static void TestCompareVersions(void)
{
/* use a 1d array to be palatable to java */
const char *testCases[] = {
/* v1 <|=|> v2 */
"0.0.0.0", "=", "0.0.0.0",
"3.1.2.0", ">", "3.0.9.0",
"3.2.8.6", "<", "3.4",
"4.0", ">", "3.2",
NULL, NULL, NULL
};
const char *v1str;
const char *opstr;
const char *v2str;
int32_t op, invop, got, invgot;
UVersionInfo v1, v2;
int32_t j;
log_verbose("Testing memcmp()\n");
for(j=0;testCases[j]!=NULL;j+=3) {
v1str = testCases[j+0];
opstr = testCases[j+1];
v2str = testCases[j+2];
switch(opstr[0]) {
case '-':
case '<': op = -1; break;
case '0':
case '=': op = 0; break;
case '+':
case '>': op = 1; break;
default: log_err("Bad operator at j/3=%d\n", (j/3)); return;
}
invop = 0-op; /* inverse operation: with v1 and v2 switched */
u_versionFromString(v1, v1str);
u_versionFromString(v2, v2str);
got = memcmp(v1, v2, sizeof(UVersionInfo));
invgot = memcmp(v2, v1, sizeof(UVersionInfo)); /* Opposite */
if((got <= 0 && op <= 0) || (got >= 0 && op >= 0)) {
log_verbose("%d: %s %s %s, OK\n", (j/3), v1str, opstr, v2str);
} else {
log_err("%d: %s %s %s: wanted values of the same sign, %d got %d\n", (j/3), v1str, opstr, v2str, op, got);
}
if((invgot >= 0 && invop >= 0) || (invgot <= 0 && invop <= 0)) {
log_verbose("%d: %s (%d) %s, OK (inverse)\n", (j/3), v2str, invop, v1str);
} else {
log_err("%d: %s (%d) %s: wanted values of the same sign, %d got %d\n", (j/3), v2str, invop, v1str, invop, invgot);
}
}
}
extern void
setUnicodeVersion(const char *v) {
UVersionInfo version;
u_versionFromString(version, v);
uprv_memcpy(dataInfo.dataVersion, version, 4);
ucdVersion=findUnicodeVersion(version);
}
static void U_CALLCONV
normalizationCorrectionsLineFn(void *context,
char *fields[][2], int32_t fieldCount,
UErrorCode *pErrorCode) {
uint32_t mapping[40];
char *end, *s;
uint32_t code;
int32_t length;
UVersionInfo version;
UVersionInfo thisVersion;
/* get the character code, field 0 */
code=(uint32_t)uprv_strtoul(fields[0][0], &end, 16);
if(U_FAILURE(*pErrorCode)) {
fprintf(stderr, "gensprep: error parsing NormalizationCorrections.txt mapping at %s\n", fields[0][0]);
exit(*pErrorCode);
}
/* Original (erroneous) decomposition */
s = fields[1][0];
/* parse the mapping string */
length=u_parseCodePoints(s, mapping, sizeof(mapping)/4, pErrorCode);
/* ignore corrected decomposition */
u_versionFromString(version,fields[3][0] );
u_versionFromString(thisVersion, "3.2.0");
if(U_FAILURE(*pErrorCode)) {
fprintf(stderr, "gensprep error parsing NormalizationCorrections.txt of U+%04lx - %s\n",
(long)code, u_errorName(*pErrorCode));
exit(*pErrorCode);
}
/* store the mapping */
if( version[0] > thisVersion[0] ||
((version[0]==thisVersion[0]) && (version[1] > thisVersion[1]))
){
storeMapping(code,mapping, length, USPREP_MAP, pErrorCode);
}
setUnicodeVersionNC(version);
}
void
Normalizer2DataBuilder::setUnicodeVersion(const char *v) {
UVersionInfo nullVersion= { 0, 0, 0, 0 };
UVersionInfo version;
u_versionFromString(version, v);
if( 0!=memcmp(version, unicodeVersion, U_MAX_VERSION_LENGTH) &&
0!=memcmp(nullVersion, unicodeVersion, U_MAX_VERSION_LENGTH)
) {
char buffer[U_MAX_VERSION_STRING_LENGTH];
u_versionToString(unicodeVersion, buffer);
fprintf(stderr, "gennorm2 error: multiple inconsistent Unicode version numbers %s vs. %s\n",
buffer, v);
exit(U_ILLEGAL_ARGUMENT_ERROR);
}
memcpy(unicodeVersion, version, U_MAX_VERSION_LENGTH);
}
static void TestVersion(void)
{
UVersionInfo versionArray = {0x01, 0x00, 0x02, 0x02};
UVersionInfo versionArray2 = {0x01, 0x00, 0x02, 0x02};
char versionString[17]; /* xxx.xxx.xxx.xxx\0 */
UChar versionUString[] = { 0x0031, 0x002E, 0x0030, 0x002E,
0x0032, 0x002E, 0x0038, 0x0000 }; /* 1.0.2.8 */
UBool isModified = FALSE;
UVersionInfo version;
UErrorCode status = U_ZERO_ERROR;
log_verbose("Testing the API u_versionToString().....\n");
u_versionToString(versionArray, versionString);
if(strcmp(versionString, "1.0.2.2") != 0){
log_err("ERROR: u_versionToString() failed. Expected: 1.0.2.2, Got=%s\n", versionString);
