From 36d22d82aa202bb199967e9512281e9a53db42c9 Mon Sep 17 00:00:00 2001 From: Daniel Baumann Date: Sun, 7 Apr 2024 21:33:14 +0200 Subject: Adding upstream version 115.7.0esr. Signed-off-by: Daniel Baumann --- intl/icu/source/common/brkeng.h | 277 ++++++++++++++++++++++++++++++++++++++++ 1 file changed, 277 insertions(+) create mode 100644 intl/icu/source/common/brkeng.h (limited to 'intl/icu/source/common/brkeng.h') diff --git a/intl/icu/source/common/brkeng.h b/intl/icu/source/common/brkeng.h new file mode 100644 index 0000000000..240dc8f4d3 --- /dev/null +++ b/intl/icu/source/common/brkeng.h @@ -0,0 +1,277 @@ +// © 2016 and later: Unicode, Inc. and others. +// License & terms of use: http://www.unicode.org/copyright.html +/** + ************************************************************************************ + * Copyright (C) 2006-2012, International Business Machines Corporation and others. * + * All Rights Reserved. * + ************************************************************************************ + */ + +#ifndef BRKENG_H +#define BRKENG_H + +#include "unicode/utypes.h" +#include "unicode/uobject.h" +#include "unicode/utext.h" +#include "unicode/uscript.h" + +U_NAMESPACE_BEGIN + +class UnicodeSet; +class UStack; +class UVector32; +class DictionaryMatcher; + +/******************************************************************* + * LanguageBreakEngine + */ + +/** + *

LanguageBreakEngines implement language-specific knowledge for + * finding text boundaries within a run of characters belonging to a + * specific set. The boundaries will be of a specific kind, e.g. word, + * line, etc.

+ * + *

LanguageBreakEngines should normally be implemented so as to + * be shared between threads without locking.

+ */ +class LanguageBreakEngine : public UMemory { + public: + + /** + *

Default constructor.

+ * + */ + LanguageBreakEngine(); + + /** + *

Virtual destructor.

+ */ + virtual ~LanguageBreakEngine(); + + /** + *

Indicate whether this engine handles a particular character for + * a particular kind of break.

+ * + * @param c A character which begins a run that the engine might handle + * @return true if this engine handles the particular character and break + * type. + */ + virtual UBool handles(UChar32 c) const = 0; + + /** + *

Find any breaks within a run in the supplied text.

+ * + * @param text A UText representing the text. The + * iterator is left at the end of the run of characters which the engine + * is capable of handling. + * @param startPos The start of the run within the supplied text. + * @param endPos The end of the run within the supplied text. + * @param foundBreaks A Vector of int32_t to receive the breaks. + * @param status Information on any errors encountered. + * @return The number of breaks found. + */ + virtual int32_t findBreaks( UText *text, + int32_t startPos, + int32_t endPos, + UVector32 &foundBreaks, + UBool isPhraseBreaking, + UErrorCode &status) const = 0; + +}; + +/******************************************************************* + * LanguageBreakFactory + */ + +/** + *

LanguageBreakFactorys find and return a LanguageBreakEngine + * that can determine breaks for characters in a specific set, if + * such an object can be found.

+ * + *

If a LanguageBreakFactory is to be shared between threads, + * appropriate synchronization must be used; there is none internal + * to the factory.

+ * + *

A LanguageBreakEngine returned by a LanguageBreakFactory can + * normally be shared between threads without synchronization, unless + * the specific subclass of LanguageBreakFactory indicates otherwise.

+ * + *

A LanguageBreakFactory is responsible for deleting any LanguageBreakEngine + * it returns when it itself is deleted, unless the specific subclass of + * LanguageBreakFactory indicates otherwise. Naturally, the factory should + * not be deleted until the LanguageBreakEngines it has returned are no + * longer needed.

+ */ +class LanguageBreakFactory : public UMemory { + public: + + /** + *

Default constructor.

+ * + */ + LanguageBreakFactory(); + + /** + *

Virtual destructor.

