/src/icu/icu4c/source/common/characterproperties.cpp

Source (jump to first uncovered line)
// © 2018 and later: Unicode, Inc. and others.
// License & terms of use: http://www.unicode.org/copyright.html

// characterproperties.cpp
// created: 2018sep03 Markus W. Scherer

#include "unicode/utypes.h"
#include "unicode/localpointer.h"
#include "unicode/uchar.h"
#include "unicode/ucpmap.h"
#include "unicode/ucptrie.h"
#include "unicode/umutablecptrie.h"
#include "unicode/uniset.h"
#include "unicode/uscript.h"
#include "unicode/uset.h"
#include "cmemory.h"
#include "emojiprops.h"
#include "mutex.h"
#include "normalizer2impl.h"
#include "uassert.h"
#include "ubidi_props.h"
#include "ucase.h"
#include "ucln_cmn.h"
#include "umutex.h"
#include "uprops.h"

using icu::LocalPointer;
#if !UCONFIG_NO_NORMALIZATION
using icu::Normalizer2Factory;
using icu::Normalizer2Impl;
#endif
using icu::UInitOnce;
using icu::UnicodeSet;

namespace {

UBool U_CALLCONV characterproperties_cleanup();

constexpr int32_t NUM_INCLUSIONS = UPROPS_SRC_COUNT + (UCHAR_INT_LIMIT - UCHAR_INT_START);

struct Inclusion {
    UnicodeSet  *fSet = nullptr;
    UInitOnce    fInitOnce {};
};
Inclusion gInclusions[NUM_INCLUSIONS]; // cached getInclusions()

UnicodeSet *sets[UCHAR_BINARY_LIMIT] = {};

UCPMap *maps[UCHAR_INT_LIMIT - UCHAR_INT_START] = {};

icu::UMutex cpMutex;

//----------------------------------------------------------------
// Inclusions list
//----------------------------------------------------------------

// USetAdder implementation
// Does not use uset.h to reduce code dependencies
void U_CALLCONV
_set_add(USet *set, UChar32 c) {
    reinterpret_cast<UnicodeSet*>(set)->add(c);
}

void U_CALLCONV
_set_addRange(USet *set, UChar32 start, UChar32 end) {
    reinterpret_cast<UnicodeSet*>(set)->add(start, end);
}

void U_CALLCONV
_set_addString(USet *set, const char16_t *str, int32_t length) {
    reinterpret_cast<UnicodeSet*>(set)->add(icu::UnicodeString(static_cast<UBool>(length < 0), str, length));
}

UBool U_CALLCONV characterproperties_cleanup() {
    for (Inclusion &in: gInclusions) {
        delete in.fSet;
        in.fSet = nullptr;
        in.fInitOnce.reset();
    }
    for (int32_t i = 0; i < UPRV_LENGTHOF(sets); ++i) {
        delete sets[i];
        sets[i] = nullptr;
    }
    for (int32_t i = 0; i < UPRV_LENGTHOF(maps); ++i) {
        ucptrie_close(reinterpret_cast<UCPTrie *>(maps[i]));
        maps[i] = nullptr;
    }
    return true;
}

void U_CALLCONV initInclusion(UPropertySource src, UErrorCode &errorCode) {
    // This function is invoked only via umtx_initOnce().
    U_ASSERT(0 <= src && src < UPROPS_SRC_COUNT);
    if (src == UPROPS_SRC_NONE) {
        errorCode = U_INTERNAL_PROGRAM_ERROR;
        return;
    }
    U_ASSERT(gInclusions[src].fSet == nullptr);

    LocalPointer<UnicodeSet> incl(new UnicodeSet());
    if (incl.isNull()) {
        errorCode = U_MEMORY_ALLOCATION_ERROR;
        return;
    }
    USetAdder sa = {
        reinterpret_cast<USet*>(incl.getAlias()),
        _set_add,
        _set_addRange,
        _set_addString,
        nullptr, // don't need remove()
        nullptr // don't need removeRange()
    };

