You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@doris.apache.org by ji...@apache.org on 2024/02/02 09:22:50 UTC
(doris-thirdparty) 01/01: Revert "[Pick 2.0](clucene) pick from clucene for unitest case (#188)"
This is an automated email from the ASF dual-hosted git repository.
jianliangqi pushed a commit to branch revert-188-pick-branch-2.0
in repository https://gitbox.apache.org/repos/asf/doris-thirdparty.git
commit a934452dbde6f023d267c75ea0456d1006df2ba3
Author: qiye <ji...@gmail.com>
AuthorDate: Fri Feb 2 17:22:45 2024 +0800
Revert "[Pick 2.0](clucene) pick from clucene for unitest case (#188)"
This reverts commit 6d4e2181744b4910dc5aa614e351b2162fd7f8f7.
---
.../CLucene/analysis/LanguageBasedAnalyzer.cpp | 5 +-
.../CLucene/analysis/cjk/CJKAnalyzer.cpp | 122 ++-
.../CLucene/analysis/cjk/CJKAnalyzer.h | 3 +-
src/test/CMakeLists.txt | 14 -
src/test/analysis/TestAnalysis.cpp | 3 +-
src/test/contribs-lib/analysis/testChinese.cpp | 867 +++++++++------------
src/test/document/TestDocument.cpp | 172 ++--
src/test/test.h | 2 +-
src/test/tests.cpp | 32 +
src/test/util/TestBKD.cpp | 486 ++++--------
src/test/util/TestBKD.h | 21 +-
src/test/util/TestMSBRadixSorter.cpp | 12 +-
12 files changed, 697 insertions(+), 1042 deletions(-)
diff --git a/src/contribs-lib/CLucene/analysis/LanguageBasedAnalyzer.cpp b/src/contribs-lib/CLucene/analysis/LanguageBasedAnalyzer.cpp
index 6adfcf1e34..04a1594c8e 100644
--- a/src/contribs-lib/CLucene/analysis/LanguageBasedAnalyzer.cpp
+++ b/src/contribs-lib/CLucene/analysis/LanguageBasedAnalyzer.cpp
@@ -44,7 +44,10 @@ LanguageBasedAnalyzer::~LanguageBasedAnalyzer() {
}
bool LanguageBasedAnalyzer::isSDocOpt() {
- return true;
+ if (_tcscmp(lang, _T("chinese")) == 0) {
+ return true;
+ }
+ return false;
}
void LanguageBasedAnalyzer::setStopWords(const TCHAR** stopwords) {
diff --git a/src/contribs-lib/CLucene/analysis/cjk/CJKAnalyzer.cpp b/src/contribs-lib/CLucene/analysis/cjk/CJKAnalyzer.cpp
index 23d7b08c71..b19ed77dd0 100644
--- a/src/contribs-lib/CLucene/analysis/cjk/CJKAnalyzer.cpp
+++ b/src/contribs-lib/CLucene/analysis/cjk/CJKAnalyzer.cpp
@@ -8,23 +8,26 @@
#include "CJKAnalyzer.h"
#include "CLucene/util/CLStreams.h"
-CL_NS_DEF2(analysis, cjk)
+CL_NS_DEF2(analysis,cjk)
CL_NS_USE(analysis)
CL_NS_USE(util)
+
const TCHAR* CJKTokenizer::tokenTypeSingle = _T("single");
const TCHAR* CJKTokenizer::tokenTypeDouble = _T("double");
-CJKTokenizer::CJKTokenizer(Reader* in) : Tokenizer(in) {
- tokenType = Token::getDefaultType();
- offset = 0;
- bufferIndex = 0;
- dataLen = 0;
- preIsTokened = false;
- ignoreSurrogates = true;
+CJKTokenizer::CJKTokenizer(Reader* in):
+ Tokenizer(in)
+{
+ tokenType = Token::getDefaultType();
+ offset = 0;
+ bufferIndex = 0;
+ dataLen = 0;
+ preIsTokened = false;
+ ignoreSurrogates = true;
}
-CL_NS(analysis)::Token* CJKTokenizer::next(Token* token) {
+CL_NS(analysis)::Token* CJKTokenizer::next(Token* token){
/** how many character(s) has been stored in buffer */
int32_t length = 0;
@@ -34,17 +37,13 @@ CL_NS(analysis)::Token* CJKTokenizer::next(Token* token) {
while (true) {
/** current character */
clunichar c;
- int charlen = 1;
+ int charlen = 1;
offset++;
if (bufferIndex >= dataLen) {
- dataLen = input->read((const void**)&cbuffer, 1, LUCENE_IO_BUFFER_SIZE);
+ dataLen = input->read((const void**)&ioBuffer, 1, LUCENE_IO_BUFFER_SIZE);
bufferIndex = 0;
- if (dataLen > 0) {
- lucene_utf8towcs(ioBuffer, cbuffer, LUCENE_MAX_WORD_LEN);
- dataLen = _tcslen(ioBuffer);
- }
}
if (dataLen == -1) {
@@ -63,33 +62,33 @@ CL_NS(analysis)::Token* CJKTokenizer::next(Token* token) {
c = ioBuffer[bufferIndex++];
}
- //to support surrogates, we'll need to convert the incoming utf16 into
- //ucs4(c variable). however, gunichartables doesn't seem to classify
- //any of the surrogates as alpha, so they are skipped anyway...
- //so for now we just convert to ucs4 so that we dont corrupt the input.
- if (c >= 0xd800 || c <= 0xdfff) {
- clunichar c2 = ioBuffer[bufferIndex];
- if (c2 >= 0xdc00 && c2 <= 0xdfff) {
- bufferIndex++;
- offset++;
- charlen = 2;
-
- c = (((c & 0x03ffL) << 10) | ((c2 & 0x03ffL) << 0)) + 0x00010000L;
- }
- }
+ //to support surrogates, we'll need to convert the incoming utf16 into
+ //ucs4(c variable). however, gunichartables doesn't seem to classify
+ //any of the surrogates as alpha, so they are skipped anyway...
+ //so for now we just convert to ucs4 so that we dont corrupt the input.
+ if ( c >= 0xd800 || c <= 0xdfff ){
+ clunichar c2 = ioBuffer[bufferIndex];
+ if ( c2 >= 0xdc00 && c2 <= 0xdfff ){
+ bufferIndex++;
+ offset++;
+ charlen=2;
+
+ c = (((c & 0x03ffL) << 10) | ((c2 & 0x03ffL) << 0)) + 0x00010000L;
+ }
+ }
//if the current character is ASCII or Extend ASCII
- if ((c <= 0xFF) //is BASIC_LATIN
- || (c >= 0xFF00 && c <= 0xFFEF) //ascii >0x74 cast to unsigned...
- ) {
+ if ((c <= 0xFF) //is BASIC_LATIN
+ || (c>=0xFF00 && c<=0xFFEF) //ascii >0x74 cast to unsigned...
+ ) {
if (c >= 0xFF00) {
- //todo: test this... only happens on platforms where char is signed, i think...
+ //todo: test this... only happens on platforms where char is signed, i think...
/** convert HALFWIDTH_AND_FULLWIDTH_FORMS to BASIC_LATIN */
c -= 0xFEE0;
}
// if the current character is a letter or "_" "+" "#"
- if (_istalnum(c) || ((c == '_') || (c == '+') || (c == '#'))) {
+ if (_istalnum(c) || ((c == '_') || (c == '+') || (c == '#')) ) {
if (length == 0) {
// "javaC1C2C3C4linux" <br>
// ^--: the current character begin to token the ASCII
@@ -99,8 +98,8 @@ CL_NS(analysis)::Token* CJKTokenizer::next(Token* token) {
// "javaC1C2C3C4linux" <br>
// ^--: the previous non-ASCII
// : the current character
- offset -= charlen;
- bufferIndex -= charlen;
+ offset-=charlen;
+ bufferIndex-=charlen;
tokenType = tokenTypeSingle;
if (preIsTokened == true) {
@@ -116,7 +115,7 @@ CL_NS(analysis)::Token* CJKTokenizer::next(Token* token) {
// store the LowerCase(c) in the buffer
buffer[length++] = _totlower((TCHAR)c);
- tokenType = tokenTypeSingle;
+ tokenType = tokenTypeSingle;
// break the procedure if buffer overflowed!
