organicmaps/icu
13
Fork 0
mirror of https://github.com/unicode-org/icu.git synced 2025-04-06 14:05:32 +00:00
Code Issues 1 Projects Releases Wiki Activity Actions 9

ICU-4790 more spoof detection

Browse source 
X-SVN-Rev: 25921
This commit is contained in:
Andy Heninger 2009-04-27 06:53:38 +00:00
parent c480d63da5
commit 25be8362b9
3 changed files with 104 additions and 34 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

107
icu4c/source/i18n/uspoof.cpp
View file

@ -234,48 +234,89 @@ uspoof_check(const USpoofChecker *sc,
}
}
// TODO: add USPOOF_INVISIBLE check
if (This->fChecks & (USPOOF_WHOLE_SCRIPT_CONFUSABLE | USPOOF_MIXED_SCRIPT_CONFUSABLE)) {
// The basic test is the same for both whole and mixed script confusables.
// Compute the set of scripts that every input character has a confusable in.
// For this computation an input character is always considered to be
// confusable with itself in its own script.
// If the number of such scripts is two or more, and the input consisted of
// characters all from a single script, we have a whole script confusable.
// (The two scripts will be the original script and the one that is confusable)
// If the number of such scripts >= one, and the original input contained characters from
// more than one script, we have a mixed script confusable. (We can transform
// some of the characters, and end up with a visually similar string all in
// one script.)
if (This->fChecks &
(USPOOF_WHOLE_SCRIPT_CONFUSABLE | USPOOF_MIXED_SCRIPT_CONFUSABLE | USPOOF_INVISIBLE)) {
// These are the checks that need to be done on NFKD input
NFKDBuffer normalizedInput(text, length, *status);
const UChar *nfkdText = normalizedInput.getBuffer();
int32_t nfkdLength = normalizedInput.getLength();
if (scriptCount == -1) {
int32_t t;
scriptCount = This->scriptScan(text, length, t, *status);
if (This->fChecks & USPOOF_INVISIBLE) {
// scan for more than one occurence of the same non-spacing mark
// in a sequence of non-spacing marks.
int32_t i;
UChar32 c;
UChar32 firstNonspacingMark = 0;
UBool haveMultipleMarks = FALSE;
UnicodeSet marksSeenSoFar; // Set of combining marks in a single combining sequence.
for (i=0; i<length ;) {
U16_NEXT(nfkdText, i, nfkdLength, c);
if (u_charType(c) != U_NON_SPACING_MARK) {
firstNonspacingMark = 0;
if (haveMultipleMarks) {
marksSeenSoFar.clear();
haveMultipleMarks = FALSE;
}
continue;
}
if (firstNonspacingMark == 0) {
firstNonspacingMark = c;
continue;
}
if (!haveMultipleMarks) {
marksSeenSoFar.add(firstNonspacingMark);
haveMultipleMarks = TRUE;
}
if (marksSeenSoFar.contains(c)) {
// report the error, and stop scanning.
// No need to find more than the first failure.
result |= USPOOF_INVISIBLE;
failPos = i;
break;
}
marksSeenSoFar.add(c);
}
}
ScriptSet scripts;
This->wholeScriptCheck(nfkdText, nfkdLength, &scripts, *status);
int32_t confusableScriptCount = scripts.countMembers();
//printf("confusableScriptCount = %d\n", confusableScriptCount);
if (This->fChecks & (USPOOF_WHOLE_SCRIPT_CONFUSABLE | USPOOF_MIXED_SCRIPT_CONFUSABLE)) {
// The basic test is the same for both whole and mixed script confusables.
// Compute the set of scripts that every input character has a confusable in.
// For this computation an input character is always considered to be
// confusable with itself in its own script.
// If the number of such scripts is two or more, and the input consisted of
// characters all from a single script, we have a whole script confusable.
// (The two scripts will be the original script and the one that is confusable)
// If the number of such scripts >= one, and the original input contained characters from
// more than one script, we have a mixed script confusable. (We can transform
// some of the characters, and end up with a visually similar string all in
// one script.)
if (scriptCount == -1) {
int32_t t;
scriptCount = This->scriptScan(text, length, t, *status);
}
ScriptSet scripts;
This->wholeScriptCheck(nfkdText, nfkdLength, &scripts, *status);
int32_t confusableScriptCount = scripts.countMembers();
//printf("confusableScriptCount = %d\n", confusableScriptCount);
if ((This->fChecks & USPOOF_WHOLE_SCRIPT_CONFUSABLE) &&
confusableScriptCount >= 2 &&
scriptCount == 1) {
result |= USPOOF_WHOLE_SCRIPT_CONFUSABLE;
}
if ((This->fChecks & USPOOF_WHOLE_SCRIPT_CONFUSABLE) &&
confusableScriptCount >= 2 &&
scriptCount == 1) {
result |= USPOOF_WHOLE_SCRIPT_CONFUSABLE;
}
if ((This->fChecks & USPOOF_MIXED_SCRIPT_CONFUSABLE) &&
confusableScriptCount >= 1 &&
scriptCount > 1) {
result |= USPOOF_MIXED_SCRIPT_CONFUSABLE;
if ((This->fChecks & USPOOF_MIXED_SCRIPT_CONFUSABLE) &&
confusableScriptCount >= 1 &&
scriptCount > 1) {
result |= USPOOF_MIXED_SCRIPT_CONFUSABLE;
}
}
}
if (position != NULL && failPos != 0x7fffffff) {
*position = failPos;
}