}
log_verbose("Testing the API u_versionToString().....with versionArray=NULL\n");
u_versionToString(NULL, versionString);
if(strcmp(versionString, "") != 0){
log_err("ERROR: u_versionToString() failed. with versionArray=NULL. It should just return\n");
}
log_verbose("Testing the API u_versionToString().....with versionArray=NULL\n");
u_versionToString(NULL, versionString);
if(strcmp(versionString, "") != 0){
log_err("ERROR: u_versionToString() failed . It should just return\n");
}
log_verbose("Testing the API u_versionToString().....with versionString=NULL\n");
u_versionToString(versionArray, NULL);
if(strcmp(versionString, "") != 0){
log_err("ERROR: u_versionToString() failed. with versionArray=NULL It should just return\n");
}
versionArray[0] = 0x0a;
log_verbose("Testing the API u_versionToString().....\n");
u_versionToString(versionArray, versionString);
if(strcmp(versionString, "10.0.2.2") != 0){
log_err("ERROR: u_versionToString() failed. Expected: 10.0.2.2, Got=%s\n", versionString);
}
versionArray[0] = 0xa0;
u_versionToString(versionArray, versionString);
if(strcmp(versionString, "160.0.2.2") != 0){
log_err("ERROR: u_versionToString() failed. Expected: 160.0.2.2, Got=%s\n", versionString);
}
versionArray[0] = 0xa0;
versionArray[1] = 0xa0;
u_versionToString(versionArray, versionString);
if(strcmp(versionString, "160.160.2.2") != 0){
log_err("ERROR: u_versionToString() failed. Expected: 160.160.2.2, Got=%s\n", versionString);
}
versionArray[0] = 0x01;
versionArray[1] = 0x0a;
u_versionToString(versionArray, versionString);
if(strcmp(versionString, "1.10.2.2") != 0){
log_err("ERROR: u_versionToString() failed. Expected: 160.160.2.2, Got=%s\n", versionString);
}
log_verbose("Testing the API u_versionFromString() ....\n");
u_versionFromString(versionArray, "1.3.5.6");
u_versionToString(versionArray, versionString);
if(strcmp(versionString, "1.3.5.6") != 0){
log_err("ERROR: u_getVersion() failed. Expected: 1.3.5.6, Got=%s\n", versionString);
}
log_verbose("Testing the API u_versionFromString() where versionArray=NULL....\n");
u_versionFromString(NULL, "1.3.5.6");
u_versionToString(versionArray, versionString);
if(strcmp(versionString, "1.3.5.6") != 0){
log_err("ERROR: u_getVersion() failed. Expected: 1.3.5.6, Got=%s\n", versionString);
}
log_verbose("Testing the API u_getVersion().....\n");
u_getVersion(versionArray);
u_versionToString(versionArray, versionString);
if(strcmp(versionString, U_ICU_VERSION) != 0){
log_err("ERROR: u_getVersion() failed. Got=%s, expected %s\n", versionString, U_ICU_VERSION);
}
/* test unicode */
log_verbose("Testing u_versionFromUString...\n");
u_versionFromString(versionArray,"1.0.2.8");
u_versionFromUString(versionArray2, versionUString);
u_versionToString(versionArray2, versionString);
if(memcmp(versionArray, versionArray2, sizeof(UVersionInfo))) {
log_err("FAIL: u_versionFromUString produced a different result - not 1.0.2.8 but %s [%x.%x.%x.%x]\n",
versionString,
(int)versionArray2[0],
(int)versionArray2[1],
(int)versionArray2[2],
(int)versionArray2[3]);
}
else {
log_verbose(" from UString: %s\n", versionString);
}
/* Test the data version API for better code coverage */
u_getDataVersion(version, &status);
if (U_FAILURE(status)) {
log_data_err("ERROR: Unable to get data version. %s\n", u_errorName(status));
} else {
u_isDataOlder(version, &isModified, &status);
if (U_FAILURE(status)) {
log_err("ERROR: Unable to compare data version. %s\n", u_errorName(status));
}
}
}
UnicodeSet&
UnicodeSet::applyPropertyAlias(const UnicodeString& prop,
const UnicodeString& value,
UErrorCode& ec) {
if (U_FAILURE(ec) || isFrozen()) return *this;
// prop and value used to be converted to char * using the default
// converter instead of the invariant conversion.
// This should not be necessary because all Unicode property and value
// names use only invariant characters.
// If there are any variant characters, then we won't find them anyway.