if (length == LUCENE_MAX_WORD_LEN) {
@@ -132,39 +131,39 @@ CL_NS(analysis)::Token* CJKTokenizer::next(Token* token) {
}
} else {
// non-ASCII letter, eg."C1C2C3C4"
- if (_istalpha(c) || (!ignoreSurrogates && c >= 0x10000)) {
+ if ( _istalpha(c) || (!ignoreSurrogates && c>=0x10000) ) {
if (length == 0) {
start = offset - 1;
-
- if (c < 0x00010000L)
- buffer[length++] = (TCHAR)c;
- else {
- clunichar ucs4 = c - 0x00010000L;
- buffer[length++] = (TCHAR)((ucs4 >> 10) & 0x3ff) | 0xd800;
- buffer[length++] = (TCHAR)((ucs4 >> 0) & 0x3ff) | 0xdc00;
- }
+
+ if ( c < 0x00010000L )
+ buffer[length++] = (TCHAR)c;
+ else{
+ clunichar ucs4 = c - 0x00010000L;
+ buffer[length++] = (TCHAR)((ucs4 >> 10) & 0x3ff) | 0xd800;
+ buffer[length++] = (TCHAR)((ucs4 >> 0) & 0x3ff) | 0xdc00;
+ }
tokenType = tokenTypeDouble;
} else {
if (tokenType == tokenTypeSingle) {
- offset -= charlen;
- bufferIndex -= charlen;
+ offset-=charlen;
+ bufferIndex-=charlen;
//return the previous ASCII characters
break;
} else {
- if (c < 0x00010000L)
- buffer[length++] = (TCHAR)c;
- else {
- clunichar ucs4 = c - 0x00010000L;
- buffer[length++] = (TCHAR)((ucs4 >> 10) & 0x3ff) | 0xd800;
- buffer[length++] = (TCHAR)((ucs4 >> 0) & 0x3ff) | 0xdc00;
- }
- tokenType = tokenTypeDouble;
+ if ( c < 0x00010000L )
+ buffer[length++] = (TCHAR)c;
+ else{
+ clunichar ucs4 = c - 0x00010000L;
+ buffer[length++] = (TCHAR)((ucs4 >> 10) & 0x3ff) | 0xd800;
+ buffer[length++] = (TCHAR)((ucs4 >> 0) & 0x3ff) | 0xdc00;
+ }
+ tokenType = tokenTypeDouble;
if (length >= 2) {
- offset -= charlen;
- bufferIndex -= charlen;
+ offset-=charlen;
+ bufferIndex-=charlen;
preIsTokened = true;
break;
@@ -183,10 +182,9 @@ CL_NS(analysis)::Token* CJKTokenizer::next(Token* token) {
}
}
- buffer[length] = '\0';
- std::string term = lucene_wcstoutf8string(buffer, length);
- token->set(term.c_str(), 0, term.length(), tokenType);
- return token;
+ buffer[length]='\0';
+ token->set(buffer,start, start+length, tokenType);
+ return token;
}
CL_NS_END2
diff --git a/src/contribs-lib/CLucene/analysis/cjk/CJKAnalyzer.h b/src/contribs-lib/CLucene/analysis/cjk/CJKAnalyzer.h
index ccba70804e..978ad81a7f 100644
--- a/src/contribs-lib/CLucene/analysis/cjk/CJKAnalyzer.h
+++ b/src/contribs-lib/CLucene/analysis/cjk/CJKAnalyzer.h
@@ -40,13 +40,12 @@ private:
* the returned Token
*/
TCHAR buffer[LUCENE_MAX_WORD_LEN];
- char* cbuffer;
/**
* I/O buffer, used to store the content of the input(one of the <br>
* members of Tokenizer)
*/
- TCHAR ioBuffer[LUCENE_MAX_WORD_LEN]{};
+ const TCHAR* ioBuffer;
/** word type: single=>ASCII double=>non-ASCII word=>default */
const TCHAR* tokenType;
diff --git a/src/test/CMakeLists.txt b/src/test/CMakeLists.txt
index 88c7c229dd..736a5f7ea4 100644
--- a/src/test/CMakeLists.txt
+++ b/src/test/CMakeLists.txt
@@ -274,20 +274,6 @@ IF (BUILD_STATIC_LIBRARIES)
TARGET_LINK_LIBRARIES(cl_test clucene-core-static clucene-shared-static ic gtest ${EXTRA_LIBS} ${Roaring_LIBRARY})
ENDIF (UNIX)
- SET(DATA_SOURCE_DIR ${clucene_SOURCE_DIR}/src/test/data)
- SET(DATA_TARGET_DIR "${EXECUTABLE_OUTPUT_PATH}/data")
-
- ADD_CUSTOM_COMMAND(TARGET cl_test POST_BUILD
- COMMAND ${CMAKE_COMMAND} -E copy_directory ${DATA_SOURCE_DIR} ${DATA_TARGET_DIR}
- COMMENT "Copying ${DATA_SOURCE_DIR} to ${DATA_TARGET_DIR}/data")
-
- SET(DICT_SOURCE_DIR ${clucene_SOURCE_DIR}/src/contribs-lib/CLucene/analysis/jieba/dict)
- SET(DICT_TARGET_DIR "${EXECUTABLE_OUTPUT_PATH}/dict")
-
- ADD_CUSTOM_COMMAND(TARGET cl_test POST_BUILD
- COMMAND ${CMAKE_COMMAND} -E copy_directory ${DICT_SOURCE_DIR} ${DICT_TARGET_DIR}
- COMMENT "Copying ${DATA_SOURCE_DIR} to ${DATA_TARGET_DIR}/dict")
-
ENDIF (BUILD_STATIC_LIBRARIES)
############################
diff --git a/src/test/analysis/TestAnalysis.cpp b/src/test/analysis/TestAnalysis.cpp
index a29b65ed04..da51aa06f4 100644
--- a/src/test/analysis/TestAnalysis.cpp
+++ b/src/test/analysis/TestAnalysis.cpp
@@ -5,7 +5,6 @@
* the GNU Lesser General Public License, as specified in the COPYING file.
------------------------------------------------------------------------------*/
#include "test.h"
-#include <memory>
void test(CuTest *tc, Reader *reader, bool verbose, int64_t bytes) {
StandardAnalyzer analyzer;
@@ -64,7 +63,7 @@ void testTokenStreamField(CuTest *tc) {
TokenStream *stream = analyzer.reusableTokenStream(L"field1", &reader);
int field_config = lucene::document::Field::STORE_NO | lucene::document::Field::INDEX_TOKENIZED;
- auto field = std::make_unique<Field>(L"field1", field_config);
+ auto field = _CLNEW Field(L"field1", field_config);
field->setValue(stream);
auto s = field->tokenStreamValue();
int32_t count = 0;
diff --git a/src/test/contribs-lib/analysis/testChinese.cpp b/src/test/contribs-lib/analysis/testChinese.cpp
index ab36589523..7e47aa2d5d 100644
--- a/src/test/contribs-lib/analysis/testChinese.cpp
+++ b/src/test/contribs-lib/analysis/testChinese.cpp
@@ -4,7 +4,6 @@
* Distributable under the terms of either the Apache License (Version 2.0) or
* the GNU Lesser General Public License, as specified in the COPYING file.
------------------------------------------------------------------------------*/
-#include <memory>
#include "test.h"
#include "CLucene/analysis/cjk/CJKAnalyzer.h"
#include "CLucene/analysis/LanguageBasedAnalyzer.h"
@@ -23,7 +22,7 @@
CL_NS_USE2(analysis, cjk)
-void test(CuTest* tc, char* orig, Reader* reader, bool verbose, int64_t bytes) {
+void test(CuTest *tc, char* orig, Reader* reader, bool verbose, int64_t bytes) {
StandardAnalyzer analyzer;
TokenStream* stream = analyzer.tokenStream(NULL, reader);
@@ -35,8 +34,7 @@ void test(CuTest* tc, char* orig, Reader* reader, bool verbose, int64_t bytes) {
TCHAR ttmp[LUCENE_MAX_WORD_LEN + 1];
for (; stream->next(&t);) {
if (verbose) {
- CuMessage(tc, _T("Text=%s start=%d end=%d\n"), t.termBuffer<TCHAR>(), t.startOffset(),
- t.endOffset());
+ CuMessage(tc, _T("Text=%s start=%d end=%d\n"), t.termBuffer<TCHAR>(), t.startOffset(), t.endOffset());
}
int len = t.termLength<TCHAR>();
@@ -48,8 +46,7 @@ void test(CuTest* tc, char* orig, Reader* reader, bool verbose, int64_t bytes) {
if (_tcsncmp(t.termBuffer<TCHAR>(), ttmp, len) != 0) {
TCHAR err[1024];
- _sntprintf(err, 1024, _T("token '%s' didnt match original text at %d-%d"),
- t.termBuffer<TCHAR>(), t.startOffset(), t.endOffset());
+ _sntprintf(err, 1024, _T("token '%s' didnt match original text at %d-%d"), t.termBuffer<TCHAR>(), t.startOffset(), t.endOffset());
CuAssert(tc, err, false);
}
@@ -67,7 +64,7 @@ void test(CuTest* tc, char* orig, Reader* reader, bool verbose, int64_t bytes) {
_CLDELETE(stream);
}
-void _testFile(CuTest* tc, const char* fname, bool verbose) {
+void _testFile(CuTest *tc, const char* fname, bool verbose) {
struct fileStat buf;
fileStat(fname, &buf);
int64_t bytes = buf.st_size;
@@ -81,12 +78,13 @@ void _testFile(CuTest* tc, const char* fname, bool verbose) {