    switch(src) {
    case UPROPS_SRC_CHAR:
        uchar_addPropertyStarts(&sa, &errorCode);
        break;
    case UPROPS_SRC_PROPSVEC:
        upropsvec_addPropertyStarts(&sa, &errorCode);
        break;
    case UPROPS_SRC_CHAR_AND_PROPSVEC:
        uchar_addPropertyStarts(&sa, &errorCode);
        upropsvec_addPropertyStarts(&sa, &errorCode);
        break;
#if !UCONFIG_NO_NORMALIZATION
    case UPROPS_SRC_CASE_AND_NORM: {
        const Normalizer2Impl *impl=Normalizer2Factory::getNFCImpl(errorCode);
        if(U_SUCCESS(errorCode)) {
            impl->addPropertyStarts(&sa, errorCode);
        }
        ucase_addPropertyStarts(&sa, &errorCode);
        break;
    }
    case UPROPS_SRC_NFC: {
        const Normalizer2Impl *impl=Normalizer2Factory::getNFCImpl(errorCode);
        if(U_SUCCESS(errorCode)) {
            impl->addPropertyStarts(&sa, errorCode);
        }
        break;
    }
    case UPROPS_SRC_NFKC: {
        const Normalizer2Impl *impl=Normalizer2Factory::getNFKCImpl(errorCode);
        if(U_SUCCESS(errorCode)) {
            impl->addPropertyStarts(&sa, errorCode);
        }
        break;
    }
    case UPROPS_SRC_NFKC_CF: {
        const Normalizer2Impl *impl=Normalizer2Factory::getNFKC_CFImpl(errorCode);
        if(U_SUCCESS(errorCode)) {
            impl->addPropertyStarts(&sa, errorCode);
        }
        break;
    }
    case UPROPS_SRC_NFC_CANON_ITER: {
        const Normalizer2Impl *impl=Normalizer2Factory::getNFCImpl(errorCode);
        if(U_SUCCESS(errorCode)) {
            impl->addCanonIterPropertyStarts(&sa, errorCode);
        }
        break;
    }
#endif
    case UPROPS_SRC_CASE:
        ucase_addPropertyStarts(&sa, &errorCode);
        break;
    case UPROPS_SRC_BIDI:
        ubidi_addPropertyStarts(&sa, &errorCode);
        break;
    case UPROPS_SRC_INPC:
    case UPROPS_SRC_INSC:
    case UPROPS_SRC_VO:
        uprops_addPropertyStarts(src, &sa, &errorCode);
        break;
    case UPROPS_SRC_EMOJI: {
        const icu::EmojiProps *ep = icu::EmojiProps::getSingleton(errorCode);
        if (U_SUCCESS(errorCode)) {
            ep->addPropertyStarts(&sa, errorCode);
        }
        break;
    }
    case UPROPS_SRC_IDSU:
        // New in Unicode 15.1 for just two characters.
        sa.add(sa.set, 0x2FFE);
        sa.add(sa.set, 0x2FFF + 1);
        break;
    case UPROPS_SRC_ID_COMPAT_MATH:
    case UPROPS_SRC_MCM:
        uprops_addPropertyStarts(src, &sa, &errorCode);
        break;
    case UPROPS_SRC_BLOCK:
        ublock_addPropertyStarts(&sa, errorCode);
        break;
    default:
        errorCode = U_INTERNAL_PROGRAM_ERROR;
        break;
    }

    if (U_FAILURE(errorCode)) {
        return;
    }
    if (incl->isBogus()) {
        errorCode = U_MEMORY_ALLOCATION_ERROR;
        return;
    }
    // Compact for caching.
    incl->compact();
    gInclusions[src].fSet = incl.orphan();
    ucln_common_registerCleanup(UCLN_COMMON_CHARACTERPROPERTIES, characterproperties_cleanup);
}

const UnicodeSet *getInclusionsForSource(UPropertySource src, UErrorCode &errorCode) {
    if (U_FAILURE(errorCode)) { return nullptr; }
    if (src < 0 || UPROPS_SRC_COUNT <= src) {
        errorCode = U_ILLEGAL_ARGUMENT_ERROR;
        return nullptr;
    }
    Inclusion &i = gInclusions[src];
    umtx_initOnce(i.fInitOnce, &initInclusion, src, errorCode);
    return i.fSet;
}

void U_CALLCONV initIntPropInclusion(UProperty prop, UErrorCode &errorCode) {
    // This function is invoked only via umtx_initOnce().
    U_ASSERT(UCHAR_INT_START <= prop && prop < UCHAR_INT_LIMIT);
    int32_t inclIndex = UPROPS_SRC_COUNT + (prop - UCHAR_INT_START);
    U_ASSERT(gInclusions[inclIndex].fSet == nullptr);
    UPropertySource src = uprops_getSource(prop);
    const UnicodeSet *incl = getInclusionsForSource(src, errorCode);
    if (U_FAILURE(errorCode)) {
        return;
    }

