+ * the following trail surrogate as well.
+ * If the offset points to a trail surrogate, then that itself
+ * will be returned as the code point.
+ * The result is undefined if the offset points to a single, unpaired lead surrogate.
+ *
+ * @param s const UChar * string
+ * @param i string offset
+ * @param c output UChar32 variable
+ * @see U16_NEXT
+ * @stable ICU 2.4
+ */
+#define U16_NEXT_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \
+    (c)=(s)[(i)++]; \
+    if(U16_IS_LEAD(c)) { \
+        (c)=U16_GET_SUPPLEMENTARY((c), (s)[(i)++]); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Get a code point from a string at a code point boundary offset,
+ * and advance the offset to the next code point boundary.
+ * (Post-incrementing forward iteration.)
+ * "Safe" macro, handles unpaired surrogates and checks for string boundaries.
+ *
+ * The length can be negative for a NUL-terminated string.
+ *
+ * The offset may point to the lead surrogate unit
+ * for a supplementary code point, in which case the macro will read
+ * the following trail surrogate as well.
+ * If the offset points to a trail surrogate or
+ * to a single, unpaired lead surrogate, then c is set to that unpaired surrogate.
+ *
+ * @param s const UChar * string
+ * @param i string offset, must be i>10)+0xd7c0); \
+        (s)[(i)++]=(uint16_t)(((c)&0x3ff)|0xdc00); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Append a code point to a string, overwriting 1 or 2 code units.
+ * The offset points to the current end of the string contents
+ * and is advanced (post-increment).
+ * "Safe" macro, checks for a valid code point.
+ * If a surrogate pair is written, checks for sufficient space in the string.
+ * If the code point is not valid or a trail surrogate does not fit,
+ * then isError is set to true.
+ *
+ * @param s const UChar * string buffer
+ * @param i string offset, must be i>10)+0xd7c0); \
+        (s)[(i)++]=(uint16_t)(((c)&0x3ff)|0xdc00); \
+    } else /* c>0x10ffff or not enough space */ { \
+        (isError)=true; \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Advance the string offset from one code point boundary to the next.
+ * (Post-incrementing iteration.)
+ * "Unsafe" macro, assumes well-formed UTF-16.
+ *
+ * @param s const UChar * string
+ * @param i string offset
+ * @see U16_FWD_1
+ * @stable ICU 2.4
+ */
+#define U16_FWD_1_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \
+    if(U16_IS_LEAD((s)[(i)++])) { \
+        ++(i); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Advance the string offset from one code point boundary to the next.
+ * (Post-incrementing iteration.)
+ * "Safe" macro, handles unpaired surrogates and checks for string boundaries.
+ *
+ * The length can be negative for a NUL-terminated string.
+ *
+ * @param s const UChar * string
+ * @param i string offset, must be i0) { \
+        U16_FWD_1_UNSAFE(s, i); \
+        --__N; \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Advance the string offset from one code point boundary to the n-th next one,
+ * i.e., move forward by n code points.
+ * (Post-incrementing iteration.)
+ * "Safe" macro, handles unpaired surrogates and checks for string boundaries.
+ *
+ * The length can be negative for a NUL-terminated string.
+ *
+ * @param s const UChar * string
+ * @param i int32_t string offset, must be i0 && ((i)<(length) || ((length)<0 && (s)[i]!=0))) { \
+        U16_FWD_1(s, i, length); \
+        --__N; \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Adjust a random-access offset to a code point boundary
+ * at the start of a code point.
+ * If the offset points to the trail surrogate of a surrogate pair,
+ * then the offset is decremented.
+ * Otherwise, it is not modified.
+ * "Unsafe" macro, assumes well-formed UTF-16.
+ *
+ * @param s const UChar * string
+ * @param i string offset
+ * @see U16_SET_CP_START
+ * @stable ICU 2.4
+ */
+#define U16_SET_CP_START_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \
+    if(U16_IS_TRAIL((s)[i])) { \
+        --(i); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Adjust a random-access offset to a code point boundary
+ * at the start of a code point.
+ * If the offset points to the trail surrogate of a surrogate pair,
+ * then the offset is decremented.
+ * Otherwise, it is not modified.
+ * "Safe" macro, handles unpaired surrogates and checks for string boundaries.
+ *
+ * @param s const UChar * string
+ * @param start starting string offset (usually 0)
+ * @param i string offset, must be start<=i
+ * @see U16_SET_CP_START_UNSAFE
+ * @stable ICU 2.4
+ */
+#define U16_SET_CP_START(s, start, i) UPRV_BLOCK_MACRO_BEGIN { \
+    if(U16_IS_TRAIL((s)[i]) && (i)>(start) && U16_IS_LEAD((s)[(i)-1])) { \
+        --(i); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/* definitions with backward iteration -------------------------------------- */
+
+/**
+ * Move the string offset from one code point boundary to the previous one
+ * and get the code point between them.
+ * (Pre-decrementing backward iteration.)
+ * "Unsafe" macro, assumes well-formed UTF-16.
+ *
+ * The input offset may be the same as the string length.
