+// © 2016 and later: Unicode, Inc. and others.
+// License & terms of use: http://www.unicode.org/copyright.html
/*
********************************************************************************
-* Copyright (C) 1999-2010 International Business Machines Corporation and
+* Copyright (C) 1999-2016 International Business Machines Corporation and
* others. All Rights Reserved.
********************************************************************************
* Date Name Description
#include "unicode/parsepos.h"
#include "unicode/symtable.h"
#include "unicode/uversion.h"
+#include "cmemory.h"
#include "hash.h"
-#define LENGTHOF(array) (int32_t)(sizeof(array)/sizeof((array)[0]))
-
#define TEST_ASSERT_SUCCESS(status) {if (U_FAILURE(status)) { \
dataerrln("fail in file \"%s\", line %d: \"%s\"", __FILE__, __LINE__, \
u_errorName(status));}}
CASE(21,TestFreezable);
CASE(22,TestSpan);
CASE(23,TestStringSpan);
+ CASE(24,TestUCAUnsafeBackwards);
default: name = ""; break;
}
}
if (set != exp) { errln("FAIL: retain('s')"); return; }
uint16_t buf[32];
- int32_t slen = set.serialize(buf, sizeof(buf)/sizeof(buf[0]), status);
+ int32_t slen = set.serialize(buf, UPRV_LENGTHOF(buf), status);
if (U_FAILURE(status)) { errln("FAIL: serialize"); return; }
if (slen != 3 || buf[0] != 2 || buf[1] != 0x73 || buf[2] != 0x74) {
errln("FAIL: serialize");
// Script_Extensions, new in Unicode 6.0
"[:scx=Arab:]",
"\\u061E\\u061F\\u0620\\u0621\\u063F\\u0640\\u0650\\u065E\\uFDF1\\uFDF2\\uFDF3",
- "\\u061D\\u065F\\uFDEF\\uFDFE",
+ "\\u061D\\uFDEF\\uFDFE",
// U+FDF2 has Script=Arabic and also Arab in its Script_Extensions,
// so scx-sc is missing U+FDF2.
"[[:Script_Extensions=Arabic:]-[:Arab:]]",
- "\\u0640\\u064B\\u0650\\u0655\\uFDFD",
+ "\\u0640\\u064B\\u0650\\u0655",
"\\uFDF2"
};
- static const int32_t DATA_LEN = sizeof(DATA)/sizeof(DATA[0]);
+ static const int32_t DATA_LEN = UPRV_LENGTHOF(DATA);
for (int32_t i=0; i<DATA_LEN; i+=3) {
expectContainment(UnicodeString(DATA[i], -1, US_INV), CharsToUnicodeString(DATA[i+1]),
const char exp[] =
"[\\u200A-\\u200E\\uFEFF\\U0001D173-\\U0001D17A\\U000F0000-\\U000FFFFD]";
// We test this with two passes; in the second pass we
- // pre-unescape the pattern. Since U+200E is rule whitespace,
+ // pre-unescape the pattern. Since U+200E is Pattern_White_Space,
// this fails -- which is what we expect.
for (int32_t pass=1; pass<=2; ++pass) {
UErrorCode ec = U_ZERO_ERROR;
(UChar32)-1, 8, 0, 8,
8, 0x110000, 8, 0x10FFFF
};
- const int32_t DATA_LENGTH = sizeof(DATA)/sizeof(DATA[0]);
+ const int32_t DATA_LENGTH = UPRV_LENGTHOF(DATA);
UnicodeString pat;
int32_t i;
b = set.contains(start, end);
b = set.containsNone(start, end);
b = set.containsSome(start, end);
+ (void)b; // Suppress set but not used warning.
