ConfusabledataBuilder::~ConfusabledataBuilder() { uprv_free(fInput); uregex_close(fParseLine); uregex_close(fParseHexNum); uhash_close(fTable); delete fKeySet; delete fKeyVec; delete fStringTable; delete fValueVec; delete stringPool; }
static void _close(JNIEnv* env, jclass clazz, RegExData* data) { if (data->regex != NULL) { uregex_close(data->regex); } if (data->text != &EMPTY_STRING) { delete[] data->text; } free(data); }
static void TestBug4315(void) { UErrorCode theICUError = U_ZERO_ERROR; URegularExpression *theRegEx; UChar *textBuff; const char *thePattern; UChar theString[100]; UChar *destFields[24]; int32_t neededLength1; int32_t neededLength2; int32_t wordCount = 0; int32_t destFieldsSize = 24; thePattern = "ck "; u_uastrcpy(theString, "The quick brown fox jumped over the slow black turtle."); /* open a regex */ theRegEx = uregex_openC(thePattern, 0, NULL, &theICUError); TEST_ASSERT_SUCCESS(theICUError); /* set the input string */ uregex_setText(theRegEx, theString, u_strlen(theString), &theICUError); TEST_ASSERT_SUCCESS(theICUError); /* split */ /*explicitly pass NULL and 0 to force the overflow error -> this is where the * error occurs! */ wordCount = uregex_split(theRegEx, NULL, 0, &neededLength1, destFields, destFieldsSize, &theICUError); TEST_ASSERT(theICUError == U_BUFFER_OVERFLOW_ERROR); TEST_ASSERT(wordCount==3); if(theICUError == U_BUFFER_OVERFLOW_ERROR) { theICUError = U_ZERO_ERROR; textBuff = (UChar *) malloc(sizeof(UChar) * (neededLength1 + 1)); wordCount = uregex_split(theRegEx, textBuff, neededLength1+1, &neededLength2, destFields, destFieldsSize, &theICUError); TEST_ASSERT(wordCount==3); TEST_ASSERT_SUCCESS(theICUError); TEST_ASSERT(neededLength1 == neededLength2); TEST_ASSERT_STRING("The qui", destFields[0], TRUE); TEST_ASSERT_STRING("brown fox jumped over the slow bla", destFields[1], TRUE); TEST_ASSERT_STRING("turtle.", destFields[2], TRUE); TEST_ASSERT(destFields[3] == NULL); free(textBuff); } uregex_close(theRegEx); }
/* ** Function to delete compiled regexp objects. Registered as ** a destructor function with sqlite3_set_auxdata(). */ static void icuRegexpDelete(void *p){ URegularExpression *pExpr = (URegularExpression *)p; uregex_close(pExpr); }
static void TestRegexCAPI(void) { UErrorCode status = U_ZERO_ERROR; URegularExpression *re; UChar pat[200]; UChar *minus1; memset(&minus1, -1, sizeof(minus1)); /* Mimimalist open/close */ u_uastrncpy(pat, "abc*", sizeof(pat)/2); re = uregex_open(pat, -1, 0, 0, &status); TEST_ASSERT_SUCCESS(status); uregex_close(re); /* Open with all flag values set */ status = U_ZERO_ERROR; re = uregex_open(pat, -1, UREGEX_CASE_INSENSITIVE | UREGEX_COMMENTS | UREGEX_DOTALL | UREGEX_MULTILINE | UREGEX_UWORD, 0, &status); TEST_ASSERT_SUCCESS(status); uregex_close(re); /* Open with an invalid flag */ status = U_ZERO_ERROR; re = uregex_open(pat, -1, 0x40000000, 0, &status); TEST_ASSERT(status == U_REGEX_INVALID_FLAG); uregex_close(re); /* openC with an invalid parameter */ status = U_ZERO_ERROR; re = uregex_openC(NULL, UREGEX_CASE_INSENSITIVE | UREGEX_COMMENTS | UREGEX_DOTALL | UREGEX_MULTILINE | UREGEX_UWORD, 0, &status); TEST_ASSERT(status == U_ILLEGAL_ARGUMENT_ERROR && re == NULL); /* openC with an invalid parameter */ status = U_USELESS_COLLATOR_ERROR; re = uregex_openC(NULL, UREGEX_CASE_INSENSITIVE | UREGEX_COMMENTS | UREGEX_DOTALL | UREGEX_MULTILINE | UREGEX_UWORD, 0, &status); TEST_ASSERT(status == U_USELESS_COLLATOR_ERROR && re == NULL); /* openC open from a C string */ { const UChar *p; int32_t len; status = U_ZERO_ERROR; re = uregex_openC("abc*", 0, 0, &status); TEST_ASSERT_SUCCESS(status); p = uregex_pattern(re, &len, &status); TEST_ASSERT_SUCCESS(status); /* The TEST_ASSERT_SUCCESS above should change too... */ if(U_SUCCESS(status)) { u_uastrncpy(pat, "abc*", sizeof(pat)/2); TEST_ASSERT(u_strcmp(pat, p) == 0); TEST_ASSERT(len==(int32_t)strlen("abc*")); } uregex_close(re); /* TODO: Open with ParseError parameter */ } /* * clone */ { URegularExpression *clone1; URegularExpression *clone2; URegularExpression *clone3; UChar testString1[30]; UChar testString2[30]; UBool result; status = U_ZERO_ERROR; re = uregex_openC("abc*", 0, 0, &status); TEST_ASSERT_SUCCESS(status); clone1 = uregex_clone(re, &status); TEST_ASSERT_SUCCESS(status); TEST_ASSERT(clone1 != NULL); status = U_ZERO_ERROR; clone2 = uregex_clone(re, &status); TEST_ASSERT_SUCCESS(status); TEST_ASSERT(clone2 != NULL); uregex_close(re); status = U_ZERO_ERROR; clone3 = uregex_clone(clone2, &status); TEST_ASSERT_SUCCESS(status); TEST_ASSERT(clone3 != NULL); u_uastrncpy(testString1, "abcccd", sizeof(pat)/2); u_uastrncpy(testString2, "xxxabcccd", sizeof(pat)/2); status = U_ZERO_ERROR; uregex_setText(clone1, testString1, -1, &status); TEST_ASSERT_SUCCESS(status); result = uregex_lookingAt(clone1, 0, &status); TEST_ASSERT_SUCCESS(status); TEST_ASSERT(result==TRUE); status = U_ZERO_ERROR; uregex_setText(clone2, testString2, -1, &status); TEST_ASSERT_SUCCESS(status); result = uregex_lookingAt(clone2, 0, &status); TEST_ASSERT_SUCCESS(status); TEST_ASSERT(result==FALSE); result = uregex_find(clone2, 0, &status); TEST_ASSERT_SUCCESS(status); TEST_ASSERT(result==TRUE); uregex_close(clone1); uregex_close(clone2); uregex_close(clone3); } /* * pattern() */ { const UChar *resultPat; int32_t resultLen; u_uastrncpy(pat, "hello", sizeof(pat)/2); status = U_ZERO_ERROR; re = uregex_open(pat, -1, 