    LocalPointer<UnicodeSet> intPropIncl(new UnicodeSet(0, 0));
    if (intPropIncl.isNull()) {
        errorCode = U_MEMORY_ALLOCATION_ERROR;
        return;
    }
    int32_t numRanges = incl->getRangeCount();
    int32_t prevValue = 0;
    for (int32_t i = 0; i < numRanges; ++i) {
        UChar32 rangeEnd = incl->getRangeEnd(i);
        for (UChar32 c = incl->getRangeStart(i); c <= rangeEnd; ++c) {
            // TODO: Get a UCharacterProperty.IntProperty to avoid the property dispatch.
            int32_t value = u_getIntPropertyValue(c, prop);
            if (value != prevValue) {
                intPropIncl->add(c);
                prevValue = value;
            }
        }
    }

    if (intPropIncl->isBogus()) {
        errorCode = U_MEMORY_ALLOCATION_ERROR;
        return;
    }
    // Compact for caching.
    intPropIncl->compact();
    gInclusions[inclIndex].fSet = intPropIncl.orphan();
    ucln_common_registerCleanup(UCLN_COMMON_CHARACTERPROPERTIES, characterproperties_cleanup);
}

}  // namespace

U_NAMESPACE_BEGIN

const UnicodeSet *CharacterProperties::getInclusionsForProperty(
        UProperty prop, UErrorCode &errorCode) {
    if (U_FAILURE(errorCode)) { return nullptr; }
    if (UCHAR_INT_START <= prop && prop < UCHAR_INT_LIMIT) {
        int32_t inclIndex = UPROPS_SRC_COUNT + (prop - UCHAR_INT_START);
        Inclusion &i = gInclusions[inclIndex];
        umtx_initOnce(i.fInitOnce, &initIntPropInclusion, prop, errorCode);
        return i.fSet;
    } else {
        UPropertySource src = uprops_getSource(prop);
        return getInclusionsForSource(src, errorCode);
    }
}

U_NAMESPACE_END

namespace {

UnicodeSet *makeSet(UProperty property, UErrorCode &errorCode) {
    if (U_FAILURE(errorCode)) { return nullptr; }
    LocalPointer<UnicodeSet> set(new UnicodeSet());
    if (set.isNull()) {
        errorCode = U_MEMORY_ALLOCATION_ERROR;
        return nullptr;
    }
    if (UCHAR_BASIC_EMOJI <= property && property <= UCHAR_RGI_EMOJI) {
        // property of strings
        const icu::EmojiProps *ep = icu::EmojiProps::getSingleton(errorCode);
        if (U_FAILURE(errorCode)) { return nullptr; }
        USetAdder sa = {
            reinterpret_cast<USet*>(set.getAlias()),
            _set_add,
            _set_addRange,
            _set_addString,
            nullptr, // don't need remove()
            nullptr // don't need removeRange()
        };
        ep->addStrings(&sa, property, errorCode);
        if (property != UCHAR_BASIC_EMOJI && property != UCHAR_RGI_EMOJI) {
            // property of _only_ strings
            set->freeze();
            return set.orphan();
        }
    }

    const UnicodeSet *inclusions =
        icu::CharacterProperties::getInclusionsForProperty(property, errorCode);
    if (U_FAILURE(errorCode)) { return nullptr; }
    int32_t numRanges = inclusions->getRangeCount();
    UChar32 startHasProperty = -1;

    for (int32_t i = 0; i < numRanges; ++i) {
        UChar32 rangeEnd = inclusions->getRangeEnd(i);
        for (UChar32 c = inclusions->getRangeStart(i); c <= rangeEnd; ++c) {
            // TODO: Get a UCharacterProperty.BinaryProperty to avoid the property dispatch.
            if (u_hasBinaryProperty(c, property)) {
                if (startHasProperty < 0) {
                    // Transition from false to true.
                    startHasProperty = c;
                }
            } else if (startHasProperty >= 0) {
                // Transition from true to false.
                set->add(startHasProperty, c - 1);
                startHasProperty = -1;
            }
        }
    }
    if (startHasProperty >= 0) {
        set->add(startHasProperty, 0x10FFFF);
    }
    set->freeze();
    return set.orphan();
}

UCPMap *makeMap(UProperty property, UErrorCode &errorCode) {
    if (U_FAILURE(errorCode)) { return nullptr; }
    uint32_t nullValue = property == UCHAR_SCRIPT ? USCRIPT_UNKNOWN : 0;
    icu::LocalUMutableCPTriePointer mutableTrie(
        umutablecptrie_open(nullValue, nullValue, &errorCode));
    const UnicodeSet *inclusions =
        icu::CharacterProperties::getInclusionsForProperty(property, errorCode);
    if (U_FAILURE(errorCode)) { return nullptr; }
    int32_t numRanges = inclusions->getRangeCount();
    UChar32 start = 0;
    uint32_t value = nullValue;

    for (int32_t i = 0; i < numRanges; ++i) {
        UChar32 rangeEnd = inclusions->getRangeEnd(i);
        for (UChar32 c = inclusions->getRangeStart(i); c <= rangeEnd; ++c) {
            // TODO: Get a UCharacterProperty.IntProperty to avoid the property dispatch.
            uint32_t nextValue = u_getIntPropertyValue(c, property);
            if (value != nextValue) {
                if (value != nullValue) {
                    umutablecptrie_setRange(mutableTrie.getAlias(), start, c - 1, value, &errorCode);
                }
                start = c;
                value = nextValue;
            }
        }
    }
    if (value != 0) {
        umutablecptrie_setRange(mutableTrie.getAlias(), start, 0x10FFFF, value, &errorCode);
    }