// Checking first avoids assertion failures in the conversion.
if( !uprv_isInvariantUString(prop.getBuffer(), prop.length()) ||
!uprv_isInvariantUString(value.getBuffer(), value.length())
) {
FAIL(ec);
}
CharString pname, vname;
pname.appendInvariantChars(prop, ec);
vname.appendInvariantChars(value, ec);
if (U_FAILURE(ec)) return *this;
UProperty p;
int32_t v;
UBool mustNotBeEmpty = FALSE, invert = FALSE;
if (value.length() > 0) {
p = u_getPropertyEnum(pname.data());
if (p == UCHAR_INVALID_CODE) FAIL(ec);
// Treat gc as gcm
if (p == UCHAR_GENERAL_CATEGORY) {
p = UCHAR_GENERAL_CATEGORY_MASK;
}
if ((p >= UCHAR_BINARY_START && p < UCHAR_BINARY_LIMIT) ||
(p >= UCHAR_INT_START && p < UCHAR_INT_LIMIT) ||
(p >= UCHAR_MASK_START && p < UCHAR_MASK_LIMIT)) {
v = u_getPropertyValueEnum(p, vname.data());
if (v == UCHAR_INVALID_CODE) {
// Handle numeric CCC
if (p == UCHAR_CANONICAL_COMBINING_CLASS ||
p == UCHAR_TRAIL_CANONICAL_COMBINING_CLASS ||
p == UCHAR_LEAD_CANONICAL_COMBINING_CLASS) {
char* end;
double value = uprv_strtod(vname.data(), &end);
v = (int32_t) value;
if (v != value || v < 0 || *end != 0) {
// non-integral or negative value, or trailing junk
FAIL(ec);
}
// If the resultant set is empty then the numeric value
// was invalid.
mustNotBeEmpty = TRUE;
} else {
FAIL(ec);
}
}
}
else {
switch (p) {
case UCHAR_NUMERIC_VALUE:
{
char* end;
double value = uprv_strtod(vname.data(), &end);
if (*end != 0) {
FAIL(ec);
}
applyFilter(numericValueFilter, &value, UPROPS_SRC_CHAR, ec);
return *this;
}
case UCHAR_NAME:
{
// Must munge name, since u_charFromName() does not do
// 'loose' matching.
char buf[128]; // it suffices that this be > uprv_getMaxCharNameLength
if (!mungeCharName(buf, vname.data(), sizeof(buf))) FAIL(ec);
UChar32 ch = u_charFromName(U_EXTENDED_CHAR_NAME, buf, &ec);
if (U_SUCCESS(ec)) {
clear();
add(ch);
return *this;
} else {
FAIL(ec);
}
}
case UCHAR_UNICODE_1_NAME:
// ICU 49 deprecates the Unicode_1_Name property APIs.
FAIL(ec);
case UCHAR_AGE:
{
// Must munge name, since u_versionFromString() does not do
// 'loose' matching.
char buf[128];
if (!mungeCharName(buf, vname.data(), sizeof(buf))) FAIL(ec);
UVersionInfo version;
u_versionFromString(version, buf);
applyFilter(versionFilter, &version, UPROPS_SRC_PROPSVEC, ec);
return *this;
}
case UCHAR_SCRIPT_EXTENSIONS:
v = u_getPropertyValueEnum(UCHAR_SCRIPT, vname.data());
if (v == UCHAR_INVALID_CODE) {
FAIL(ec);
}
// fall through to calling applyIntPropertyValue()
break;
default:
// p is a non-binary, non-enumerated property that we
// don't support (yet).
FAIL(ec);
}
}
}
else {
// value is empty. Interpret as General Category, Script, or
// Binary property.
p = UCHAR_GENERAL_CATEGORY_MASK;
v = u_getPropertyValueEnum(p, pname.data());
if (v == UCHAR_INVALID_CODE) {
p = UCHAR_SCRIPT;
v = u_getPropertyValueEnum(p, pname.data());
if (v == UCHAR_INVALID_CODE) {
p = u_getPropertyEnum(pname.data());
if (p >= UCHAR_BINARY_START && p < UCHAR_BINARY_LIMIT) {
v = 1;
} else if (0 == uprv_comparePropertyNames(ANY, pname.data())) {
set(MIN_VALUE, MAX_VALUE);
return *this;
} else if (0 == uprv_comparePropertyNames(ASCII, pname.data())) {
set(0, 0x7F);
return *this;
} else if (0 == uprv_comparePropertyNames(ASSIGNED, pname.data())) {
// [:Assigned:]=[:^Cn:]
p = UCHAR_GENERAL_CATEGORY_MASK;
v = U_GC_CN_MASK;
invert = TRUE;
} else {
FAIL(ec);
}
}
}
}
applyIntPropertyValue(p, v, ec);
if(invert) {
complement();
}
if (U_SUCCESS(ec) && (mustNotBeEmpty && isEmpty())) {
// mustNotBeEmpty is set to true if an empty set indicates
// invalid input.
ec = U_ILLEGAL_ARGUMENT_ERROR;
}
if (isBogus() && U_SUCCESS(ec)) {
// We likely ran out of memory. AHHH!