0, NULL, &status); resultPat = uregex_pattern(re, &resultLen, &status); TEST_ASSERT_SUCCESS(status); /* The TEST_ASSERT_SUCCESS above should change too... */ if (U_SUCCESS(status)) { TEST_ASSERT(resultLen == -1); TEST_ASSERT(u_strcmp(resultPat, pat) == 0); } uregex_close(re); status = U_ZERO_ERROR; re = uregex_open(pat, 3, 0, NULL, &status); resultPat = uregex_pattern(re, &resultLen, &status); TEST_ASSERT_SUCCESS(status); TEST_ASSERT_SUCCESS(status); /* The TEST_ASSERT_SUCCESS above should change too... */ if (U_SUCCESS(status)) { TEST_ASSERT(resultLen == 3); TEST_ASSERT(u_strncmp(resultPat, pat, 3) == 0); TEST_ASSERT(u_strlen(resultPat) == 3); } uregex_close(re); } /* * flags() */ { int32_t t; status = U_ZERO_ERROR; re = uregex_open(pat, -1, 0, NULL, &status); t = uregex_flags(re, &status); TEST_ASSERT_SUCCESS(status); TEST_ASSERT(t == 0); uregex_close(re); status = U_ZERO_ERROR; re = uregex_open(pat, -1, 0, NULL, &status); t = uregex_flags(re, &status); TEST_ASSERT_SUCCESS(status); TEST_ASSERT(t == 0); uregex_close(re); status = U_ZERO_ERROR; re = uregex_open(pat, -1, UREGEX_CASE_INSENSITIVE | UREGEX_DOTALL, NULL, &status); t = uregex_flags(re, &status); TEST_ASSERT_SUCCESS(status); TEST_ASSERT(t == (UREGEX_CASE_INSENSITIVE | UREGEX_DOTALL)); uregex_close(re); } /* * setText() and lookingAt() */ { UChar text1[50]; UChar text2[50]; UBool result; u_uastrncpy(text1, "abcccd", sizeof(text1)/2); u_uastrncpy(text2, "abcccxd", sizeof(text2)/2); status = U_ZERO_ERROR; u_uastrncpy(pat, "abc*d", sizeof(pat)/2); re = uregex_open(pat, -1, 0, NULL, &status); TEST_ASSERT_SUCCESS(status); /* Operation before doing a setText should fail... */ status = U_ZERO_ERROR; uregex_lookingAt(re, 0, &status); TEST_ASSERT( status== U_REGEX_INVALID_STATE); status = U_ZERO_ERROR; uregex_setText(re, text1, -1, &status); result = uregex_lookingAt(re, 0, &status); TEST_ASSERT(result == TRUE); TEST_ASSERT_SUCCESS(status); status = U_ZERO_ERROR; uregex_setText(re, text2, -1, &status); result = uregex_lookingAt(re, 0, &status); TEST_ASSERT(result == FALSE); TEST_ASSERT_SUCCESS(status); status = U_ZERO_ERROR; uregex_setText(re, text1, -1, &status); result = uregex_lookingAt(re, 0, &status); TEST_ASSERT(result == TRUE); TEST_ASSERT_SUCCESS(status); status = U_ZERO_ERROR; uregex_setText(re, text1, 5, &status); result = uregex_lookingAt(re, 0, &status); TEST_ASSERT(result == FALSE); TEST_ASSERT_SUCCESS(status); status = U_ZERO_ERROR; uregex_setText(re, text1, 6, &status); result = uregex_lookingAt(re, 0, &status); TEST_ASSERT(result == TRUE); TEST_ASSERT_SUCCESS(status); uregex_close(re); } /* * getText() */ { UChar text1[50]; UChar text2[50]; const UChar *result; int32_t textLength; u_uastrncpy(text1, "abcccd", sizeof(text1)/2); u_uastrncpy(text2, "abcccxd", sizeof(text2)/2); status = U_ZERO_ERROR; u_uastrncpy(pat, "abc*d", sizeof(pat)/2); re = uregex_open(pat, -1, 0, NULL, &status); uregex_setText(re, text1, -1, &status); result = uregex_getText(re, &textLength, &status); TEST_ASSERT(result == text1); TEST_ASSERT(textLength == -1); TEST_ASSERT_SUCCESS(status); status = U_ZERO_ERROR; uregex_setText(re, text2, 7, &status); result = uregex_getText(re, &textLength, &status); TEST_ASSERT(result == text2); TEST_ASSERT(textLength == 7); TEST_ASSERT_SUCCESS(status); status = U_ZERO_ERROR; uregex_setText(re, text2, 4, &status); result = uregex_getText(re, &textLength, &status); TEST_ASSERT(result == text2); TEST_ASSERT(textLength == 4); TEST_ASSERT_SUCCESS(status); uregex_close(re); } /* * matches() */ { UChar text1[50]; UBool result; int len; UChar nullString[] = {0,0,0}; u_uastrncpy(text1, "abcccde", sizeof(text1)/2); status = U_ZERO_ERROR; u_uastrncpy(pat, "abc*d", sizeof(pat)/2); re = uregex_open(pat, -1, 0, NULL, &status); uregex_setText(re, text1, -1, &status); result = uregex_matches(re, 0, &status); TEST_ASSERT(result == FALSE); TEST_ASSERT_SUCCESS(status); status = U_ZERO_ERROR; uregex_setText(re, text1, 6, &status); result = uregex_matches(re, 0, &status); TEST_ASSERT(result == TRUE); TEST_ASSERT_SUCCESS(status); status = U_ZERO_ERROR; uregex_setText(re, text1, 6, &status); result = uregex_matches(re, 1, &status); TEST_ASSERT(result == FALSE); TEST_ASSERT_SUCCESS(status); uregex_close(re); status = U_ZERO_ERROR; re = uregex_openC(".?", 0, NULL, &status); uregex_setText(re, text1, -1, &status); len = u_strlen(text1); result = uregex_matches(re, len, &status); TEST_ASSERT(result == TRUE); TEST_ASSERT_SUCCESS(status); status = U_ZERO_ERROR; uregex_setText(re, nullString, -1, &status); TEST_ASSERT_SUCCESS(status); result = uregex_matches(re, 0, &status); TEST_ASSERT(result == TRUE); TEST_ASSERT_SUCCESS(status); uregex_close(re); } /* * lookingAt() Used in setText test. */ /* * find(), findNext, start, end, reset */ { UChar text1[50]; UBool result; u_uastrncpy(text1, "012rx5rx890rxrx...", sizeof(text1)/2); status = U_ZERO_ERROR; re = uregex_openC("rx", 0, NULL, &status); uregex_setText(re, text1, -1, &status); result = uregex_find(re, 0, &status); TEST_ASSERT(result == TRUE); TEST_ASSERT(uregex_start(re, 0, &status) == 3); TEST_ASSERT(uregex_end(re, 0, &status) == 5); TEST_ASSERT_SUCCESS(status); result = uregex_find(re, 9, &status); TEST_ASSERT(result == TRUE); TEST_ASSERT(uregex_start(re, 0, &status) == 11); TEST_ASSERT(uregex_end(re, 0, &status) == 13); TEST_ASSERT_SUCCESS(status); result = uregex_find(re, 14, &status); TEST_ASSERT(result == FALSE); TEST_ASSERT_SUCCESS(status); status = U_ZERO_ERROR; uregex_reset(re, 0, &status); result = uregex_findNext(re, &status); TEST_ASSERT(result == TRUE); TEST_ASSERT(uregex_start(re, 0, &status) == 3); TEST_ASSERT(uregex_end(re, 0, &status) == 5); TEST_ASSERT_SUCCESS(status); result = uregex_findNext(re, &status); TEST_ASSERT(result == TRUE); TEST_ASSERT(uregex_start(re, 0, &status) == 6); TEST_ASSERT(uregex_end(re, 0, &status) == 8); TEST_ASSERT_SUCCESS(status); status = U_ZERO_ERROR; uregex_reset(re, 12, &status); result = uregex_findNext(re, &status); TEST_ASSERT(result == TRUE); TEST_ASSERT(uregex_start(re, 0, &status) == 13); TEST_ASSERT(uregex_end(re, 0, &status) == 15); TEST_ASSERT_SUCCESS(status); result = uregex_findNext(re, &status); TEST_ASSERT(result == FALSE); TEST_ASSERT_SUCCESS(status); uregex_close(re); } /* * groupCount */ { int32_t result; status = U_ZERO_ERROR; re = uregex_openC("abc", 0, NULL, &status); result = uregex_groupCount(re, &status); TEST_ASSERT_SUCCESS(status); TEST_ASSERT(result == 0); uregex_close(re); status = U_ZERO_ERROR; re = uregex_openC("abc(def)(ghi(j))", 0, NULL, &status); result = uregex_groupCount(re, &status); TEST_ASSERT_SUCCESS(status); TEST_ASSERT(result == 3); uregex_close(re); } /* * group() */ { UChar text1[80]; UChar buf[80]; UBool result; int32_t resultSz; u_uastrncpy(text1, "noise abc interior def, and this is off the end", sizeof(text1)/2); status = U_ZERO_ERROR; re = uregex_openC("abc(.*?)def", 0, NULL, &status); TEST_ASSERT_SUCCESS(status); uregex_setText(re, text1, -1, &status); result = uregex_find(re, 0, &status); TEST_ASSERT(result==TRUE); /* Capture Group 0, the full match. Should succeed. */ status = U_ZERO_ERROR; resultSz = uregex_group(re, 0, buf, sizeof(buf)/2, &status); TEST_ASSERT_SUCCESS(status); TEST_ASSERT_STRING("abc interior def", buf, TRUE); TEST_ASSERT(resultSz == (int32_t)strlen("abc interior def")); /* Capture group #1. Should succeed. */ status = U_ZERO_ERROR; resultSz = uregex_group(re, 1, buf, sizeof(buf)/2, &status); TEST_ASSERT_SUCCESS(status); TEST_ASSERT_STRING(" interior ", buf, TRUE); TEST_ASSERT(resultSz == (int32_t)strlen(" interior ")); /* Capture group out of range. Error. */ status = U_ZERO_ERROR; uregex_group(re, 2, buf, sizeof(buf)/2, &status); TEST_ASSERT(status == U_INDEX_OUTOFBOUNDS_ERROR); /* NULL buffer, pure pre-flight */ status = U_ZERO_ERROR; resultSz = uregex_group(re, 0, NULL, 0, &status); TEST_ASSERT(status == U_BUFFER_OVERFLOW_ERROR); TEST_ASSERT(resultSz == (int32_t)strlen("abc interior def")); /* Too small buffer, truncated string */ status = U_ZERO_ERROR; memset(buf, -1, sizeof(buf)); resultSz = uregex_group(re, 0, buf, 5, &status); TEST_ASSERT(status == U_BUFFER_OVERFLOW_ERROR); TEST_ASSERT_STRING("abc i", buf, FALSE); TEST_ASSERT(buf[5] == (UChar)0xffff); TEST_ASSERT(resultSz == (int32_t)strlen("abc interior def")); /* Output string just fits buffer, no NUL term. */ status = U_ZERO_ERROR; resultSz = uregex_group(re, 0, buf, (int32_t)strlen("abc interior def"), &status); TEST_ASSERT(status == U_STRING_NOT_TERMINATED_WARNING); TEST_ASSERT_STRING("abc interior def", buf, FALSE); TEST_ASSERT(resultSz == (int32_t)strlen("abc interior def")); TEST_ASSERT(buf[strlen("abc interior def")] == (UChar)0xffff); uregex_close(re); } /* * Regions */ /* SetRegion(), getRegion() do something */ TEST_SETUP(".*", "0123456789ABCDEF", 0) UChar resultString[40]; TEST_ASSERT(uregex_regionStart(re, &status) == 0); TEST_ASSERT(uregex_regionEnd(re, &status) == 16); uregex_setRegion(re, 3, 6, &status); TEST_ASSERT(uregex_regionStart(re, &status) == 3); TEST_ASSERT(uregex_regionEnd(re, &status) == 6); TEST_ASSERT(uregex_findNext(re, &status)); TEST_ASSERT(uregex_group(re, 0, resultString, sizeof(resultString)/2, &status) == 3) TEST_ASSERT_STRING("345", resultString, TRUE); TEST_TEARDOWN; /* find(start=-1) uses regions */ TEST_SETUP(".*", "0123456789ABCDEF", 0); uregex_setRegion(re, 4, 6, &status); TEST_ASSERT(uregex_find(re, -1, &status) == TRUE); TEST_ASSERT(uregex_start(re, 0, &status) == 4); TEST_ASSERT(uregex_end(re, 0, &status) == 6); TEST_TEARDOWN; /* find (start >=0) does not use regions */ TEST_SETUP(".*", "0123456789ABCDEF", 0); uregex_setRegion(re, 4, 6, &status); TEST_ASSERT(uregex_find(re, 0, &status) == TRUE); TEST_ASSERT(uregex_start(re, 0, &status) == 0); TEST_ASSERT(uregex_end(re, 0, &status) == 16); TEST_TEARDOWN; /* findNext() obeys regions */ TEST_SETUP(".", "0123456789ABCDEF", 0); uregex_setRegion(re, 4, 6, &status); TEST_ASSERT(uregex_findNext(re,&status) == TRUE); TEST_ASSERT(uregex_start(re, 0, &status) == 4); TEST_ASSERT(uregex_findNext(re, &status) == TRUE); TEST_ASSERT(uregex_start(re, 0, &status) == 5); TEST_ASSERT(uregex_findNext(re, &status) == FALSE); TEST_TEARDOWN; /* matches(start=-1) uses regions */ /* Also, verify that non-greedy *? succeeds in finding the full match. */ TEST_SETUP(".*?", "0123456789ABCDEF", 0); uregex_setRegion(re, 4, 6, &status); TEST_ASSERT(uregex_matches(re, -1, &status) == TRUE); TEST_ASSERT(uregex_start(re, 0, &status) == 4); TEST_ASSERT(uregex_end(re, 0, &status) == 6); TEST_TEARDOWN; /* matches (start >=0) does not use regions */ TEST_SETUP(".*?", "0123456789ABCDEF", 0); uregex_setRegion(re, 4, 6, &status); TEST_ASSERT(uregex_matches(re, 0, &status) == TRUE); TEST_ASSERT(uregex_start(re, 0, &status) == 0); TEST_ASSERT(uregex_end(re, 0, &status) == 16); TEST_TEARDOWN; /* lookingAt(start=-1) uses regions */ /* Also, verify that non-greedy *? finds the first (shortest) match. */ TEST_SETUP(".