CuMessageA(tc, " Reading test file containing %d bytes.\n", bytes);
jstreams::FileReader fr(fname, "ASCII");
- const TCHAR* start;
+ const TCHAR *start;
size_t total = 0;
int32_t numRead;
do {
numRead = fr.read((const void**)&start, 1, 0);
- if (numRead == -1) break;
+ if (numRead == -1)
+ break;
total += numRead;
} while (numRead >= 0);
@@ -97,32 +95,29 @@ void _testFile(CuTest* tc, const char* fname, bool verbose) {
_CLDELETE_CaARRAY(orig);
}
-void testFile(CuTest* tc) {
+void testFile(CuTest *tc) {
char loc[1024];
strcpy(loc, clucene_data_location);
strcat(loc, "/reuters-21578/feldman-cia-worldfactbook-data.txt");
- CuAssert(tc, _T("reuters-21578/feldman-cia-worldfactbook-data.txt does not exist"),
- Misc::dir_Exists(loc));
+ CuAssert(tc, _T("reuters-21578/feldman-cia-worldfactbook-data.txt does not exist"), Misc::dir_Exists(loc));
_testFile(tc, loc, false);
}
-void _testCJK(CuTest* tc, const char* astr, const char** results, bool ignoreSurrogates = true) {
- //SimpleInputStreamReader r(new AStringReader(astr), SimpleInputStreamReader::UTF8);
- auto r = std::make_unique<lucene::util::SStringReader<char>>(astr, strlen(astr), false);
+void _testCJK(CuTest *tc, const char* astr, const char** results, bool ignoreSurrogates = true) {
+ SimpleInputStreamReader r(new AStringReader(astr), SimpleInputStreamReader::UTF8);
- CJKTokenizer* tokenizer = _CLNEW CJKTokenizer(r.get());
+ CJKTokenizer* tokenizer = _CLNEW CJKTokenizer(&r);
tokenizer->setIgnoreSurrogates(ignoreSurrogates);
int pos = 0;
Token tok;
- //TCHAR tres[LUCENE_MAX_WORD_LEN];
+ TCHAR tres[LUCENE_MAX_WORD_LEN];
while (results[pos] != NULL) {
CLUCENE_ASSERT(tokenizer->next(&tok) != NULL);
- //lucene_utf8towcs(tres, results[pos], LUCENE_MAX_WORD_LEN);
- //cout << results[pos] << " actual " << std::string(tok.termBuffer<char>() ,tok.termLength<char>())<< std::endl;
- CLUCENE_ASSERT(strncmp(tok.termBuffer<char>(), results[pos], tok.termLength<char>()) == 0);
- //CuAssertStrEquals(tc, "unexpected token value", tres, tok.termBuffer<char>());
+ lucene_utf8towcs(tres, results[pos], LUCENE_MAX_WORD_LEN);
+ //wcout << tres << " actual " << tok.termBuffer<TCHAR>() << std::endl;
+ CuAssertStrEquals(tc, _T("unexpected token value"), tres, tok.termBuffer<TCHAR>());
pos++;
}
CLUCENE_ASSERT(!tokenizer->next(&tok));
@@ -130,30 +125,35 @@ void _testCJK(CuTest* tc, const char* astr, const char** results, bool ignoreSur
_CLDELETE(tokenizer);
}
-void testCJK(CuTest* tc) {
+void testCJK(CuTest *tc) {
//utf16 test
//we have a very large unicode character:
//xEFFFF = utf8(F3 AF BF BF) = utf16(DB7F DFFF) = utf8(ED AD BF, ED BF BF)
- static const char* exp4[4] = {"我爱", "爱你", "", NULL};
- _testCJK(tc, "我爱你", exp4, false);
+ static const char* exp4[4] = {"我爱","爱你","", NULL};
+ _testCJK(tc, "我爱你",exp4, false);
- static const char* exp3[4] = {"\xED\xAD\xBF\xED\xBF\xBF\xe5\x95\xa4",
- "\xe5\x95\xa4\xED\xAD\xBF\xED\xBF\xBF", "", NULL};
+ static const char* exp3[4] = {"\xED\xAD\xBF\xED\xBF\xBF\xe5\x95\xa4", "\xe5\x95\xa4\xED\xAD\xBF\xED\xBF\xBF", "", NULL};
_testCJK(tc, "\xED\xAD\xBF\xED\xBF\xBF\xe5\x95\xa4\xED\xAD\xBF\xED\xBF\xBF", exp3, false);
static const char* exp1[5] = {"test", "t\xc3\xbcrm", "values", NULL};
_testCJK(tc, "test t\xc3\xbcrm values", exp1);
- static const char* exp2[6] = {
- "a", "\xe5\x95\xa4\xe9\x85\x92", "\xe9\x85\x92\xe5\x95\xa4", "", "x", NULL};
+ static const char* exp2[6] = {"a", "\xe5\x95\xa4\xe9\x85\x92", "\xe9\x85\x92\xe5\x95\xa4", "", "x", NULL};
_testCJK(tc, "a\xe5\x95\xa4\xe9\x85\x92\xe5\x95\xa4x", exp2);
}
+std::string get_dict_path() {
+ if(const char* env_p = std::getenv("DICT_PATH")) {
+ return env_p;
+ }
+ return "";
+}
+
void testSimpleJiebaSearchModeTokenizer2(CuTest* tc) {
LanguageBasedAnalyzer a;
const char* field_value_data = "冰咒龙";
- auto stringReader = std::make_unique<lucene::util::SStringReader<char>>(
- field_value_data, strlen(field_value_data), false);
+ auto stringReader =
+ _CLNEW lucene::util::SStringReader<char>(field_value_data, strlen(field_value_data), false);
TokenStream* ts;
Token t;
@@ -161,13 +161,13 @@ void testSimpleJiebaSearchModeTokenizer2(CuTest* tc) {
a.setLanguage(_T("chinese"));
a.setStem(false);
a.setMode(lucene::analysis::AnalyzerMode::Search);
- a.initDict("./dict");
- ts = a.tokenStream(_T("contents"), stringReader.get());
+ a.initDict(get_dict_path());
+ ts = a.tokenStream(_T("contents"), stringReader);
CLUCENE_ASSERT(ts->next(&t) != NULL);
- CLUCENE_ASSERT(strncmp(t.termBuffer<char>(), "冰咒", t.termLength<char>()) == 0);
+ CLUCENE_ASSERT(_tcscmp(t.termBuffer<TCHAR>(), _T("冰咒")) == 0);
CLUCENE_ASSERT(ts->next(&t) != NULL);
- CLUCENE_ASSERT(strncmp(t.termBuffer<char>(), "龙", t.termLength<char>()) == 0);
+ CLUCENE_ASSERT(_tcscmp(t.termBuffer<TCHAR>(), _T("龙")) == 0);
CLUCENE_ASSERT(ts->next(&t) == NULL);
_CLDELETE(ts);
}
@@ -175,8 +175,8 @@ void testSimpleJiebaSearchModeTokenizer2(CuTest* tc) {
void testSimpleJiebaAllModeTokenizer2(CuTest* tc) {
LanguageBasedAnalyzer a;
const char* field_value_data = "冰咒龙";
- auto stringReader = std::make_unique<lucene::util::SStringReader<char>>(
- field_value_data, strlen(field_value_data), false);
+ auto stringReader =
+ _CLNEW lucene::util::SStringReader<char>(field_value_data, strlen(field_value_data), false);
TokenStream* ts;
Token t;
@@ -184,15 +184,15 @@ void testSimpleJiebaAllModeTokenizer2(CuTest* tc) {
a.setLanguage(_T("chinese"));
a.setStem(false);
a.setMode(lucene::analysis::AnalyzerMode::All);
- a.initDict("./dict");
- ts = a.tokenStream(_T("contents"), stringReader.get());
+ a.initDict(get_dict_path());
+ ts = a.tokenStream(_T("contents"), stringReader);
CLUCENE_ASSERT(ts->next(&t) != NULL);
- CLUCENE_ASSERT(strncmp(t.termBuffer<char>(), "冰", t.termLength<char>()) == 0);
+ CLUCENE_ASSERT(_tcscmp(t.termBuffer<TCHAR>(), _T("冰")) == 0);
CLUCENE_ASSERT(ts->next(&t) != NULL);
- CLUCENE_ASSERT(strncmp(t.termBuffer<char>(), "咒", t.termLength<char>()) == 0);
+ CLUCENE_ASSERT(_tcscmp(t.termBuffer<TCHAR>(), _T("咒")) == 0);
CLUCENE_ASSERT(ts->next(&t) != NULL);
- CLUCENE_ASSERT(strncmp(t.termBuffer<char>(), "龙", t.termLength<char>()) == 0);
+ CLUCENE_ASSERT(_tcscmp(t.termBuffer<TCHAR>(), _T("龙")) == 0);
CLUCENE_ASSERT(ts->next(&t) == NULL);
_CLDELETE(ts);
}
@@ -200,8 +200,8 @@ void testSimpleJiebaAllModeTokenizer2(CuTest* tc) {
void testSimpleJiebaAllModeTokenizer(CuTest* tc) {
LanguageBasedAnalyzer a;
const char* field_value_data = "我来到北京清华大学";
- auto stringReader = std::make_unique<lucene::util::SStringReader<char>>(
- field_value_data, strlen(field_value_data), false);
+ auto stringReader =
+ _CLNEW lucene::util::SStringReader<char>(field_value_data, strlen(field_value_data), false);
TokenStream* ts;
Token t;
@@ -209,23 +209,21 @@ void testSimpleJiebaAllModeTokenizer(CuTest* tc) {
a.setLanguage(_T("chinese"));
a.setStem(false);
a.setMode(lucene::analysis::AnalyzerMode::All);
- a.initDict("./dict");