+ * If the offset is behind a trail surrogate unit
+ * for a supplementary code point, then the macro will read
+ * the preceding lead surrogate as well.
+ * If the offset is behind a lead surrogate, then that itself
+ * will be returned as the code point.
+ * The result is undefined if the offset is behind a single, unpaired trail surrogate.
+ *
+ * @param s const UChar * string
+ * @param i string offset
+ * @param c output UChar32 variable
+ * @see U16_PREV
+ * @stable ICU 2.4
+ */
+#define U16_PREV_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \
+    (c)=(s)[--(i)]; \
+    if(U16_IS_TRAIL(c)) { \
+        (c)=U16_GET_SUPPLEMENTARY((s)[--(i)], (c)); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Move the string offset from one code point boundary to the previous one
+ * and get the code point between them.
+ * (Pre-decrementing backward iteration.)
+ * "Safe" macro, handles unpaired surrogates and checks for string boundaries.
+ *
+ * The input offset may be the same as the string length.
+ * If the offset is behind a trail surrogate unit
+ * for a supplementary code point, then the macro will read
+ * the preceding lead surrogate as well.
+ * If the offset is behind a lead surrogate or behind a single, unpaired
+ * trail surrogate, then c is set to that unpaired surrogate.
+ *
+ * @param s const UChar * string
+ * @param start starting string offset (usually 0)
+ * @param i string offset, must be start(start) && U16_IS_LEAD(__c2=(s)[(i)-1])) { \
+            --(i); \
+            (c)=U16_GET_SUPPLEMENTARY(__c2, (c)); \
+        } \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Move the string offset from one code point boundary to the previous one
+ * and get the code point between them.
+ * (Pre-decrementing backward iteration.)
+ * "Safe" macro, handles unpaired surrogates and checks for string boundaries.
+ *
+ * The input offset may be the same as the string length.
+ * If the offset is behind a trail surrogate unit
+ * for a supplementary code point, then the macro will read
+ * the preceding lead surrogate as well.
+ * If the offset is behind a lead surrogate or behind a single, unpaired
+ * trail surrogate, then c is set to U+FFFD.
+ *
+ * @param s const UChar * string
+ * @param start starting string offset (usually 0)
+ * @param i string offset, must be start(start) && U16_IS_LEAD(__c2=(s)[(i)-1])) { \
+            --(i); \
+            (c)=U16_GET_SUPPLEMENTARY(__c2, (c)); \
+        } else { \
+            (c)=0xfffd; \
+        } \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Move the string offset from one code point boundary to the previous one.
+ * (Pre-decrementing backward iteration.)
+ * The input offset may be the same as the string length.
+ * "Unsafe" macro, assumes well-formed UTF-16.
+ *
+ * @param s const UChar * string
+ * @param i string offset
+ * @see U16_BACK_1
+ * @stable ICU 2.4
+ */
+#define U16_BACK_1_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \
+    if(U16_IS_TRAIL((s)[--(i)])) { \
+        --(i); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Move the string offset from one code point boundary to the previous one.
+ * (Pre-decrementing backward iteration.)
+ * The input offset may be the same as the string length.
+ * "Safe" macro, handles unpaired surrogates and checks for string boundaries.
+ *
+ * @param s const UChar * string
+ * @param start starting string offset (usually 0)
+ * @param i string offset, must be start(start) && U16_IS_LEAD((s)[(i)-1])) { \
+        --(i); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Move the string offset from one code point boundary to the n-th one before it,
+ * i.e., move backward by n code points.
+ * (Pre-decrementing backward iteration.)
+ * The input offset may be the same as the string length.
+ * "Unsafe" macro, assumes well-formed UTF-16.
+ *
+ * @param s const UChar * string
+ * @param i string offset
+ * @param n number of code points to skip
+ * @see U16_BACK_N
+ * @stable ICU 2.4
+ */
+#define U16_BACK_N_UNSAFE(s, i, n) UPRV_BLOCK_MACRO_BEGIN { \
+    int32_t __N=(n); \
+    while(__N>0) { \
+        U16_BACK_1_UNSAFE(s, i); \
+        --__N; \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Move the string offset from one code point boundary to the n-th one before it,
+ * i.e., move backward by n code points.
+ * (Pre-decrementing backward iteration.)
+ * The input offset may be the same as the string length.
+ * "Safe" macro, handles unpaired surrogates and checks for string boundaries.
+ *
+ * @param s const UChar * string
+ * @param start start of string
+ * @param i string offset, must be start0 && (i)>(start)) { \
+        U16_BACK_1(s, start, i); \
+        --__N; \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Adjust a random-access offset to a code point boundary after a code point.
+ * If the offset is behind the lead surrogate of a surrogate pair,
+ * then the offset is incremented.
+ * Otherwise, it is not modified.
+ * The input offset may be the same as the string length.
+ * "Unsafe" macro, assumes well-formed UTF-16.
+ *
+ * @param s const UChar * string
+ * @param i string offset
+ * @see U16_SET_CP_LIMIT
+ * @stable ICU 2.4
+ */
+#define U16_SET_CP_LIMIT_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \
+    if(U16_IS_LEAD((s)[(i)-1])) { \
+        ++(i); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Adjust a random-access offset to a code point boundary after a code point.