*?", "0123456789ABCDEF", 0); uregex_setRegion(re, 4, 6, &status); TEST_ASSERT(uregex_lookingAt(re, -1, &status) == TRUE); TEST_ASSERT(uregex_start(re, 0, &status) == 4); TEST_ASSERT(uregex_end(re, 0, &status) == 4); TEST_TEARDOWN; /* lookingAt (start >=0) does not use regions */ TEST_SETUP(".*?", "0123456789ABCDEF", 0); uregex_setRegion(re, 4, 6, &status); TEST_ASSERT(uregex_lookingAt(re, 0, &status) == TRUE); TEST_ASSERT(uregex_start(re, 0, &status) == 0); TEST_ASSERT(uregex_end(re, 0, &status) == 0); TEST_TEARDOWN; /* hitEnd() */ TEST_SETUP("[a-f]*", "abcdefghij", 0); TEST_ASSERT(uregex_find(re, 0, &status) == TRUE); TEST_ASSERT(uregex_hitEnd(re, &status) == FALSE); TEST_TEARDOWN; TEST_SETUP("[a-f]*", "abcdef", 0); TEST_ASSERT(uregex_find(re, 0, &status) == TRUE); TEST_ASSERT(uregex_hitEnd(re, &status) == TRUE); TEST_TEARDOWN; /* requireEnd */ TEST_SETUP("abcd", "abcd", 0); TEST_ASSERT(uregex_find(re, 0, &status) == TRUE); TEST_ASSERT(uregex_requireEnd(re, &status) == FALSE); TEST_TEARDOWN; TEST_SETUP("abcd$", "abcd", 0); TEST_ASSERT(uregex_find(re, 0, &status) == TRUE); TEST_ASSERT(uregex_requireEnd(re, &status) == TRUE); TEST_TEARDOWN; /* anchoringBounds */ TEST_SETUP("abc$", "abcdef", 0); TEST_ASSERT(uregex_hasAnchoringBounds(re, &status) == TRUE); uregex_useAnchoringBounds(re, FALSE, &status); TEST_ASSERT(uregex_hasAnchoringBounds(re, &status) == FALSE); TEST_ASSERT(uregex_find(re, -1, &status) == FALSE); uregex_useAnchoringBounds(re, TRUE, &status); uregex_setRegion(re, 0, 3, &status); TEST_ASSERT(uregex_find(re, -1, &status) == TRUE); TEST_ASSERT(uregex_end(re, 0, &status) == 3); TEST_TEARDOWN; /* Transparent Bounds */ TEST_SETUP("abc(?=def)", "abcdef", 0); TEST_ASSERT(uregex_hasTransparentBounds(re, &status) == FALSE); uregex_useTransparentBounds(re, TRUE, &status); TEST_ASSERT(uregex_hasTransparentBounds(re, &status) == TRUE); uregex_useTransparentBounds(re, FALSE, &status); TEST_ASSERT(uregex_find(re, -1, &status) == TRUE); /* No Region */ uregex_setRegion(re, 0, 3, &status); TEST_ASSERT(uregex_find(re, -1, &status) == FALSE); /* with region, opaque bounds */ uregex_useTransparentBounds(re, TRUE, &status); TEST_ASSERT(uregex_find(re, -1, &status) == TRUE); /* with region, transparent bounds */ TEST_ASSERT(uregex_end(re, 0, &status) == 3); TEST_TEARDOWN; /* * replaceFirst() */ { UChar text1[80]; UChar text2[80]; UChar replText[80]; UChar buf[80]; int32_t resultSz; u_uastrncpy(text1, "Replace xaax x1x x...x.", sizeof(text1)/2); u_uastrncpy(text2, "No match here.", sizeof(text2)/2); u_uastrncpy(replText, "<$1>", sizeof(replText)/2); status = U_ZERO_ERROR; re = uregex_openC("x(.*?)x", 0, NULL, &status); TEST_ASSERT_SUCCESS(status); /* Normal case, with match */ uregex_setText(re, text1, -1, &status); resultSz = uregex_replaceFirst(re, replText, -1, buf, sizeof(buf)/2, &status); TEST_ASSERT_SUCCESS(status); TEST_ASSERT_STRING("Replace <aa> x1x x...x.", buf, TRUE); TEST_ASSERT(resultSz == (int32_t)strlen("Replace xaax x1x x...x.")); /* No match. Text should copy to output with no changes. */ status = U_ZERO_ERROR; uregex_setText(re, text2, -1, &status); resultSz = uregex_replaceFirst(re, replText, -1, buf, sizeof(buf)/2, &status); TEST_ASSERT_SUCCESS(status); TEST_ASSERT_STRING("No match here.", buf, TRUE); TEST_ASSERT(resultSz == (int32_t)strlen("No match here.")); /* Match, output just fills buffer, no termination warning. */ status = U_ZERO_ERROR; uregex_setText(re, text1, -1, &status); memset(buf, -1, sizeof(buf)); resultSz = uregex_replaceFirst(re, replText, -1, buf, strlen("Replace <aa> x1x x...x."), &status); TEST_ASSERT(status == U_STRING_NOT_TERMINATED_WARNING); TEST_ASSERT_STRING("Replace <aa> x1x x...x.", buf, FALSE); TEST_ASSERT(resultSz == (int32_t)strlen("Replace xaax x1x x...x.")); TEST_ASSERT(buf[resultSz] == (UChar)0xffff); /* Do the replaceFirst again, without first resetting anything. * Should give the same results. */ status = U_ZERO_ERROR; memset(buf, -1, sizeof(buf)); resultSz = uregex_replaceFirst(re, replText, -1, buf, strlen("Replace <aa> x1x x...x."), &status); TEST_ASSERT(status == U_STRING_NOT_TERMINATED_WARNING); TEST_ASSERT_STRING("Replace <aa> x1x x...x.", buf, FALSE); TEST_ASSERT(resultSz == (int32_t)strlen("Replace xaax x1x x...x.")); TEST_ASSERT(buf[resultSz] == (UChar)0xffff); /* NULL buffer, zero buffer length */ status = U_ZERO_ERROR; resultSz = uregex_replaceFirst(re, replText, -1, NULL, 0, &status); TEST_ASSERT(status == U_BUFFER_OVERFLOW_ERROR); TEST_ASSERT(resultSz == (int32_t)strlen("Replace xaax x1x x...x.")); /* Buffer too small by one */ status = U_ZERO_ERROR; memset(buf, -1, sizeof(buf)); resultSz = uregex_replaceFirst(re, replText, -1, buf, strlen("Replace <aa> x1x x...x.")-1, &status); TEST_ASSERT(status == U_BUFFER_OVERFLOW_ERROR); TEST_ASSERT_STRING("Replace <aa> x1x x...x", buf, FALSE); TEST_ASSERT(resultSz == (int32_t)strlen("Replace xaax x1x x...x.")); TEST_ASSERT(buf[resultSz] == (UChar)0xffff); uregex_close(re); } /* * replaceAll() */ { UChar text1[80]; UChar text2[80]; UChar replText[80]; UChar buf[80]; int32_t resultSz; int32_t expectedResultSize; int32_t i; u_uastrncpy(text1, "Replace xaax x1x x...x.", sizeof(text1)/2); u_uastrncpy(text2, "No match here.", sizeof(text2)/2); u_uastrncpy(replText, "<$1>", sizeof(replText)/2); expectedResultSize = u_strlen(text1); status = U_ZERO_ERROR; re = uregex_openC("x(.