- ts = a.tokenStream(_T("contents"), stringReader.get());
+ a.initDict(get_dict_path());
+ ts = a.tokenStream(_T("contents"), stringReader);
CLUCENE_ASSERT(ts->next(&t) != NULL);
- CLUCENE_ASSERT(strncmp(t.termBuffer<char>(), "我", t.termLength<char>()) == 0);
- CLUCENE_ASSERT(ts->next(&t) != NULL);
- CLUCENE_ASSERT(strncmp(t.termBuffer<char>(), "来到", t.termLength<char>()) == 0);
+ CLUCENE_ASSERT(_tcscmp(t.termBuffer<TCHAR>(), _T("来到")) == 0);
CLUCENE_ASSERT(ts->next(&t) != NULL);
- CLUCENE_ASSERT(strncmp(t.termBuffer<char>(), "北京", t.termLength<char>()) == 0);
+ CLUCENE_ASSERT(_tcscmp(t.termBuffer<TCHAR>(), _T("北京")) == 0);
CLUCENE_ASSERT(ts->next(&t) != NULL);
- CLUCENE_ASSERT(strncmp(t.termBuffer<char>(), "清华", t.termLength<char>()) == 0);
+ CLUCENE_ASSERT(_tcscmp(t.termBuffer<TCHAR>(), _T("清华")) == 0);
CLUCENE_ASSERT(ts->next(&t) != NULL);
- CLUCENE_ASSERT(strncmp(t.termBuffer<char>(), "清华大学", t.termLength<char>()) == 0);
+ CLUCENE_ASSERT(_tcscmp(t.termBuffer<TCHAR>(), _T("清华大学")) == 0);
CLUCENE_ASSERT(ts->next(&t) != NULL);
- CLUCENE_ASSERT(strncmp(t.termBuffer<char>(), "华大", t.termLength<char>()) == 0);
+ CLUCENE_ASSERT(_tcscmp(t.termBuffer<TCHAR>(), _T("华大")) == 0);
CLUCENE_ASSERT(ts->next(&t) != NULL);
- CLUCENE_ASSERT(strncmp(t.termBuffer<char>(), "大学", t.termLength<char>()) == 0);
+ CLUCENE_ASSERT(_tcscmp(t.termBuffer<TCHAR>(), _T("大学")) == 0);
CLUCENE_ASSERT(ts->next(&t) == NULL);
_CLDELETE(ts);
}
@@ -233,8 +231,8 @@ void testSimpleJiebaAllModeTokenizer(CuTest* tc) {
void testSimpleJiebaDefaultModeTokenizer2(CuTest* tc) {
LanguageBasedAnalyzer a;
const char* field_value_data = "中国的科技发展在世界上处于领先";
- auto stringReader = std::make_unique<lucene::util::SStringReader<char>>(
- field_value_data, strlen(field_value_data), false);
+ auto stringReader =
+ _CLNEW lucene::util::SStringReader<char>(field_value_data, strlen(field_value_data), false);
TokenStream* ts;
Token t;
@@ -242,8 +240,8 @@ void testSimpleJiebaDefaultModeTokenizer2(CuTest* tc) {
a.setLanguage(_T("chinese"));
a.setStem(false);
a.setMode(lucene::analysis::AnalyzerMode::Default);
- a.initDict("./dict");
- ts = a.tokenStream(_T("contents"), stringReader.get());
+ a.initDict(get_dict_path());
+ ts = a.tokenStream(_T("contents"), stringReader);
/*char tmp[255] = {};
while(ts->next(&t) != nullptr) {
@@ -252,17 +250,17 @@ void testSimpleJiebaDefaultModeTokenizer2(CuTest* tc) {
}*/
CLUCENE_ASSERT(ts->next(&t) != NULL);
- CLUCENE_ASSERT(strncmp(t.termBuffer<char>(), "中国", t.termLength<char>()) == 0);
+ CLUCENE_ASSERT(_tcscmp(t.termBuffer<TCHAR>(), _T("中国")) == 0);
CLUCENE_ASSERT(ts->next(&t) != NULL);
- CLUCENE_ASSERT(strncmp(t.termBuffer<char>(), "科技", t.termLength<char>()) == 0);
+ CLUCENE_ASSERT(_tcscmp(t.termBuffer<TCHAR>(), _T("科技")) == 0);
CLUCENE_ASSERT(ts->next(&t) != NULL);
- CLUCENE_ASSERT(strncmp(t.termBuffer<char>(), "发展", t.termLength<char>()) == 0);
+ CLUCENE_ASSERT(_tcscmp(t.termBuffer<TCHAR>(), _T("发展")) == 0);
CLUCENE_ASSERT(ts->next(&t) != NULL);
- CLUCENE_ASSERT(strncmp(t.termBuffer<char>(), "在世界上", t.termLength<char>()) == 0);
+ CLUCENE_ASSERT(_tcscmp(t.termBuffer<TCHAR>(), _T("在世界上")) == 0);
CLUCENE_ASSERT(ts->next(&t) != NULL);
- CLUCENE_ASSERT(strncmp(t.termBuffer<char>(), "处于", t.termLength<char>()) == 0);
+ CLUCENE_ASSERT(_tcscmp(t.termBuffer<TCHAR>(), _T("处于")) == 0);
CLUCENE_ASSERT(ts->next(&t) != NULL);
- CLUCENE_ASSERT(strncmp(t.termBuffer<char>(), "领先", t.termLength<char>()) == 0);
+ CLUCENE_ASSERT(_tcscmp(t.termBuffer<TCHAR>(), _T("领先")) == 0);
CLUCENE_ASSERT(ts->next(&t) == NULL);
_CLDELETE(ts);
}
@@ -270,8 +268,8 @@ void testSimpleJiebaDefaultModeTokenizer2(CuTest* tc) {
void testSimpleJiebaDefaultModeTokenizer(CuTest* tc) {
LanguageBasedAnalyzer a;
const char* field_value_data = "我来到北京清华大学";
- auto stringReader = std::make_unique<lucene::util::SStringReader<char>>(
- field_value_data, strlen(field_value_data), false);
+ auto stringReader =
+ _CLNEW lucene::util::SStringReader<char>(field_value_data, strlen(field_value_data), false);
TokenStream* ts;
Token t;
@@ -279,17 +277,15 @@ void testSimpleJiebaDefaultModeTokenizer(CuTest* tc) {
a.setLanguage(_T("chinese"));
a.setStem(false);
a.setMode(lucene::analysis::AnalyzerMode::Default);
- a.initDict("./dict");
- ts = a.tokenStream(_T("contents"), stringReader.get());
+ a.initDict(get_dict_path());
+ ts = a.tokenStream(_T("contents"), stringReader);
CLUCENE_ASSERT(ts->next(&t) != NULL);
- CLUCENE_ASSERT(strncmp(t.termBuffer<char>(), "我", t.termLength<char>()) == 0);
+ CLUCENE_ASSERT(_tcscmp(t.termBuffer<TCHAR>(), _T("来到")) == 0);
CLUCENE_ASSERT(ts->next(&t) != NULL);
- CLUCENE_ASSERT(strncmp(t.termBuffer<char>(), "来到", t.termLength<char>()) == 0);
+ CLUCENE_ASSERT(_tcscmp(t.termBuffer<TCHAR>(), _T("北京")) == 0);
CLUCENE_ASSERT(ts->next(&t) != NULL);
- CLUCENE_ASSERT(strncmp(t.termBuffer<char>(), "北京", t.termLength<char>()) == 0);
- CLUCENE_ASSERT(ts->next(&t) != NULL);
- CLUCENE_ASSERT(strncmp(t.termBuffer<char>(), "清华大学", t.termLength<char>()) == 0);
+ CLUCENE_ASSERT(_tcscmp(t.termBuffer<TCHAR>(), _T("清华大学")) == 0);
CLUCENE_ASSERT(ts->next(&t) == NULL);
_CLDELETE(ts);
}
@@ -297,8 +293,8 @@ void testSimpleJiebaDefaultModeTokenizer(CuTest* tc) {
void testSimpleJiebaSearchModeTokenizer(CuTest* tc) {
LanguageBasedAnalyzer a;
const char* field_value_data = "我来到北京清华大学";
- auto stringReader = std::make_unique<lucene::util::SStringReader<char>>(
- field_value_data, strlen(field_value_data), false);
+ auto stringReader =
+ _CLNEW lucene::util::SStringReader<char>(field_value_data, strlen(field_value_data), false);
TokenStream* ts;
Token t;
@@ -306,23 +302,23 @@ void testSimpleJiebaSearchModeTokenizer(CuTest* tc) {
a.setLanguage(_T("chinese"));
a.setStem(false);
a.setMode(lucene::analysis::AnalyzerMode::Search);
- a.initDict("./dict");
- ts = a.tokenStream(_T("contents"), stringReader.get());
+ a.initDict(get_dict_path());
+ ts = a.tokenStream(_T("contents"), stringReader);
CLUCENE_ASSERT(ts->next(&t) != NULL);
- CLUCENE_ASSERT(strncmp(t.termBuffer<char>(), "我", t.termLength<char>()) == 0);
+ CLUCENE_ASSERT(_tcscmp(t.termBuffer<TCHAR>(), _T("我")) == 0);
CLUCENE_ASSERT(ts->next(&t) != NULL);
- CLUCENE_ASSERT(strncmp(t.termBuffer<char>(), "来到", t.termLength<char>()) == 0);
+ CLUCENE_ASSERT(_tcscmp(t.termBuffer<TCHAR>(), _T("来到")) == 0);
CLUCENE_ASSERT(ts->next(&t) != NULL);
- CLUCENE_ASSERT(strncmp(t.termBuffer<char>(), "北京", t.termLength<char>()) == 0);
+ CLUCENE_ASSERT(_tcscmp(t.termBuffer<TCHAR>(), _T("北京")) == 0);
CLUCENE_ASSERT(ts->next(&t) != NULL);
- CLUCENE_ASSERT(strncmp(t.termBuffer<char>(), "清华", t.termLength<char>()) == 0);
+ CLUCENE_ASSERT(_tcscmp(t.termBuffer<TCHAR>(), _T("清华")) == 0);
CLUCENE_ASSERT(ts->next(&t) != NULL);
- CLUCENE_ASSERT(strncmp(t.termBuffer<char>(), "华大", t.termLength<char>()) == 0);
+ CLUCENE_ASSERT(_tcscmp(t.termBuffer<TCHAR>(), _T("华大")) == 0);