+ * If the offset is behind the lead surrogate of a surrogate pair,
+ * then the offset is incremented.
+ * Otherwise, it is not modified.
+ * The input offset may be the same as the string length.
+ * "Safe" macro, handles unpaired surrogates and checks for string boundaries.
+ *
+ * The length can be negative for a NUL-terminated string.
+ *
+ * @param s const UChar * string
+ * @param start int32_t starting string offset (usually 0)
+ * @param i int32_t string offset, start<=i<=length
+ * @param length int32_t string length
+ * @see U16_SET_CP_LIMIT_UNSAFE
+ * @stable ICU 2.4
+ */
+#define U16_SET_CP_LIMIT(s, start, i, length) UPRV_BLOCK_MACRO_BEGIN { \
+    if((start)<(i) && ((i)<(length) || (length)<0) && U16_IS_LEAD((s)[(i)-1]) && U16_IS_TRAIL((s)[i])) { \
+        ++(i); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+#endif
diff --git a/third_party/icu4c/ndk_headers/unicode/utf8.h b/third_party/icu4c/ndk_headers/unicode/utf8.h
new file mode 100644
index 0000000000000000000000000000000000000000..9d7bd37ce4979388e8e999b17ed1a6bce255c888
--- /dev/null
+++ b/third_party/icu4c/ndk_headers/unicode/utf8.h
@@ -0,0 +1,760 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*
+*   Copyright (C) 1999-2015, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*
+*******************************************************************************
+*   file name:  utf8.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   created on: 1999sep13
+*   created by: Markus W. Scherer
+*/
+
+/**
+ * \file
+ * \brief C API: 8-bit Unicode handling macros
+ * 
+ * This file defines macros to deal with 8-bit Unicode (UTF-8) code units (bytes) and strings.
+ *
+ * For more information see utf.h and the ICU User Guide Strings chapter
+ * (https://unicode-org.github.io/icu/userguide/strings).
+ *
+ * Usage:
+ * ICU coding guidelines for if() statements should be followed when using these macros.
+ * Compound statements (curly braces {}) must be used  for if-else-while... 
+ * bodies and all macro statements should be terminated with semicolon.
+ */
+
+#ifndef __UTF8_H__
+#define __UTF8_H__
+
+#include 
+#include "unicode/umachine.h"
+#ifndef __UTF_H__
+#   include "unicode/utf.h"
+#endif
+
+/* single-code point definitions -------------------------------------------- */
+
+/**
+ * Does this code unit (byte) encode a code point by itself (US-ASCII 0..0x7f)?
+ * @param c 8-bit code unit (byte)
+ * @return true or false
+ * @stable ICU 2.4
+ */
+#define U8_IS_SINGLE(c) (((c)&0x80)==0)
+
+/**
+ * Is this code unit (byte) a UTF-8 lead byte? (0xC2..0xF4)
+ * @param c 8-bit code unit (byte)
+ * @return true or false
+ * @stable ICU 2.4
+ */
+#define U8_IS_LEAD(c) ((uint8_t)((c)-0xc2)<=0x32)
+// 0x32=0xf4-0xc2
+
+/**
+ * Is this code unit (byte) a UTF-8 trail byte? (0x80..0xBF)
+ * @param c 8-bit code unit (byte)
+ * @return true or false
+ * @stable ICU 2.4
+ */
+#define U8_IS_TRAIL(c) ((int8_t)(c)<-0x40)
+
+/**
+ * How many code units (bytes) are used for the UTF-8 encoding
+ * of this Unicode code point?
+ * @param c 32-bit code point
+ * @return 1..4, or 0 if c is a surrogate or not a Unicode code point
+ * @stable ICU 2.4
+ */
+#define U8_LENGTH(c) \
+    ((uint32_t)(c)<=0x7f ? 1 : \
+        ((uint32_t)(c)<=0x7ff ? 2 : \
+            ((uint32_t)(c)<=0xd7ff ? 3 : \
+                ((uint32_t)(c)<=0xdfff || (uint32_t)(c)>0x10ffff ? 0 : \
+                    ((uint32_t)(c)<=0xffff ? 3 : 4)\
+                ) \
+            ) \
+        ) \
+    )
+
+/**
+ * The maximum number of UTF-8 code units (bytes) per Unicode code point (U+0000..U+10ffff).
+ * @return 4
+ * @stable ICU 2.4
+ */
+#define U8_MAX_LENGTH 4
+
+/**
+ * Get a code point from a string at a random-access offset,
+ * without changing the offset.
+ * The offset may point to either the lead byte or one of the trail bytes
+ * for a code point, in which case the macro will read all of the bytes
+ * for the code point.
+ * The result is undefined if the offset points to an illegal UTF-8
+ * byte sequence.
+ * Iteration through a string is more efficient with U8_NEXT_UNSAFE or U8_NEXT.