*?)x", 0, NULL, &status); TEST_ASSERT_SUCCESS(status); /* Normal case, with match */ uregex_setText(re, text1, -1, &status); resultSz = uregex_replaceAll(re, replText, -1, buf, sizeof(buf)/2, &status); TEST_ASSERT_SUCCESS(status); TEST_ASSERT_STRING("Replace <aa> <1> <...>.", buf, TRUE); TEST_ASSERT(resultSz == (int32_t)strlen("Replace xaax x1x x...x.")); /* No match. Text should copy to output with no changes. */ status = U_ZERO_ERROR; uregex_setText(re, text2, -1, &status); resultSz = uregex_replaceAll(re, replText, -1, buf, sizeof(buf)/2, &status); TEST_ASSERT_SUCCESS(status); TEST_ASSERT_STRING("No match here.", buf, TRUE); TEST_ASSERT(resultSz == (int32_t)strlen("No match here.")); /* Match, output just fills buffer, no termination warning. */ status = U_ZERO_ERROR; uregex_setText(re, text1, -1, &status); memset(buf, -1, sizeof(buf)); resultSz = uregex_replaceAll(re, replText, -1, buf, strlen("Replace xaax x1x x...x."), &status); TEST_ASSERT(status == U_STRING_NOT_TERMINATED_WARNING); TEST_ASSERT_STRING("Replace <aa> <1> <...>.", buf, FALSE); TEST_ASSERT(resultSz == (int32_t)strlen("Replace <aa> <1> <...>.")); TEST_ASSERT(buf[resultSz] == (UChar)0xffff); /* Do the replaceFirst again, without first resetting anything. * Should give the same results. */ status = U_ZERO_ERROR; memset(buf, -1, sizeof(buf)); resultSz = uregex_replaceAll(re, replText, -1, buf, strlen("Replace xaax x1x x...x."), &status); TEST_ASSERT(status == U_STRING_NOT_TERMINATED_WARNING); TEST_ASSERT_STRING("Replace <aa> <1> <...>.", buf, FALSE); TEST_ASSERT(resultSz == (int32_t)strlen("Replace <aa> <1> <...>.")); TEST_ASSERT(buf[resultSz] == (UChar)0xffff); /* NULL buffer, zero buffer length */ status = U_ZERO_ERROR; resultSz = uregex_replaceAll(re, replText, -1, NULL, 0, &status); TEST_ASSERT(status == U_BUFFER_OVERFLOW_ERROR); TEST_ASSERT(resultSz == (int32_t)strlen("Replace <aa> <1> <...>.")); /* Buffer too small. Try every size, which will tickle edge cases * in uregex_appendReplacement (used by replaceAll) */ for (i=0; i<expectedResultSize; i++) { char expected[80]; status = U_ZERO_ERROR; memset(buf, -1, sizeof(buf)); resultSz = uregex_replaceAll(re, replText, -1, buf, i, &status); TEST_ASSERT(status == U_BUFFER_OVERFLOW_ERROR); strcpy(expected, "Replace <aa> <1> <...>."); expected[i] = 0; TEST_ASSERT_STRING(expected, buf, FALSE); TEST_ASSERT(resultSz == expectedResultSize); TEST_ASSERT(buf[i] == (UChar)0xffff); } uregex_close(re); } /* * appendReplacement() */ { UChar text[100]; UChar repl[100]; UChar buf[100]; UChar *bufPtr; int32_t bufCap; status = U_ZERO_ERROR; re = uregex_openC(".*", 0, 0, &status); TEST_ASSERT_SUCCESS(status); u_uastrncpy(text, "whatever", sizeof(text)/2); u_uastrncpy(repl, "some other", sizeof(repl)/2); uregex_setText(re, text, -1, &status); /* match covers whole target string */ uregex_find(re, 0, &status); TEST_ASSERT_SUCCESS(status); bufPtr = buf; bufCap = sizeof(buf) / 2; uregex_appendReplacement(re, repl, -1, &bufPtr, &bufCap, &status); TEST_ASSERT_SUCCESS(status); TEST_ASSERT_STRING("some other", buf, TRUE); /* Match has \u \U escapes */ uregex_find(re, 0, &status); TEST_ASSERT_SUCCESS(status); bufPtr = buf; bufCap = sizeof(buf) / 2; u_uastrncpy(repl, "abc\\u0041\\U00000042 \\\\ $ \\abc", sizeof(repl)/2); uregex_appendReplacement(re, repl, -1, &bufPtr, &bufCap, &status); TEST_ASSERT_SUCCESS(status); TEST_ASSERT_STRING("abcAB \\ $ abc", buf, TRUE); uregex_close(re); } /* * appendTail(). Checked in ReplaceFirst(), replaceAll(). */ /* * split() */ { UChar textToSplit[80]; UChar text2[80]; UChar buf[200]; UChar *fields[10]; int32_t numFields; int32_t requiredCapacity; int32_t spaceNeeded; int32_t sz; u_uastrncpy(textToSplit, "first : second: third", sizeof(textToSplit)/2); u_uastrncpy(text2, "No match here.", sizeof(text2)/2); status = U_ZERO_ERROR; re = uregex_openC(":", 0, NULL, &status); /* Simple split */ uregex_setText(re, textToSplit, -1, &status); TEST_ASSERT_SUCCESS(status); /* The TEST_ASSERT_SUCCESS call above should change too... */ if (U_SUCCESS(status)) { memset(fields, -1, sizeof(fields)); numFields = uregex_split(re, buf, sizeof(buf)/2, &requiredCapacity, fields, 10, &status); TEST_ASSERT_SUCCESS(status); /* The TEST_ASSERT_SUCCESS call above should change too... */ if(U_SUCCESS(status)) { TEST_ASSERT(numFields == 3); TEST_ASSERT_STRING("first ", fields[0], TRUE); TEST_ASSERT_STRING(" second", fields[1], TRUE); TEST_ASSERT_STRING(" third", fields[2], TRUE); TEST_ASSERT(fields[3] == NULL); spaceNeeded = u_strlen(textToSplit) - (numFields - 1) + /* Field delimiters do not appear in output */ numFields; /* Each field gets a NUL terminator */ TEST_ASSERT(spaceNeeded == requiredCapacity); } } uregex_close(re); /* Split with too few output strings available */ status = U_ZERO_ERROR; re = uregex_openC(":", 0, NULL, &status); uregex_setText(re, textToSplit, -1, &status); TEST_ASSERT_SUCCESS(status); /* The TEST_ASSERT_SUCCESS call above should change too... */ if(U_SUCCESS(status)) { memset(fields, -1, sizeof(fields)); numFields = uregex_split(re, buf, sizeof(buf)/2, &requiredCapacity, fields, 2, &status); TEST_ASSERT_SUCCESS(status); /* The TEST_ASSERT_SUCCESS call above should change too... */ if(U_SUCCESS(status)) { TEST_ASSERT(numFields == 2); TEST_ASSERT_STRING("first ", fields[0], TRUE); TEST_ASSERT_STRING(" second: third", fields[1], TRUE); TEST_ASSERT(!memcmp(&fields[2],&minus1,sizeof(UChar*))); spaceNeeded = u_strlen(textToSplit) - (numFields - 1) + /* Field delimiters do not appear in output */ numFields; /* Each field gets a NUL terminator */ TEST_ASSERT(spaceNeeded == requiredCapacity); /* Split with a range of output buffer sizes. */ spaceNeeded = u_strlen(textToSplit) - (numFields - 1) + /* Field delimiters do not appear in output */ numFields; /* Each field gets a NUL terminator */ for (sz=0; sz < spaceNeeded+1; sz++) { memset(fields, -1, sizeof(fields)); status = U_ZERO_ERROR; numFields = uregex_split(re, buf, sz, &requiredCapacity, fields, 10, &status); if (sz >= spaceNeeded) { TEST_ASSERT_SUCCESS(status); TEST_ASSERT_STRING("first ", fields[0], TRUE); TEST_ASSERT_STRING(" second", fields[1], TRUE); TEST_ASSERT_STRING(" third", fields[2], TRUE); } else { TEST_ASSERT(status == U_BUFFER_OVERFLOW_ERROR); } TEST_ASSERT(numFields == 3); TEST_ASSERT(fields[3] == NULL); TEST_ASSERT(spaceNeeded == requiredCapacity); } } } uregex_close(re); } /* Split(), part 2. Patterns with capture groups. The capture group text * comes out as additional fields. */ { UChar textToSplit[80]; UChar buf[200]; UChar *fields[10]; int32_t numFields; int32_t requiredCapacity; int32_t spaceNeeded; int32_t sz; u_uastrncpy(textToSplit, "first <tag-a> second<tag-b> third", sizeof(textToSplit)/2); status = U_ZERO_ERROR; re = uregex_openC("<(.*?)>", 0, NULL, &status); uregex_setText(re, textToSplit, -1, &status); TEST_ASSERT_SUCCESS(status); /* The TEST_ASSERT_SUCCESS call above should change too... */ if(U_SUCCESS(status)) { memset(fields, -1, sizeof(fields)); numFields = uregex_split(re, buf, sizeof(buf)/2, &requiredCapacity, fields, 10, &status); TEST_ASSERT_SUCCESS(status); /* The TEST_ASSERT_SUCCESS call above should change too... */ if(U_SUCCESS(status)) { TEST_ASSERT(numFields == 5); TEST_ASSERT_STRING("first ", fields[0], TRUE); TEST_ASSERT_STRING("tag-a", fields[1], TRUE); TEST_ASSERT_STRING(" second", fields[2], TRUE); TEST_ASSERT_STRING("tag-b", fields[3], TRUE); TEST_ASSERT_STRING(" third", fields[4], TRUE); TEST_ASSERT(fields[5] == NULL); spaceNeeded = strlen("first .tag-a. second.tag-b. third."); /* "." at NUL positions */ TEST_ASSERT(spaceNeeded == requiredCapacity); } } /* Split with too few output strings available (2) */ status = U_ZERO_ERROR; memset(fields, -1, sizeof(fields)); numFields = uregex_split(re, buf, sizeof(buf)/2, &requiredCapacity, fields, 2, &status); TEST_ASSERT_SUCCESS(status); /* The TEST_ASSERT_SUCCESS call above should change too... */ if(U_SUCCESS(status)) { TEST_ASSERT(numFields == 2); TEST_ASSERT_STRING("first ", fields[0], TRUE); TEST_ASSERT_STRING(" second<tag-b> third", fields[1], TRUE); TEST_ASSERT(!memcmp(&fields[2],&minus1,sizeof(UChar*))); spaceNeeded = strlen("first . second<tag-b> third."); /* "." at NUL positions */ TEST_ASSERT(spaceNeeded == requiredCapacity); } /* Split with too few output strings available (3) */ status = U_ZERO_ERROR; memset(fields, -1, sizeof(fields)); numFields = uregex_split(re, buf, sizeof(buf)/2, &requiredCapacity, fields, 3, &status); TEST_ASSERT_SUCCESS(status); /* The TEST_ASSERT_SUCCESS call above should change too... */ if(U_SUCCESS(status)) { TEST_ASSERT(numFields == 3); TEST_ASSERT_STRING("first ", fields[0], TRUE); TEST_ASSERT_STRING("tag-a", fields[1], TRUE); TEST_ASSERT_STRING(" second<tag-b> third", fields[2], TRUE); TEST_ASSERT(!memcmp(&fields[3],&minus1,sizeof(UChar*))); spaceNeeded = strlen("first .tag-a. second<tag-b> third."); /* "." at NUL positions */ TEST_ASSERT(spaceNeeded == requiredCapacity); } /* Split with just enough output strings available (5) */ status = U_ZERO_ERROR; memset(fields, -1, sizeof(fields)); numFields = uregex_split(re, buf, sizeof(buf)/2, &requiredCapacity, fields, 5, &status); TEST_ASSERT_SUCCESS(status); /* The TEST_ASSERT_SUCCESS call above should change too... */ if(U_SUCCESS(status)) { TEST_ASSERT(numFields == 5); TEST_ASSERT_STRING("first ", fields[0], TRUE); TEST_ASSERT_STRING("tag-a", fields[1], TRUE); TEST_ASSERT_STRING(" second", fields[2], TRUE); TEST_ASSERT_STRING("tag-b", fields[3], TRUE); TEST_ASSERT_STRING(" third", fields[4], TRUE); TEST_ASSERT(!memcmp(&fields[5],&minus1,sizeof(UChar*))); spaceNeeded = strlen("first .tag-a. second.tag-b. third."); /* "." at NUL positions */ TEST_ASSERT(spaceNeeded == requiredCapacity); } /* Split, end of text is a field delimiter. */ status = U_ZERO_ERROR; sz = strlen("first <tag-a> second<tag-b>"); uregex_setText(re, textToSplit, sz, &status); TEST_ASSERT_SUCCESS(status); /* The TEST_ASSERT_SUCCESS call above should change too... */ if(U_SUCCESS(status)) { memset(fields, -1, sizeof(fields)); numFields = uregex_split(re, buf, sizeof(buf)/2, &requiredCapacity, fields, 9, &status); TEST_ASSERT_SUCCESS(status); /* The TEST_ASSERT_SUCCESS call above should change too... */ if(U_SUCCESS(status)) { TEST_ASSERT(numFields == 4); TEST_ASSERT_STRING("first ", fields[0], TRUE); TEST_ASSERT_STRING("tag-a", fields[1], TRUE); TEST_ASSERT_STRING(" second", fields[2], TRUE); TEST_ASSERT_STRING("tag-b", fields[3], TRUE); TEST_ASSERT(fields[4] == NULL); TEST_ASSERT(fields[8] == NULL); TEST_ASSERT(!memcmp(&fields[9],&minus1,sizeof(UChar*))); spaceNeeded = strlen("first .tag-a. second.tag-b."); /* "." at NUL positions */ TEST_ASSERT(spaceNeeded == requiredCapacity); } } uregex_close(re); } }