/*int32_t index = set.indexOf(start);*/
(UChar32)-1,
0x110000
};
- const int32_t DATA2_LENGTH = sizeof(DATA2)/sizeof(DATA2[0]);
+ const int32_t DATA2_LENGTH = UPRV_LENGTHOF(DATA2);
for (i=0; i<DATA2_LENGTH; ++i) {
UChar32 c = DATA2[i], end = 0x10FFFF;
Hashtable contents;
TokenSymbolTable(UErrorCode& ec) : contents(FALSE, ec) {
- contents.setValueDeleter(uhash_deleteUnicodeString);
+ contents.setValueDeleter(uprv_deleteUObject);
}
~TokenSymbolTable() {}
errln((UnicodeString)"FAIL: " + UnicodeString(DATA[i], -1, US_INV) + ".size() == " +
set.size() + ", expected 4");
}
+
+ {
+ UErrorCode subErr = U_ZERO_ERROR;
+ checkRoundTrip(set);
+ checkSerializeRoundTrip(set, subErr);
+ }
}
}
logln((UnicodeString)"Testing " + i + ", " + x);
_testComplement(i, x, y);
+ UnicodeSet &toTest = bitsToSet(i, aa);
+
// AS LONG AS WE ARE HERE, check roundtrip
- checkRoundTrip(bitsToSet(i, aa));
+ checkRoundTrip(toTest);
+ UErrorCode ec = U_ZERO_ERROR;
+ checkSerializeRoundTrip(toTest, ec);
for (int32_t j = 0; j < limit; ++j) {
_testAdd(i,j, x,y,z);
* get the same thing back
*/
void UnicodeSetTest::checkRoundTrip(const UnicodeSet& s) {
- UErrorCode ec = U_ZERO_ERROR;
+ {
+ UnicodeSet t(s);
+ checkEqual(s, t, "copy ct");
+ }
- UnicodeSet t(s);
- checkEqual(s, t, "copy ct");
+ {
+ UnicodeSet t(0xabcd, 0xdef0); // dummy contents should be overwritten
+ t = s;
+ checkEqual(s, t, "operator=");
+ }
- t = s;
- checkEqual(s, t, "operator=");
+ {
+ UnicodeSet t;
+ copyWithIterator(t, s, FALSE);
+ checkEqual(s, t, "iterator roundtrip");
+ }
- copyWithIterator(t, s, FALSE);
- checkEqual(s, t, "iterator roundtrip");
+ {
+ UnicodeSet t;
+ copyWithIterator(t, s, TRUE); // try range
+ checkEqual(s, t, "iterator roundtrip");
+ }
- copyWithIterator(t, s, TRUE); // try range
- checkEqual(s, t, "iterator roundtrip");
-
- UnicodeString pat; s.toPattern(pat, FALSE);
- t.applyPattern(pat, ec);
- if (U_FAILURE(ec)) {
- errln("FAIL: applyPattern");
- return;
- } else {
- checkEqual(s, t, "toPattern(false)");
+ {
+ UnicodeSet t;
+ UnicodeString pat;
+ UErrorCode ec = U_ZERO_ERROR;
+ s.toPattern(pat, FALSE);
+ t.applyPattern(pat, ec);
+ if (U_FAILURE(ec)) {
+ errln("FAIL: toPattern(escapeUnprintable=FALSE), applyPattern - %s", u_errorName(ec));
+ return;
+ } else {
+ checkEqual(s, t, "toPattern(false)");
+ }
}
-
- s.toPattern(pat, TRUE);
- t.applyPattern(pat, ec);
- if (U_FAILURE(ec)) {
- errln("FAIL: applyPattern");
- return;
- } else {
- checkEqual(s, t, "toPattern(true)");
+
+ {
+ UnicodeSet t;
+ UnicodeString pat;
+ UErrorCode ec = U_ZERO_ERROR;
+ s.toPattern(pat, TRUE);
+ t.applyPattern(pat, ec);
+ if (U_FAILURE(ec)) {
+ errln("FAIL: toPattern(escapeUnprintable=TRUE), applyPattern - %s", u_errorName(ec));
+ return;
+ } else {
+ checkEqual(s, t, "toPattern(true)");
+ }
}
}
-
+
+void UnicodeSetTest::checkSerializeRoundTrip(const UnicodeSet& t, UErrorCode &status) {
+ if(U_FAILURE(status)) return;
+ int32_t len = t.serialize(serializeBuffer.getAlias(), serializeBuffer.getCapacity(), status);
+ if(status == U_BUFFER_OVERFLOW_ERROR) {
+ status = U_ZERO_ERROR;
+ serializeBuffer.resize(len);
+ len = t.serialize(serializeBuffer.getAlias(), serializeBuffer.getCapacity(), status);
+ // let 2nd error stand
+ }
+ if(U_FAILURE(status)) {
+ errln("checkSerializeRoundTrip: error %s serializing buffer\n", u_errorName(status));
+ return;
+ }
+ UnicodeSet deserialized(serializeBuffer.getAlias(), len, UnicodeSet::kSerialized, status);
+ if(U_FAILURE(status)) {
+ errln("checkSerializeRoundTrip: error %s deserializing buffer: buf %p len %d, original %d\n", u_errorName(status), serializeBuffer.getAlias(), len, t.getRangeCount());
+ return;
+ }
+
+ checkEqual(t, deserialized, "Set was unequal when deserialized");
+}
+
void UnicodeSetTest::copyWithIterator(UnicodeSet& t, const UnicodeSet& s, UBool withRange) {
t.clear();
UnicodeSetIterator it(s);
}
UBool UnicodeSetTest::checkEqual(const UnicodeSet& s, const UnicodeSet& t, const char* message) {
+ assertEquals(UnicodeString("RangeCount: ","") + message, s.getRangeCount(), t.getRangeCount());
+ assertEquals(UnicodeString("size: ","") + message, s.size(), t.size());
UnicodeString source; s.toPattern(source, TRUE);
UnicodeString result; t.toPattern(result, TRUE);
if (s != t) {
const UnicodeString *s;
char *s8=utf8;
int32_t length8, utf8Count=0;
- while(iter.nextRange() && stringsLength<LENGTHOF(strings)) {
+ while(iter.nextRange() && stringsLength<UPRV_LENGTHOF(strings)) {
if(iter.isString()) {
// Store the pointer to the set's string element
// which we happen to know is a stable pointer.