+ *
+ * @param s const uint8_t * string
+ * @param i string offset
+ * @param c output UChar32 variable
+ * @see U8_GET
+ * @stable ICU 2.4
+ */
+#define U8_GET_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \
+    int32_t _u8_get_unsafe_index=(int32_t)(i); \
+    U8_SET_CP_START_UNSAFE(s, _u8_get_unsafe_index); \
+    U8_NEXT_UNSAFE(s, _u8_get_unsafe_index, c); \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Get a code point from a string at a random-access offset,
+ * without changing the offset.
+ * The offset may point to either the lead byte or one of the trail bytes
+ * for a code point, in which case the macro will read all of the bytes
+ * for the code point.
+ *
+ * The length can be negative for a NUL-terminated string.
+ *
+ * If the offset points to an illegal UTF-8 byte sequence, then
+ * c is set to a negative value.
+ * Iteration through a string is more efficient with U8_NEXT_UNSAFE or U8_NEXT.
+ *
+ * @param s const uint8_t * string
+ * @param start int32_t starting string offset
+ * @param i int32_t string offset, must be start<=i=0xe0 ? \
+                ((c)<0xf0 ?  /* U+0800..U+FFFF except surrogates */ \
+                    U8_LEAD3_T1_BITS[(c)&=0xf]&(1<<((__t=(s)[i])>>5)) && \
+                    (__t&=0x3f, 1) \
+                :  /* U+10000..U+10FFFF */ \
+                    ((c)-=0xf0)<=4 && \
+                    U8_LEAD4_T1_BITS[(__t=(s)[i])>>4]&(1<<(c)) && \
+                    ((c)=((c)<<6)|(__t&0x3f), ++(i)!=(length)) && \
+                    (__t=(s)[i]-0x80)<=0x3f) && \
+                /* valid second-to-last trail byte */ \
+                ((c)=((c)<<6)|__t, ++(i)!=(length)) \
+            :  /* U+0080..U+07FF */ \
+                (c)>=0xc2 && ((c)&=0x1f, 1)) && \
+            /* last trail byte */ \
+            (__t=(s)[i]-0x80)<=0x3f && \
+            ((c)=((c)<<6)|__t, ++(i), 1)) { \
+        } else { \
+            (c)=(sub);  /* ill-formed*/ \
+        } \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Append a code point to a string, overwriting 1 to 4 bytes.
+ * The offset points to the current end of the string contents
+ * and is advanced (post-increment).
+ * "Unsafe" macro, assumes a valid code point and sufficient space in the string.
+ * Otherwise, the result is undefined.
+ *
+ * @param s const uint8_t * string buffer
+ * @param i string offset
+ * @param c code point to append
+ * @see U8_APPEND
+ * @stable ICU 2.4
+ */
+#define U8_APPEND_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \
+    uint32_t __uc=(c); \
+    if(__uc<=0x7f) { \
+        (s)[(i)++]=(uint8_t)__uc; \
+    } else { \
+        if(__uc<=0x7ff) { \
+            (s)[(i)++]=(uint8_t)((__uc>>6)|0xc0); \
+        } else { \
+            if(__uc<=0xffff) { \
+                (s)[(i)++]=(uint8_t)((__uc>>12)|0xe0); \
+            } else { \
+                (s)[(i)++]=(uint8_t)((__uc>>18)|0xf0); \
+                (s)[(i)++]=(uint8_t)(((__uc>>12)&0x3f)|0x80); \
+            } \
+            (s)[(i)++]=(uint8_t)(((__uc>>6)&0x3f)|0x80); \
+        } \
+        (s)[(i)++]=(uint8_t)((__uc&0x3f)|0x80); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Append a code point to a string, overwriting 1 to 4 bytes.
+ * The offset points to the current end of the string contents
+ * and is advanced (post-increment).
+ * "Safe" macro, checks for a valid code point.
+ * If a non-ASCII code point is written, checks for sufficient space in the string.
+ * If the code point is not valid or trail bytes do not fit,
+ * then isError is set to true.
+ *
+ * @param s const uint8_t * string buffer
+ * @param i int32_t string offset, must be i>6)|0xc0); \
+        (s)[(i)++]=(uint8_t)((__uc&0x3f)|0x80); \
+    } else if((__uc<=0xd7ff || (0xe000<=__uc && __uc<=0xffff)) && (i)+2<(capacity)) { \
+        (s)[(i)++]=(uint8_t)((__uc>>12)|0xe0); \
+        (s)[(i)++]=(uint8_t)(((__uc>>6)&0x3f)|0x80); \
+        (s)[(i)++]=(uint8_t)((__uc&0x3f)|0x80); \
+    } else if(0xffff<__uc && __uc<=0x10ffff && (i)+3<(capacity)) { \
+        (s)[(i)++]=(uint8_t)((__uc>>18)|0xf0); \
+        (s)[(i)++]=(uint8_t)(((__uc>>12)&0x3f)|0x80); \
+        (s)[(i)++]=(uint8_t)(((__uc>>6)&0x3f)|0x80); \
+        (s)[(i)++]=(uint8_t)((__uc&0x3f)|0x80); \
+    } else { \
+        (isError)=true; \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Advance the string offset from one code point boundary to the next.
+ * (Post-incrementing iteration.)
+ * "Unsafe" macro, assumes well-formed UTF-8.
