C++ UnicodeSet::getRangeStart方法代码示例

//------------------------------------------------------------------------
//
//   build          Build the list of non-overlapping character ranges
//                  from the Unicode Sets.
//
//------------------------------------------------------------------------
void RBBISetBuilder::build() {
    RBBINode        *usetNode;
    RangeDescriptor *rlRange;

    if (fRB->fDebugEnv && uprv_strstr(fRB->fDebugEnv, "usets")) {
        printSets();
    }

    //
    //  Initialize the process by creating a single range encompassing all characters
    //  that is in no sets.
    //
    fRangeList                = new RangeDescriptor(*fStatus); // will check for status here
    fRangeList->fStartChar    = 0;
    fRangeList->fEndChar      = 0x10ffff;

    if (U_FAILURE(*fStatus)) {
        return;
    }

    //
    //  Find the set of non-overlapping ranges of characters
    //
    int  ni;
    for (ni=0; ; ni++) {
        usetNode = (RBBINode *)this->fRB->fUSetNodes->elementAt(ni);
        if (usetNode==NULL) {
            break;
        }

        UnicodeSet      *inputSet             = usetNode->fInputSet;
        int32_t          inputSetRangeCount   = inputSet->getRangeCount();
        int              inputSetRangeIndex   = 0;
        rlRange              = fRangeList;

        for (;;) {
            if (inputSetRangeIndex >= inputSetRangeCount) {
                break;
            }
            UChar32      inputSetRangeBegin  = inputSet->getRangeStart(inputSetRangeIndex);
            UChar32      inputSetRangeEnd    = inputSet->getRangeEnd(inputSetRangeIndex);

            // skip over ranges from the range list that are completely
            //   below the current range from the input unicode set.
            while (rlRange->fEndChar < inputSetRangeBegin) {
                rlRange = rlRange->fNext;
            }

            // If the start of the range from the range list is before with
            //   the start of the range from the unicode set, split the range list range
            //   in two, with one part being before (wholly outside of) the unicode set
            //   and the other containing the rest.
            //   Then continue the loop; the post-split current range will then be skipped
            //     over
            if (rlRange->fStartChar < inputSetRangeBegin) {
                rlRange->split(inputSetRangeBegin, *fStatus);
                if (U_FAILURE(*fStatus)) {
                    return;
                }
                continue;
            }

            // Same thing at the end of the ranges...
            // If the end of the range from the range list doesn't coincide with
            //   the end of the range from the unicode set, split the range list
            //   range in two.  The first part of the split range will be
            //   wholly inside the Unicode set.
            if (rlRange->fEndChar > inputSetRangeEnd) {
                rlRange->split(inputSetRangeEnd+1, *fStatus);
                if (U_FAILURE(*fStatus)) {
                    return;
                }
            }

            // The current rlRange is now entirely within the UnicodeSet range.
            // Add this unicode set to the list of sets for this rlRange
            if (rlRange->fIncludesSets->indexOf(usetNode) == -1) {
                rlRange->fIncludesSets->addElement(usetNode, *fStatus);
                if (U_FAILURE(*fStatus)) {
                    return;
                }
            }

            // Advance over ranges that we are finished with.
            if (inputSetRangeEnd == rlRange->fEndChar) {
                inputSetRangeIndex++;
            }
            rlRange = rlRange->fNext;
        }
    }

    if (fRB->fDebugEnv && uprv_strstr(fRB->fDebugEnv, "range")) {
        printRanges();
    }
//.........这里部分代码省略.........

//------------------------------------------------------------------------
//
//   build          Build the list of non-overlapping character ranges
//                  from the Unicode Sets.
//
//------------------------------------------------------------------------
void RBBISetBuilder::build() {
    RBBINode        *usetNode;
    RangeDescriptor *rlRange;

    if (fRB->fDebugEnv && uprv_strstr(fRB->fDebugEnv, "usets")) {printSets();}

    //
    //  Initialize the process by creating a single range encompassing all characters
    //  that is in no sets.
    //
    fRangeList                = new RangeDescriptor(*fStatus); // will check for status here
    fRangeList->fStartChar    = 0;
    fRangeList->fEndChar      = 0x10ffff;

    if (U_FAILURE(*fStatus)) {
        return;
    }

    //
    //  Find the set of non-overlapping ranges of characters
    //
    int  ni;
    for (ni=0; ; ni++) {        // Loop over each of the UnicodeSets encountered in the input rules
        usetNode = (RBBINode *)this->fRB->fUSetNodes->elementAt(ni);
        if (usetNode==NULL) {
            break;
        }

        UnicodeSet      *inputSet             = usetNode->fInputSet;
        int32_t          inputSetRangeCount   = inputSet->getRangeCount();
        int              inputSetRangeIndex   = 0;
                         rlRange              = fRangeList;

        for (;;) {
            if (inputSetRangeIndex >= inputSetRangeCount) {
                break;
            }
            UChar32      inputSetRangeBegin  = inputSet->getRangeStart(inputSetRangeIndex);
            UChar32      inputSetRangeEnd    = inputSet->getRangeEnd(inputSetRangeIndex);

            // skip over ranges from the range list that are completely
            //   below the current range from the input unicode set.
            while (rlRange->fEndChar < inputSetRangeBegin) {
                rlRange = rlRange->fNext;
            }

