File Manager

X7ROOT File Manager

Current Path: /opt/alt/libicu65/usr/include/unicode

opt / alt / libicu65 / usr / include / unicode /

ðŸ“ ..
ðŸ“„ alphaindex.h(26.48 KB)
ðŸ“„ appendable.h(8.49 KB)
ðŸ“„ basictz.h(9.15 KB)
ðŸ“„ brkiter.h(27.8 KB)
ðŸ“„ bytestream.h(9.6 KB)
ðŸ“„ bytestrie.h(20.77 KB)
ðŸ“„ bytestriebuilder.h(7.08 KB)
ðŸ“„ calendar.h(105.74 KB)
ðŸ“„ caniter.h(7.43 KB)
ðŸ“„ casemap.h(25.33 KB)
ðŸ“„ char16ptr.h(7.22 KB)
ðŸ“„ chariter.h(24.05 KB)
ðŸ“„ choicfmt.h(23.91 KB)
ðŸ“„ coleitr.h(13.76 KB)
ðŸ“„ coll.h(56.23 KB)
ðŸ“„ compactdecimalformat.h(6.88 KB)
ðŸ“„ curramt.h(3.76 KB)
ðŸ“„ currpinf.h(7.3 KB)
ðŸ“„ currunit.h(4.05 KB)
ðŸ“„ datefmt.h(40.67 KB)
ðŸ“„ dbbi.h(1.19 KB)
ðŸ“„ dcfmtsym.h(20.13 KB)
ðŸ“„ decimfmt.h(87.38 KB)
ðŸ“„ docmain.h(6.97 KB)
ðŸ“„ dtfmtsym.h(37.7 KB)
ðŸ“„ dtintrv.h(3.84 KB)
ðŸ“„ dtitvfmt.h(46.63 KB)
ðŸ“„ dtitvinf.h(18.51 KB)
ðŸ“„ dtptngen.h(25.08 KB)
ðŸ“„ dtrule.h(8.68 KB)
ðŸ“„ edits.h(20.74 KB)
ðŸ“„ enumset.h(2.08 KB)
ðŸ“„ errorcode.h(4.84 KB)
ðŸ“„ fieldpos.h(8.69 KB)
ðŸ“„ filteredbrk.h(5.37 KB)
ðŸ“„ fmtable.h(24.42 KB)
ðŸ“„ format.h(12.5 KB)
ðŸ“„ formattedvalue.h(10.27 KB)
ðŸ“„ fpositer.h(3.04 KB)
ðŸ“„ gender.h(3.33 KB)
ðŸ“„ gregocal.h(31.71 KB)
ðŸ“„ icudataver.h(1.03 KB)
ðŸ“„ icuplug.h(11.88 KB)
ðŸ“„ idna.h(12.7 KB)
ðŸ“„ listformatter.h(9.47 KB)
ðŸ“„ localebuilder.h(11.27 KB)
ðŸ“„ localematcher.h(22.5 KB)
ðŸ“„ localpointer.h(19.69 KB)
ðŸ“„ locdspnm.h(7.12 KB)
ðŸ“„ locid.h(47.4 KB)
ðŸ“„ measfmt.h(11.33 KB)
ðŸ“„ measunit.h(93.31 KB)
ðŸ“„ measure.h(4.32 KB)
ðŸ“„ messagepattern.h(33.71 KB)
ðŸ“„ msgfmt.h(44.11 KB)
ðŸ“„ normalizer2.h(34.03 KB)
ðŸ“„ normlzr.h(30.94 KB)