+ *
+ * @param s const uint8_t * string
+ * @param i string offset
+ * @see U8_FWD_1
+ * @stable ICU 2.4
+ */
+#define U8_FWD_1_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \
+    (i)+=1+U8_COUNT_TRAIL_BYTES_UNSAFE((s)[i]); \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Advance the string offset from one code point boundary to the next.
+ * (Post-incrementing iteration.)
+ * "Safe" macro, checks for illegal sequences and for string boundaries.
+ *
+ * The length can be negative for a NUL-terminated string.
+ *
+ * @param s const uint8_t * string
+ * @param i int32_t string offset, must be i=0xf0 */ { \
+            if(U8_IS_VALID_LEAD4_AND_T1(__b, __t1) && \
+                    ++(i)!=(length) && U8_IS_TRAIL((s)[i]) && \
+                    ++(i)!=(length) && U8_IS_TRAIL((s)[i])) { \
+                ++(i); \
+            } \
+        } \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Advance the string offset from one code point boundary to the n-th next one,
+ * i.e., move forward by n code points.
+ * (Post-incrementing iteration.)
+ * "Unsafe" macro, assumes well-formed UTF-8.
+ *
+ * @param s const uint8_t * string
+ * @param i string offset
+ * @param n number of code points to skip
+ * @see U8_FWD_N
+ * @stable ICU 2.4
+ */
+#define U8_FWD_N_UNSAFE(s, i, n) UPRV_BLOCK_MACRO_BEGIN { \
+    int32_t __N=(n); \
+    while(__N>0) { \
+        U8_FWD_1_UNSAFE(s, i); \
+        --__N; \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Advance the string offset from one code point boundary to the n-th next one,
+ * i.e., move forward by n code points.
+ * (Post-incrementing iteration.)
+ * "Safe" macro, checks for illegal sequences and for string boundaries.
+ *
+ * The length can be negative for a NUL-terminated string.
+ *
+ * @param s const uint8_t * string
+ * @param i int32_t string offset, must be i0 && ((i)<(length) || ((length)<0 && (s)[i]!=0))) { \
+        U8_FWD_1(s, i, length); \
+        --__N; \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Adjust a random-access offset to a code point boundary
+ * at the start of a code point.
+ * If the offset points to a UTF-8 trail byte,
+ * then the offset is moved backward to the corresponding lead byte.
+ * Otherwise, it is not modified.
+ * "Unsafe" macro, assumes well-formed UTF-8.
+ *
+ * @param s const uint8_t * string
+ * @param i string offset
+ * @see U8_SET_CP_START
+ * @stable ICU 2.4
+ */
+#define U8_SET_CP_START_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \
+    while(U8_IS_TRAIL((s)[i])) { --(i); } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Adjust a random-access offset to a code point boundary
+ * at the start of a code point.
+ * If the offset points to a UTF-8 trail byte,
+ * then the offset is moved backward to the corresponding lead byte.
+ * Otherwise, it is not modified.
+ *
+ * "Safe" macro, checks for illegal sequences and for string boundaries.
+ * Unlike U8_TRUNCATE_IF_INCOMPLETE(), this macro always reads s[i].
+ *
+ * @param s const uint8_t * string
+ * @param start int32_t starting string offset (usually 0)
+ * @param i int32_t string offset, must be start<=i
+ * @see U8_SET_CP_START_UNSAFE
+ * @see U8_TRUNCATE_IF_INCOMPLETE
+ * @stable ICU 2.4
+ */
+#define U8_SET_CP_START(s, start, i) UPRV_BLOCK_MACRO_BEGIN { \
+    if(U8_IS_TRAIL((s)[(i)])) { \
+        (i)=utf8_back1SafeBody(s, start, (i)); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * If the string ends with a UTF-8 byte sequence that is valid so far
+ * but incomplete, then reduce the length of the string to end before
+ * the lead byte of that incomplete sequence.
+ * For example, if the string ends with E1 80, the length is reduced by 2.
+ *
+ * In all other cases (the string ends with a complete sequence, or it is not
+ * possible for any further trail byte to extend the trailing sequence)
+ * the length remains unchanged.
+ *
+ * Useful for processing text split across multiple buffers
+ * (save the incomplete sequence for later)
+ * and for optimizing iteration
+ * (check for string length only once per character).
+ *
+ * "Safe" macro, checks for illegal sequences and for string boundaries.
+ * Unlike U8_SET_CP_START(), this macro never reads s[length].
+ *
+ * (In UTF-16, simply check for U16_IS_LEAD(last code unit).)
