+
+static void TestBug11665(void) {
+ // The problem was with the incorrect breaking of Japanese text beginning
+ // with Katakana characters when no prior Japanese or Chinese text had been
+ // encountered.
+ //
+ // Tested here in cintltst, rather than in intltest, because only cintltst
+ // tests have the ability to reset ICU, which is needed to get the bug
+ // to manifest itself.
+
+ static UChar japaneseText[] = {0x30A2, 0x30EC, 0x30EB, 0x30AE, 0x30FC, 0x6027, 0x7D50, 0x819C, 0x708E};
+ int32_t boundaries[10] = {0};
+ UBreakIterator *bi = NULL;
+ int32_t brk;
+ int32_t brkIdx = 0;
+ int32_t totalBreaks = 0;
+ UErrorCode status = U_ZERO_ERROR;
+
+ ctest_resetICU();
+ bi = ubrk_open(UBRK_WORD, "en_US", japaneseText, UPRV_LENGTHOF(japaneseText), &status);
+ TEST_ASSERT_SUCCESS(status);
+ if (!bi) {
+ return;
+ }
+ for (brk=ubrk_first(bi); brk != UBRK_DONE; brk=ubrk_next(bi)) {
+ boundaries[brkIdx] = brk;
+ if (++brkIdx >= UPRV_LENGTHOF(boundaries) - 1) {
+ break;
+ }
+ }
+ if (brkIdx <= 2 || brkIdx >= UPRV_LENGTHOF(boundaries)) {
+ log_err("%s:%d too few or many breaks found.\n", __FILE__, __LINE__);
+ } else {
+ totalBreaks = brkIdx;
+ brkIdx = 0;
+ for (brk=ubrk_first(bi); brk != UBRK_DONE; brk=ubrk_next(bi)) {
+ if (brk != boundaries[brkIdx]) {
+ log_err("%s:%d Break #%d differs between first and second iteration.\n", __FILE__, __LINE__, brkIdx);
+ break;
+ }
+ if (++brkIdx >= UPRV_LENGTHOF(boundaries) - 1) {
+ log_err("%s:%d Too many breaks.\n", __FILE__, __LINE__);
+ break;
+ }
+ }
+ if (totalBreaks != brkIdx) {
+ log_err("%s:%d Number of breaks differ between first and second iteration.\n", __FILE__, __LINE__);
+ }
+ }
+ ubrk_close(bi);
+}
+
+/*
+ * expOffset is the set of expected offsets, ending with '-1'.
+ * "Expected expOffset -1" means "expected the end of the offsets"
+ */
+
+static const char testSentenceSuppressionsEn[] = "Mr. Jones comes home. Dr. Smith Ph.D. is out. In the U.S.A. it is hot.";
+static const int32_t testSentSuppFwdOffsetsEn[] = { 22, 46, 70, -1 }; /* With suppressions */
+static const int32_t testSentFwdOffsetsEn[] = { 4, 22, 26, 46, 70, -1 }; /* Without suppressions */
+static const int32_t testSentSuppRevOffsetsEn[] = { 46, 22, 0, -1 }; /* With suppressions */
+static const int32_t testSentRevOffsetsEn[] = { 46, 26, 22, 4, 0, -1 }; /* Without suppressions */
+
+static const char testSentenceSuppressionsDe[] = "Wenn ich schon h\\u00F6re zu Guttenberg kommt evtl. zur\\u00FCck.";
+// "Wenn ich schon höre zu Guttenberg kommt evtl. zurück."