    UCPTrieType type;
    if (property == UCHAR_BIDI_CLASS || property == UCHAR_GENERAL_CATEGORY) {
        type = UCPTRIE_TYPE_FAST;
    } else {
        type = UCPTRIE_TYPE_SMALL;
    }
    UCPTrieValueWidth valueWidth;
    // TODO: UCharacterProperty.IntProperty
    int32_t max = u_getIntPropertyMaxValue(property);
    if (max <= 0xff) {
        valueWidth = UCPTRIE_VALUE_BITS_8;
    } else if (max <= 0xffff) {
        valueWidth = UCPTRIE_VALUE_BITS_16;
    } else {
        valueWidth = UCPTRIE_VALUE_BITS_32;
    }
    return reinterpret_cast<UCPMap *>(
        umutablecptrie_buildImmutable(mutableTrie.getAlias(), type, valueWidth, &errorCode));
}

}  // namespace

U_NAMESPACE_BEGIN

const UnicodeSet *CharacterProperties::getBinaryPropertySet(UProperty property, UErrorCode &errorCode) {
    if (U_FAILURE(errorCode)) { return nullptr; }
    if (property < 0 || UCHAR_BINARY_LIMIT <= property) {
        errorCode = U_ILLEGAL_ARGUMENT_ERROR;
        return nullptr;
    }
    Mutex m(&cpMutex);
    UnicodeSet *set = sets[property];
    if (set == nullptr) {
        sets[property] = set = makeSet(property, errorCode);
    }
    return set;
}

U_NAMESPACE_END

U_NAMESPACE_USE

U_CAPI const USet * U_EXPORT2
u_getBinaryPropertySet(UProperty property, UErrorCode *pErrorCode) {
    const UnicodeSet *set = CharacterProperties::getBinaryPropertySet(property, *pErrorCode);
    return U_SUCCESS(*pErrorCode) ? set->toUSet() : nullptr;
}

U_CAPI const UCPMap * U_EXPORT2
u_getIntPropertyMap(UProperty property, UErrorCode *pErrorCode) {
    if (U_FAILURE(*pErrorCode)) { return nullptr; }
    if (property < UCHAR_INT_START || UCHAR_INT_LIMIT <= property) {
        *pErrorCode = U_ILLEGAL_ARGUMENT_ERROR;
        return nullptr;
    }
    Mutex m(&cpMutex);
    UCPMap *map = maps[property - UCHAR_INT_START];
    if (map == nullptr) {
        maps[property - UCHAR_INT_START] = map = makeMap(property, *pErrorCode);
    }
    return map;
}