ec = U_MEMORY_ALLOCATION_ERROR;
}
return *this;
}
static void TestPUtilAPI(void){
double n1=0.0, y1=0.0, expn1, expy1;
double value1 = 0.021;
UVersionInfo versionArray = {0x01, 0x00, 0x02, 0x02};
char versionString[17]; /* xxx.xxx.xxx.xxx\0 */
char *str=0;
UBool isTrue=FALSE;
log_verbose("Testing the API uprv_modf()\n");
y1 = uprv_modf(value1, &n1);
expn1=0;
expy1=0.021;
if(y1 != expy1 || n1 != expn1){
log_err("Error in uprv_modf. Expected IntegralValue=%f, Got=%f, \n Expected FractionalValue=%f, Got=%f\n",
expn1, n1, expy1, y1);
}
if(VERBOSITY){
log_verbose("[float] x = %f n = %f y = %f\n", value1, n1, y1);
}
log_verbose("Testing the API uprv_fmod()\n");
expn1=uprv_fmod(30.50, 15.00);
doAssert(expn1, 0.5, "uprv_fmod(30.50, 15.00) failed.");
log_verbose("Testing the API uprv_ceil()\n");
expn1=uprv_ceil(value1);
doAssert(expn1, 1, "uprv_ceil(0.021) failed.");
log_verbose("Testing the API uprv_floor()\n");
expn1=uprv_floor(value1);
doAssert(expn1, 0, "uprv_floor(0.021) failed.");
log_verbose("Testing the API uprv_fabs()\n");
expn1=uprv_fabs((2.02-1.345));
doAssert(expn1, 0.675, "uprv_fabs(2.02-1.345) failed.");
log_verbose("Testing the API uprv_fmax()\n");
doAssert(uprv_fmax(2.4, 1.2), 2.4, "uprv_fmax(2.4, 1.2) failed.");
log_verbose("Testing the API uprv_fmax() with x value= NaN\n");
expn1=uprv_fmax(uprv_getNaN(), 1.2);
doAssert(expn1, uprv_getNaN(), "uprv_fmax(uprv_getNaN(), 1.2) failed. when one parameter is NaN");
log_verbose("Testing the API uprv_fmin()\n");
doAssert(uprv_fmin(2.4, 1.2), 1.2, "uprv_fmin(2.4, 1.2) failed.");
log_verbose("Testing the API uprv_fmin() with x value= NaN\n");
expn1=uprv_fmin(uprv_getNaN(), 1.2);
doAssert(expn1, uprv_getNaN(), "uprv_fmin(uprv_getNaN(), 1.2) failed. when one parameter is NaN");
log_verbose("Testing the API uprv_max()\n");
doAssert(uprv_max(4, 2), 4, "uprv_max(4, 2) failed.");
log_verbose("Testing the API uprv_min()\n");
doAssert(uprv_min(-4, 2), -4, "uprv_min(-4, 2) failed.");
log_verbose("Testing the API uprv_trunc()\n");
doAssert(uprv_trunc(12.3456), 12, "uprv_trunc(12.3456) failed.");
doAssert(uprv_trunc(12.234E2), 1223, "uprv_trunc(12.234E2) failed.");
doAssert(uprv_trunc(uprv_getNaN()), uprv_getNaN(), "uprv_trunc(uprv_getNaN()) failed. with parameter=NaN");
doAssert(uprv_trunc(uprv_getInfinity()), uprv_getInfinity(), "uprv_trunc(uprv_getInfinity()) failed. with parameter=Infinity");
log_verbose("Testing the API uprv_pow10()\n");
doAssert(uprv_pow10(4), 10000, "uprv_pow10(4) failed.");
log_verbose("Testing the API uprv_log10()\n");
doAssert(uprv_log10(3456), 3, "uprv_log10(3456) failed.");
#ifdef OS390
doAssert(uprv_log10(1.0e55), 55, "uprv_log10(1.0e55) failed.");
#else
doAssert(uprv_log10(1.0e300), 300, "uprv_log10(1.0e300) failed.");
#endif
log_verbose("Testing the API uprv_isNegativeInfinity()\n");
isTrue=uprv_isNegativeInfinity(uprv_getInfinity() * -1);
if(isTrue != TRUE){
log_err("ERROR: uprv_isNegativeInfinity failed.\n");
}
log_verbose("Testing the API uprv_isPositiveInfinity()\n");
isTrue=uprv_isPositiveInfinity(uprv_getInfinity());
if(isTrue != TRUE){
log_err("ERROR: uprv_isPositiveInfinity failed.\n");
}
log_verbose("Testing the API uprv_isInfinite()\n");
isTrue=uprv_isInfinite(uprv_getInfinity());
if(isTrue != TRUE){
log_err("ERROR: uprv_isInfinite failed.\n");
}
#if 0
log_verbose("Testing the API uprv_digitsAfterDecimal()....\n");
doAssert(uprv_digitsAfterDecimal(value1), 3, "uprv_digitsAfterDecimal() failed.");
doAssert(uprv_digitsAfterDecimal(1.2345E2), 2, "uprv_digitsAfterDecimal(1.2345E2) failed.");
doAssert(uprv_digitsAfterDecimal(1.2345E-2), 6, "uprv_digitsAfterDecimal(1.2345E-2) failed.");
doAssert(uprv_digitsAfterDecimal(1.2345E2), 2, "uprv_digitsAfterDecimal(1.2345E2) failed.");
doAssert(uprv_digitsAfterDecimal(-1.2345E-20), 24, "uprv_digitsAfterDecimal(1.2345E-20) failed.");
doAssert(uprv_digitsAfterDecimal(1.2345E20), 0, "uprv_digitsAfterDecimal(1.2345E20) failed.");
doAssert(uprv_digitsAfterDecimal(-0.021), 3, "uprv_digitsAfterDecimal(-0.021) failed.");
doAssert(uprv_digitsAfterDecimal(23.0), 0, "uprv_digitsAfterDecimal(23.0) failed.");
doAssert(uprv_digitsAfterDecimal(0.022223333321), 9, "uprv_digitsAfterDecimal(0.022223333321) failed.");
#endif
log_verbose("Testing the API u_versionToString().....\n");
u_versionToString(versionArray, versionString);
if(strcmp(versionString, "1.0.2.2") != 0){
log_err("ERROR: u_versionToString() failed. Expected: 1.0.2.2, Got=%s\n", versionString);
}
log_verbose("Testing the API u_versionToString().....with versionArray=NULL\n");
u_versionToString(NULL, versionString);
if(strcmp(versionString, "") != 0){
log_err("ERROR: u_versionToString() failed. with versionArray=NULL. It should just return\n");