+static const int32_t testSentSuppFwdOffsetsDe[] = { 53, -1 }; /* With suppressions */
+static const int32_t testSentFwdOffsetsDe[] = { 53, -1 }; /* Without suppressions; no break in evtl. zur due to casing */
+static const int32_t testSentSuppRevOffsetsDe[] = { 0, -1 }; /* With suppressions */
+static const int32_t testSentRevOffsetsDe[] = { 0, -1 }; /* Without suppressions */
+
+static const char testSentenceSuppressionsEs[] = "Te esperamos todos los miercoles en Bravo 416, Col. El Pueblo a las 7 PM.";
+static const int32_t testSentSuppFwdOffsetsEs[] = { 73, -1 }; /* With suppressions */
+static const int32_t testSentFwdOffsetsEs[] = { 52, 73, -1 }; /* Without suppressions */
+static const int32_t testSentSuppRevOffsetsEs[] = { 0, -1 }; /* With suppressions */
+static const int32_t testSentRevOffsetsEs[] = { 52, 0, -1 }; /* Without suppressions */
+
+static const char testSentenceSuppressionsE1[] = "Add or detract. The world will little note.";
+static const char testSentenceSuppressionsE1u[] = "ADD OR DETRACT. THE WORLD WILL LITTLE NOTE.";
+static const int32_t testSentFwdOffsetsE1[] = { 16, 43, -1 }; /* Suppressions and case should make no difference */
+static const int32_t testSentRevOffsetsE1[] = { 16, 0, -1 }; /* Suppressions and case should make no difference */
+
+static const char testSentenceSuppressionsE2[] = "Coming up, the sprints at NCAA. Are you watching?";
+static const char testSentenceSuppressionsE2u[] = "COMING UP, THE SPRINTS AT NCAA. ARE YOU WATCHING?";
+static const int32_t testSentFwdOffsetsE2[] = { 32, 49, -1 }; /* Suppressions and case should make no difference */
+static const int32_t testSentRevOffsetsE2[] = { 32, 0, -1 }; /* Suppressions and case should make no difference */
+
+static const char testSentenceSuppressionsFr[] = "Tr\\u00E8s bonne prise de parole de M. Junod, municipal \\u00E0 la culture de Lausanne.";
+// "Très bonne prise de parole de M. Junod, municipal à la culture de Lausanne."
+static const int32_t testSentFwdOffsetsFr[] = { 33, 75, -1 }; /* Without suppressions */
+static const int32_t testSentSuppFwdOffsetsFr[] = { 75, -1 }; /* With suppressions */
+static const int32_t testSentRevOffsetsFr[] = { 33, 0, -1 }; /* Without suppressions */
+static const int32_t testSentSuppRevOffsetsFr[] = { 0, -1 }; /* With suppressions */
+
+static const char testSentenceSuppressionsE3[] = "G8 countries e.g. U.K., Japan. Sanctions i.e. restrictions. Test E. Xx G. Xx I. Xx.";
+static const char testSentenceSuppressionsE3u[] = "G8 COUNTRIES E.G. U.K., JAPAN. SANCTIONS I.E. RESTRICTIONS. TEST E. XX G. XX I. XX.";
+static const int32_t testSentSuppFwdOffsetsE3[] = { 31, 60, 83, -1 }; /* With suppressions */
+static const int32_t testSentSuppRevOffsetsE3[] = { 60, 31, 0, -1 }; /* With suppressions */
+static const int32_t testSentFwdOffsetsE3[] = { 18, 31, 60, 68, 74, 80, 83, -1 }; /* Without suppressions */
+static const int32_t testSentRevOffsetsE3[] = { 80, 74, 68, 60, 31, 18, 0, -1 }; /* Without suppressions */
+static const int32_t testSentFwdOffsetsE3u[] = { 18, 31, 46, 60, 68, 74, 80, 83, -1 }; /* Without suppressions */
+static const int32_t testSentRevOffsetsE3u[] = { 80, 74, 68, 60, 46, 31, 18, 0, -1 }; /* Without suppressions */
+
+enum { kTextULenMax = 128, kTextBLenMax = 192 };
+
+typedef struct {
+ const char * locale;
+ const char * text;
+ const int32_t * expFwdOffsets;
+ const int32_t * expRevOffsets;
+} TestBISuppressionsItem;
+