            // If the start of the range from the range list is before with
            //   the start of the range from the unicode set, split the range list range
            //   in two, with one part being before (wholly outside of) the unicode set
            //   and the other containing the rest.
            //   Then continue the loop; the post-split current range will then be skipped
            //     over
            if (rlRange->fStartChar < inputSetRangeBegin) {
                rlRange->split(inputSetRangeBegin, *fStatus);
                if (U_FAILURE(*fStatus)) {
                    return;
                }
                continue;
            }

            // Same thing at the end of the ranges...
            // If the end of the range from the range list doesn't coincide with
            //   the end of the range from the unicode set, split the range list
            //   range in two.  The first part of the split range will be
            //   wholly inside the Unicode set.
            if (rlRange->fEndChar > inputSetRangeEnd) {
                rlRange->split(inputSetRangeEnd+1, *fStatus);
                if (U_FAILURE(*fStatus)) {
                    return;
                }
            }

            // The current rlRange is now entirely within the UnicodeSet range.
            // Add this unicode set to the list of sets for this rlRange
            if (rlRange->fIncludesSets->indexOf(usetNode) == -1) {
                rlRange->fIncludesSets->addElement(usetNode, *fStatus);
                if (U_FAILURE(*fStatus)) {
                    return;
                }
            }

            // Advance over ranges that we are finished with.
            if (inputSetRangeEnd == rlRange->fEndChar) {
                inputSetRangeIndex++;
            }
            rlRange = rlRange->fNext;
        }
    }

    if (fRB->fDebugEnv && uprv_strstr(fRB->fDebugEnv, "range")) { printRanges();}

    //
    //  Group the above ranges, with each group consisting of one or more
    //    ranges that are in exactly the same set of original UnicodeSets.
//.........这里部分代码省略.........

//.........这里部分代码省略.........
                    duplicateCount++;
                }
                // But this doesn't get all.  We need to fix the TRIE.
            }
        }
        // printf("Number of distinct script sets: %d\n", rtScriptSetsCount);
    }

    

    // Update the Trie values to be reflect the run time script indexes (after duplicate merging).
    //    (Trie Values 0 and 1 are reserved, and the corresponding slots in scriptSets
    //     are unused, which is why the loop index starts at 2.)
    {
        for (int32_t i=2; i<scriptSets->size(); i++) {
            BuilderScriptSet *bSet = static_cast<BuilderScriptSet *>(scriptSets->elementAt(i));
            if (bSet->rindex != (uint32_t)i) {
                utrie2_set32(bSet->trie, bSet->codePoint, bSet->rindex, &status);
            }
        }
    }

    // For code points with script==Common or script==Inherited,
    //   Set the reserved value of 1 into both Tries.  These characters do not participate
    //   in Whole Script Confusable detection; this reserved value is the means
    //   by which they are detected.
    {
        UnicodeSet ignoreSet;
        ignoreSet.applyIntPropertyValue(UCHAR_SCRIPT, USCRIPT_COMMON, status);
        UnicodeSet inheritedSet;
        inheritedSet.applyIntPropertyValue(UCHAR_SCRIPT, USCRIPT_INHERITED, status);
        ignoreSet.addAll(inheritedSet);
        for (int32_t rn=0; rn<ignoreSet.getRangeCount(); rn++) {
            UChar32 rangeStart = ignoreSet.getRangeStart(rn);
            UChar32 rangeEnd   = ignoreSet.getRangeEnd(rn);
            utrie2_setRange32(anyCaseTrie,   rangeStart, rangeEnd, 1, TRUE, &status);
            utrie2_setRange32(lowerCaseTrie, rangeStart, rangeEnd, 1, TRUE, &status);
        }
    }

    // Serialize the data to the Spoof Detector
    {
        utrie2_freeze(anyCaseTrie,   UTRIE2_16_VALUE_BITS, &status);
        int32_t size = utrie2_serialize(anyCaseTrie, NULL, 0, &status);
        // printf("Any case Trie size: %d\n", size);
        if (status != U_BUFFER_OVERFLOW_ERROR) {
            goto cleanup;
        }
        status = U_ZERO_ERROR;
        spImpl->fSpoofData->fRawData->fAnyCaseTrie = spImpl->fSpoofData->fMemLimit;
        spImpl->fSpoofData->fRawData->fAnyCaseTrieLength = size;
        spImpl->fSpoofData->fAnyCaseTrie = anyCaseTrie;
        void *where = spImpl->fSpoofData->reserveSpace(size, status);
        utrie2_serialize(anyCaseTrie, where, size, &status);
        
        utrie2_freeze(lowerCaseTrie, UTRIE2_16_VALUE_BITS, &status);
        size = utrie2_serialize(lowerCaseTrie, NULL, 0, &status);
        // printf("Lower case Trie size: %d\n", size);
        if (status != U_BUFFER_OVERFLOW_ERROR) {
            goto cleanup;
        }
        status = U_ZERO_ERROR;
        spImpl->fSpoofData->fRawData->fLowerCaseTrie = spImpl->fSpoofData->fMemLimit;
        spImpl->fSpoofData->fRawData->fLowerCaseTrieLength = size;
        spImpl->fSpoofData->fLowerCaseTrie = lowerCaseTrie;
        where = spImpl->fSpoofData->reserveSpace(size, status);