}
log_verbose("Testing the API u_versionToString().....with versionArray=NULL\n");
u_versionToString(NULL, versionString);
if(strcmp(versionString, "") != 0){
log_err("ERROR: u_versionToString() failed . It should just return\n");
}
log_verbose("Testing the API u_versionToString().....with versionString=NULL\n");
u_versionToString(versionArray, NULL);
if(strcmp(versionString, "") != 0){
log_err("ERROR: u_versionToString() failed. with versionArray=NULL It should just return\n");
}
versionArray[0] = 0x0a;
log_verbose("Testing the API u_versionToString().....\n");
u_versionToString(versionArray, versionString);
if(strcmp(versionString, "10.0.2.2") != 0){
log_err("ERROR: u_versionToString() failed. Expected: 10.0.2.2, Got=%s\n", versionString);
}
versionArray[0] = 0xa0;
u_versionToString(versionArray, versionString);
if(strcmp(versionString, "160.0.2.2") != 0){
log_err("ERROR: u_versionToString() failed. Expected: 160.0.2.2, Got=%s\n", versionString);
}
versionArray[0] = 0xa0;
versionArray[1] = 0xa0;
u_versionToString(versionArray, versionString);
if(strcmp(versionString, "160.160.2.2") != 0){
log_err("ERROR: u_versionToString() failed. Expected: 160.160.2.2, Got=%s\n", versionString);
}
versionArray[0] = 0x01;
versionArray[1] = 0x0a;
u_versionToString(versionArray, versionString);
if(strcmp(versionString, "1.10.2.2") != 0){
log_err("ERROR: u_versionToString() failed. Expected: 160.160.2.2, Got=%s\n", versionString);
}
log_verbose("Testing the API u_versionFromString() ....\n");
u_versionFromString(versionArray, "1.3.5.6");
u_versionToString(versionArray, versionString);
if(strcmp(versionString, "1.3.5.6") != 0){
log_err("ERROR: u_getVersion() failed. Expected: 1.3.5.6, Got=%s\n", versionString);
}
log_verbose("Testing the API u_versionFromString() where versionArray=NULL....\n");
u_versionFromString(NULL, "1.3.5.6");
u_versionToString(versionArray, versionString);
if(strcmp(versionString, "1.3.5.6") != 0){
log_err("ERROR: u_getVersion() failed. Expected: 1.3.5.6, Got=%s\n", versionString);
}
log_verbose("Testing the API u_getVersion().....\n");
u_getVersion(versionArray);
u_versionToString(versionArray, versionString);
if(strcmp(versionString, U_ICU_VERSION) != 0){
log_err("ERROR: u_getVersion() failed. Got=%s, expected %s\n", versionString, U_ICU_VERSION);
}
log_verbose("Testing the API u_errorName()...\n");
str=(char*)u_errorName((UErrorCode)0);
if(strcmp(str, "U_ZERO_ERROR") != 0){
log_err("ERROR: u_getVersion() failed. Expected: U_ZERO_ERROR Got=%s\n", str);
}
log_verbose("Testing the API u_errorName()...\n");
str=(char*)u_errorName((UErrorCode)-127);
if(strcmp(str, "U_USING_DEFAULT_WARNING") != 0){
log_err("ERROR: u_getVersion() failed. Expected: U_USING_DEFAULT_WARNING Got=%s\n", str);
}
log_verbose("Testing the API u_errorName().. with BOGUS ERRORCODE...\n");
str=(char*)u_errorName((UErrorCode)200);
if(strcmp(str, "[BOGUS UErrorCode]") != 0){
log_err("ERROR: u_getVersion() failed. Expected: [BOGUS UErrorCode] Got=%s\n", str);
}
{
const char* dataDirectory;
UChar *udataDir=0;
UChar temp[100];
char *charvalue=0;
log_verbose("Testing chars to UChars\n");
/* This cannot really work on a japanese system. u_uastrcpy will have different results than */
/* u_charsToUChars when there is a backslash in the string! */
/*dataDirectory=u_getDataDirectory();*/
dataDirectory="directory1"; /*no backslashes*/
udataDir=(UChar*)malloc(sizeof(UChar) * (strlen(dataDirectory) + 1));
u_charsToUChars(dataDirectory, udataDir, (strlen(dataDirectory)+1));
u_uastrcpy(temp, dataDirectory);
if(u_strcmp(temp, udataDir) != 0){
log_err("ERROR: u_charsToUChars failed. Expected %s, Got %s\n", austrdup(temp), austrdup(udataDir));
}
log_verbose("Testing UChars to chars\n");
charvalue=(char*)malloc(sizeof(char) * (u_strlen(udataDir) + 1));
u_UCharsToChars(udataDir, charvalue, (u_strlen(udataDir)+1));
if(strcmp(charvalue, dataDirectory) != 0){
log_err("ERROR: u_UCharsToChars failed. Expected %s, Got %s\n", charvalue, dataDirectory);
}
free(charvalue);
free(udataDir);
}
log_verbose("Testing uprv_timezone()....\n");
{
int32_t tzoffset = uprv_timezone();
log_verbose("Value returned from uprv_timezone = %d\n", tzoffset);
if (tzoffset != 28800) {
log_verbose("***** WARNING: If testing in the PST timezone, t_timezone should return 28800! *****");
}
if ((tzoffset % 1800 != 0)) {
log_err("FAIL: t_timezone may be incorrect. It is not a multiple of 30min.");
}
tzoffset=uprv_getUTCtime();
}
}
U_CAPI void U_EXPORT2
u_getVersion(UVersionInfo versionArray) {
u_versionFromString(versionArray, U_ICU_VERSION);
}
// Returns TRUE for "ok to continue parsing fields".