+static const TestBISuppressionsItem testBISuppressionsItems[] = {
+ { "en@ss=standard", testSentenceSuppressionsEn, testSentSuppFwdOffsetsEn, testSentSuppRevOffsetsEn },
+ { "en", testSentenceSuppressionsEn, testSentFwdOffsetsEn, testSentRevOffsetsEn },
+ { "en_CA", testSentenceSuppressionsEn, testSentFwdOffsetsEn, testSentRevOffsetsEn },
+ { "en_CA@ss=standard", testSentenceSuppressionsEn, testSentSuppFwdOffsetsEn, testSentSuppRevOffsetsEn },
+ { "fr@ss=standard", testSentenceSuppressionsEn, testSentFwdOffsetsEn, testSentRevOffsetsEn },
+ { "af@ss=standard", testSentenceSuppressionsEn, testSentFwdOffsetsEn, testSentRevOffsetsEn }, /* no brkiter data => nosuppressions? */
+ { "af_ZA@ss=standard", testSentenceSuppressionsEn, testSentFwdOffsetsEn, testSentRevOffsetsEn }, /* no brkiter data => nosuppressions? */
+ { "zh@ss=standard", testSentenceSuppressionsEn, testSentFwdOffsetsEn, testSentRevOffsetsEn }, /* brkiter data, no suppressions data => no suppressions */
+ { "zh_Hant@ss=standard", testSentenceSuppressionsEn, testSentFwdOffsetsEn, testSentRevOffsetsEn }, /* brkiter data, no suppressions data => no suppressions */
+ { "fi@ss=standard", testSentenceSuppressionsEn, testSentFwdOffsetsEn, testSentRevOffsetsEn }, /* brkiter data, no suppressions data => no suppressions */
+ { "ja@ss=standard", testSentenceSuppressionsEn, testSentFwdOffsetsEn, testSentRevOffsetsEn }, /* brkiter data, no suppressions data => no suppressions */
+ { "de@ss=standard", testSentenceSuppressionsDe, testSentSuppFwdOffsetsDe, testSentSuppRevOffsetsDe },
+ { "de", testSentenceSuppressionsDe, testSentFwdOffsetsDe, testSentRevOffsetsDe },
+ { "es@ss=standard", testSentenceSuppressionsEs, testSentSuppFwdOffsetsEs, testSentSuppRevOffsetsEs },
+ { "es", testSentenceSuppressionsEs, testSentFwdOffsetsEs, testSentRevOffsetsEs },
+ { "en", testSentenceSuppressionsE1, testSentFwdOffsetsE1, testSentRevOffsetsE1 },
+ { "en@ss=standard", testSentenceSuppressionsE1, testSentFwdOffsetsE1, testSentRevOffsetsE1 },
+ { "en", testSentenceSuppressionsE1u, testSentFwdOffsetsE1, testSentRevOffsetsE1 },
+ { "en@ss=standard", testSentenceSuppressionsE1u, testSentFwdOffsetsE1, testSentRevOffsetsE1 },
+ { "en", testSentenceSuppressionsE2, testSentFwdOffsetsE2, testSentRevOffsetsE2 },
+ { "en@ss=standard", testSentenceSuppressionsE2, testSentFwdOffsetsE2, testSentRevOffsetsE2 },
+ { "en", testSentenceSuppressionsE2u, testSentFwdOffsetsE2, testSentRevOffsetsE2 },
+ { "en@ss=standard", testSentenceSuppressionsE2u, testSentFwdOffsetsE2, testSentRevOffsetsE2 },
+ { "fr", testSentenceSuppressionsFr, testSentFwdOffsetsFr, testSentRevOffsetsFr },
+ { "fr@ss=standard", testSentenceSuppressionsFr, testSentSuppFwdOffsetsFr, testSentSuppRevOffsetsFr },
+ { "en@ss=standard", testSentenceSuppressionsE3, testSentSuppFwdOffsetsE3, testSentSuppRevOffsetsE3 },
+ { "en", testSentenceSuppressionsE3, testSentFwdOffsetsE3, testSentRevOffsetsE3 },
+ { "en@ss=standard", testSentenceSuppressionsE3u, testSentSuppFwdOffsetsE3, testSentSuppRevOffsetsE3 },
+ { "en", testSentenceSuppressionsE3u, testSentFwdOffsetsE3u, testSentRevOffsetsE3u },
+ { NULL, NULL, NULL }
+};
+
+static void TestBreakIteratorSuppressions(void) {
+ const TestBISuppressionsItem * itemPtr;
+
+ for (itemPtr = testBISuppressionsItems; itemPtr->locale != NULL; itemPtr++) {
+ UChar textU[kTextULenMax + 1];
+ char textB[kTextBLenMax];
+ int32_t textULen = u_unescape(itemPtr->text, textU, kTextULenMax);