char utf8[1024];
int32_t utf8Lengths[20];
-
- int32_t nextStringIndex;
- int32_t nextUTF8Start;
};
class UnicodeSetWithStringsIterator {
UChar32 c;
int32_t start=0, prev;
while((prev=start)<length) {
- U8_NEXT(s, start, length, c);
- if(c<0) {
- c=0xfffd;
- }
+ U8_NEXT_OR_FFFD(s, start, length, c);
if(realSet.contains(c)!=spanCondition) {
break;
}
UChar32 c;
int32_t start, next;
for(start=next=0; start<length;) {
- U8_NEXT(s, next, length, c);
- if(c<0) {
- c=0xfffd;
- }
+ U8_NEXT_OR_FFFD(s, next, length, c);
if(realSet.contains(c)) {
break;
}
UChar32 c;
int32_t start, next, maxSpanLimit=0;
for(start=next=0; start<length;) {
- U8_NEXT(s, next, length, c);
- if(c<0) {
- c=0xfffd;
- }
+ U8_NEXT_OR_FFFD(s, next, length, c);
if(!realSet.contains(c)) {
next=start; // Do not span this single, not-contained code point.
}
UChar32 c;
int32_t prev=length;
do {
- U8_PREV(s, 0, length, c);
- if(c<0) {
- c=0xfffd;
- }
+ U8_PREV_OR_FFFD(s, 0, length, c);
if(realSet.contains(c)!=spanCondition) {
break;
}
UChar32 c;
int32_t prev=length;
do {
- U8_PREV(s, 0, length, c);
- if(c<0) {
- c=0xfffd;
- }
+ U8_PREV_OR_FFFD(s, 0, length, c);
if(realSet.contains(c)) {
break;
}
UChar32 c;
int32_t prev=length, minSpanStart=length;
do {
- U8_PREV(s, 0, length, c);
- if(c<0) {
- c=0xfffd;
- }
+ U8_PREV_OR_FFFD(s, 0, length, c);
if(!realSet.contains(c)) {
length=prev; // Do not span this single, not-contained code point.
}
s, length, isUTF16,
whichSpans,
type, typeName,
- limits, LENGTHOF(limits), expectCount);
+ limits, UPRV_LENGTHOF(limits), expectCount);
if(typeName[0]==0) {
break; // All types tried.
}
}
if(expectCount<0) {
expectCount=limitsCount;
- if(limitsCount>LENGTHOF(limits)) {
+ if(limitsCount>UPRV_LENGTHOF(limits)) {
errln("FAIL: %s[0x%lx].%s.%s span count=%ld > %ld capacity - too many spans",
- testName, (long)index, setNames[i], typeName, (long)limitsCount, (long)LENGTHOF(limits));
+ testName, (long)index, setNames[i], typeName, (long)limitsCount, (long)UPRV_LENGTHOF(limits));
return;
}
memcpy(expectLimits, limits, limitsCount*4);
UChar32 c, first;
for(first=c=0;; c=nextCodePoint(c)) {
- if(c>0x10ffff || length>(LENGTHOF(s)-U16_MAX_LENGTH)) {
+ if(c>0x10ffff || length>(UPRV_LENGTHOF(s)-U16_MAX_LENGTH)) {
localWhichSpans=whichSpans;
if(stringContainsUnpairedSurrogate(s, length) && inconsistentSurrogates) {
localWhichSpans&=~SPAN_UTF8;
return;
}
testSpan(sets, s, -1, TRUE, (whichSpans&~SPAN_UTF8), testName, 0);
- testSpan(sets, s, LENGTHOF(s)-1, TRUE, (whichSpans&~SPAN_UTF8), testName, 1);
+ testSpan(sets, s, UPRV_LENGTHOF(s)-1, TRUE, (whichSpans&~SPAN_UTF8), testName, 1);
}
void UnicodeSetTest::testSpanUTF8String(const UnicodeSetWithStrings *sets[4], uint32_t whichSpans, const char *testName) {
return;
}
testSpan(sets, s, -1, FALSE, (whichSpans&~SPAN_UTF16), testName, 0);
- testSpan(sets, s, LENGTHOF(s)-1, FALSE, (whichSpans&~SPAN_UTF16), testName, 1);
+ testSpan(sets, s, UPRV_LENGTHOF(s)-1, FALSE, (whichSpans&~SPAN_UTF16), testName, 1);
}
// Take a set of span options and multiply them so that
char *testNameLimit=testName;
int32_t i, j;
- for(i=0; i<LENGTHOF(testdata); ++i) {
+ for(i=0; i<UPRV_LENGTHOF(testdata); ++i) {
const char *s=testdata[i];
if(s[0]=='[') {
// Create new test sets from this pattern.