Coverage Report

Created: 2025-06-13 06:34

Line	Count	Source (jump to first uncovered line)
1		// © 2018 and later: Unicode, Inc. and others.
2		// License & terms of use: http://www.unicode.org/copyright.html
3
4		// characterproperties.cpp
5		// created: 2018sep03 Markus W. Scherer
6
7		#include "unicode/utypes.h"
8		#include "unicode/localpointer.h"
9		#include "unicode/uchar.h"
10		#include "unicode/ucpmap.h"
11		#include "unicode/ucptrie.h"
12		#include "unicode/umutablecptrie.h"
13		#include "unicode/uniset.h"
14		#include "unicode/uscript.h"
15		#include "unicode/uset.h"
16		#include "cmemory.h"
17		#include "emojiprops.h"
18		#include "mutex.h"
19		#include "normalizer2impl.h"
20		#include "uassert.h"
21		#include "ubidi_props.h"
22		#include "ucase.h"
23		#include "ucln_cmn.h"
24		#include "umutex.h"
25		#include "uprops.h"
26
27		using icu::LocalPointer;
28		#if !UCONFIG_NO_NORMALIZATION
29		using icu::Normalizer2Factory;
30		using icu::Normalizer2Impl;
31		#endif
32		using icu::UInitOnce;
33		using icu::UnicodeSet;
34
35		namespace {
36
37		UBool U_CALLCONV characterproperties_cleanup();
38
39		constexpr int32_t NUM_INCLUSIONS = UPROPS_SRC_COUNT + (UCHAR_INT_LIMIT - UCHAR_INT_START);
40
41		struct Inclusion {
42		UnicodeSet *fSet = nullptr;
43		UInitOnce fInitOnce {};
44		};
45		Inclusion gInclusions[NUM_INCLUSIONS]; // cached getInclusions()
46
47		UnicodeSet *sets[UCHAR_BINARY_LIMIT] = {};
48
49		UCPMap *maps[UCHAR_INT_LIMIT - UCHAR_INT_START] = {};
50
51		icu::UMutex cpMutex;
52
53		//----------------------------------------------------------------
54		// Inclusions list
55		//----------------------------------------------------------------
56
57		// USetAdder implementation
58		// Does not use uset.h to reduce code dependencies
59		void U_CALLCONV
60	0	_set_add(USet *set, UChar32 c) {
61	0	reinterpret_cast<UnicodeSet*>(set)->add(c);
62	0	}
63
64		void U_CALLCONV
65	0	_set_addRange(USet *set, UChar32 start, UChar32 end) {
66	0	reinterpret_cast<UnicodeSet*>(set)->add(start, end);
67	0	}
68
69		void U_CALLCONV
70	0	_set_addString(USet set, const char16_t str, int32_t length) {
71	0	reinterpret_cast<UnicodeSet*>(set)->add(icu::UnicodeString(static_cast<UBool>(length < 0), str, length));
72	0	}
73
74	0	UBool U_CALLCONV characterproperties_cleanup() {
75	0	for (Inclusion &in: gInclusions) {
76	0	delete in.fSet;
77	0	in.fSet = nullptr;
78	0	in.fInitOnce.reset();
79	0	}
80	0	for (int32_t i = 0; i < UPRV_LENGTHOF(sets); ++i) {
81	0	delete sets[i];
82	0	sets[i] = nullptr;
83	0	}
84	0	for (int32_t i = 0; i < UPRV_LENGTHOF(maps); ++i) {
85	0	ucptrie_close(reinterpret_cast<UCPTrie *>(maps[i]));
86	0	maps[i] = nullptr;
87	0	}
88	0	return true;
89	0	}
90
91	0	void U_CALLCONV initInclusion(UPropertySource src, UErrorCode &errorCode) {
92		// This function is invoked only via umtx_initOnce().
93	0	U_ASSERT(0 <= src && src < UPROPS_SRC_COUNT);
94	0	if (src == UPROPS_SRC_NONE) {
95	0	errorCode = U_INTERNAL_PROGRAM_ERROR;
96	0	return;
97	0	}
98	0	U_ASSERT(gInclusions[src].fSet == nullptr);
99
100	0	LocalPointer<UnicodeSet> incl(new UnicodeSet());
101	0	if (incl.isNull()) {
102	0	errorCode = U_MEMORY_ALLOCATION_ERROR;
103	0	return;
104	0	}
105	0	USetAdder sa = {
106	0	reinterpret_cast<USet*>(incl.getAlias()),
107	0	_set_add,
108	0	_set_addRange,
109	0	_set_addString,
110	0	nullptr, // don't need remove()
111	0	nullptr // don't need removeRange()
112	0	};
113
114	0	switch(src) {
115	0	case UPROPS_SRC_CHAR:
116	0	uchar_addPropertyStarts(&sa, &errorCode);
117	0	break;
118	0	case UPROPS_SRC_PROPSVEC:
119	0	upropsvec_addPropertyStarts(&sa, &errorCode);
120	0	break;
121	0	case UPROPS_SRC_CHAR_AND_PROPSVEC:
122	0	uchar_addPropertyStarts(&sa, &errorCode);
123	0	upropsvec_addPropertyStarts(&sa, &errorCode);
124	0	break;
125	0	#if !UCONFIG_NO_NORMALIZATION
126	0	case UPROPS_SRC_CASE_AND_NORM: {
127	0	const Normalizer2Impl *impl=Normalizer2Factory::getNFCImpl(errorCode);
128	0	if(U_SUCCESS(errorCode)) {
129	0	impl->addPropertyStarts(&sa, errorCode);
130	0	}
131	0	ucase_addPropertyStarts(&sa, &errorCode);
132	0	break;
133	0	}
134	0	case UPROPS_SRC_NFC: {
135	0	const Normalizer2Impl *impl=Normalizer2Factory::getNFCImpl(errorCode);
136	0	if(U_SUCCESS(errorCode)) {
137	0	impl->addPropertyStarts(&sa, errorCode);
138	0	}
139	0	break;
140	0	}
141	0	case UPROPS_SRC_NFKC: {
142	0	const Normalizer2Impl *impl=Normalizer2Factory::getNFKCImpl(errorCode);
143	0	if(U_SUCCESS(errorCode)) {
144	0	impl->addPropertyStarts(&sa, errorCode);
145	0	}
146	0	break;
147	0	}
148	0	case UPROPS_SRC_NFKC_CF: {
149	0	const Normalizer2Impl *impl=Normalizer2Factory::getNFKC_CFImpl(errorCode);
150	0	if(U_SUCCESS(errorCode)) {
151	0	impl->addPropertyStarts(&sa, errorCode);
152	0	}
153	0	break;
154	0	}
155	0	case UPROPS_SRC_NFC_CANON_ITER: {
156	0	const Normalizer2Impl *impl=Normalizer2Factory::getNFCImpl(errorCode);
157	0	if(U_SUCCESS(errorCode)) {
158	0	impl->addCanonIterPropertyStarts(&sa, errorCode);
159	0	}
160	0	break;
161	0	}
162	0	#endif
163	0	case UPROPS_SRC_CASE:
164	0	ucase_addPropertyStarts(&sa, &errorCode);
165	0	break;
166	0	case UPROPS_SRC_BIDI:
167	0	ubidi_addPropertyStarts(&sa, &errorCode);
168	0	break;
169	0	case UPROPS_SRC_INPC:
170	0	case UPROPS_SRC_INSC:
171	0	case UPROPS_SRC_VO:
172	0	uprops_addPropertyStarts(src, &sa, &errorCode);
173	0	break;
174	0	case UPROPS_SRC_EMOJI: {
175	0	const icu::EmojiProps *ep = icu::EmojiProps::getSingleton(errorCode);
176	0	if (U_SUCCESS(errorCode)) {
177	0	ep->addPropertyStarts(&sa, errorCode);
178	0	}
179	0	break;
180	0	}
181	0	case UPROPS_SRC_IDSU:
182		// New in Unicode 15.1 for just two characters.
183	0	sa.add(sa.set, 0x2FFE);
184	0	sa.add(sa.set, 0x2FFF + 1);
185	0	break;
186	0	case UPROPS_SRC_ID_COMPAT_MATH:
187	0	case UPROPS_SRC_MCM:
188	0	uprops_addPropertyStarts(src, &sa, &errorCode);
189	0	break;
190	0	case UPROPS_SRC_BLOCK:
191	0	ublock_addPropertyStarts(&sa, errorCode);
192	0	break;
193	0	default:
194	0	errorCode = U_INTERNAL_PROGRAM_ERROR;
195	0	break;
196	0	}
197
198	0	if (U_FAILURE(errorCode)) {
199	0	return;
200	0	}
201	0	if (incl->isBogus()) {
202	0	errorCode = U_MEMORY_ALLOCATION_ERROR;
203	0	return;
204	0	}
205		// Compact for caching.
206	0	incl->compact();
207	0	gInclusions[src].fSet = incl.orphan();
208	0	ucln_common_registerCleanup(UCLN_COMMON_CHARACTERPROPERTIES, characterproperties_cleanup);
209	0	}
210
211	0	const UnicodeSet *getInclusionsForSource(UPropertySource src, UErrorCode &errorCode) {
212	0	if (U_FAILURE(errorCode)) { return nullptr; }
213	0	if (src < 0 \|\| UPROPS_SRC_COUNT <= src) {
214	0	errorCode = U_ILLEGAL_ARGUMENT_ERROR;
215	0	return nullptr;
216	0	}
217	0	Inclusion &i = gInclusions[src];
218	0	umtx_initOnce(i.fInitOnce, &initInclusion, src, errorCode);
219	0	return i.fSet;
220	0	}
221
222	0	void U_CALLCONV initIntPropInclusion(UProperty prop, UErrorCode &errorCode) {
223		// This function is invoked only via umtx_initOnce().
224	0	U_ASSERT(UCHAR_INT_START <= prop && prop < UCHAR_INT_LIMIT);
225	0	int32_t inclIndex = UPROPS_SRC_COUNT + (prop - UCHAR_INT_START);
226	0	U_ASSERT(gInclusions[inclIndex].fSet == nullptr);
227	0	UPropertySource src = uprops_getSource(prop);