+ textU[kTextULenMax] = 0; // ensure zero termination
+ UErrorCode status = U_ZERO_ERROR;
+ UBreakIterator *bi = ubrk_open(UBRK_SENTENCE, itemPtr->locale, textU, textULen, &status);
+ log_verbose("#%d: %s\n", (itemPtr-testBISuppressionsItems), itemPtr->locale);
+ if (U_SUCCESS(status)) {
+ int32_t offset, start;
+ const int32_t * expOffsetPtr;
+ const int32_t * expOffsetStart;
+ u_austrcpy(textB, textU);
+
+ expOffsetStart = expOffsetPtr = itemPtr->expFwdOffsets;
+ ubrk_first(bi);
+ for (; (offset = ubrk_next(bi)) != UBRK_DONE && *expOffsetPtr >= 0; expOffsetPtr++) {
+ if (offset != *expOffsetPtr) {
+ log_err("FAIL: ubrk_next loc \"%s\", expected %d, got %d, text \"%s\"\n",
+ itemPtr->locale, *expOffsetPtr, offset, textB);
+ }
+ }
+ if (offset != UBRK_DONE || *expOffsetPtr >= 0) {
+ log_err("FAIL: ubrk_next loc \"%s\", expected UBRK_DONE & expOffset -1, got %d and %d, text \"%s\"\n",
+ itemPtr->locale, offset, *expOffsetPtr, textB);
+ }
+
+ expOffsetStart = expOffsetPtr = itemPtr->expFwdOffsets;
+ start = ubrk_first(bi) + 1;
+ for (; (offset = ubrk_following(bi, start)) != UBRK_DONE && *expOffsetPtr >= 0; expOffsetPtr++) {
+ if (offset != *expOffsetPtr) {
+ log_err("FAIL: ubrk_following(%d) loc \"%s\", expected %d, got %d, text \"%s\"\n",
+ start, itemPtr->locale, *expOffsetPtr, offset, textB);
+ }
+ start = *expOffsetPtr + 1;
+ }
+ if (offset != UBRK_DONE || *expOffsetPtr >= 0) {
+ log_err("FAIL: ubrk_following(%d) loc \"%s\", expected UBRK_DONE & expOffset -1, got %d and %d, text \"%s\"\n",
+ start, itemPtr->locale, offset, *expOffsetPtr, textB);
+ }
+
+ expOffsetStart = expOffsetPtr = itemPtr->expRevOffsets;
+ offset = ubrk_last(bi);
+ log_verbose("___ @%d ubrk_last\n", offset);
+ if(offset == 0) {
+ log_err("FAIL: ubrk_last loc \"%s\" unexpected %d\n", itemPtr->locale, offset);
+ }
+ for (; (offset = ubrk_previous(bi)) != UBRK_DONE && *expOffsetPtr >= 0; expOffsetPtr++) {
+ if (offset != *expOffsetPtr) {
+ log_err("FAIL: ubrk_previous loc \"%s\", expected %d, got %d, text \"%s\"\n",
+ itemPtr->locale, *expOffsetPtr, offset, textB);
+ } else {
+ log_verbose("[%d] @%d ubrk_previous()\n", (expOffsetPtr - expOffsetStart), offset);
+ }
+ }
+ if (offset != UBRK_DONE || *expOffsetPtr >= 0) {
+ log_err("FAIL: ubrk_previous loc \"%s\", expected UBRK_DONE & expOffset[%d] -1, got %d and %d, text \"%s\"\n",
+ itemPtr->locale, expOffsetPtr - expOffsetStart, offset, *expOffsetPtr, textB);
+ }
+
+ expOffsetStart = expOffsetPtr = itemPtr->expRevOffsets;
+ start = ubrk_last(bi) - 1;
+ for (; (offset = ubrk_preceding(bi, start)) != UBRK_DONE && *expOffsetPtr >= 0; expOffsetPtr++) {
+ if (offset != *expOffsetPtr) {
+ log_err("FAIL: ubrk_preceding(%d) loc \"%s\", expected %d, got %d, text \"%s\"\n",
+ start, itemPtr->locale, *expOffsetPtr, offset, textB);
+ }
+ start = *expOffsetPtr - 1;
+ }
+ if (start >=0 && (offset != UBRK_DONE || *expOffsetPtr >= 0)) {
+ log_err("FAIL: ubrk_preceding loc(%d) \"%s\", expected UBRK_DONE & expOffset -1, got %d and %d, text \"%s\"\n",
+ start, itemPtr->locale, offset, *expOffsetPtr, textB);
+ }
+
+ ubrk_close(bi);
+ } else {
+ log_data_err("FAIL: ubrk_open(UBRK_SENTENCE, \"%s\", ...) status %s (Are you missing data?)\n",
+ itemPtr->locale, u_errorName(status));
+ }
+ }
+}
+
+#if APPLE_ADDITIONS
+#include <stdio.h>
+#include <unistd.h>
+#include <mach/mach_time.h>
+#include "unicode/urbtok.h"
+#include "cstring.h"
+
+static const char testRulesFilePath[] = "../testdata/word_urbTokTest.txt";