+ */ + virtual ~LanguageBreakFactory(); + + /** + *

Find and return a LanguageBreakEngine that can find the desired + * kind of break for the set of characters to which the supplied + * character belongs. It is up to the set of available engines to + * determine what the sets of characters are.

+ * + * @param c A character that begins a run for which a LanguageBreakEngine is + * sought. + * @return A LanguageBreakEngine with the desired characteristics, or 0. + */ + virtual const LanguageBreakEngine *getEngineFor(UChar32 c) = 0; + +}; + +/******************************************************************* + * UnhandledEngine + */ + +/** + *

UnhandledEngine is a special subclass of LanguageBreakEngine that + * handles characters that no other LanguageBreakEngine is available to + * handle. It is told the character and the type of break; at its + * discretion it may handle more than the specified character (e.g., + * the entire script to which that character belongs.

+ * + *

UnhandledEngines may not be shared between threads without + * external synchronization.

+ */ + +class UnhandledEngine : public LanguageBreakEngine { + private: + + /** + * The sets of characters handled. + * @internal + */ + + UnicodeSet *fHandled; + + public: + + /** + *

Default constructor.

+ * + */ + UnhandledEngine(UErrorCode &status); + + /** + *

Virtual destructor.

+ */ + virtual ~UnhandledEngine(); + + /** + *

Indicate whether this engine handles a particular character for + * a particular kind of break.

+ * + * @param c A character which begins a run that the engine might handle + * @return true if this engine handles the particular character and break + * type. + */ + virtual UBool handles(UChar32 c) const override; + + /** + *

Find any breaks within a run in the supplied text.

+ * + * @param text A UText representing the text (TODO: UText). The + * iterator is left at the end of the run of characters which the engine + * is capable of handling. + * @param startPos The start of the run within the supplied text. + * @param endPos The end of the run within the supplied text. + * @param foundBreaks An allocated C array of the breaks found, if any + * @param status Information on any errors encountered. + * @return The number of breaks found. + */ + virtual int32_t findBreaks( UText *text, + int32_t startPos, + int32_t endPos, + UVector32 &foundBreaks, + UBool isPhraseBreaking, + UErrorCode &status) const override; + + /** + *

Tell the engine to handle a particular character and break type.

+ * + * @param c A character which the engine should handle + */ + virtual void handleCharacter(UChar32 c); + +}; + +/******************************************************************* + * ICULanguageBreakFactory + */ + +/** + *

ICULanguageBreakFactory is the default LanguageBreakFactory for + * ICU. It creates dictionary-based LanguageBreakEngines from dictionary + * data in the ICU data file.

+ */ +class ICULanguageBreakFactory : public LanguageBreakFactory { + private: + + /** + * The stack of break engines created by this factory + * @internal + */ + + UStack *fEngines; + + public: + + /** + *

Standard constructor.

+ * + */ + ICULanguageBreakFactory(UErrorCode &status); + + /** + *

Virtual destructor.

+ */ + virtual ~ICULanguageBreakFactory(); + + /** + *

Find and return a LanguageBreakEngine that can find the desired + * kind of break for the set of characters to which the supplied + * character belongs. It is up to the set of available engines to + * determine what the sets of characters are.

+ * + * @param c A character that begins a run for which a LanguageBreakEngine is + * sought. + * @return A LanguageBreakEngine with the desired characteristics, or 0. + */ + virtual const LanguageBreakEngine *getEngineFor(UChar32 c) override; + +protected: + /** + *

Create a LanguageBreakEngine for the set of characters to which + * the supplied character belongs, for the specified break type.

+ * + * @param c A character that begins a run for which a LanguageBreakEngine is + * sought. + * @return A LanguageBreakEngine with the desired characteristics, or 0. + */ + virtual const LanguageBreakEngine *loadEngineFor(UChar32 c); + + /** + *

Create a DictionaryMatcher for the specified script and break type.

+ * @param script An ISO 15924 script code that identifies the dictionary to be + * created. + * @return A DictionaryMatcher with the desired characteristics, or nullptr. + */ + virtual DictionaryMatcher *loadDictionaryMatcherFor(UScriptCode script); +}; + +U_NAMESPACE_END + + /* BRKENG_H */ +#endif -- cgit v1.2.3