diff options
author | Daniel Baumann <daniel.baumann@progress-linux.org> | 2024-04-07 19:33:14 +0000 |
---|---|---|
committer | Daniel Baumann <daniel.baumann@progress-linux.org> | 2024-04-07 19:33:14 +0000 |
commit | 36d22d82aa202bb199967e9512281e9a53db42c9 (patch) | |
tree | 105e8c98ddea1c1e4784a60a5a6410fa416be2de /intl/icu/source/i18n/uitercollationiterator.h | |
parent | Initial commit. (diff) | |
download | firefox-esr-36d22d82aa202bb199967e9512281e9a53db42c9.tar.xz firefox-esr-36d22d82aa202bb199967e9512281e9a53db42c9.zip |
Adding upstream version 115.7.0esr.upstream/115.7.0esrupstream
Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>
Diffstat (limited to 'intl/icu/source/i18n/uitercollationiterator.h')
-rw-r--r-- | intl/icu/source/i18n/uitercollationiterator.h | 162 |
1 files changed, 162 insertions, 0 deletions
diff --git a/intl/icu/source/i18n/uitercollationiterator.h b/intl/icu/source/i18n/uitercollationiterator.h new file mode 100644 index 0000000000..23d8562a87 --- /dev/null +++ b/intl/icu/source/i18n/uitercollationiterator.h @@ -0,0 +1,162 @@ +// © 2016 and later: Unicode, Inc. and others. +// License & terms of use: http://www.unicode.org/copyright.html +/* +******************************************************************************* +* Copyright (C) 2012-2016, International Business Machines +* Corporation and others. All Rights Reserved. +******************************************************************************* +* uitercollationiterator.h +* +* created on: 2012sep23 (from utf16collationiterator.h) +* created by: Markus W. Scherer +*/ + +#ifndef __UITERCOLLATIONITERATOR_H__ +#define __UITERCOLLATIONITERATOR_H__ + +#include "unicode/utypes.h" + +#if !UCONFIG_NO_COLLATION + +#include "unicode/uiter.h" +#include "cmemory.h" +#include "collation.h" +#include "collationdata.h" +#include "collationiterator.h" +#include "normalizer2impl.h" + +U_NAMESPACE_BEGIN + +/** + * UCharIterator-based collation element and character iterator. + * Handles normalized text inline, with length or NUL-terminated. + * Unnormalized text is handled by a subclass. + */ +class U_I18N_API UIterCollationIterator : public CollationIterator { +public: + UIterCollationIterator(const CollationData *d, UBool numeric, UCharIterator &ui) + : CollationIterator(d, numeric), iter(ui) {} + + virtual ~UIterCollationIterator(); + + virtual void resetToOffset(int32_t newOffset) override; + + virtual int32_t getOffset() const override; + + virtual UChar32 nextCodePoint(UErrorCode &errorCode) override; + + virtual UChar32 previousCodePoint(UErrorCode &errorCode) override; + +protected: + virtual uint32_t handleNextCE32(UChar32 &c, UErrorCode &errorCode) override; + + virtual char16_t handleGetTrailSurrogate() override; + + virtual void forwardNumCodePoints(int32_t num, UErrorCode &errorCode) override; + + virtual void backwardNumCodePoints(int32_t num, UErrorCode &errorCode) override; + + UCharIterator &iter; +}; + +/** + * Incrementally checks the input text for FCD and normalizes where necessary. + */ +class U_I18N_API FCDUIterCollationIterator : public UIterCollationIterator { +public: + FCDUIterCollationIterator(const CollationData *data, UBool numeric, UCharIterator &ui, int32_t startIndex) + : UIterCollationIterator(data, numeric, ui), + state(ITER_CHECK_FWD), start(startIndex), + nfcImpl(data->nfcImpl) {} + + virtual ~FCDUIterCollationIterator(); + + virtual void resetToOffset(int32_t newOffset) override; + + virtual int32_t getOffset() const override; + + virtual UChar32 nextCodePoint(UErrorCode &errorCode) override; + + virtual UChar32 previousCodePoint(UErrorCode &errorCode) override; + +protected: + virtual uint32_t handleNextCE32(UChar32 &c, UErrorCode &errorCode) override; + + virtual char16_t handleGetTrailSurrogate() override; + + + virtual void forwardNumCodePoints(int32_t num, UErrorCode &errorCode) override; + + virtual void backwardNumCodePoints(int32_t num, UErrorCode &errorCode) override; + +private: + /** + * Switches to forward checking if possible. + */ + void switchToForward(); + + /** + * Extends the FCD text segment forward or normalizes around pos. + * @return true if success + */ + UBool nextSegment(UErrorCode &errorCode); + + /** + * Switches to backward checking. + */ + void switchToBackward(); + + /** + * Extends the FCD text segment backward or normalizes around pos. + * @return true if success + */ + UBool previousSegment(UErrorCode &errorCode); + + UBool normalize(const UnicodeString &s, UErrorCode &errorCode); + + enum State { + /** + * The input text [start..(iter index)[ passes the FCD check. + * Moving forward checks incrementally. + * pos & limit are undefined. + */ + ITER_CHECK_FWD, + /** + * The input text [(iter index)..limit[ passes the FCD check. + * Moving backward checks incrementally. + * start & pos are undefined. + */ + ITER_CHECK_BWD, + /** + * The input text [start..limit[ passes the FCD check. + * pos tracks the current text index. + */ + ITER_IN_FCD_SEGMENT, + /** + * The input text [start..limit[ failed the FCD check and was normalized. + * pos tracks the current index in the normalized string. + * The text iterator is at the limit index. + */ + IN_NORM_ITER_AT_LIMIT, + /** + * The input text [start..limit[ failed the FCD check and was normalized. + * pos tracks the current index in the normalized string. + * The text iterator is at the start index. + */ + IN_NORM_ITER_AT_START + }; + + State state; + + int32_t start; + int32_t pos; + int32_t limit; + + const Normalizer2Impl &nfcImpl; + UnicodeString normalized; +}; + +U_NAMESPACE_END + +#endif // !UCONFIG_NO_COLLATION +#endif // __UITERCOLLATIONITERATOR_H__ |