+static const UChar textToTokenize[] = {
+/*
+"Short phrase! Another (with parens); done.\n
+At 4:00, tea-time.\n
+He wouldn't've wanted y'all to ... come at 3:30pm for $3 coffee @funman :)\n
+x3:30 -- x1.0"
+*/
+ 0x53,0x68,0x6F,0x72,0x74,0x20,0x70,0x68,0x72,0x61,0x73,0x65,0x21,0x20,
+ 0x41,0x6E,0x6F,0x74,0x68,0x65,0x72,0x20,0x28,0x77,0x69,0x74,0x68,0x20,0x70,0x61,0x72,0x65,0x6E,0x73,0x29,0x3B,0x20,0x64,0x6F,0x6E,0x65,0x2E,0x0A,
+ 0x41,0x74,0x20,0x34,0x3A,0x30,0x30,0x2C,0x20,0x74,0x65,0x61,0x2D,0x74,0x69,0x6D,0x65,0x2E,0x0A,
+ 0x48,0x65,0x20,0x77,0x6F,0x75,0x6C,0x64,0x6E,0x27,0x74,0x27,0x76,0x65,0x20,0x77,0x61,0x6E,0x74,0x65,0x64,0x20,
+ 0x79,0x27,0x61,0x6C,0x6C,0x20,0x74,0x6F,0x20,0x2E,0x2E,0x2E,0x20, 0x63,0x6F,0x6D,0x65,0x20,0x61,0x74,0x20,
+ 0x33,0x3A,0x33,0x30,0x70,0x6D,0x20,0x66,0x6F,0x72,0x20,0x24,0x33,0x20,0x63,0x6F,0x66,0x66,0x65,0x65,0x20,
+ 0x40,0x66,0x75,0x6E,0x6D,0x61,0x6E,0x20,0x3A,0x29,0x0A,
+ 0x78,0x33,0x3A,0x33,0x30,0x20,0x2D,0x2D,0x20,0x78,0x31,0x2E,0x30,0
+};
+typedef struct {
+ RuleBasedTokenRange token;
+ unsigned long flags;
+} RBTokResult;
+static const RBTokResult expectedResults[] = { // 66 tokens
+ { { 0, 5 }, 0xC8 }, // Short
+ { { 5, 1 }, 0x01 }, // _sp_
+ { { 6, 6 }, 0xC8 }, // phrase
+ { { 12, 1 }, 0x00 }, // !
+ { { 13, 1 }, 0x01 }, // _sp_
+ { { 14, 7 }, 0xC8 }, // Another
+ { { 21, 1 }, 0x01 }, // _sp_
+ { { 22, 1 }, 0x00 }, // (
+ { { 23, 4 }, 0xC8 }, // with
+ { { 27, 1 }, 0x01 }, // _sp_
+ { { 28, 6 }, 0xC8 }, // parens
+ { { 34, 1 }, 0x00 }, // )
+ { { 35, 1 }, 0x00 }, // ;
+ { { 36, 1 }, 0x01 }, // _sp_
+ { { 37, 4 }, 0xC8 }, // done
+ { { 41, 1 }, 0x14 }, // .
+ { { 42, 1 }, 0x00 }, // _nl_
+
+ { { 43, 2 }, 0xC8 }, // At
+ { { 45, 1 }, 0x01 }, // _sp_
+ { { 46, 4 }, 0x76 }, // 4:00 ** here RBBI has x64
+ { { 50, 1 }, 0x00 }, // ,
+ { { 51, 1 }, 0x01 }, // _sp_
+ { { 52, 3 }, 0xC8 }, // tea
+ { { 55, 1 }, 0x15 }, // -
+ { { 56, 4 }, 0xC8 }, // time
+ { { 60, 1 }, 0x14 }, // .
+ { { 61, 1 }, 0x00 }, // _nl_
+
+ { { 62, 2 }, 0xC8 }, // He
+ { { 64, 1 }, 0x01 }, // _sp_
+ { { 65, 8 }, 0xCA }, // wouldn't
+ { { 73, 1 }, 0x16 }, // '
+ { { 74, 2 }, 0xC8 }, // ve
+ { { 76, 1 }, 0x01 }, // _sp_
+ { { 77, 6 }, 0xC8 }, // wanted
+ { { 83, 1 }, 0x01 }, // _sp_
+ { { 84, 5 }, 0xCA }, // y'all
+ { { 89, 1 }, 0x01 }, // _sp_
+ { { 90, 2 }, 0xC8 }, // to
+ { { 92, 1 }, 0x01 }, // _sp_
+ { { 93, 3 }, 0x3C }, // ... ** here RBBI has 0x28
+ { { 96, 1 }, 0x01 }, // _sp_
+ { { 97, 4 }, 0xC8 }, // come
+ { { 101, 1 }, 0x01 }, // _sp_
+ { { 102, 2 }, 0xC8 }, // at
+ { { 104, 1 }, 0x01 }, // _sp_
+ { { 105, 6 }, 0xC8 }, // 3:30pm
+ { { 111, 1 }, 0x01 }, // _sp_
+ { { 112, 3 }, 0xC8 }, // for
+ { { 115, 1 }, 0x01 }, // _sp_
+ { { 116, 1 }, 0x00 }, // $
+ { { 117, 1 }, 0x64 }, // 3
+ { { 118, 1 }, 0x01 }, // _sp_
+ { { 119, 6 }, 0xC8 }, // coffee
+ { { 125, 1 }, 0x01 }, // _sp_
+ { { 126, 7 }, 0xDF }, // @funman ** here RBBI has 0xC8
+ { { 133, 1 }, 0x01 }, // _sp_
+ { { 134, 2 }, 0x20 }, // :)
+ { { 136, 1 }, 0x00 }, // _nl_
+
+ // ** incorrect ranges (and flags) currently produced by RBTok
+ { { 137, 2 }, 0xEC }, // x3
+ { { 139, 1 }, 0x00 }, // :
+ { { 140, 2 }, 0x64 }, // 30
+ // ** for the above, RBBI has
+ //{ { 137, 1 }, 0x64 }, // x
+ //{ { 138, 4 }, 0x64 }, // 3:30
+ //
+ { { 142, 1 }, 0x01 }, // _sp_
+ { { 143, 2 }, 0x3D }, // -- ** here RBBI has 0x28
+ { { 145, 1 }, 0x01 }, // _sp_
+ { { 146, 2 }, 0xEC }, // x1 ** here RBBI has 0xC8
+ { { 148, 1 }, 0x14 }, // .
