BytesTrieDictLookup(const DictionaryTriePerfTest &perfTest)
: DictLookup(perfTest), trie(NULL), noDict(FALSE) {
IcuToolErrorCode errorCode("BytesTrieDictLookup()");
builder=new BytesTrieBuilder(errorCode);
CharString str;
const ULine *lines=perf.getCachedLines();
int32_t numLines=perf.getNumLines();
for(int32_t i=0; i<numLines; ++i) {
// Skip comment lines (start with a character below 'A').
if(lines[i].name[0]<0x41) {
continue;
}
if(!thaiWordToBytes(lines[i].name, lines[i].len, str.clear(), errorCode)) {
fprintf(stderr, "thaiWordToBytes(): failed for word %ld (0-based)\n", (long)i);
noDict=TRUE;
break;
}
builder->add(str.toStringPiece(), 0, errorCode);
}
if(!noDict) {
int32_t length=builder->buildStringPiece(USTRINGTRIE_BUILD_SMALL, errorCode).length();
printf("size of BytesTrie: %6ld bytes\n", (long)length);
trie=builder->build(USTRINGTRIE_BUILD_SMALL, errorCode);
}
}
/**
* Return a string form of this number.
* Format is as defined by the decNumber library, for interchange of
* decimal numbers.
*/
void DigitList::getDecimal(CharString &str, UErrorCode &status) {
if (U_FAILURE(status)) {
return;
}
// A decimal number in string form can, worst case, be 14 characters longer
// than the number of digits. So says the decNumber library doc.
int32_t maxLength = fDecNumber->digits + 14;
int32_t capacity = 0;
char *buffer = str.clear().getAppendBuffer(maxLength, 0, capacity, status);
if (U_FAILURE(status)) {
return; // Memory allocation error on growing the string.
}
U_ASSERT(capacity >= maxLength);
uprv_decNumberToString(this->fDecNumber, buffer);
U_ASSERT((int32_t)uprv_strlen(buffer) <= maxLength);
str.append(buffer, -1, status);
}
void
PreparsedUCD::parseScriptExtensions(const char *s, UnicodeSet &scx, UErrorCode &errorCode) {
if(U_FAILURE(errorCode)) { return; }
scx.clear();
CharString scString;
for(;;) {
const char *scs;
const char *scLimit=strchr(s, ' ');
if(scLimit!=NULL) {
scs=scString.clear().append(s, (int32_t)(scLimit-s), errorCode).data();
if(U_FAILURE(errorCode)) { return; }
} else {
scs=s;
}
int32_t script=pnames->getPropertyValueEnum(UCHAR_SCRIPT, scs);
if(script==UCHAR_INVALID_CODE) {
fprintf(stderr,
"error in preparsed UCD: '%s' is not a valid script code on line %ld\n",
scs, (long)lineNumber);
errorCode=U_PARSE_ERROR;
return;
} else if(scx.contains(script)) {
fprintf(stderr,
"error in preparsed UCD: scx has duplicate '%s' codes on line %ld\n",
scs, (long)lineNumber);
errorCode=U_PARSE_ERROR;
return;
} else {
scx.add(script);
}
if(scLimit!=NULL) {
s=scLimit+1;
} else {
break;
}
}
if(scx.isEmpty()) {
fprintf(stderr, "error in preparsed UCD: empty scx= on line %ld\n", (long)lineNumber);
errorCode=U_PARSE_ERROR;
}
}