228	0	const UnicodeSet *incl = getInclusionsForSource(src, errorCode);
229	0	if (U_FAILURE(errorCode)) {
230	0	return;
231	0	}
232
233	0	LocalPointer<UnicodeSet> intPropIncl(new UnicodeSet(0, 0));
234	0	if (intPropIncl.isNull()) {
235	0	errorCode = U_MEMORY_ALLOCATION_ERROR;
236	0	return;
237	0	}
238	0	int32_t numRanges = incl->getRangeCount();
239	0	int32_t prevValue = 0;
240	0	for (int32_t i = 0; i < numRanges; ++i) {
241	0	UChar32 rangeEnd = incl->getRangeEnd(i);
242	0	for (UChar32 c = incl->getRangeStart(i); c <= rangeEnd; ++c) {
243		// TODO: Get a UCharacterProperty.IntProperty to avoid the property dispatch.
244	0	int32_t value = u_getIntPropertyValue(c, prop);
245	0	if (value != prevValue) {
246	0	intPropIncl->add(c);
247	0	prevValue = value;
248	0	}
249	0	}
250	0	}
251
252	0	if (intPropIncl->isBogus()) {
253	0	errorCode = U_MEMORY_ALLOCATION_ERROR;
254	0	return;
255	0	}
256		// Compact for caching.
257	0	intPropIncl->compact();
258	0	gInclusions[inclIndex].fSet = intPropIncl.orphan();
259	0	ucln_common_registerCleanup(UCLN_COMMON_CHARACTERPROPERTIES, characterproperties_cleanup);
260	0	}
261
262		} // namespace
263
264		U_NAMESPACE_BEGIN
265
266		const UnicodeSet *CharacterProperties::getInclusionsForProperty(
267	0	UProperty prop, UErrorCode &errorCode) {
268	0	if (U_FAILURE(errorCode)) { return nullptr; }
269	0	if (UCHAR_INT_START <= prop && prop < UCHAR_INT_LIMIT) {
270	0	int32_t inclIndex = UPROPS_SRC_COUNT + (prop - UCHAR_INT_START);
271	0	Inclusion &i = gInclusions[inclIndex];
272	0	umtx_initOnce(i.fInitOnce, &initIntPropInclusion, prop, errorCode);
273	0	return i.fSet;
274	0	} else {
275	0	UPropertySource src = uprops_getSource(prop);
276	0	return getInclusionsForSource(src, errorCode);
277	0	}
278	0	}
279
280		U_NAMESPACE_END
281
282		namespace {
283
284	0	UnicodeSet *makeSet(UProperty property, UErrorCode &errorCode) {
285	0	if (U_FAILURE(errorCode)) { return nullptr; }
286	0	LocalPointer<UnicodeSet> set(new UnicodeSet());
287	0	if (set.isNull()) {
288	0	errorCode = U_MEMORY_ALLOCATION_ERROR;
289	0	return nullptr;
290	0	}
291	0	if (UCHAR_BASIC_EMOJI <= property && property <= UCHAR_RGI_EMOJI) {
292		// property of strings
293	0	const icu::EmojiProps *ep = icu::EmojiProps::getSingleton(errorCode);
294	0	if (U_FAILURE(errorCode)) { return nullptr; }
295	0	USetAdder sa = {
296	0	reinterpret_cast<USet*>(set.getAlias()),
297	0	_set_add,
298	0	_set_addRange,
299	0	_set_addString,
300	0	nullptr, // don't need remove()
301	0	nullptr // don't need removeRange()
302	0	};
303	0	ep->addStrings(&sa, property, errorCode);
304	0	if (property != UCHAR_BASIC_EMOJI && property != UCHAR_RGI_EMOJI) {
305		// property of _only_ strings
306	0	set->freeze();
307	0	return set.orphan();
308	0	}
309	0	}
310
311	0	const UnicodeSet *inclusions =
312	0	icu::CharacterProperties::getInclusionsForProperty(property, errorCode);
313	0	if (U_FAILURE(errorCode)) { return nullptr; }
314	0	int32_t numRanges = inclusions->getRangeCount();
315	0	UChar32 startHasProperty = -1;
316
317	0	for (int32_t i = 0; i < numRanges; ++i) {
318	0	UChar32 rangeEnd = inclusions->getRangeEnd(i);
319	0	for (UChar32 c = inclusions->getRangeStart(i); c <= rangeEnd; ++c) {
320		// TODO: Get a UCharacterProperty.BinaryProperty to avoid the property dispatch.
321	0	if (u_hasBinaryProperty(c, property)) {
322	0	if (startHasProperty < 0) {
323		// Transition from false to true.
324	0	startHasProperty = c;
325	0	}
326	0	} else if (startHasProperty >= 0) {
327		// Transition from true to false.
328	0	set->add(startHasProperty, c - 1);