+ { { 149, 1 }, 0x64 }, // 0
+};
+enum {
+ kNumTokensExpected = UPRV_LENGTHOF(expectedResults), // 66
+ kMaxTokens = 96
+};
+
+static void TestRuleBasedTokenizer(void) {
+ FILE * testRulesFile;
+ char * testRulesUTF8Buf;
+ UChar* testRulesUTF16Buf = NULL;
+ long testRulesFileSize, testRulesFileRead = 0;
+ long testRulesUTF8Offset = 0;
+ int32_t testRulesUTF16Size;
+ UErrorCode status = U_ZERO_ERROR;
+
+ testRulesFile = fopen(testRulesFilePath, "r");
+ if (testRulesFile == NULL) {
+ log_data_err("FAIL: fopen fails for: %s\n", testRulesFilePath);
+ return;
+ }
+ fseek(testRulesFile, 0, SEEK_END);
+ testRulesFileSize = ftell(testRulesFile);
+ rewind(testRulesFile);
+
+ testRulesUTF8Buf = (char *)uprv_malloc(testRulesFileSize);
+ if (testRulesUTF8Buf != NULL) {
+ testRulesFileRead = fread(testRulesUTF8Buf, 1, testRulesFileSize, testRulesFile);
+ }
+ fclose(testRulesFile);
+ if (testRulesUTF8Buf == NULL) {
+ log_data_err("FAIL: uprv_malloc fails for testRulesUTF8Buf[%ld]\n", testRulesFileSize);
+ return;
+ }
+ if (testRulesFileRead < testRulesFileSize) {
+ log_data_err("FAIL: fread fails for %s, read %ld of %ld\n", testRulesFile, testRulesFileRead, testRulesFileSize);
+ uprv_free(testRulesUTF8Buf);
+ return;
+ }
+ /* done with file, UTF8 rules in testRulesUTF8Buf. Handle UTF8 BOM: */
+ if (uprv_strncmp(testRulesUTF8Buf, "\xEF\xBB\xBF", 3) == 0) {
+ testRulesUTF8Offset = 3;
+ testRulesFileSize -= testRulesUTF8Offset;
+ }
+
+ u_strFromUTF8(NULL, 0, &testRulesUTF16Size, testRulesUTF8Buf+testRulesUTF8Offset, testRulesFileSize, &status); /* preflight */
+ if (status == U_BUFFER_OVERFLOW_ERROR) { /* expected for preflight */
+ status = U_ZERO_ERROR;
+ }
+ if (U_FAILURE(status)) {
+ log_data_err("FAIL: u_strFromUTF8 preflight fails: %s\n", u_errorName(status));
+ } else {
+ testRulesUTF16Buf = (UChar *)uprv_malloc(testRulesUTF16Size*sizeof(UChar));
+ if (testRulesUTF16Buf == NULL) {
+ log_data_err("FAIL: uprv_malloc fails for testRulesUTF16Buf[%ld]\n", testRulesUTF16Size*sizeof(UChar));
+ } else {
+ u_strFromUTF8(testRulesUTF16Buf, testRulesUTF16Size, &testRulesUTF16Size, testRulesUTF8Buf+testRulesUTF8Offset, testRulesFileSize, &status);
+ }
+ }
+ uprv_free(testRulesUTF8Buf);
+ if (testRulesUTF16Buf == NULL) {
+ return;
+ }
+ if (U_FAILURE(status)) {
+ log_data_err("FAIL: u_strFromUTF8 fails: %s\n", u_errorName(status));
+ } else {
+ mach_timebase_info_data_t info;
+ uint64_t start, duration;
+ UParseError parseErr;
+ UBreakIterator *brkFromRules;
+
+ mach_timebase_info(&info);
+
+ start = mach_absolute_time();
+ brkFromRules = urbtok_openRules(testRulesUTF16Buf, testRulesUTF16Size, &parseErr, &status);
+ duration = ((mach_absolute_time() - start) * info.numer)/info.denom;
+ if (U_FAILURE(status)) {
+ log_err("FAIL: urbtok_openRules status: %s, line %d, col %d\n", u_errorName(status), parseErr.line, parseErr.offset);
+ } else {
+ uint8_t *rulesBinaryBuf;
+ uint32_t rulesBinarySize;
+ log_info("urbtok_openRules nsec %llu\n", duration);
+ rulesBinarySize = urbtok_getBinaryRules(brkFromRules, NULL, 0, &status);
+ if (U_FAILURE(status)) {