ðŸ“„ nounit.h(2.69 KB)
ðŸ“„ numberformatter.h(86.31 KB)
ðŸ“„ numberrangeformatter.h(30.14 KB)
ðŸ“„ numfmt.h(49.81 KB)
ðŸ“„ numsys.h(7.19 KB)
ðŸ“„ parseerr.h(3.08 KB)
ðŸ“„ parsepos.h(5.56 KB)
ðŸ“„ platform.h(28.08 KB)
ðŸ“„ plurfmt.h(25.2 KB)
ðŸ“„ plurrule.h(18.39 KB)
ðŸ“„ ptypes.h(3.49 KB)
ðŸ“„ putil.h(6.33 KB)
ðŸ“„ rbbi.h(26.58 KB)
ðŸ“„ rbnf.h(48.73 KB)
ðŸ“„ rbtz.h(15.6 KB)
ðŸ“„ regex.h(84.36 KB)
ðŸ“„ region.h(9.18 KB)
ðŸ“„ reldatefmt.h(22.62 KB)
ðŸ“„ rep.h(9.37 KB)
ðŸ“„ resbund.h(18.07 KB)
ðŸ“„ schriter.h(6.32 KB)
ðŸ“„ scientificnumberformatter.h(6.4 KB)
ðŸ“„ search.h(22.22 KB)
ðŸ“„ selfmt.h(14.3 KB)
ðŸ“„ simpleformatter.h(12.59 KB)
ðŸ“„ simpletz.h(45.44 KB)
ðŸ“„ smpdtfmt.h(70.97 KB)
ðŸ“„ sortkey.h(11.18 KB)
ðŸ“„ std_string.h(1.05 KB)
ðŸ“„ strenum.h(9.92 KB)
ðŸ“„ stringoptions.h(5.79 KB)
ðŸ“„ stringpiece.h(7.38 KB)
ðŸ“„ stringtriebuilder.h(15.33 KB)
ðŸ“„ stsearch.h(21.3 KB)
ðŸ“„ symtable.h(4.27 KB)
ðŸ“„ tblcoll.h(36.61 KB)
ðŸ“„ timezone.h(41.02 KB)
ðŸ“„ tmunit.h(3.38 KB)
ðŸ“„ tmutamt.h(4.9 KB)
ðŸ“„ tmutfmt.h(7.85 KB)
ðŸ“„ translit.h(65.82 KB)
ðŸ“„ tzfmt.h(42.89 KB)
ðŸ“„ tznames.h(16.85 KB)
ðŸ“„ tzrule.h(35.37 KB)
ðŸ“„ tztrans.h(6.12 KB)
ðŸ“„ ubidi.h(89.56 KB)
ðŸ“„ ubiditransform.h(12.65 KB)
ðŸ“„ ubrk.h(23.97 KB)
ðŸ“„ ucal.h(56.9 KB)
ðŸ“„ ucasemap.h(15.18 KB)
ðŸ“„ ucat.h(5.36 KB)
ðŸ“„ uchar.h(140.56 KB)
ðŸ“„ ucharstrie.h(22.58 KB)
ðŸ“„ ucharstriebuilder.h(7.21 KB)
ðŸ“„ uchriter.h(13.2 KB)
ðŸ“„ uclean.h(11.21 KB)
ðŸ“„ ucnv.h(83.09 KB)
ðŸ“„ ucnv_cb.h(6.59 KB)
ðŸ“„ ucnv_err.h(20.99 KB)
ðŸ“„ ucnvsel.h(6.14 KB)
ðŸ“„ ucol.h(61.46 KB)
ðŸ“„ ucoleitr.h(9.46 KB)
ðŸ“„ uconfig.h(12.07 KB)
ðŸ“„ ucpmap.h(5.53 KB)