+ *
+ * @param s const uint8_t * string
+ * @param start int32_t starting string offset (usually 0)
+ * @param length int32_t string length (usually start<=length)
+ * @see U8_SET_CP_START
+ * @stable ICU 61
+ */
+#define U8_TRUNCATE_IF_INCOMPLETE(s, start, length) UPRV_BLOCK_MACRO_BEGIN { \
+    if((length)>(start)) { \
+        uint8_t __b1=s[(length)-1]; \
+        if(U8_IS_SINGLE(__b1)) { \
+            /* common ASCII character */ \
+        } else if(U8_IS_LEAD(__b1)) { \
+            --(length); \
+        } else if(U8_IS_TRAIL(__b1) && ((length)-2)>=(start)) { \
+            uint8_t __b2=s[(length)-2]; \
+            if(0xe0<=__b2 && __b2<=0xf4) { \
+                if(__b2<0xf0 ? U8_IS_VALID_LEAD3_AND_T1(__b2, __b1) : \
+                        U8_IS_VALID_LEAD4_AND_T1(__b2, __b1)) { \
+                    (length)-=2; \
+                } \
+            } else if(U8_IS_TRAIL(__b2) && ((length)-3)>=(start)) { \
+                uint8_t __b3=s[(length)-3]; \
+                if(0xf0<=__b3 && __b3<=0xf4 && U8_IS_VALID_LEAD4_AND_T1(__b3, __b2)) { \
+                    (length)-=3; \
+                } \
+            } \
+        } \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/* definitions with backward iteration -------------------------------------- */
+
+/**
+ * Move the string offset from one code point boundary to the previous one
+ * and get the code point between them.
+ * (Pre-decrementing backward iteration.)
+ * "Unsafe" macro, assumes well-formed UTF-8.
+ *
+ * The input offset may be the same as the string length.
+ * If the offset is behind a multi-byte sequence, then the macro will read
+ * the whole sequence.
+ * If the offset is behind a lead byte, then that itself
+ * will be returned as the code point.
+ * The result is undefined if the offset is behind an illegal UTF-8 sequence.
+ *
+ * @param s const uint8_t * string
+ * @param i string offset
+ * @param c output UChar32 variable
+ * @see U8_PREV
+ * @stable ICU 2.4
+ */
+#define U8_PREV_UNSAFE(s, i, c) UPRV_BLOCK_MACRO_BEGIN { \
+    (c)=(uint8_t)(s)[--(i)]; \
+    if(U8_IS_TRAIL(c)) { \
+        uint8_t __b, __count=1, __shift=6; \
+\
+        /* c is a trail byte */ \
+        (c)&=0x3f; \
+        for(;;) { \
+            __b=(s)[--(i)]; \
+            if(__b>=0xc0) { \
+                U8_MASK_LEAD_BYTE(__b, __count); \
+                (c)|=(UChar32)__b<<__shift; \
+                break; \
+            } else { \
+                (c)|=(UChar32)(__b&0x3f)<<__shift; \
+                ++__count; \
+                __shift+=6; \
+            } \
+        } \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Move the string offset from one code point boundary to the previous one
+ * and get the code point between them.
+ * (Pre-decrementing backward iteration.)
+ * "Safe" macro, checks for illegal sequences and for string boundaries.
+ *
+ * The input offset may be the same as the string length.
+ * If the offset is behind a multi-byte sequence, then the macro will read
+ * the whole sequence.
+ * If the offset is behind a lead byte, then that itself
+ * will be returned as the code point.
+ * If the offset is behind an illegal UTF-8 sequence, then c is set to a negative value.
+ *
+ * @param s const uint8_t * string
+ * @param start int32_t starting string offset (usually 0)
+ * @param i int32_t string offset, must be start0) { \
+        U8_BACK_1_UNSAFE(s, i); \
+        --__N; \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Move the string offset from one code point boundary to the n-th one before it,
+ * i.e., move backward by n code points.
+ * (Pre-decrementing backward iteration.)
+ * The input offset may be the same as the string length.
+ * "Safe" macro, checks for illegal sequences and for string boundaries.
+ *
+ * @param s const uint8_t * string
+ * @param start int32_t index of the start of the string
+ * @param i int32_t string offset, must be start0 && (i)>(start)) { \
+        U8_BACK_1(s, start, i); \
+        --__N; \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Adjust a random-access offset to a code point boundary after a code point.
+ * If the offset is behind a partial multi-byte sequence,
+ * then the offset is incremented to behind the whole sequence.
+ * Otherwise, it is not modified.
+ * The input offset may be the same as the string length.
+ * "Unsafe" macro, assumes well-formed UTF-8.
+ *
+ * @param s const uint8_t * string
+ * @param i string offset
+ * @see U8_SET_CP_LIMIT
+ * @stable ICU 2.4
+ */
+#define U8_SET_CP_LIMIT_UNSAFE(s, i) UPRV_BLOCK_MACRO_BEGIN { \
+    U8_BACK_1_UNSAFE(s, i); \
+    U8_FWD_1_UNSAFE(s, i); \
+} UPRV_BLOCK_MACRO_END
+
+/**
+ * Adjust a random-access offset to a code point boundary after a code point.
+ * If the offset is behind a partial multi-byte sequence,
+ * then the offset is incremented to behind the whole sequence.
+ * Otherwise, it is not modified.
+ * The input offset may be the same as the string length.
+ * "Safe" macro, checks for illegal sequences and for string boundaries.
+ *
+ * The length can be negative for a NUL-terminated string.