+ log_err("FAIL: urbtok_getBinaryRules preflight status: %s, rulesBinarySize %u\n", u_errorName(status), rulesBinarySize);
+ } else {
+ rulesBinaryBuf = (uint8_t *)uprv_malloc(rulesBinarySize);
+ if (rulesBinaryBuf == NULL) {
+ log_data_err("FAIL: uprv_malloc fails for rulesBinaryBuf[%ld]\n", rulesBinarySize);
+ } else {
+ start = mach_absolute_time();
+ rulesBinarySize = urbtok_getBinaryRules(brkFromRules, rulesBinaryBuf, rulesBinarySize, &status);
+ duration = ((mach_absolute_time() - start) * info.numer)/info.denom;
+ if (U_FAILURE(status)) {
+ log_err("FAIL: urbtok_getBinaryRules status: %s, rulesBinarySize %u\n", u_errorName(status), rulesBinarySize);
+ } else {
+ UBreakIterator *brkFromBinary;
+ log_info("urbtok_getBinaryRules nsec %llu\n", duration);
+ start = mach_absolute_time();
+ brkFromBinary = urbtok_openBinaryRules(rulesBinaryBuf, &status);
+ duration = ((mach_absolute_time() - start) * info.numer)/info.denom;
+ if (U_FAILURE(status)) {
+ log_err("FAIL: urbtok_openBinaryRules status: %s\n", u_errorName(status));
+ } else {
+ RuleBasedTokenRange tokens[kMaxTokens];
+ unsigned long flags[kMaxTokens];
+ int32_t iToken, numTokens = 0;
+
+ log_info("urbtok_openBinaryRules nsec %llu\n", duration);
+
+ status = U_ZERO_ERROR;
+ ubrk_setText(brkFromRules, textToTokenize, -1, &status);
+ if (U_FAILURE(status)) {
+ log_err("FAIL: ubrk_setText brkFromRules status: %s\n", u_errorName(status));
+ } else {
+ start = mach_absolute_time();
+ numTokens = urbtok_tokenize(brkFromRules, kMaxTokens, tokens, flags);
+ duration = ((mach_absolute_time() - start) * info.numer)/info.denom;
+ UBool fail = (numTokens != kNumTokensExpected);
+ for (iToken = 0; !fail && iToken < numTokens; iToken++) {
+ if ( tokens[iToken].location != expectedResults[iToken].token.location ||
+ tokens[iToken].length != expectedResults[iToken].token.length ||
+ flags[iToken] != expectedResults[iToken].flags ) {
+ fail = TRUE;
+ }
+ }
+ if (fail) {
+ log_err("FAIL: urbtok_tokenize bulk brkFromRules expected %d tokens, got %d\n", kNumTokensExpected, numTokens);
+ printf("# expect get\n");
+ printf("# loc len flags loc len flags\n");
+ int32_t maxTokens = (numTokens >= kNumTokensExpected)? numTokens: kNumTokensExpected;
+ for (iToken = 0; iToken < maxTokens; iToken++) {
+ if (iToken < kNumTokensExpected) {
+ printf(" %3ld %3ld 0x%03lX", expectedResults[iToken].token.location,
+ expectedResults[iToken].token.length, expectedResults[iToken].flags);
+ } else {
+ printf(" ");
+ }
+ if (iToken < numTokens) {
+ printf(" %3ld %3ld 0x%03lX\n", tokens[iToken].location, tokens[iToken].length, flags[iToken] );
+ } else {
+ printf("\n");
+ }
+ }
+ } else {
+ log_info("Latn urbtok_tokenize bulk brkFromRules nsec %llu\n", duration);
+ }
+ }
+
+ status = U_ZERO_ERROR;
+ ubrk_setText(brkFromBinary, textToTokenize, -1, &status);
+ if (U_FAILURE(status)) {
+ log_err("FAIL: ubrk_setText brkFromBinary status: %s\n", u_errorName(status));
+ } else {
+ start = mach_absolute_time();
+ numTokens = urbtok_tokenize(brkFromBinary, kMaxTokens, tokens, flags);
+ duration = ((mach_absolute_time() - start) * info.numer)/info.denom;