UBool
PreparsedUCD::parseProperty(UniProps &props, const char *field, UnicodeSet &newValues,
UErrorCode &errorCode) {
CharString pBuffer;
const char *p=field;
const char *v=strchr(p, '=');
int binaryValue;
if(*p=='-') {
if(v!=NULL) {
fprintf(stderr,
"error in preparsed UCD: mix of binary-property-no and "
"enum-property syntax '%s' on line %ld\n",
field, (long)lineNumber);
errorCode=U_PARSE_ERROR;
return FALSE;
}
binaryValue=0;
++p;
} else if(v==NULL) {
binaryValue=1;
} else {
binaryValue=-1;
// Copy out the property name rather than modifying the field (writing a NUL).
pBuffer.append(p, (int32_t)(v-p), errorCode);
p=pBuffer.data();
++v;
}
int32_t prop=pnames->getPropertyEnum(p);
if(prop<0) {
for(int32_t i=0;; ++i) {
if(i==UPRV_LENGTHOF(ppucdProperties)) {
// Ignore unknown property names.
return TRUE;
}
if(0==uprv_stricmp(p, ppucdProperties[i].name)) {
prop=ppucdProperties[i].prop;
U_ASSERT(prop>=0);
break;
}
}
}
if(prop<UCHAR_BINARY_LIMIT) {
if(binaryValue>=0) {
props.binProps[prop]=(UBool)binaryValue;
} else {
// No binary value for a binary property.
fprintf(stderr,
"error in preparsed UCD: enum-property syntax '%s' "
"for binary property on line %ld\n",
field, (long)lineNumber);
errorCode=U_PARSE_ERROR;
}
} else if(binaryValue>=0) {
// Binary value for a non-binary property.
fprintf(stderr,
"error in preparsed UCD: binary-property syntax '%s' "
"for non-binary property on line %ld\n",
field, (long)lineNumber);
errorCode=U_PARSE_ERROR;
} else if (prop < UCHAR_INT_START) {
fprintf(stderr,
"error in preparsed UCD: prop value is invalid: '%d' for line %ld\n",
prop, (long)lineNumber);
errorCode=U_PARSE_ERROR;
} else if(prop<UCHAR_INT_LIMIT) {
int32_t value=pnames->getPropertyValueEnum(prop, v);
if(value==UCHAR_INVALID_CODE && prop==UCHAR_CANONICAL_COMBINING_CLASS) {
// TODO: Make getPropertyValueEnum(UCHAR_CANONICAL_COMBINING_CLASS, v) work.
char *end;
unsigned long ccc=uprv_strtoul(v, &end, 10);
if(v<end && *end==0 && ccc<=254) {
value=(int32_t)ccc;
}
}
if(value==UCHAR_INVALID_CODE) {
fprintf(stderr,
"error in preparsed UCD: '%s' is not a valid value on line %ld\n",
field, (long)lineNumber);
errorCode=U_PARSE_ERROR;
} else {
props.intProps[prop-UCHAR_INT_START]=value;
}
} else if(*v=='<') {
// Do not parse default values like <code point>, just set null values.
switch(prop) {
case UCHAR_BIDI_MIRRORING_GLYPH:
props.bmg=U_SENTINEL;
break;
case UCHAR_BIDI_PAIRED_BRACKET:
props.bpb=U_SENTINEL;
break;
case UCHAR_SIMPLE_CASE_FOLDING:
props.scf=U_SENTINEL;
break;
case UCHAR_SIMPLE_LOWERCASE_MAPPING:
props.slc=U_SENTINEL;
break;
case UCHAR_SIMPLE_TITLECASE_MAPPING:
props.stc=U_SENTINEL;
break;
case UCHAR_SIMPLE_UPPERCASE_MAPPING:
props.suc=U_SENTINEL;
break;
case UCHAR_CASE_FOLDING:
props.cf.remove();
break;
case UCHAR_LOWERCASE_MAPPING:
props.lc.remove();
break;
case UCHAR_TITLECASE_MAPPING:
props.tc.remove();
break;
case UCHAR_UPPERCASE_MAPPING:
props.uc.remove();
break;
case UCHAR_SCRIPT_EXTENSIONS:
props.scx.clear();
break;
default:
fprintf(stderr,
"error in preparsed UCD: '%s' is not a valid default value on line %ld\n",
field, (long)lineNumber);
errorCode=U_PARSE_ERROR;
}
} else {
char c;
switch(prop) {
case UCHAR_NUMERIC_VALUE:
props.numericValue=v;
c=*v;
if('0'<=c && c<='9' && v[1]==0) {
props.digitValue=c-'0';
} else {
props.digitValue=-1;
}
break;
case UCHAR_NAME:
props.name=v;
break;
case UCHAR_AGE:
u_versionFromString(props.age, v); // Writes 0.0.0.0 if v is not numeric.