// Build the Whole Script Confusable data // // TODO: Reorganize. Either get rid of the WSConfusableDataBuilder class, // because everything is local to this one build function anyhow, // OR // break this function into more reasonably sized pieces, with // state in WSConfusableDataBuilder. // void buildWSConfusableData(SpoofImpl *spImpl, const char * confusablesWS, int32_t confusablesWSLen, UParseError *pe, UErrorCode &status) { if (U_FAILURE(status)) { return; } URegularExpression *parseRegexp = NULL; int32_t inputLen = 0; UChar *input = NULL; int32_t lineNum = 0; UVector *scriptSets = NULL; uint32_t rtScriptSetsCount = 2; UTrie2 *anyCaseTrie = NULL; UTrie2 *lowerCaseTrie = NULL; anyCaseTrie = utrie2_open(0, 0, &status); lowerCaseTrie = utrie2_open(0, 0, &status); // The scriptSets vector provides a mapping from TRIE values to the set of scripts. // // Reserved TRIE values: // 0: Code point has no whole script confusables. // 1: Code point is of script Common or Inherited. // These code points do not participate in whole script confusable detection. // (This is logically equivalent to saying that they contain confusables in // all scripts) // // Because Trie values are indexes into the ScriptSets vector, pre-fill // vector positions 0 and 1 to avoid conflicts with the reserved values. scriptSets = new UVector(status); if (scriptSets == NULL) { status = U_MEMORY_ALLOCATION_ERROR; goto cleanup; } scriptSets->addElement((void *)NULL, status); scriptSets->addElement((void *)NULL, status); // Convert the user input data from UTF-8 to UChar (UTF-16) u_strFromUTF8(NULL, 0, &inputLen, confusablesWS, confusablesWSLen, &status); if (status != U_BUFFER_OVERFLOW_ERROR) { goto cleanup; } status = U_ZERO_ERROR; input = static_cast<UChar *>(uprv_malloc((inputLen+1) * sizeof(UChar))); if (input == NULL) { status = U_MEMORY_ALLOCATION_ERROR; goto cleanup; } u_strFromUTF8(input, inputLen+1, NULL, confusablesWS, confusablesWSLen, &status); parseRegexp = uregex_openC(parseExp, 0, NULL, &status); // Zap any Byte Order Mark at the start of input. Changing it to a space is benign // given the syntax of the input. if (*input == 0xfeff) { *input = 0x20; } // Parse the input, one line per iteration of this loop. uregex_setText(parseRegexp, input, inputLen, &status); while (uregex_findNext(parseRegexp, &status)) { lineNum++; UChar line[200]; uregex_group(parseRegexp, 0, line, 200, &status); if (uregex_start(parseRegexp, 1, &status) >= 0) { // this was a blank or comment line. continue; } if (uregex_start(parseRegexp, 8, &status) >= 0) { // input file syntax error. status = U_PARSE_ERROR; goto cleanup; } if (U_FAILURE(status)) { goto cleanup; } // Pick up the start and optional range end code points from the parsed line. UChar32 startCodePoint = SpoofImpl::ScanHex( input, uregex_start(parseRegexp, 2, &status), uregex_end(parseRegexp, 2, &status), status); UChar32 endCodePoint = startCodePoint; if (uregex_start(parseRegexp, 3, &status) >=0) { endCodePoint = SpoofImpl::ScanHex( input, uregex_start(parseRegexp, 3, &status), uregex_end(parseRegexp, 3, &status), status); } // Extract the two script names from the source line. We need these in an 8 bit // default encoding (will be EBCDIC on IBM mainframes) in order to pass them on // to the ICU u_getPropertyValueEnum() function. Ugh. char srcScriptName[20]; char targScriptName[20]; extractGroup(parseRegexp, 4, srcScriptName, sizeof(srcScriptName), status); extractGroup(parseRegexp, 5, targScriptName, sizeof(targScriptName), status); UScriptCode srcScript = static_cast<UScriptCode>(u_getPropertyValueEnum(UCHAR_SCRIPT, srcScriptName)); UScriptCode targScript = static_cast<UScriptCode>(u_getPropertyValueEnum(UCHAR_SCRIPT, targScriptName)); if (U_FAILURE(status)) { goto cleanup; } if (srcScript == USCRIPT_INVALID_CODE || targScript == USCRIPT_INVALID_CODE) { status = U_INVALID_FORMAT_ERROR; goto cleanup; } // select the table - (A) any case or (L) lower case only UTrie2 *table = anyCaseTrie; if (uregex_start(parseRegexp, 7, &status) >= 0) { table = lowerCaseTrie; } // Build the set of scripts containing confusable characters for // the code point(s) specified in this input line. // Sanity check that the script of the source code point is the same // as the source script indicated in the input file. Failure of this check is // an error in the input file. // Include the source script in the set (needed for Mixed Script Confusable detection). // UChar32 cp; for (cp=startCodePoint; cp<=endCodePoint; cp++) { int32_t setIndex = utrie2_get32(table, cp); BuilderScriptSet *bsset = NULL; if (setIndex > 0) { U_ASSERT(setIndex < scriptSets->size()); bsset = static_cast<BuilderScriptSet *>(scriptSets->elementAt(setIndex)); } else { bsset = new BuilderScriptSet(); if (bsset == NULL) { status = U_MEMORY_ALLOCATION_ERROR; goto cleanup; } bsset->codePoint = cp; bsset->trie = table; bsset->sset = new ScriptSet(); setIndex = scriptSets->size(); bsset->index = setIndex; bsset->rindex = 0; if (bsset->sset == NULL) { status = U_MEMORY_ALLOCATION_ERROR; goto cleanup; } scriptSets->addElement(bsset, status); utrie2_set32(table, cp, setIndex, &status); } bsset->sset->Union(targScript); bsset->sset->Union(srcScript); if (U_FAILURE(status)) { goto cleanup; } UScriptCode cpScript = uscript_getScript(cp, &status); if (cpScript != srcScript) { status = U_INVALID_FORMAT_ERROR; goto cleanup; } } } // Eliminate duplicate script sets. At