CLUCENE_ASSERT(ts->next(&t) != NULL);
- CLUCENE_ASSERT(strncmp(t.termBuffer<char>(), "大学", t.termLength<char>()) == 0);
+ CLUCENE_ASSERT(_tcscmp(t.termBuffer<TCHAR>(), _T("大学")) == 0);
CLUCENE_ASSERT(ts->next(&t) != NULL);
- CLUCENE_ASSERT(strncmp(t.termBuffer<char>(), "清华大学", t.termLength<char>()) == 0);
+ CLUCENE_ASSERT(_tcscmp(t.termBuffer<TCHAR>(), _T("清华大学")) == 0);
CLUCENE_ASSERT(ts->next(&t) == NULL);
_CLDELETE(ts);
}
@@ -330,8 +326,8 @@ void testSimpleJiebaSearchModeTokenizer(CuTest* tc) {
void testSimpleJiebaTokenizer(CuTest* tc) {
LanguageBasedAnalyzer a;
const char* field_value_data = "我爱你中国";
- auto stringReader = std::make_unique<lucene::util::SStringReader<char>>(
- field_value_data, strlen(field_value_data), false);
+ auto stringReader =
+ _CLNEW lucene::util::SStringReader<char>(field_value_data, strlen(field_value_data), false);
TokenStream* ts;
Token t;
@@ -339,13 +335,13 @@ void testSimpleJiebaTokenizer(CuTest* tc) {
a.setLanguage(_T("chinese"));
a.setStem(false);
a.setMode(lucene::analysis::AnalyzerMode::Default);
- a.initDict("./dict");
- ts = a.tokenStream(_T("contents"), stringReader.get());
+ a.initDict(get_dict_path());
+ ts = a.tokenStream(_T("contents"), stringReader);
CLUCENE_ASSERT(ts->next(&t) != NULL);
- CLUCENE_ASSERT(strncmp(t.termBuffer<char>(), "我爱你", t.termLength<char>()) == 0);
+ CLUCENE_ASSERT(_tcscmp(t.termBuffer<TCHAR>(), _T("我爱你")) == 0);
CLUCENE_ASSERT(ts->next(&t) != NULL);
- CLUCENE_ASSERT(strncmp(t.termBuffer<char>(), "中国", t.termLength<char>()) == 0);
+ CLUCENE_ASSERT(_tcscmp(t.termBuffer<TCHAR>(), _T("中国")) == 0);
CLUCENE_ASSERT(ts->next(&t) == NULL);
_CLDELETE(ts);
}
@@ -353,8 +349,8 @@ void testSimpleJiebaTokenizer(CuTest* tc) {
void testSimpleJiebaTokenizer2(CuTest* tc) {
LanguageBasedAnalyzer a;
const char* field_value_data = "人民可以得到更多实惠";
- auto stringReader = std::make_unique<lucene::util::SStringReader<char>>(
- field_value_data, strlen(field_value_data), false);
+ auto stringReader =
+ _CLNEW lucene::util::SStringReader<char>(field_value_data, strlen(field_value_data), false);
TokenStream* ts;
Token t;
@@ -362,20 +358,16 @@ void testSimpleJiebaTokenizer2(CuTest* tc) {
a.setLanguage(_T("chinese"));
a.setStem(false);
a.setMode(lucene::analysis::AnalyzerMode::Default);
- ts = a.tokenStream(_T("contents"), stringReader.get());
+ ts = a.tokenStream(_T("contents"), stringReader);
CLUCENE_ASSERT(ts->next(&t) != NULL);
- CLUCENE_ASSERT(strncmp(t.termBuffer<char>(), "人民", t.termLength<char>()) == 0);
- CLUCENE_ASSERT(ts->next(&t) != NULL);
- CLUCENE_ASSERT(strncmp(t.termBuffer<char>(), "可以", t.termLength<char>()) == 0);
+ CLUCENE_ASSERT(_tcscmp(t.termBuffer<TCHAR>(), _T("人民")) == 0);
CLUCENE_ASSERT(ts->next(&t) != NULL);
- CLUCENE_ASSERT(strncmp(t.termBuffer<char>(), "得到", t.termLength<char>()) == 0);
+ CLUCENE_ASSERT(_tcscmp(t.termBuffer<TCHAR>(), _T("得到")) == 0);
CLUCENE_ASSERT(ts->next(&t) != NULL);
- CLUCENE_ASSERT(strncmp(t.termBuffer<char>(), "更", t.termLength<char>()) == 0);
+ CLUCENE_ASSERT(_tcscmp(t.termBuffer<TCHAR>(), _T("更")) == 0);
CLUCENE_ASSERT(ts->next(&t) != NULL);
- CLUCENE_ASSERT(strncmp(t.termBuffer<char>(), "多", t.termLength<char>()) == 0);
- CLUCENE_ASSERT(ts->next(&t) != NULL);
- CLUCENE_ASSERT(strncmp(t.termBuffer<char>(), "实惠", t.termLength<char>()) == 0);
+ CLUCENE_ASSERT(_tcscmp(t.termBuffer<TCHAR>(), _T("实惠")) == 0);
CLUCENE_ASSERT(ts->next(&t) == NULL);
_CLDELETE(ts);
}
@@ -383,8 +375,8 @@ void testSimpleJiebaTokenizer2(CuTest* tc) {
void testSimpleJiebaTokenizer3(CuTest* tc) {
LanguageBasedAnalyzer a;
const char* field_value_data = "中国人民银行";
- auto stringReader = std::make_unique<lucene::util::SStringReader<char>>(
- field_value_data, strlen(field_value_data), false);
+ auto stringReader =
+ _CLNEW lucene::util::SStringReader<char>(field_value_data, strlen(field_value_data), false);
TokenStream* ts;
Token t;
@@ -393,10 +385,10 @@ void testSimpleJiebaTokenizer3(CuTest* tc) {
a.setLanguage(_T("chinese"));
a.setStem(false);
a.setMode(lucene::analysis::AnalyzerMode::Default);
- ts = a.tokenStream(_T("contents"), stringReader.get());
+ ts = a.tokenStream(_T("contents"), stringReader);
CLUCENE_ASSERT(ts->next(&t) != NULL);
- CLUCENE_ASSERT(strncmp(t.termBuffer<char>(), "中国人民银行", t.termLength<char>()) == 0);
+ CLUCENE_ASSERT(_tcscmp(t.termBuffer<TCHAR>(), _T("中国人民银行")) == 0);
CLUCENE_ASSERT(ts->next(&t) == NULL);
_CLDELETE(ts);
}
@@ -404,46 +396,44 @@ void testSimpleJiebaTokenizer3(CuTest* tc) {
void testSimpleJiebaTokenizer4(CuTest* tc) {
LanguageBasedAnalyzer a;
const char* field_value_data = "人民，银行";
- auto stringReader = std::make_unique<lucene::util::SStringReader<char>>(
- field_value_data, strlen(field_value_data), false);
+ auto stringReader =
+ _CLNEW lucene::util::SStringReader<char>(field_value_data, strlen(field_value_data), false);
TokenStream* ts;
Token t;
//test with chinese
a.setLanguage(_T("chinese"));
a.setStem(false);
- ts = a.tokenStream(_T("contents"), stringReader.get());
+ ts = a.tokenStream(_T("contents"), stringReader);
CLUCENE_ASSERT(ts->next(&t) != NULL);
- CLUCENE_ASSERT(strncmp(t.termBuffer<char>(), "人民", t.termLength<char>()) == 0);
+ CLUCENE_ASSERT(_tcscmp(t.termBuffer<TCHAR>(), _T("人民")) == 0);
CLUCENE_ASSERT(ts->next(&t) != NULL);
- CLUCENE_ASSERT(strncmp(t.termBuffer<char>(), "银行", t.termLength<char>()) == 0);
+ CLUCENE_ASSERT(_tcscmp(t.termBuffer<TCHAR>(), _T("银行")) == 0);
CLUCENE_ASSERT(ts->next(&t) == NULL);
_CLDELETE(ts);
}
void testChineseAnalyzer(CuTest* tc) {
LanguageBasedAnalyzer a;
- //CL_NS(util)::StringReader reader(_T("我爱你"));
- auto reader =
- std::make_unique<lucene::util::SStringReader<char>>("我爱你", strlen("我爱你"), false);
- //reader->mark(50);
+ CL_NS(util)::StringReader reader(_T("我爱你"));
+ reader.mark(50);
TokenStream* ts;
Token t;
//test with cjk
a.setLanguage(_T("cjk"));
a.setStem(false);
- ts = a.tokenStream(_T("contents"), reader.get());
+ ts = a.tokenStream(_T("contents"), &reader);
CLUCENE_ASSERT(ts->next(&t) != NULL);
- CLUCENE_ASSERT(strncmp(t.termBuffer<char>(), "我爱", t.termLength<char>()) == 0);
+ CLUCENE_ASSERT(_tcscmp(t.termBuffer<TCHAR>(), _T("我爱")) == 0);
CLUCENE_ASSERT(ts->next(&t) != NULL);
- CLUCENE_ASSERT(strncmp(t.termBuffer<char>(), "爱你", t.termLength<char>()) == 0);
+ CLUCENE_ASSERT(_tcscmp(t.termBuffer<TCHAR>(), _T("爱你")) == 0);
_CLDELETE(ts);
}
-void testChinese(CuTest* tc) {
+void testChinese(CuTest *tc) {
RAMDirectory dir;
auto analyzer = _CLNEW lucene::analysis::LanguageBasedAnalyzer();
@@ -456,260 +446,229 @@ void testChinese(CuTest* tc) {
auto field = _CLNEW Field(field_name, Field::INDEX_TOKENIZED | Field::STORE_NO);
doc.add(*field);
+
const char* field_value_data = "人民可以得到更多实惠";
- auto stringReader = _CLNEW lucene::util::SimpleInputStreamReader(
- new lucene::util::AStringReader(field_value_data),
- lucene::util::SimpleInputStreamReader::UTF8);
+ auto stringReader =