break;
case UCHAR_BIDI_MIRRORING_GLYPH:
props.bmg=parseCodePoint(v, errorCode);
break;
case UCHAR_BIDI_PAIRED_BRACKET:
props.bpb=parseCodePoint(v, errorCode);
break;
case UCHAR_SIMPLE_CASE_FOLDING:
props.scf=parseCodePoint(v, errorCode);
break;
case UCHAR_SIMPLE_LOWERCASE_MAPPING:
props.slc=parseCodePoint(v, errorCode);
break;
case UCHAR_SIMPLE_TITLECASE_MAPPING:
props.stc=parseCodePoint(v, errorCode);
break;
case UCHAR_SIMPLE_UPPERCASE_MAPPING:
props.suc=parseCodePoint(v, errorCode);
break;
case UCHAR_CASE_FOLDING:
parseString(v, props.cf, errorCode);
break;
case UCHAR_LOWERCASE_MAPPING:
parseString(v, props.lc, errorCode);
break;
case UCHAR_TITLECASE_MAPPING:
parseString(v, props.tc, errorCode);
break;
case UCHAR_UPPERCASE_MAPPING:
parseString(v, props.uc, errorCode);
break;
case PPUCD_NAME_ALIAS:
props.nameAlias=v;
break;
case PPUCD_CONDITIONAL_CASE_MAPPINGS:
case PPUCD_TURKIC_CASE_FOLDING:
// No need to parse their values: They are hardcoded in the runtime library.
break;
case UCHAR_SCRIPT_EXTENSIONS:
parseScriptExtensions(v, props.scx, errorCode);
break;
default:
// Ignore unhandled properties.
return TRUE;
}
}
if(U_SUCCESS(errorCode)) {
newValues.add((UChar32)prop);
return TRUE;
} else {
return FALSE;
}
}
PreparsedUCD::LineType
PreparsedUCD::readLine(UErrorCode &errorCode) {
if(U_FAILURE(errorCode)) { return NO_LINE; }
// Select the next available line buffer.
while(!isLineBufferAvailable(lineIndex)) {
++lineIndex;
if (lineIndex == kNumLineBuffers) {
lineIndex = 0;
}
}
char *line=lines[lineIndex];
*line=0;
lineLimit=fieldLimit=line;
lineType=NO_LINE;
char *result=fgets(line, sizeof(lines[0]), file);
if(result==NULL) {
if(ferror(file)) {
perror("error reading preparsed UCD");
fprintf(stderr, "error reading preparsed UCD before line %ld\n", (long)lineNumber);
errorCode=U_FILE_ACCESS_ERROR;
}
return NO_LINE;
}
++lineNumber;
if(*line=='#') {
fieldLimit=strchr(line, 0);
return lineType=EMPTY_LINE;
}
// Remove trailing /r/n.
char c;
char *limit=strchr(line, 0);
while(line<limit && ((c=*(limit-1))=='\n' || c=='\r')) { --limit; }
// Remove trailing white space.
while(line<limit && ((c=*(limit-1))==' ' || c=='\t')) { --limit; }
*limit=0;
lineLimit=limit;
if(line==limit) {
fieldLimit=limit;
return lineType=EMPTY_LINE;
}
// Split by ';'.
char *semi=line;
while((semi=strchr(semi, ';'))!=NULL) { *semi++=0; }
fieldLimit=strchr(line, 0);
// Determine the line type.
int32_t type;
for(type=EMPTY_LINE+1;; ++type) {
if(type==LINE_TYPE_COUNT) {
fprintf(stderr,
"error in preparsed UCD: unknown line type (first field) '%s' on line %ld\n",
line, (long)lineNumber);
errorCode=U_PARSE_ERROR;
return NO_LINE;
}
if(0==strcmp(line, lineTypeStrings[type])) {
break;
}
}
lineType=(LineType)type;
if(lineType==UNICODE_VERSION_LINE && fieldLimit<lineLimit) {
u_versionFromString(ucdVersion, fieldLimit+1);
}
return lineType;
}
extern int
main(int argc, char* argv[]) {
UVersionInfo version;
Options moreOptions={ TRUE, FALSE, TRUE };
UErrorCode errorCode = U_ZERO_ERROR;
U_MAIN_INIT_ARGS(argc, argv);
/* Initialize ICU */
u_init(&errorCode);
if (U_FAILURE(errorCode) && errorCode != U_FILE_ACCESS_ERROR) {
/* Note: u_init() will try to open ICU property data.
* failures here are expected when building ICU from scratch.
* ignore them.
*/
fprintf(stderr, "%s: can not initialize ICU. errorCode = %s\n",
argv[0], u_errorName(errorCode));
exit(1);
}
/* preset then read command line options */
options[DESTDIR].value=u_getDataDirectory();
options[UNICODE].value="4.1";
argc=u_parseArgs(argc, argv, LENGTHOF(options), options);
/* error handling, printing usage message */
if(argc<0) {
fprintf(stderr,
"error in command line argument \"%s\"\n",
argv[-argc]);
} else if(argc<2) {
argc=-1;
}
if(argc<0 || options[HELP_H].doesOccur || options[HELP_QUESTION_MARK].doesOccur) {
/*
* Broken into chucks because the C89 standard says the minimum
* required supported string length is 509 bytes.