ðŸ“„ ucptrie.h(22.46 KB)
ðŸ“„ ucsdet.h(14.67 KB)
ðŸ“„ ucurr.h(16.12 KB)
ðŸ“„ udat.h(60.88 KB)
ðŸ“„ udata.h(15.56 KB)
ðŸ“„ udateintervalformat.h(10.03 KB)
ðŸ“„ udatpg.h(26.01 KB)
ðŸ“„ udisplaycontext.h(5.89 KB)
ðŸ“„ uenum.h(7.78 KB)
ðŸ“„ ufieldpositer.h(4.36 KB)
ðŸ“„ uformattable.h(10.94 KB)
ðŸ“„ uformattedvalue.h(12.14 KB)
ðŸ“„ ugender.h(2 KB)
ðŸ“„ uidna.h(33.37 KB)
ðŸ“„ uiter.h(22.77 KB)
ðŸ“„ uldnames.h(10.45 KB)
ðŸ“„ ulistformatter.h(8.83 KB)
ðŸ“„ uloc.h(52.54 KB)
ðŸ“„ ulocdata.h(11.26 KB)
ðŸ“„ umachine.h(14.53 KB)
ðŸ“„ umisc.h(1.33 KB)
ðŸ“„ umsg.h(24.23 KB)
ðŸ“„ umutablecptrie.h(8.24 KB)
ðŸ“„ unifilt.h(3.96 KB)
ðŸ“„ unifunct.h(4.04 KB)
ðŸ“„ unimatch.h(6.1 KB)
ðŸ“„ unirepl.h(3.38 KB)
ðŸ“„ uniset.h(64.9 KB)
ðŸ“„ unistr.h(170.43 KB)
ðŸ“„ unorm.h(20.52 KB)
ðŸ“„ unorm2.h(24.66 KB)
ðŸ“„ unum.h(53.62 KB)
ðŸ“„ unumberformatter.h(25.36 KB)
ðŸ“„ unumsys.h(7.21 KB)
ðŸ“„ uobject.h(10.68 KB)
ðŸ“„ upluralrules.h(7.88 KB)
ðŸ“„ uregex.h(72.05 KB)
ðŸ“„ uregion.h(9.84 KB)
ðŸ“„ ureldatefmt.h(17.26 KB)
ðŸ“„ urename.h(130.97 KB)
ðŸ“„ urep.h(5.38 KB)
ðŸ“„ ures.h(36.54 KB)
ðŸ“„ uscript.h(26.87 KB)
ðŸ“„ usearch.h(38.12 KB)
ðŸ“„ uset.h(40 KB)
ðŸ“„ usetiter.h(9.55 KB)
ðŸ“„ ushape.h(18 KB)
ðŸ“„ uspoof.h(65.9 KB)
ðŸ“„ usprep.h(8.14 KB)
ðŸ“„ ustdio.h(38.54 KB)
ðŸ“„ ustream.h(1.89 KB)
ðŸ“„ ustring.h(72.47 KB)
ðŸ“„ ustringtrie.h(3.15 KB)
ðŸ“„ utext.h(58.13 KB)
ðŸ“„ utf.h(7.86 KB)
ðŸ“„ utf16.h(23.32 KB)
ðŸ“„ utf32.h(763 B)
ðŸ“„ utf8.h(30.96 KB)
ðŸ“„ utf_old.h(45.83 KB)
ðŸ“„ utmscale.h(13.78 KB)
ðŸ“„ utrace.h(15.73 KB)
ðŸ“„ utrans.h(25.52 KB)
ðŸ“„ utypes.h(30.74 KB)
ðŸ“„ uvernum.h(6.67 KB)
ðŸ“„ uversion.h(6 KB)
ðŸ“„ vtzone.h(20.3 KB)