this point we have a separate // script set for every code point that had data in the input file. // // We eliminate underlying ScriptSet objects, not the BuildScriptSets that wrap them // // printf("Number of scriptSets: %d\n", scriptSets->size()); { int32_t duplicateCount = 0; rtScriptSetsCount = 2; for (int32_t outeri=2; outeri<scriptSets->size(); outeri++) { BuilderScriptSet *outerSet = static_cast<BuilderScriptSet *>(scriptSets->elementAt(outeri)); if (outerSet->index != static_cast<uint32_t>(outeri)) { // This set was already identified as a duplicate. // It will not be allocated a position in the runtime array of ScriptSets. continue; } outerSet->rindex = rtScriptSetsCount++; for (int32_t inneri=outeri+1; inneri<scriptSets->size(); inneri++) { BuilderScriptSet *innerSet = static_cast<BuilderScriptSet *>(scriptSets->elementAt(inneri)); if (*(outerSet->sset) == *(innerSet->sset) && outerSet->sset != innerSet->sset) { delete innerSet->sset; innerSet->scriptSetOwned = FALSE; innerSet->sset = outerSet->sset; innerSet->index = outeri; innerSet->rindex = outerSet->rindex; 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); utrie2_serialize(lowerCaseTrie, where, size, &status); spImpl->fSpoofData->fRawData->fScriptSets = spImpl->fSpoofData->fMemLimit; spImpl->fSpoofData->fRawData->fScriptSetsLength = rtScriptSetsCount; ScriptSet *rtScriptSets = static_cast<ScriptSet *> (spImpl->fSpoofData->reserveSpace(rtScriptSetsCount * sizeof(ScriptSet), status)); uint32_t rindex = 2; for (int32_t i=2; i<scriptSets->size(); i++) { BuilderScriptSet *bSet = static_cast<BuilderScriptSet *>(scriptSets->elementAt(i)); if (bSet->rindex < rindex) { // We have already copied this script set to the serialized data. continue; } U_ASSERT(rindex == bSet->rindex); rtScriptSets[rindex] = *bSet->sset; // Assignment of a ScriptSet just copies the bits. rindex++; } } // Open new utrie2s from the serialized data. We don't want to keep the ones // we just built because we would then have two copies of the data, one internal to // the utries that we have already constructed, and one in the serialized data area. // An alternative would be to not pre-serialize the Trie data, but that makes the // spoof detector data different, depending on how the detector was constructed. // It's simpler to keep the data always the same. spImpl->fSpoofData->fAnyCaseTrie = utrie2_openFromSerialized( UTRIE2_16_VALUE_BITS, (const char *)spImpl->fSpoofData->fRawData + spImpl->fSpoofData->fRawData->fAnyCaseTrie, spImpl->fSpoofData->fRawData->fAnyCaseTrieLength, NULL, &status); spImpl->fSpoofData->fLowerCaseTrie = utrie2_openFromSerialized( UTRIE2_16_VALUE_BITS, (const char *)spImpl->fSpoofData->fRawData + spImpl->fSpoofData->fRawData->fLowerCaseTrie, spImpl->fSpoofData->fRawData->fAnyCaseTrieLength, NULL, &status); cleanup: if (U_FAILURE(status)) { pe->line = lineNum; } uregex_close(parseRegexp); uprv_free(input); int32_t i; for (i=0; i<scriptSets->size(); i++) { BuilderScriptSet *bsset = static_cast<BuilderScriptSet *>(scriptSets->elementAt(i)); delete bsset; } delete scriptSets; utrie2_close(anyCaseTrie); utrie2_close(lowerCaseTrie); return; }
void __hs_uregex_close(URegularExpression *regexp) { return uregex_close(regexp); }
int main(void) { int ret; int32_t l, u; UErrorCode status; UFILE *ustdout, *ustderr; URegularExpression *uregex; UBool case_insensitive = FALSE; UParseError pe = {-1, -1, {0}, {0}}; UChar pattern[] = { 0x0028, // 28, ( 0x005C, // 5C, backslash 0x0070, // 70, p 0x007B, // 7B, { 0x004C, // 4C, L 0x007D, // 7D, } 0x0029, // 29, ) 0x0028, // 28, ( 0x005C, // 5C, backslash 0x0070, // 70, p 0x007B, // 7B, { 0x004E, // 4E, N 0x0064, // 64, d 0x007D, // 7D, } 0x0029, // 29, ) 0 }; UChar string[] = { 0xD835, 0xDE3C, // A 0xD835, 0xDE3C, // A 0xD835, 0xDFE2, // 0 0xD835, 0xDE3D, // B 0xD835, 0xDE3D, // B 0xD835, 0xDFE3, // 1 0xD835, 0xDE3E, // C 0xD835, 0xDE3E, // C 0xD835, 0xDFE4, // 2 0 }; uregex = NULL; ret = EXIT_SUCCESS; status = U_ZERO_ERROR; ustdout = u_finit(stdout, NULL, NULL); ustderr = u_finit(stderr, NULL, NULL); uregex = uregex_open(pattern, -1, case_insensitive ? UREGEX_CASE_INSENSITIVE : 0, &pe, &status); if (U_FAILURE(status)) { if (U_REGEX_RULE_SYNTAX == status) { u_fprintf(ustderr, "Invalid pattern: error at offset %d\n\t%S\n\t%*c\n", pe.offset, pattern, pe.offset, '^'); } goto end; } uregex_setText(uregex, string, -1, &status); if (U_FAILURE(status)) { goto end; } while (uregex_findNext(uregex, &status)) { l = uregex_start(uregex, 0, &status); if (U_FAILURE(status)) { goto end; } u = uregex_end(uregex, 0, &status); if (U_FAILURE(status)) { goto end; } // $0 u_fprintf(ustdout, "Match found at %d position (to %d): %.*S\n", l, u, u - l, string + l); // $1 à $2 (on pourrait aussi utiliser uregex_group avec un groupNum à valeur 0 pour récupérer $0) { UChar buffer[1024]; int32_t i, l, g; l = uregex_groupCount(uregex, &status); if (U_FAILURE(status)) { icu_error(status, "uregex_groupCount"); } for (i = 1; i <= l; i++) { g = uregex_group(uregex, i, buffer, USTRING_SIZE(buffer), &status); if (U_FAILURE(status)) { icu_error(status, "uregex_group"); } u_fprintf(ustdout, "$%d : %S\n", i, buffer); } } } if (U_FAILURE(status)) { goto end; } if (FALSE) { end: ret = EXIT_FAILURE; } if (NULL != ustderr) { u_fclose(ustderr); } if (NULL != ustdout) { u_fclose(ustdout); } if (NULL != uregex) { uregex_close(uregex); } u_cleanup(); return ret; }