29
icu4c/source/test/intltest/itspoof.cpp
View file

@ -73,6 +73,12 @@ void IntlTestSpoof::runIndexedTest( int32_t index, UBool exec, const char* &name
testAreConfusable();
}
break;
case 3:
name = "TestInvisible";
if (exec) {
testInvisible();
}
break;
default: name=""; break;
}
}
@ -206,7 +212,6 @@ void IntlTestSpoof::checkSkeleton(const USpoofChecker *sc, uint32_t type,
}
void IntlTestSpoof::testAreConfusable() {
UErrorCode status = U_ZERO_ERROR;
TEST_SETUP
UnicodeString s1("A long string that will overflow stack buffers. A long string that will overflow stack buffers. "
"A long string that will overflow stack buffers. A long string that will overflow stack buffers. ");
@ -218,5 +223,27 @@ void IntlTestSpoof::testAreConfusable() {
TEST_TEARDOWN;
}
void IntlTestSpoof::testInvisible() {
TEST_SETUP
UnicodeString s = UnicodeString("abcd\\u0301ef").unescape();
int32_t position = -42;
TEST_ASSERT_EQ(0, uspoof_checkUnicodeString(sc, s, &position, &status));
TEST_ASSERT_SUCCESS(status);
TEST_ASSERT(position == -42);
UnicodeString s2 = UnicodeString("abcd\\u0301\\u0302\\u0301ef").unescape();
TEST_ASSERT_EQ(USPOOF_INVISIBLE, uspoof_checkUnicodeString(sc, s2, &position, &status));
TEST_ASSERT_SUCCESS(status);
TEST_ASSERT_EQ(7, position);
// Tow acute accents, one from the composed a with acute accent, \u00e1,
// and one separate.
position = -42;
UnicodeString s3 = UnicodeString("abcd\\u00e1\\u0301xyz").unescape();
TEST_ASSERT_EQ(USPOOF_INVISIBLE, uspoof_checkUnicodeString(sc, s3, &position, &status));
TEST_ASSERT_SUCCESS(status);
TEST_ASSERT_EQ(7, position);
TEST_TEARDOWN;
}
#endif /* #if !UCONFIG_NO_SPOOF_DETECT*/

2
icu4c/source/test/intltest/itspoof.h
View file

@ -29,6 +29,8 @@ public:
void testAreConfusable();
void testInvisible();
// Internal function to run a single skeleton test case.
void checkSkeleton(const USpoofChecker *sc, uint32_t flags,
const char *input, const char *expected, int32_t lineNum);