Editing: ucharstrie.h

// © 2016 and later: Unicode, Inc. and others.
// License & terms of use: http://www.unicode.org/copyright.html
/*
*******************************************************************************
*   Copyright (C) 2010-2012, International Business Machines
*   Corporation and others.  All Rights Reserved.
*******************************************************************************
*   file name:  ucharstrie.h
*   encoding:   UTF-8
*   tab size:   8 (not used)
*   indentation:4
*
*   created on: 2010nov14
*   created by: Markus W. Scherer
*/

#ifndef __UCHARSTRIE_H__
#define __UCHARSTRIE_H__

/**
 * \file
 * \brief C++ API: Trie for mapping Unicode strings (or 16-bit-unit sequences)
 *                 to integer values.
 */

#include "unicode/utypes.h"

#if U_SHOW_CPLUSPLUS_API

#include "unicode/unistr.h"
#include "unicode/uobject.h"
#include "unicode/ustringtrie.h"

U_NAMESPACE_BEGIN

class Appendable;
class UCharsTrieBuilder;
class UVector32;

/**
 * Light-weight, non-const reader class for a UCharsTrie.
 * Traverses a char16_t-serialized data structure with minimal state,
 * for mapping strings (16-bit-unit sequences) to non-negative integer values.
 *
 * This class owns the serialized trie data only if it was constructed by
 * the builder's build() method.
 * The public constructor and the copy constructor only alias the data (only copy the pointer).
 * There is no assignment operator.
 *
 * This class is not intended for public subclassing.
 * @stable ICU 4.8
 */
class U_COMMON_API UCharsTrie : public UMemory {
public:
    /**
     * Constructs a UCharsTrie reader instance.
     *
     * The trieUChars must contain a copy of a char16_t sequence from the UCharsTrieBuilder,
     * starting with the first char16_t of that sequence.
     * The UCharsTrie object will not read more char16_ts than
     * the UCharsTrieBuilder generated in the corresponding build() call.
     *
     * The array is not copied/cloned and must not be modified while
     * the UCharsTrie object is in use.
     *
     * @param trieUChars The char16_t array that contains the serialized trie.
     * @stable ICU 4.8
     */
    UCharsTrie(ConstChar16Ptr trieUChars)
            : ownedArray_(NULL), uchars_(trieUChars),
              pos_(uchars_), remainingMatchLength_(-1) {}

/**
     * Destructor.
     * @stable ICU 4.8
     */
    ~UCharsTrie();

/**
     * Copy constructor, copies the other trie reader object and its state,
     * but not the char16_t array which will be shared. (Shallow copy.)
     * @param other Another UCharsTrie object.
     * @stable ICU 4.8
     */
    UCharsTrie(const UCharsTrie &other)
            : ownedArray_(NULL), uchars_(other.uchars_),
              pos_(other.pos_), remainingMatchLength_(other.remainingMatchLength_) {}

/**
     * Resets this trie to its initial state.
     * @return *this
     * @stable ICU 4.8
     */
    UCharsTrie &reset() {
        pos_=uchars_;
        remainingMatchLength_=-1;
        return *this;
    }

#ifndef U_HIDE_DRAFT_API
    /**
     * Returns the state of this trie as a 64-bit integer.
     * The state value is never 0.
     *
     * @return opaque state value
     * @see resetToState64
     * @draft ICU 65
     */
    uint64_t getState64() const {
        return (static_cast<uint64_t>(remainingMatchLength_ + 2) << kState64RemainingShift) |
            (uint64_t)(pos_ - uchars_);
    }

/**
     * Resets this trie to the saved state.
     * Unlike resetToState(State), the 64-bit state value
     * must be from getState64() from the same trie object or
     * from one initialized the exact same way.
     * Because of no validation, this method is faster.
     *
     * @param state The opaque trie state value from getState64().
     * @return *this
     * @see getState64
     * @see resetToState
     * @see reset
     * @draft ICU 65
     */
    UCharsTrie &resetToState64(uint64_t state) {
        remainingMatchLength_ = static_cast<int32_t>(state >> kState64RemainingShift) - 2;
        pos_ = uchars_ + (state & kState64PosMask);
        return *this;
    }
#endif  /* U_HIDE_DRAFT_API */

/**
     * UCharsTrie state object, for saving a trie's current state
     * and resetting the trie back to this state later.
     * @stable ICU 4.8
     */
    class State : public UMemory {
    public:
        /**
         * Constructs an empty State.
         * @stable ICU 4.8
         */
        State() { uchars=NULL; }
    private:
        friend class UCharsTrie;

const char16_t *uchars;
        const char16_t *pos;
        int32_t remainingMatchLength;
    };

/**
     * Saves the state of this trie.
     * @param state The State object to hold the trie's state.
     * @return *this
     * @see resetToState
     * @stable ICU 4.8
     */
    const UCharsTrie &saveState(State &state) const {
        state.uchars=uchars_;
        state.pos=pos_;
        state.remainingMatchLength=remainingMatchLength_;
        return *this;
    }

/**
     * Resets this trie to the saved state.
     * If the state object contains no state, or the state of a different trie,
     * then this trie remains unchanged.
     * @param state The State object which holds a saved trie state.
     * @return *this
     * @see saveState
     * @see reset
     * @stable ICU 4.8
     */
    UCharsTrie &resetToState(const State &state) {
        if(uchars_==state.uchars && uchars_!=NULL) {
            pos_=state.pos;
            remainingMatchLength_=state.remainingMatchLength;
        }
        return *this;
    }

/**
     * Determines whether the string so far matches, whether it has a value,
     * and whether another input char16_t can continue a matching string.
     * @return The match/value Result.
     * @stable ICU 4.8
     */
    UStringTrieResult current() const;