+ *
+ * @param s const uint8_t * string
+ * @param start int32_t starting string offset (usually 0)
+ * @param i int32_t string offset, must be start<=i<=length
+ * @param length int32_t string length
+ * @see U8_SET_CP_LIMIT_UNSAFE
+ * @stable ICU 2.4
+ */
+#define U8_SET_CP_LIMIT(s, start, i, length) UPRV_BLOCK_MACRO_BEGIN { \
+    if((start)<(i) && ((i)<(length) || (length)<0)) { \
+        U8_BACK_1(s, start, i); \
+        U8_FWD_1(s, i, length); \
+    } \
+} UPRV_BLOCK_MACRO_END
+
+#endif
diff --git a/third_party/icu4c/ndk_headers/unicode/utypes.h b/third_party/icu4c/ndk_headers/unicode/utypes.h
index 8293ca70e7caafc051c959bd87c1d8b532341dca..1df6d0ec202da9f1b3574057df4de9002c12a9aa 100644
--- a/third_party/icu4c/ndk_headers/unicode/utypes.h
+++ b/third_party/icu4c/ndk_headers/unicode/utypes.h
@@ -36,6 +36,7 @@
 
 
 #include "unicode/umachine.h"
+#include "unicode/uversion.h"
 #include "unicode/uconfig.h"
 #include 
 
diff --git a/third_party/icu4c/ndk_headers/unicode/uvernum.h b/third_party/icu4c/ndk_headers/unicode/uvernum.h
new file mode 100644
index 0000000000000000000000000000000000000000..e77cd6195c80f05fdce228e7b71fb65d61b92f56
--- /dev/null
+++ b/third_party/icu4c/ndk_headers/unicode/uvernum.h
@@ -0,0 +1,128 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*   Copyright (C) 2000-2016, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*******************************************************************************
+*
+*   file name:  uvernum.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   Created by: Vladimir Weinstein
+*   Updated by: Steven R. Loomis
+*
+*/
+
+/**
+ * \file
+ * \brief C API: definitions of ICU version numbers
+ *
+ * This file is included by uversion.h and other files. This file contains only
+ * macros and definitions. The actual version numbers are defined here.
+ */
+
+ /*
+  * IMPORTANT: When updating version, the following things need to be done:
+  * source/common/unicode/uvernum.h - this file: update major, minor,
+  *        patchlevel, suffix, version, short version constants, namespace,
+  *                    renaming macro, and copyright
+  *
+  * The following files need to be updated as well, which can be done
+  *  by running the UNIX makefile target 'update-windows-makefiles' in icu4c/source.
+  *
+  * source/allinone/Build.Windows.IcuVersion.props - Update the IcuMajorVersion
+  * source/data/makedata.mak - change U_ICUDATA_NAME so that it contains
+  *                            the new major/minor combination, and UNICODE_VERSION
+  *                            for the Unicode version.
+  */
+
+#ifndef UVERNUM_H
+#define UVERNUM_H
+
+/** The standard copyright notice that gets compiled into each library.
+ *  This value will change in the subsequent releases of ICU
+ *  @stable ICU 2.4
+ */
+#define U_COPYRIGHT_STRING \
+  " Copyright (C) 2016 and later: Unicode, Inc. and others. License & terms of use: http://www.unicode.org/copyright.html "
+
+/** The current ICU major version as an integer.
+ *  This value will change in the subsequent releases of ICU
+ *  @stable ICU 2.4
+ */
+#define U_ICU_VERSION_MAJOR_NUM 72
+
+/** The current ICU minor version as an integer.
+ *  This value will change in the subsequent releases of ICU
+ *  @stable ICU 2.6
+ */
+#define U_ICU_VERSION_MINOR_NUM 1
+
+/** The current ICU patchlevel version as an integer.
+ *  This value will change in the subsequent releases of ICU
+ *  @stable ICU 2.4
+ */
+#define U_ICU_VERSION_PATCHLEVEL_NUM 0
+
+/** The current ICU build level version as an integer.
+ *  This value is for use by ICU clients. It defaults to 0.
+ *  @stable ICU 4.0
+ */
+#ifndef U_ICU_VERSION_BUILDLEVEL_NUM
+#define U_ICU_VERSION_BUILDLEVEL_NUM 0
+#endif
+
+/** Glued version suffix for renamers
+ *  This value will change in the subsequent releases of ICU
+ *  @stable ICU 2.6
+ */
+#define U_ICU_VERSION_SUFFIX _72
+
+/** The current ICU library version as a dotted-decimal string. The patchlevel
+ *  only appears in this string if it non-zero.
+ *  This value will change in the subsequent releases of ICU
+ *  @stable ICU 2.4
+ */
+#define U_ICU_VERSION "72.1"
+
+/**
+ * The current ICU library major version number as a string, for library name suffixes.
+ * This value will change in subsequent releases of ICU.
+ *
+ * Until ICU 4.8, this was the combination of the single-digit major and minor ICU version numbers
+ * into one string without dots ("48").
+ * Since ICU 49, it is the double-digit major ICU version number.
+ * See https://unicode-org.github.io/icu/userguide/design#version-numbers-in-icu
+ *
+ * @stable ICU 2.6
+ */
+#define U_ICU_VERSION_SHORT "72"
+
+/*===========================================================================
+ * ICU collation framework version information
+ * Version info that can be obtained from a collator is affected by these
+ * numbers in a secret and magic way. Please use collator version as whole
+ *===========================================================================
+ */
+
+/**
+ * Collation runtime version (sort key generator, strcoll).