+ _CLNEW lucene::util::SimpleInputStreamReader(new lucene::util::AStringReader(field_value_data), lucene::util::SimpleInputStreamReader::UTF8);
field->setValue(stringReader);
w.addDocument(&doc);
const char* field_value_data1 = "中国人民银行";
- auto stringReader1 = _CLNEW lucene::util::SimpleInputStreamReader(
- new lucene::util::AStringReader(field_value_data1),
- lucene::util::SimpleInputStreamReader::UTF8);
+ auto stringReader1 =
+ _CLNEW lucene::util::SimpleInputStreamReader(new lucene::util::AStringReader(field_value_data1), lucene::util::SimpleInputStreamReader::UTF8);
field->setValue(stringReader1);
w.addDocument(&doc);
const char* field_value_data2 = "洛杉矶人，洛杉矶居民";
- auto stringReader2 = _CLNEW lucene::util::SimpleInputStreamReader(
- new lucene::util::AStringReader(field_value_data2),
- lucene::util::SimpleInputStreamReader::UTF8);
+ auto stringReader2 =
+ _CLNEW lucene::util::SimpleInputStreamReader(new lucene::util::AStringReader(field_value_data2), lucene::util::SimpleInputStreamReader::UTF8);
field->setValue(stringReader2);
w.addDocument(&doc);
const char* field_value_data3 = "民族，人民";
- auto stringReader3 = _CLNEW lucene::util::SimpleInputStreamReader(
- new lucene::util::AStringReader(field_value_data3),
- lucene::util::SimpleInputStreamReader::UTF8);
+ auto stringReader3 =
+ _CLNEW lucene::util::SimpleInputStreamReader(new lucene::util::AStringReader(field_value_data3), lucene::util::SimpleInputStreamReader::UTF8);
field->setValue(stringReader3);
w.addDocument(&doc);
w.close();
IndexSearcher searcher(&dir);
- Term* t1 = _CLNEW Term(_T("chinese"), _T("人民"));
- auto* query1 = _CLNEW TermQuery(t1);
- Hits* hits1 = searcher.search(query1);
+ Term *t1 = _CLNEW Term(_T("chinese"), _T("人民"));
+ auto *query1 = _CLNEW TermQuery(t1);
+ Hits *hits1 = searcher.search(query1);
CLUCENE_ASSERT(3 == hits1->length());
- Term* t2 = _CLNEW Term(_T("chinese"), _T("民族"));
- auto* query2 = _CLNEW TermQuery(t2);
- Hits* hits2 = searcher.search(query2);
+ Term *t2 = _CLNEW Term(_T("chinese"), _T("民族"));
+ auto *query2 = _CLNEW TermQuery(t2);
+ Hits *hits2 = searcher.search(query2);
CLUCENE_ASSERT(1 == hits2->length());
doc.clear();
//_CLDELETE(field)
- _CLDELETE(hits1)
- _CLDELETE(hits2)
}
void testJiebaMatch(CuTest* tc) {
RAMDirectory dir;
+
+ auto analyzer = _CLNEW lucene::analysis::LanguageBasedAnalyzer();
+ analyzer->setLanguage(L"chinese");
+ analyzer->setMode(lucene::analysis::AnalyzerMode::Default);
+ IndexWriter w(&dir, analyzer, true);
auto field_name = lucene::util::Misc::_charToWide("chinese");
- try {
- auto analyzer = std::make_unique<lucene::analysis::LanguageBasedAnalyzer>();
- analyzer->setLanguage(L"chinese");
- analyzer->setMode(lucene::analysis::AnalyzerMode::Default);
- IndexWriter w(&dir, analyzer.get(), true);
- w.setUseCompoundFile(false);
-
- Document doc;
- auto field = _CLNEW Field(field_name, Field::INDEX_TOKENIZED | Field::STORE_NO);
- doc.add(*field);
-
- const char* field_value_data = "人民可以得到更多实惠";
- auto stringReader = std::make_unique<lucene::util::SStringReader<char>>(
- field_value_data, strlen(field_value_data), false);
- auto* stream = analyzer->tokenStream(field->name(), stringReader.get());
- field->setValue(stream);
- w.addDocument(&doc);
-
- const char* field_value_data1 = "中国人民银行";
- auto stringReader1 = std::make_unique<lucene::util::SStringReader<char>>(
- field_value_data1, strlen(field_value_data1), false);
- auto* stream1 = analyzer->tokenStream(field->name(), stringReader1.get());
- field->setValue(stream1);
- w.addDocument(&doc);
-
- const char* field_value_data2 = "洛杉矶人，洛杉矶居民";
- auto stringReader2 = std::make_unique<lucene::util::SStringReader<char>>(
- field_value_data2, strlen(field_value_data2), false);
- auto* stream2 = analyzer->tokenStream(field->name(), stringReader2.get());
- field->setValue(stream2);
- w.addDocument(&doc);
-
- const char* field_value_data3 = "民族，人民";
- auto stringReader3 = std::make_unique<lucene::util::SStringReader<char>>(
- field_value_data3, strlen(field_value_data3), false);
- auto* stream3 = analyzer->tokenStream(field->name(), stringReader3.get());
- field->setValue(stream3);
- w.addDocument(&doc);
-
- w.close();
- doc.clear();
- _CLDELETE(stream)
- _CLDELETE(stream1)
- _CLDELETE(stream2)
- _CLDELETE(stream3)
- } catch (CLuceneError& r) {
- printf("clucene error in testJiebaMatch: %s\n", r.what());
- }
+
+ Document doc;
+ auto field = _CLNEW Field(field_name, Field::INDEX_TOKENIZED | Field::STORE_NO);
+ doc.add(*field);
+
+
+ const char* field_value_data = "人民可以得到更多实惠";
+ auto stringReader =
+ _CLNEW lucene::util::SStringReader<char>(field_value_data, strlen(field_value_data), false);
+ field->setValue(stringReader);
+ w.addDocument(&doc);
+
+ const char* field_value_data1 = "中国人民银行";
+ auto stringReader1 =
+ _CLNEW lucene::util::SStringReader<char>(field_value_data1, strlen(field_value_data1), false);
+ field->setValue(stringReader1);
+ w.addDocument(&doc);
+
+ const char* field_value_data2 = "洛杉矶人，洛杉矶居民";
+ auto stringReader2 =
+ _CLNEW lucene::util::SStringReader<char>(field_value_data2, strlen(field_value_data2), false);
+ field->setValue(stringReader2);
+ w.addDocument(&doc);
+
+ const char* field_value_data3 = "民族，人民";
+ auto stringReader3 =
+ _CLNEW lucene::util::SStringReader<char>(field_value_data3, strlen(field_value_data3), false);
+ field->setValue(stringReader3);
+ w.addDocument(&doc);
+
+ w.close();
+
IndexSearcher searcher(&dir);
- std::vector<std::string> analyse_result;
+ lucene::util::Reader* reader = nullptr;
+ std::vector<std::wstring> analyse_result;
const char* value = "民族";
- auto analyzer = std::make_unique<lucene::analysis::LanguageBasedAnalyzer>(L"chinese", false);
- auto reader = std::make_unique<lucene::util::SStringReader<char>>(value, strlen(value), false);
+ analyzer = _CLNEW lucene::analysis::LanguageBasedAnalyzer(L"chinese", false);
+ reader = _CLNEW lucene::util::SStringReader<char>(value, strlen(value), false);
- lucene::analysis::TokenStream* token_stream = analyzer->tokenStream(field_name, reader.get());
+ lucene::analysis::TokenStream* token_stream = analyzer->tokenStream(field_name, reader);
lucene::analysis::Token token;
while (token_stream->next(&token)) {
- if (token.termLength<char>() != 0) {
- analyse_result.emplace_back(token.termBuffer<char>(), token.termLength<char>());
+ if(token.termLength<TCHAR>() != 0) {
+ analyse_result.emplace_back(token.termBuffer<TCHAR>(), token.termLength<TCHAR>());
}
}
if (token_stream != nullptr) {
token_stream->close();
}
- _CLDELETE(token_stream)
- auto query = std::make_unique<lucene::search::BooleanQuery>();
+ lucene::search::Query* query = _CLNEW lucene::search::BooleanQuery();
for (const auto& t : analyse_result) {
- std::wstring token_ws = StringUtil::string_to_wstring(t);
- auto* term = _CLNEW lucene::index::Term(field_name, token_ws.c_str());
- dynamic_cast<lucene::search::BooleanQuery*>(query.get())