/**
     * Traverses the trie from the initial state for this input char16_t.
     * Equivalent to reset().next(uchar).
     * @param uchar Input char value. Values below 0 and above 0xffff will never match.
     * @return The match/value Result.
     * @stable ICU 4.8
     */
    inline UStringTrieResult first(int32_t uchar) {
        remainingMatchLength_=-1;
        return nextImpl(uchars_, uchar);
    }

/**
     * Traverses the trie from the initial state for the
     * one or two UTF-16 code units for this input code point.
     * Equivalent to reset().nextForCodePoint(cp).
     * @param cp A Unicode code point 0..0x10ffff.
     * @return The match/value Result.
     * @stable ICU 4.8
     */
    UStringTrieResult firstForCodePoint(UChar32 cp);

/**
     * Traverses the trie from the current state for this input char16_t.
     * @param uchar Input char value. Values below 0 and above 0xffff will never match.
     * @return The match/value Result.
     * @stable ICU 4.8
     */
    UStringTrieResult next(int32_t uchar);

/**
     * Traverses the trie from the current state for the
     * one or two UTF-16 code units for this input code point.
     * @param cp A Unicode code point 0..0x10ffff.
     * @return The match/value Result.
     * @stable ICU 4.8
     */
    UStringTrieResult nextForCodePoint(UChar32 cp);

/**
     * Traverses the trie from the current state for this string.
     * Equivalent to
     * \code
     * Result result=current();
     * for(each c in s)
     *   if(!USTRINGTRIE_HAS_NEXT(result)) return USTRINGTRIE_NO_MATCH;
     *   result=next(c);
     * return result;
     * \endcode
     * @param s A string. Can be NULL if length is 0.
     * @param length The length of the string. Can be -1 if NUL-terminated.
     * @return The match/value Result.
     * @stable ICU 4.8
     */
    UStringTrieResult next(ConstChar16Ptr s, int32_t length);

/**
     * Returns a matching string's value if called immediately after
     * current()/first()/next() returned USTRINGTRIE_INTERMEDIATE_VALUE or USTRINGTRIE_FINAL_VALUE.
     * getValue() can be called multiple times.
     *
     * Do not call getValue() after USTRINGTRIE_NO_MATCH or USTRINGTRIE_NO_VALUE!
     * @return The value for the string so far.
     * @stable ICU 4.8
     */
    inline int32_t getValue() const {
        const char16_t *pos=pos_;
        int32_t leadUnit=*pos++;
        // U_ASSERT(leadUnit>=kMinValueLead);
        return leadUnit&kValueIsFinal ?
            readValue(pos, leadUnit&0x7fff) : readNodeValue(pos, leadUnit);
    }

/**
     * Determines whether all strings reachable from the current state
     * map to the same value.
     * @param uniqueValue Receives the unique value, if this function returns TRUE.
     *                    (output-only)
     * @return TRUE if all strings reachable from the current state
     *         map to the same value.
     * @stable ICU 4.8
     */
    inline UBool hasUniqueValue(int32_t &uniqueValue) const {
        const char16_t *pos=pos_;
        // Skip the rest of a pending linear-match node.
        return pos!=NULL && findUniqueValue(pos+remainingMatchLength_+1, FALSE, uniqueValue);
    }

/**
     * Finds each char16_t which continues the string from the current state.
     * That is, each char16_t c for which it would be next(c)!=USTRINGTRIE_NO_MATCH now.
     * @param out Each next char16_t is appended to this object.
     * @return the number of char16_ts which continue the string from here
     * @stable ICU 4.8
     */
    int32_t getNextUChars(Appendable &out) const;

/**
     * Iterator for all of the (string, value) pairs in a UCharsTrie.
     * @stable ICU 4.8
     */
    class U_COMMON_API Iterator : public UMemory {
    public:
        /**
         * Iterates from the root of a char16_t-serialized UCharsTrie.
         * @param trieUChars The trie char16_ts.
         * @param maxStringLength If 0, the iterator returns full strings.
         *                        Otherwise, the iterator returns strings with this maximum length.
         * @param errorCode Standard ICU error code. Its input value must
         *                  pass the U_SUCCESS() test, or else the function returns
         *                  immediately. Check for U_FAILURE() on output or use with
         *                  function chaining. (See User Guide for details.)
         * @stable ICU 4.8
         */
        Iterator(ConstChar16Ptr trieUChars, int32_t maxStringLength, UErrorCode &errorCode);