+ * If the version is different, sort keys for the same string could be different.
+ * This value may change in subsequent releases of ICU.
+ * @stable ICU 2.4
+ */
+#define UCOL_RUNTIME_VERSION 9
+
+/**
+ * Collation builder code version.
+ * When this is different, the same tailoring might result
+ * in assigning different collation elements to code points.
+ * This value may change in subsequent releases of ICU.
+ * @stable ICU 2.4
+ */
+#define UCOL_BUILDER_VERSION 9
+
+#endif
diff --git a/third_party/icu4c/ndk_headers/unicode/uversion.h b/third_party/icu4c/ndk_headers/unicode/uversion.h
new file mode 100644
index 0000000000000000000000000000000000000000..5b4f37375879094e9777ab78f5a2df0925803a3c
--- /dev/null
+++ b/third_party/icu4c/ndk_headers/unicode/uversion.h
@@ -0,0 +1,119 @@
+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
+/*
+*******************************************************************************
+*   Copyright (C) 2000-2011, International Business Machines
+*   Corporation and others.  All Rights Reserved.
+*******************************************************************************
+*
+*   file name:  uversion.h
+*   encoding:   UTF-8
+*   tab size:   8 (not used)
+*   indentation:4
+*
+*   Created by: Vladimir Weinstein
+*
+*  Gets included by utypes.h and Windows .rc files
+*/
+
+/**
+ * \file
+ * \brief C API: API for accessing ICU version numbers. 
+ */
+/*===========================================================================*/
+/* Main ICU version information                                              */
+/*===========================================================================*/
+
+#ifndef UVERSION_H
+#define UVERSION_H
+
+#include "unicode/umachine.h"
+
+/* Actual version info lives in uvernum.h */
+#include "unicode/uvernum.h"
+
+/** Maximum length of the copyright string.
+ *  @stable ICU 2.4
+ */
+#define U_COPYRIGHT_STRING_LENGTH  128
+
+/** An ICU version consists of up to 4 numbers from 0..255.
+ *  @stable ICU 2.4
+ */
+#define U_MAX_VERSION_LENGTH 4
+
+/** In a string, ICU version fields are delimited by dots.
+ *  @stable ICU 2.4
+ */
+#define U_VERSION_DELIMITER '.'
+
+/** The maximum length of an ICU version string.
+ *  @stable ICU 2.4
+ */
+#define U_MAX_VERSION_STRING_LENGTH 20
+
+/** The binary form of a version on ICU APIs is an array of 4 uint8_t.
+ *  To compare two versions, use memcmp(v1,v2,sizeof(UVersionInfo)).
+ *  @stable ICU 2.4
+ */
+typedef uint8_t UVersionInfo[U_MAX_VERSION_LENGTH];
+
+/*===========================================================================*/
+/* General version helper functions. Definitions in putil.c                  */
+/*===========================================================================*/
+
+/**
+ * Parse a string with dotted-decimal version information and
+ * fill in a UVersionInfo structure with the result.
+ * Definition of this function lives in putil.c
+ *
+ * @param versionArray The destination structure for the version information.
+ * @param versionString A string with dotted-decimal version information,
+ *                      with up to four non-negative number fields with
+ *                      values of up to 255 each.
+ * @stable ICU 2.4
+ */
+U_CAPI void U_EXPORT2
+u_versionFromString(UVersionInfo versionArray, const char *versionString);
+
+/**
+ * Parse a Unicode string with dotted-decimal version information and
+ * fill in a UVersionInfo structure with the result.
+ * Definition of this function lives in putil.c
+ *
+ * @param versionArray The destination structure for the version information.
+ * @param versionString A Unicode string with dotted-decimal version
+ *                      information, with up to four non-negative number
+ *                      fields with values of up to 255 each.
+ * @stable ICU 4.2
+ */
+U_CAPI void U_EXPORT2
+u_versionFromUString(UVersionInfo versionArray, const UChar *versionString);
+
+
+/**
+ * Write a string with dotted-decimal version information according
+ * to the input UVersionInfo.
+ * Definition of this function lives in putil.c
+ *
+ * @param versionArray The version information to be written as a string.
+ * @param versionString A string buffer that will be filled in with
+ *                      a string corresponding to the numeric version
+ *                      information in versionArray.
+ *                      The buffer size must be at least U_MAX_VERSION_STRING_LENGTH.
+ * @stable ICU 2.4
+ */
+U_CAPI void U_EXPORT2
+u_versionToString(const UVersionInfo versionArray, char *versionString);
+
+/**
+ * Gets the ICU release version.  The version array stores the version information
+ * for ICU.  For example, release "1.3.31.2" is then represented as 0x01031F02.
+ * Definition of this function lives in putil.c
+ *
+ * @param versionArray the version # information, the result will be filled in
+ * @stable ICU 2.0
+ */
+U_CAPI void U_EXPORT2
+u_getVersion(UVersionInfo versionArray);
+#endif