329	0	startHasProperty = -1;
330	0	}
331	0	}
332	0	}
333	0	if (startHasProperty >= 0) {
334	0	set->add(startHasProperty, 0x10FFFF);
335	0	}
336	0	set->freeze();
337	0	return set.orphan();
338	0	}
339
340	0	UCPMap *makeMap(UProperty property, UErrorCode &errorCode) {
341	0	if (U_FAILURE(errorCode)) { return nullptr; }
342	0	uint32_t nullValue = property == UCHAR_SCRIPT ? USCRIPT_UNKNOWN : 0;
343	0	icu::LocalUMutableCPTriePointer mutableTrie(
344	0	umutablecptrie_open(nullValue, nullValue, &errorCode));
345	0	const UnicodeSet *inclusions =
346	0	icu::CharacterProperties::getInclusionsForProperty(property, errorCode);
347	0	if (U_FAILURE(errorCode)) { return nullptr; }
348	0	int32_t numRanges = inclusions->getRangeCount();
349	0	UChar32 start = 0;
350	0	uint32_t value = nullValue;
351
352	0	for (int32_t i = 0; i < numRanges; ++i) {
353	0	UChar32 rangeEnd = inclusions->getRangeEnd(i);
354	0	for (UChar32 c = inclusions->getRangeStart(i); c <= rangeEnd; ++c) {
355		// TODO: Get a UCharacterProperty.IntProperty to avoid the property dispatch.
356	0	uint32_t nextValue = u_getIntPropertyValue(c, property);
357	0	if (value != nextValue) {
358	0	if (value != nullValue) {
359	0	umutablecptrie_setRange(mutableTrie.getAlias(), start, c - 1, value, &errorCode);
360	0	}
361	0	start = c;
362	0	value = nextValue;
363	0	}
364	0	}
365	0	}
366	0	if (value != 0) {
367	0	umutablecptrie_setRange(mutableTrie.getAlias(), start, 0x10FFFF, value, &errorCode);
368	0	}
369
370	0	UCPTrieType type;
371	0	if (property == UCHAR_BIDI_CLASS \|\| property == UCHAR_GENERAL_CATEGORY) {
372	0	type = UCPTRIE_TYPE_FAST;
373	0	} else {
374	0	type = UCPTRIE_TYPE_SMALL;
375	0	}
376	0	UCPTrieValueWidth valueWidth;
377		// TODO: UCharacterProperty.IntProperty
378	0	int32_t max = u_getIntPropertyMaxValue(property);
379	0	if (max <= 0xff) {
380	0	valueWidth = UCPTRIE_VALUE_BITS_8;
381	0	} else if (max <= 0xffff) {
382	0	valueWidth = UCPTRIE_VALUE_BITS_16;
383	0	} else {
384	0	valueWidth = UCPTRIE_VALUE_BITS_32;
385	0	}
386	0	return reinterpret_cast<UCPMap *>(
387	0	umutablecptrie_buildImmutable(mutableTrie.getAlias(), type, valueWidth, &errorCode));
388	0	}
389
390		} // namespace
391
392		U_NAMESPACE_BEGIN
393
394	0	const UnicodeSet *CharacterProperties::getBinaryPropertySet(UProperty property, UErrorCode &errorCode) {
395	0	if (U_FAILURE(errorCode)) { return nullptr; }
396	0	if (property < 0 \|\| UCHAR_BINARY_LIMIT <= property) {
397	0	errorCode = U_ILLEGAL_ARGUMENT_ERROR;
398	0	return nullptr;
399	0	}
400	0	Mutex m(&cpMutex);
401	0	UnicodeSet *set = sets[property];
402	0	if (set == nullptr) {
403	0	sets[property] = set = makeSet(property, errorCode);
404	0	}
405	0	return set;
406	0	}
407
408		U_NAMESPACE_END
409
410		U_NAMESPACE_USE
411
412		U_CAPI const USet * U_EXPORT2
413	0	u_getBinaryPropertySet(UProperty property, UErrorCode *pErrorCode) {
414	0	const UnicodeSet set = CharacterProperties::getBinaryPropertySet(property, pErrorCode);
415	0	return U_SUCCESS(*pErrorCode) ? set->toUSet() : nullptr;
416	0	}
417
418		U_CAPI const UCPMap * U_EXPORT2
419	0	u_getIntPropertyMap(UProperty property, UErrorCode *pErrorCode) {
420	0	if (U_FAILURE(*pErrorCode)) { return nullptr; }
421	0	if (property < UCHAR_INT_START \|\| UCHAR_INT_LIMIT <= property) {
422	0	*pErrorCode = U_ILLEGAL_ARGUMENT_ERROR;
423	0	return nullptr;
424	0	}
425	0	Mutex m(&cpMutex);
426	0	UCPMap *map = maps[property - UCHAR_INT_START];
427	0	if (map == nullptr) {
428	0	maps[property - UCHAR_INT_START] = map = makeMap(property, *pErrorCode);
429	0	}
430	0	return map;
431	0	}