/**
         * Iterates from the current state of the specified UCharsTrie.
         * @param trie The trie whose state will be copied for iteration.
         * @param maxStringLength If 0, the iterator returns full strings.
         *                        Otherwise, the iterator returns strings with this maximum length.
         * @param errorCode Standard ICU error code. Its input value must
         *                  pass the U_SUCCESS() test, or else the function returns
         *                  immediately. Check for U_FAILURE() on output or use with
         *                  function chaining. (See User Guide for details.)
         * @stable ICU 4.8
         */
        Iterator(const UCharsTrie &trie, int32_t maxStringLength, UErrorCode &errorCode);

/**
         * Destructor.
         * @stable ICU 4.8
         */
        ~Iterator();

/**
         * Resets this iterator to its initial state.
         * @return *this
         * @stable ICU 4.8
         */
        Iterator &reset();

/**
         * @return TRUE if there are more elements.
         * @stable ICU 4.8
         */
        UBool hasNext() const;

/**
         * Finds the next (string, value) pair if there is one.
         *
         * If the string is truncated to the maximum length and does not
         * have a real value, then the value is set to -1.
         * In this case, this "not a real value" is indistinguishable from
         * a real value of -1.
         * @param errorCode Standard ICU error code. Its input value must
         *                  pass the U_SUCCESS() test, or else the function returns
         *                  immediately. Check for U_FAILURE() on output or use with
         *                  function chaining. (See User Guide for details.)
         * @return TRUE if there is another element.
         * @stable ICU 4.8
         */
        UBool next(UErrorCode &errorCode);

/**
         * @return The string for the last successful next().
         * @stable ICU 4.8
         */
        const UnicodeString &getString() const { return str_; }
        /**
         * @return The value for the last successful next().
         * @stable ICU 4.8
         */
        int32_t getValue() const { return value_; }

private:
        UBool truncateAndStop() {
            pos_=NULL;
            value_=-1;  // no real value for str
            return TRUE;
        }

const char16_t *branchNext(const char16_t *pos, int32_t length, UErrorCode &errorCode);

const char16_t *uchars_;
        const char16_t *pos_;
        const char16_t *initialPos_;
        int32_t remainingMatchLength_;
        int32_t initialRemainingMatchLength_;
        UBool skipValue_;  // Skip intermediate value which was already delivered.

UnicodeString str_;
        int32_t maxLength_;
        int32_t value_;

// The stack stores pairs of integers for backtracking to another
        // outbound edge of a branch node.
        // The first integer is an offset from uchars_.
        // The second integer has the str_.length() from before the node in bits 15..0,
        // and the remaining branch length in bits 31..16.
        // (We could store the remaining branch length minus 1 in bits 30..16 and not use the sign bit,
        // but the code looks more confusing that way.)
        UVector32 *stack_;
    };

private:
    friend class UCharsTrieBuilder;

/**
     * Constructs a UCharsTrie reader instance.
     * Unlike the public constructor which just aliases an array,
     * this constructor adopts the builder's array.
     * This constructor is only called by the builder.
     */
    UCharsTrie(char16_t *adoptUChars, const char16_t *trieUChars)
            : ownedArray_(adoptUChars), uchars_(trieUChars),
              pos_(uchars_), remainingMatchLength_(-1) {}

// No assignment operator.
    UCharsTrie &operator=(const UCharsTrie &other);

inline void stop() {
        pos_=NULL;
    }

// Reads a compact 32-bit integer.
    // pos is already after the leadUnit, and the lead unit has bit 15 reset.
    static inline int32_t readValue(const char16_t *pos, int32_t leadUnit) {
        int32_t value;
        if(leadUnit<kMinTwoUnitValueLead) {
            value=leadUnit;
        } else if(leadUnit<kThreeUnitValueLead) {
            value=((leadUnit-kMinTwoUnitValueLead)<<16)|*pos;
        } else {
            value=(pos[0]<<16)|pos[1];
        }
        return value;
    }
    static inline const char16_t *skipValue(const char16_t *pos, int32_t leadUnit) {
        if(leadUnit>=kMinTwoUnitValueLead) {
            if(leadUnit<kThreeUnitValueLead) {
                ++pos;
            } else {
                pos+=2;
            }
        }
        return pos;
    }
    static inline const char16_t *skipValue(const char16_t *pos) {
        int32_t leadUnit=*pos++;
        return skipValue(pos, leadUnit&0x7fff);
    }