+ //std::wstring token_ws = std::wstring(token.begin(), token.end());
+ auto* term =
+ _CLNEW lucene::index::Term(field_name, t.c_str());
+ dynamic_cast<lucene::search::BooleanQuery*>(query)
->add(_CLNEW lucene::search::TermQuery(term), true,
lucene::search::BooleanClause::SHOULD);
_CLDECDELETE(term);
}
- Hits* hits1 = searcher.search(query.get());
+
+ Hits *hits1 = searcher.search(query);
CLUCENE_ASSERT(1 == hits1->length());
- _CLDELETE(hits1)
- _CLDELETE_ARRAY(field_name)
+
+ doc.clear();
}
void testJiebaMatch2(CuTest* tc) {
RAMDirectory dir;
- auto analyzer = std::make_unique<lucene::analysis::LanguageBasedAnalyzer>();
+ auto analyzer = _CLNEW lucene::analysis::LanguageBasedAnalyzer();
analyzer->setLanguage(L"chinese");
analyzer->setMode(lucene::analysis::AnalyzerMode::Default);
- IndexWriter w(&dir, analyzer.get(), true);
- w.setUseCompoundFile(false);
+ IndexWriter w(&dir, analyzer, true);
auto field_name = lucene::util::Misc::_charToWide("chinese");
Document doc;
auto field = _CLNEW Field(field_name, Field::INDEX_TOKENIZED | Field::STORE_NO);
doc.add(*field);
- try {
- const char* field_value_data = "人民可以得到更多实惠";
- auto stringReader = std::make_unique<lucene::util::SStringReader<char>>(
- field_value_data, strlen(field_value_data), false);
- auto* stream = analyzer->tokenStream(field->name(), stringReader.get());
- field->setValue(stream);
- w.addDocument(&doc);
-
- const char* field_value_data1 = "中国人民银行";
- auto stringReader1 = std::make_unique<lucene::util::SStringReader<char>>(
- field_value_data1, strlen(field_value_data1), false);
- auto* stream1 = analyzer->tokenStream(field->name(), stringReader1.get());
- field->setValue(stream1);
- w.addDocument(&doc);
-
- const char* field_value_data2 = "洛杉矶人，洛杉矶居民";
- auto stringReader2 = std::make_unique<lucene::util::SStringReader<char>>(
- field_value_data2, strlen(field_value_data2), false);
- auto* stream2 = analyzer->tokenStream(field->name(), stringReader2.get());
- field->setValue(stream2);
- w.addDocument(&doc);
-
- const char* field_value_data3 = "民族，人民";
- auto stringReader3 = std::make_unique<lucene::util::SStringReader<char>>(
- field_value_data3, strlen(field_value_data3), false);
- auto* stream3 = analyzer->tokenStream(field->name(), stringReader3.get());
- field->setValue(stream3);
- w.addDocument(&doc);
-
- w.close();
- doc.clear();
- _CLDELETE(stream)
- _CLDELETE(stream1)
- _CLDELETE(stream2)
- _CLDELETE(stream3)
- } catch (CLuceneError& r) {
- printf("clucene error in testJiebaMatch2: %s\n", r.what());
- }
+
+ const char* field_value_data = "人民可以得到更多实惠";
+ auto stringReader =
+ _CLNEW lucene::util::SStringReader<char>(field_value_data, strlen(field_value_data), false);
+ field->setValue(stringReader);
+ w.addDocument(&doc);
+
+ const char* field_value_data1 = "中国人民银行";
+ auto stringReader1 =
+ _CLNEW lucene::util::SStringReader<char>(field_value_data1, strlen(field_value_data1), false);
+ field->setValue(stringReader1);
+ w.addDocument(&doc);
+
+ const char* field_value_data2 = "洛杉矶人，洛杉矶居民";
+ auto stringReader2 =
+ _CLNEW lucene::util::SStringReader<char>(field_value_data2, strlen(field_value_data2), false);
+ field->setValue(stringReader2);
+ w.addDocument(&doc);
+
+ const char* field_value_data3 = "民族，人民";
+ auto stringReader3 =
+ _CLNEW lucene::util::SStringReader<char>(field_value_data3, strlen(field_value_data3), false);
+ field->setValue(stringReader3);
+ w.addDocument(&doc);
+
+ w.close();
+
IndexSearcher searcher(&dir);
- std::vector<std::string> analyse_result;
+ lucene::util::Reader* reader = nullptr;
+ std::vector<std::wstring> analyse_result;
const char* value = "人民";
- auto analyzer1 = std::make_unique<lucene::analysis::LanguageBasedAnalyzer>(L"chinese", false);
- auto reader = std::make_unique<lucene::util::SStringReader<char>>(value, strlen(value), false);
+ analyzer = _CLNEW lucene::analysis::LanguageBasedAnalyzer(L"chinese", false);
+ reader = _CLNEW lucene::util::SStringReader<char>(value, strlen(value), false);
- lucene::analysis::TokenStream* token_stream = analyzer1->tokenStream(field_name, reader.get());
+ lucene::analysis::TokenStream* token_stream = analyzer->tokenStream(field_name, reader);
lucene::analysis::Token token;
while (token_stream->next(&token)) {
- if (token.termLength<char>() != 0) {
- analyse_result.emplace_back(token.termBuffer<char>(), token.termLength<char>());
+ if(token.termLength<TCHAR>() != 0) {
+ analyse_result.emplace_back(token.termBuffer<TCHAR>(), token.termLength<TCHAR>());
}
}
if (token_stream != nullptr) {
token_stream->close();
}
- _CLDELETE(token_stream)
- auto query = std::make_unique<lucene::search::BooleanQuery>();
+
+ lucene::search::Query* query = _CLNEW lucene::search::BooleanQuery();
for (const auto& t : analyse_result) {
- std::wstring token_ws = StringUtil::string_to_wstring(t);
- auto* term = _CLNEW lucene::index::Term(field_name, token_ws.c_str());
- dynamic_cast<lucene::search::BooleanQuery*>(query.get())
+ //std::wstring token_ws = std::wstring(token.begin(), token.end());
+ auto* term =
+ _CLNEW lucene::index::Term(field_name, t.c_str());
+ dynamic_cast<lucene::search::BooleanQuery*>(query)
->add(_CLNEW lucene::search::TermQuery(term), true,
lucene::search::BooleanClause::SHOULD);
_CLDECDELETE(term);
}
- Hits* hits1 = searcher.search(query.get());
+ Hits *hits1 = searcher.search(query);
CLUCENE_ASSERT(2 == hits1->length());
- _CLDELETE(hits1)
- _CLDELETE_ARRAY(field_name)
+
+ doc.clear();
}
void testJiebaMatchHuge(CuTest* tc) {
RAMDirectory dir;
- auto analyzer = std::make_unique<lucene::analysis::LanguageBasedAnalyzer>();
+ auto analyzer = _CLNEW lucene::analysis::LanguageBasedAnalyzer();
analyzer->setLanguage(L"chinese");
analyzer->setMode(lucene::analysis::AnalyzerMode::Default);
- analyzer->initDict("./dict");
+ analyzer->initDict(get_dict_path());
- IndexWriter w(&dir, analyzer.get(), true);
- w.setUseCompoundFile(false);
+ IndexWriter w(&dir, analyzer, true);
auto field_name = lucene::util::Misc::_charToWide("chinese");
Document doc;
auto field = _CLNEW Field(field_name, Field::INDEX_TOKENIZED | Field::STORE_NO);
doc.add(*field);
- const char* field_value_data =
- "数据模型\n"
- "本文档主要从逻辑层面，描述 Doris 的数据模型，以帮助用户更好的使用 Doris "
- "应对不同的业务场景。\n"
+
+ const char* field_value_data = "数据模型\n"
+ "本文档主要从逻辑层面，描述 Doris 的数据模型，以帮助用户更好的使用 Doris 应对不同的业务场景。\n"
"\n"
"基本概念\n"
- "在 Doris 中，数据以表（Table）的形式进行逻辑上的描述。 "
- "一张表包括行（Row）和列（Column）。Row 即用户的一行数据。Column "
- "用于描述一行数据中不同的字段。\n"
+ "在 Doris 中，数据以表（Table）的形式进行逻辑上的描述。 一张表包括行（Row）和列（Column）。Row 即用户的一行数据。Column 用于描述一行数据中不同的字段。\n"
"\n"
- "Column 可以分为两大类：Key 和 Value。从业务角度看，Key 和 Value "
- "可以分别对应维度列和指标列。Doris的key列是建表语句中指定的列，建表语句中的关键字\\'"
- "unique key\\'或\\'aggregate key\\'或\\'duplicate key\\'后面的列就是 Key 列，除了 Key "
- "列剩下的就是 Value a列。\n"