string16=UNICODE_STRING_SIMPLE("byayaxya");
const UChar *s16=string16.getBuffer();
int32_t length16=string16.length();
+ (void)length16; // Suppress set but not used warning.
if( set.span(s16, 8, USET_SPAN_NOT_CONTAINED)!=4 ||
set.span(s16, 7, USET_SPAN_NOT_CONTAINED)!=4 ||
set.span(s16, 6, USET_SPAN_NOT_CONTAINED)!=4 ||
errln("FAIL: UnicodeSet(%s).spanBack(while longest match) returns the wrong value", pattern);
}
}
+
+/**
+ * Including collationroot.h fails here with
+1>c:\Program Files (x86)\Microsoft SDKs\Windows\v7.0A\include\driverspecs.h(142): error C2008: '$' : unexpected in macro definition
+ * .. so, we skip this test on Windows.
+ *
+ * the cause is that intltest builds with /Za which disables language extensions - which means
+ * windows header files can't be used.
+ */
+#if !UCONFIG_NO_COLLATION && !U_PLATFORM_HAS_WIN32_API
+#include "collationroot.h"
+#include "collationtailoring.h"
+#endif
+
+void UnicodeSetTest::TestUCAUnsafeBackwards() {
+#if U_PLATFORM_HAS_WIN32_API
+ infoln("Skipping TestUCAUnsafeBackwards() - can't include collationroot.h on Windows without language extensions!");
+#elif !UCONFIG_NO_COLLATION
+ UErrorCode errorCode = U_ZERO_ERROR;
+
+ // Get the unsafeBackwardsSet
+ const CollationCacheEntry *rootEntry = CollationRoot::getRootCacheEntry(errorCode);
+ if(U_FAILURE(errorCode)) {
+ dataerrln("FAIL: %s getting root cache entry", u_errorName(errorCode));
+ return;
+ }
+ //const UVersionInfo &version = rootEntry->tailoring->version;
+ const UnicodeSet *unsafeBackwardSet = rootEntry->tailoring->unsafeBackwardSet;
+
+ checkSerializeRoundTrip(*unsafeBackwardSet, errorCode);
+
+ if(!logKnownIssue("11891","UnicodeSet fails to round trip on CollationRoot...unsafeBackwards set")) {
+ // simple test case
+ // TODO(ticket #11891): Simplify this test function to this simple case. Rename it appropriately.
+ // TODO(ticket #11891): Port test to Java. Is this a bug there, too?
+ UnicodeSet surrogates;
+ surrogates.add(0xd83a); // a lead surrogate
+ surrogates.add(0xdc00, 0xdfff); // a range of trail surrogates
+ UnicodeString pat;
+ surrogates.toPattern(pat, FALSE); // bad: [ 0xd83a, 0xdc00, 0x2d, 0xdfff ]
+ // TODO: Probably fix either UnicodeSet::_generatePattern() or _appendToPat()
+ // so that at least one type of surrogate code points are escaped,
+ // or (minimally) so that adjacent lead+trail surrogate code points are escaped.
+ errorCode = U_ZERO_ERROR;
+ UnicodeSet s2;
+ s2.applyPattern(pat, errorCode); // looks like invalid range [ 0x1e800, 0x2d, 0xdfff ]
+ if(U_FAILURE(errorCode)) {
+ errln("FAIL: surrogates to/from pattern - %s", u_errorName(errorCode));
+ } else {
+ checkEqual(surrogates, s2, "surrogates to/from pattern");
+ }
+ // This occurs in the UCA unsafe-backwards set.
+ checkRoundTrip(*unsafeBackwardSet);
+ }
+#endif
+}
projects / apple / icu.git / blobdiff