static inline int32_t readNodeValue(const char16_t *pos, int32_t leadUnit) {
        // U_ASSERT(kMinValueLead<=leadUnit && leadUnit<kValueIsFinal);
        int32_t value;
        if(leadUnit<kMinTwoUnitNodeValueLead) {
            value=(leadUnit>>6)-1;
        } else if(leadUnit<kThreeUnitNodeValueLead) {
            value=(((leadUnit&0x7fc0)-kMinTwoUnitNodeValueLead)<<10)|*pos;
        } else {
            value=(pos[0]<<16)|pos[1];
        }
        return value;
    }
    static inline const char16_t *skipNodeValue(const char16_t *pos, int32_t leadUnit) {
        // U_ASSERT(kMinValueLead<=leadUnit && leadUnit<kValueIsFinal);
        if(leadUnit>=kMinTwoUnitNodeValueLead) {
            if(leadUnit<kThreeUnitNodeValueLead) {
                ++pos;
            } else {
                pos+=2;
            }
        }
        return pos;
    }

static inline const char16_t *jumpByDelta(const char16_t *pos) {
        int32_t delta=*pos++;
        if(delta>=kMinTwoUnitDeltaLead) {
            if(delta==kThreeUnitDeltaLead) {
                delta=(pos[0]<<16)|pos[1];
                pos+=2;
            } else {
                delta=((delta-kMinTwoUnitDeltaLead)<<16)|*pos++;
            }
        }
        return pos+delta;
    }

static const char16_t *skipDelta(const char16_t *pos) {
        int32_t delta=*pos++;
        if(delta>=kMinTwoUnitDeltaLead) {
            if(delta==kThreeUnitDeltaLead) {
                pos+=2;
            } else {
                ++pos;
            }
        }
        return pos;
    }

static inline UStringTrieResult valueResult(int32_t node) {
        return (UStringTrieResult)(USTRINGTRIE_INTERMEDIATE_VALUE-(node>>15));
    }

// Handles a branch node for both next(uchar) and next(string).
    UStringTrieResult branchNext(const char16_t *pos, int32_t length, int32_t uchar);

// Requires remainingLength_<0.
    UStringTrieResult nextImpl(const char16_t *pos, int32_t uchar);

// Helper functions for hasUniqueValue().
    // Recursively finds a unique value (or whether there is not a unique one)
    // from a branch.
    static const char16_t *findUniqueValueFromBranch(const char16_t *pos, int32_t length,
                                                  UBool haveUniqueValue, int32_t &uniqueValue);
    // Recursively finds a unique value (or whether there is not a unique one)
    // starting from a position on a node lead unit.
    static UBool findUniqueValue(const char16_t *pos, UBool haveUniqueValue, int32_t &uniqueValue);

// Helper functions for getNextUChars().
    // getNextUChars() when pos is on a branch node.
    static void getNextBranchUChars(const char16_t *pos, int32_t length, Appendable &out);

// UCharsTrie data structure
    //
    // The trie consists of a series of char16_t-serialized nodes for incremental
    // Unicode string/char16_t sequence matching. (char16_t=16-bit unsigned integer)
    // The root node is at the beginning of the trie data.
    //
    // Types of nodes are distinguished by their node lead unit ranges.
    // After each node, except a final-value node, another node follows to
    // encode match values or continue matching further units.
    //
    // Node types:
    //  - Final-value node: Stores a 32-bit integer in a compact, variable-length format.
    //    The value is for the string/char16_t sequence so far.
    //  - Match node, optionally with an intermediate value in a different compact format.
    //    The value, if present, is for the string/char16_t sequence so far.
    //
    //  Aside from the value, which uses the node lead unit's high bits:
    //
    //  - Linear-match node: Matches a number of units.
    //  - Branch node: Branches to other nodes according to the current input unit.
    //    The node unit is the length of the branch (number of units to select from)
    //    minus 1. It is followed by a sub-node:
    //    - If the length is at most kMaxBranchLinearSubNodeLength, then
    //      there are length-1 (key, value) pairs and then one more comparison unit.
    //      If one of the key units matches, then the value is either a final value for
    //      the string so far, or a "jump" delta to the next node.
    //      If the last unit matches, then matching continues with the next node.
    //      (Values have the same encoding as final-value nodes.)
    //    - If the length is greater than kMaxBranchLinearSubNodeLength, then
    //      there is one unit and one "jump" delta.
    //      If the input unit is less than the sub-node unit, then "jump" by delta to
    //      the next sub-node which will have a length of length/2.
    //      (The delta has its own compact encoding.)
    //      Otherwise, skip the "jump" delta to the next sub-node
    //      which will have a length of length-length/2.