+ UBool fail = (numTokens != kNumTokensExpected);
+ for (iToken = 0; !fail && iToken < numTokens; iToken++) {
+ if ( tokens[iToken].location != expectedResults[iToken].token.location ||
+ tokens[iToken].length != expectedResults[iToken].token.length ||
+ flags[iToken] != expectedResults[iToken].flags ) {
+ fail = TRUE;
+ }
+ }
+ if (fail) {
+ log_err("FAIL: urbtok_tokenize bulk brkFromBinary expected %d tokens, got %d\n", kNumTokensExpected, numTokens);
+ printf("# expect get\n");
+ printf("# loc len flags loc len flags\n");
+ int32_t maxTokens = (numTokens >= kNumTokensExpected)? numTokens: kNumTokensExpected;
+ for (iToken = 0; iToken < maxTokens; iToken++) {
+ if (iToken < kNumTokensExpected) {
+ printf(" %3ld %3ld 0x%03lX", expectedResults[iToken].token.location,
+ expectedResults[iToken].token.length, expectedResults[iToken].flags);
+ } else {
+ printf(" ");
+ }
+ if (iToken < numTokens) {
+ printf(" %3ld %3ld 0x%03lX\n", tokens[iToken].location, tokens[iToken].length, flags[iToken] );
+ } else {
+ printf("\n");
+ }
+ }
+ } else {
+ log_info("Latn urbtok_tokenize bulk brkFromBinary nsec %llu\n", duration);
+ }
+ }
+
+ status = U_ZERO_ERROR;
+ ubrk_setText(brkFromBinary, textToTokenize, -1, &status);
+ if (U_FAILURE(status)) {
+ log_err("FAIL: ubrk_setText brkFromBinary status: %s\n", u_errorName(status));
+ } else {
+ RuleBasedTokenRange *tokenLimit = tokens + kMaxTokens;
+ RuleBasedTokenRange *tokenP;
+ unsigned long *flagsP;
+
+ start = mach_absolute_time();
+ for (tokenP = tokens, flagsP = flags; tokenP < tokenLimit && urbtok_tokenize(brkFromBinary, 1, tokenP, flagsP) == 1; tokenP++, flagsP++) {
+ ;
+ }
+ numTokens = tokenP - tokens;
+ duration = ((mach_absolute_time() - start) * info.numer)/info.denom;
+ UBool fail = (numTokens != kNumTokensExpected);
+ for (iToken = 0; !fail && iToken < numTokens; iToken++) {
+ if ( tokens[iToken].location != expectedResults[iToken].token.location ||
+ tokens[iToken].length != expectedResults[iToken].token.length ||
+ flags[iToken] != expectedResults[iToken].flags ) {
+ fail = TRUE;
+ }
+ }
+ if (fail) {
+ log_err("FAIL: urbtok_tokenize loop brkFromBinary expected %d tokens, got %d\n", kNumTokensExpected, numTokens);
+ printf("# expect get\n");
+ printf("# loc len flags loc len flags\n");
+ int32_t maxTokens = (numTokens >= kNumTokensExpected)? numTokens: kNumTokensExpected;
+ for (iToken = 0; iToken < maxTokens; iToken++) {
+ if (iToken < kNumTokensExpected) {
+ printf(" %3ld %3ld 0x%03lX", expectedResults[iToken].token.location,
+ expectedResults[iToken].token.length, expectedResults[iToken].flags);
+ } else {
+ printf(" ");
+ }
+ if (iToken < numTokens) {
+ printf(" %3ld %3ld 0x%03lX\n", tokens[iToken].location, tokens[iToken].length, flags[iToken] );
+ } else {
+ printf("\n");
+ }
+ }
+ } else {
+ log_info("Latn urbtok_tokenize loop brkFromBinary nsec %llu\n", duration);
+ }
+ }
+
+ ubrk_close(brkFromBinary);
+ }
+ }
+ uprv_free(rulesBinaryBuf);
+ }
+ }
+ ubrk_close(brkFromRules);
+ }
+ }
+ uprv_free(testRulesUTF16Buf);
+}
+#endif
+
+
projects / apple / icu.git / blobdiff