*/
fprintf(stderr,
"Usage: %s [-1[+|-]] [-v[+|-]] [-c[+|-]] [filename_ud [filename_na]]\n"
"\n"
"Read the UnicodeData.txt file and \n"
"create a binary file " DATA_NAME "." DATA_TYPE " with the character names\n"
"\n"
"\tfilename_ud absolute path/filename for the UnicodeData.txt file\n"
"\t (default: standard input)\n"
"\tfilename_na absolute path/filename for the NameAliases.txt file\n"
"\t (default: no name aliases)\n"
"\n",
argv[0]);
fprintf(stderr,
"Options:\n"
"\t-h or -? or --help this usage text\n"
"\t-v or --verbose verbose output\n"
"\t-q or --quiet no output\n"
"\t-c or --copyright include a copyright notice\n"
"\t-d or --destdir destination directory, followed by the path\n"
"\t-u or --unicode Unicode version, followed by the version like 3.0.0\n");
fprintf(stderr,
"\t-1 or --unicode1-names store Unicode 1.0 character names\n"
"\t --no-iso-comments do not store ISO comments\n"
"\t --only-iso-comments write ucomment.icu with only ISO comments\n");
return argc<0 ? U_ILLEGAL_ARGUMENT_ERROR : U_ZERO_ERROR;
}
/* get the options values */
beVerbose=options[VERBOSE].doesOccur;
beQuiet=options[QUIET].doesOccur;
haveCopyright=options[COPYRIGHT].doesOccur;
moreOptions.store10Names=options[UNICODE1_NAMES].doesOccur;
moreOptions.storeISOComments=!options[NO_ISO_COMMENTS].doesOccur;
if(options[ONLY_ISO_COMMENTS].doesOccur) {
moreOptions.storeNames=moreOptions.store10Names=FALSE;
moreOptions.storeISOComments=TRUE;
}
/* set the Unicode version */
u_versionFromString(version, options[UNICODE].value);
uprv_memcpy(dataInfo.dataVersion, version, 4);
ucdVersion=findUnicodeVersion(version);
init();
if(argc>=3) {
parseNameAliases(argv[2], &moreOptions);
}
parseDB(argc>=2 ? argv[1] : "-", &moreOptions);
compress();
generateData(options[DESTDIR].value, &moreOptions);
u_cleanup();
return 0;
}
extern int
main(int argc, char* argv[]) {
UVersionInfo version;
UBool store10Names=FALSE;
UErrorCode errorCode = U_ZERO_ERROR;
U_MAIN_INIT_ARGS(argc, argv);
/* Initialize ICU */
u_init(&errorCode);
if (U_FAILURE(errorCode) && errorCode != U_FILE_ACCESS_ERROR) {
/* Note: u_init() will try to open ICU property data.
* failures here are expected when building ICU from scratch.
* ignore them.
*/
fprintf(stderr, "%s: can not initialize ICU. errorCode = %s\n",
argv[0], u_errorName(errorCode));
exit(1);
}
/* preset then read command line options */
options[5].value=u_getDataDirectory();
options[6].value="4.1";
argc=u_parseArgs(argc, argv, sizeof(options)/sizeof(options[0]), options);
/* error handling, printing usage message */
if(argc<0) {
fprintf(stderr,
"error in command line argument \"%s\"\n",
argv[-argc]);
} else if(argc<2) {
argc=-1;
}
if(argc<0 || options[0].doesOccur || options[1].doesOccur) {
/*
* Broken into chucks because the C89 standard says the minimum
* required supported string length is 509 bytes.
*/
fprintf(stderr,
"Usage: %s [-1[+|-]] [-v[+|-]] [-c[+|-]] filename\n"
"\n"
"Read the UnicodeData.txt file and \n"
"create a binary file " DATA_NAME "." DATA_TYPE " with the character names\n"
"\n"
"\tfilename absolute path/filename for the Unicode database text file\n"
"\t\t(default: standard input)\n"
"\n",
argv[0]);
fprintf(stderr,
"Options:\n"
"\t-h or -? or --help this usage text\n"
"\t-v or --verbose verbose output\n"
"\t-q or --quiet no output\n"
"\t-c or --copyright include a copyright notice\n"
"\t-d or --destdir destination directory, followed by the path\n"
"\t-u or --unicode Unicode version, followed by the version like 3.0.0\n"
"\t-1 or --unicode1-names store Unicode 1.0 character names\n");
return argc<0 ? U_ILLEGAL_ARGUMENT_ERROR : U_ZERO_ERROR;
}
/* get the options values */
beVerbose=options[2].doesOccur;
beQuiet=options[3].doesOccur;
haveCopyright=options[4].doesOccur;
store10Names=options[7].doesOccur;
/* set the Unicode version */
u_versionFromString(version, options[6].value);
uprv_memcpy(dataInfo.dataVersion, version, 4);
ucdVersion=findUnicodeVersion(version);
init();
parseDB(argc>=2 ? argv[1] : "-", store10Names);
compress();
generateData(options[5].value);
u_cleanup();
return 0;
}