// Match-node lead unit values, after masking off intermediate-value bits:

// 0000..002f: Branch node. If node!=0 then the length is node+1, otherwise
    // the length is one more than the next unit.

// For a branch sub-node with at most this many entries, we drop down
    // to a linear search.
    static const int32_t kMaxBranchLinearSubNodeLength=5;

// 0030..003f: Linear-match node, match 1..16 units and continue reading the next node.
    static const int32_t kMinLinearMatch=0x30;
    static const int32_t kMaxLinearMatchLength=0x10;

// Match-node lead unit bits 14..6 for the optional intermediate value.
    // If these bits are 0, then there is no intermediate value.
    // Otherwise, see the *NodeValue* constants below.
    static const int32_t kMinValueLead=kMinLinearMatch+kMaxLinearMatchLength;  // 0x0040
    static const int32_t kNodeTypeMask=kMinValueLead-1;  // 0x003f

// A final-value node has bit 15 set.
    static const int32_t kValueIsFinal=0x8000;

// Compact value: After testing and masking off bit 15, use the following thresholds.
    static const int32_t kMaxOneUnitValue=0x3fff;

static const int32_t kMinTwoUnitValueLead=kMaxOneUnitValue+1;  // 0x4000
    static const int32_t kThreeUnitValueLead=0x7fff;

static const int32_t kMaxTwoUnitValue=((kThreeUnitValueLead-kMinTwoUnitValueLead)<<16)-1;  // 0x3ffeffff

// Compact intermediate-value integer, lead unit shared with a branch or linear-match node.
    static const int32_t kMaxOneUnitNodeValue=0xff;
    static const int32_t kMinTwoUnitNodeValueLead=kMinValueLead+((kMaxOneUnitNodeValue+1)<<6);  // 0x4040
    static const int32_t kThreeUnitNodeValueLead=0x7fc0;

static const int32_t kMaxTwoUnitNodeValue=
        ((kThreeUnitNodeValueLead-kMinTwoUnitNodeValueLead)<<10)-1;  // 0xfdffff

// Compact delta integers.
    static const int32_t kMaxOneUnitDelta=0xfbff;
    static const int32_t kMinTwoUnitDeltaLead=kMaxOneUnitDelta+1;  // 0xfc00
    static const int32_t kThreeUnitDeltaLead=0xffff;

static const int32_t kMaxTwoUnitDelta=((kThreeUnitDeltaLead-kMinTwoUnitDeltaLead)<<16)-1;  // 0x03feffff

// For getState64():
    // The remainingMatchLength_ is -1..14=(kMaxLinearMatchLength=0x10)-2
    // so we need at least 5 bits for that.
    // We add 2 to store it as a positive value 1..16=kMaxLinearMatchLength.
    static constexpr int32_t kState64RemainingShift = 59;
    static constexpr uint64_t kState64PosMask = (UINT64_C(1) << kState64RemainingShift) - 1;

char16_t *ownedArray_;

// Fixed value referencing the UCharsTrie words.
    const char16_t *uchars_;

// Iterator variables.

// Pointer to next trie unit to read. NULL if no more matches.
    const char16_t *pos_;
    // Remaining length of a linear-match node, minus 1. Negative if not in such a node.
    int32_t remainingMatchLength_;
};

U_NAMESPACE_END

#endif /* U_SHOW_CPLUSPLUS_API */

#endif  // __UCHARSTRIE_H__

X7ROOT File Manager

Editing: ucharstrie.h

Upload File

Create Folder