+ "Column 可以分为两大类：Key 和 Value。从业务角度看，Key 和 Value 可以分别对应维度列和指标列。Doris的key列是建表语句中指定的列，建表语句中的关键字\\'unique key\\'或\\'aggregate key\\'或\\'duplicate key\\'后面的列就是 Key 列，除了 Key 列剩下的就是 Value a列。\n"
"\n"
"Doris 的数据模型主要分为3类:\n"
"\n"
@@ -743,8 +702,7 @@ void testJiebaMatchHuge(CuTest* tc) {
"city VARCHAR(20) COMMENT \"用户所在城市\",\n"
"age SMALLINT COMMENT \"用户年龄\",\n"
"sex TINYINT COMMENT \"用户性别\",\n"
- "last_visit_date DATETIME REPLACE DEFAULT \"1970-01-01 00:00:00\" COMMENT "
- "\"用户最后一次访问时间\",\n"
+ "last_visit_date DATETIME REPLACE DEFAULT \"1970-01-01 00:00:00\" COMMENT \"用户最后一次访问时间\",\n"
"cost BIGINT SUM DEFAULT \"0\" COMMENT \"用户总消费\",\n"
"max_dwell_time INT MAX DEFAULT \"0\" COMMENT \"用户最大停留时间\",\n"
"min_dwell_time INT MIN DEFAULT \"99999\" COMMENT \"用户最小停留时间\"\n"
@@ -755,16 +713,11 @@ void testJiebaMatchHuge(CuTest* tc) {
"\"replication_allocation\" = \"tag.location.default: 1\"\n"
");\n"
"\n"
- "可以看到，这是一个典型的用户信息和访问行为的事实表。 "
- "在一般星型模型中，用户信息和访问行为一般分别存放在维度表和事实表中。这里我们为了更加方"
- "便的解释 Doris 的数据模型，将两部分信息统一存放在一张表中。\n"
+ "可以看到，这是一个典型的用户信息和访问行为的事实表。 在一般星型模型中，用户信息和访问行为一般分别存放在维度表和事实表中。这里我们为了更加方便的解释 Doris 的数据模型，将两部分信息统一存放在一张表中。\n"
"\n"
- "表中的列按照是否设置了 AggregationType，分为 Key (维度列) 和 "
- "Value（指标列）。没有设置 AggregationType 的，如 user_id、date、age ... 等称为 "
- "Key，而设置了 AggregationType 的称为 Value。\n"
+ "表中的列按照是否设置了 AggregationType，分为 Key (维度列) 和 Value（指标列）。没有设置 AggregationType 的，如 user_id、date、age ... 等称为 Key，而设置了 AggregationType 的称为 Value。\n"
"\n"
- "当我们导入数据时，对于 Key 列相同的行会聚合成一行，而 Value 列会按照设置的 "
- "AggregationType 进行聚合。 AggregationType 目前有以下四种聚合方式：\n"
+ "当我们导入数据时，对于 Key 列相同的行会聚合成一行，而 Value 列会按照设置的 AggregationType 进行聚合。 AggregationType 目前有以下四种聚合方式：\n"
"\n"
"SUM：求和，多行的 Value 进行累加。\n"
"REPLACE：替代，下一批数据中的 Value 会替换之前导入过的行中的 Value。\n"
@@ -812,18 +765,13 @@ void testJiebaMatchHuge(CuTest* tc) {
"10003 2017-10-02 广州 32 0 2017-10-02 11:20:00 30 11 11\n"
"10004 2017-10-01 深圳 35 0 2017-10-01 10:00:15 100 3 3\n"
"10004 2017-10-03 深圳 35 0 2017-10-03 10:20:22 11 6 6\n"
- "可以看到，用户 10000 "
- "只剩下了一行聚合后的数据。而其余用户的数据和原始数据保持一致。这里先解释下用户 10000 "
- "聚合后的数据：\n"
+ "可以看到，用户 10000 只剩下了一行聚合后的数据。而其余用户的数据和原始数据保持一致。这里先解释下用户 10000 聚合后的数据：\n"
"\n"
"前5列没有变化，从第6列 last_visit_date 开始：\n"
"\n"
- "2017-10-01 07:00:00：因为 last_visit_date 列的聚合方式为 REPLACE，所以 2017-10-01 "
- "07:00:00 替换了 2017-10-01 06:00:00 保存了下来。\n"
+ "2017-10-01 07:00:00：因为 last_visit_date 列的聚合方式为 REPLACE，所以 2017-10-01 07:00:00 替换了 2017-10-01 06:00:00 保存了下来。\n"
"\n"
- "注：在同一个导入批次中的数据，对于 REPLACE "
- "这种聚合方式，替换顺序不做保证。如在这个例子中，最终保存下来的，也有可能是 2017-10-01 "
- "06:00:00。而对于不同导入批次中的数据，可以保证，后一批次的数据会替换前一批次。\n"
+ "注：在同一个导入批次中的数据，对于 REPLACE 这种聚合方式，替换顺序不做保证。如在这个例子中，最终保存下来的，也有可能是 2017-10-01 06:00:00。而对于不同导入批次中的数据，可以保证，后一批次的数据会替换前一批次。\n"
"\n"
"35：因为 cost 列的聚合类型为 SUM，所以由 20 + 15 累加获得 35。\n"
"\n"
@@ -831,9 +779,7 @@ void testJiebaMatchHuge(CuTest* tc) {
"\n"
"2：因为 min_dwell_time 列的聚合类型为 MIN，所以 10 和 2 取最小值，获得 2。\n"
"\n"
- "经过聚合，Doris "
- "中最终只会存储聚合后的数据。换句话说，即明细数据会丢失，用户不能够再查询到聚合前的明细"
- "数据了。\n"
+ "经过聚合，Doris 中最终只会存储聚合后的数据。换句话说，即明细数据会丢失，用户不能够再查询到聚合前的明细数据了。\n"
"\n"
"示例2：保留明细数据\n"
"接示例1，我们将表结构修改如下：\n"
@@ -849,13 +795,11 @@ void testJiebaMatchHuge(CuTest* tc) {
"cost BIGINT SUM 用户总消费\n"
"max_dwell_time INT MAX 用户最大停留时间\n"
"min_dwell_time INT MIN 用户最小停留时间\n"
- "即增加了一列 timestamp，记录精确到秒的数据灌入时间。 同时，将AGGREGATE "
- "KEY设置为AGGREGATE KEY(user_id, date, timestamp, city, age, sex)\n"
+ "即增加了一列 timestamp，记录精确到秒的数据灌入时间。 同时，将AGGREGATE KEY设置为AGGREGATE KEY(user_id, date, timestamp, city, age, sex)\n"
"\n"
"导入数据如下：\n"
"\n"
- "user_id date timestamp city age sex last_visit_date cost max_dwell_time "
- "min_dwell_time\n"
+ "user_id date timestamp city age sex last_visit_date cost max_dwell_time min_dwell_time\n"
"10000 2017-10-01 2017-10-01 08:00:05 北京 20 0 2017-10-01 06:00:00 20 10 10\n"
"10000 2017-10-01 2017-10-01 09:00:05 北京 20 0 2017-10-01 07:00:00 15 2 2\n"
"10001 2017-10-01 2017-10-01 18:12:10 北京 30 1 2017-10-01 17:05:45 2 22 22\n"
@@ -866,25 +810,17 @@ void testJiebaMatchHuge(CuTest* tc) {
"通过sql导入数据：\n"
"\n"
"insert into example_db.example_tbl values\n"
- "(10000,\"2017-10-01\",\"2017-10-01 08:00:05\",\"北京\",20,0,\"2017-10-01 "
- "06:00:00\",20,10,10),\n"
- "(10000,\"2017-10-01\",\"2017-10-01 09:00:05\",\"北京\",20,0,\"2017-10-01 "
- "07:00:00\",15,2,2),\n"
- "(10001,\"2017-10-01\",\"2017-10-01 18:12:10\",\"北京\",30,1,\"2017-10-01 "
- "17:05:45\",2,22,22),\n"
- "(10002,\"2017-10-02\",\"2017-10-02 13:10:00\",\"上海\",20,1,\"2017-10-02 "
- "12:59:12\",200,5,5),\n"
- "(10003,\"2017-10-02\",\"2017-10-02 13:15:00\",\"广州\",32,0,\"2017-10-02 "
- "11:20:00\",30,11,11),\n"
- "(10004,\"2017-10-01\",\"2017-10-01 12:12:48\",\"深圳\",35,0,\"2017-10-01 "
- "10:00:15\",100,3,3),\n"
- "(10004,\"2017-10-03\",\"2017-10-03 12:38:20\",\"深圳\",35,0,\"2017-10-03 "
- "10:20:22\",11,6,6);\n"
+ "(10000,\"2017-10-01\",\"2017-10-01 08:00:05\",\"北京\",20,0,\"2017-10-01 06:00:00\",20,10,10),\n"
+ "(10000,\"2017-10-01\",\"2017-10-01 09:00:05\",\"北京\",20,0,\"2017-10-01 07:00:00\",15,2,2),\n"
+ "(10001,\"2017-10-01\",\"2017-10-01 18:12:10\",\"北京\",30,1,\"2017-10-01 17:05:45\",2,22,22),\n"
+ "(10002,\"2017-10-02\",\"2017-10-02 13:10:00\",\"上海\",20,1,\"2017-10-02 12:59:12\",200,5,5),\n"
+ "(10003,\"2017-10-02\",\"2017-10-02 13:15:00\",\"广州\",32,0,\"2017-10-02 11:20:00\",30,11,11),\n"
+ "(10004,\"2017-10-01\",\"2017-10-01 12:12:48\",\"深圳\",35,0,\"2017-10-01 10:00:15\",100,3,3),\n"
+ "(10004,\"2017-10-03\",\"2017-10-03 12:38:20\",\"深圳\",35,0,\"2017-10-03 10:20:22\",11,6,6);\n"
"\n"
"那么当这批数据正确导入到 Doris 中后，Doris 中最终存储如下：\n"
"\n"
- "user_id date timestamp city age sex last_visit_date cost max_dwell_time "
- "min_dwell_time\n"
+ "user_id date timestamp city age sex last_visit_date cost max_dwell_time min_dwell_time\n"
"10000 2017-10-01 2017-10-01 08:00:05 北京 20 0 2017-10-01 06:00:00 20 10 10\n"
"10000 2017-10-01 2017-10-01 09:00:05 北京 20 0 2017-10-01 07:00:00 15 2 2\n"
"10001 2017-10-01 2017-10-01 18:12:10 北京 30 1 2017-10-01 17:05:45 2 22 22\n"
@@ -892,10 +828,7 @@ void testJiebaMatchHuge(CuTest* tc) {
"10003 2017-10-02 2017-10-02 13:15:00 广州 32 0 2017-10-02 11:20:00 30 11 11\n"
"10004 2017-10-01 2017-10-01 12:12:48 深圳 35 0 2017-10-01 10:00:15 100 3 3\n"
"10004 2017-10-03 2017-10-03 12:38:20 深圳 35 0 2017-10-03 10:20:22 11 6 6\n"
- "我们可以看到，存储的数据，和导入数据完全一样，没有发生任何聚合。这是因为，这批数据中，"
- "因为加入了 timestamp 列，所有行的 Key "
- "都不完全相同。也就是说，只要保证导入的数据中，每一行的 Key "
- "都不完全相同，那么即使在聚合模型下，Doris 也可以保存完整的明细数据。\n"
+ "我们可以看到，存储的数据，和导入数据完全一样，没有发生任何聚合。这是因为，这批数据中，因为加入了 timestamp 列，所有行的 Key 都不完全相同。也就是说，只要保证导入的数据中，每一行的 Key 都不完全相同，那么即使在聚合模型下，Doris 也可以保存完整的明细数据。\n"
"\n"
"示例3：导入数据与已有数据聚合\n"
"接示例1。假设现在表中已有数据如下：\n"
@@ -928,31 +861,17 @@ void testJiebaMatchHuge(CuTest* tc) {
"10004 2017-10-01 深圳 35 0 2017-10-01 10:00:15 100 3 3\n"
"10004 2017-10-03 深圳 35 0 2017-10-03 11:22:00 55 19 6\n"
"10005 2017-10-03 长沙 29 1 2017-10-03 18:11:02 3 1 1\n"
- "可以看到，用户 10004 的已有数据和新导入的数据发生了聚合。同时新增了 10005 "
- "用户的数据。\n"
+ "可以看到，用户 10004 的已有数据和新导入的数据发生了聚合。同时新增了 10005 用户的数据。\n"
"\n"
"数据的聚合，在 Doris 中有如下三个阶段发生：\n"
"\n"
"每一批次数据导入的 ETL 阶段。该阶段会在每一批次导入的数据内部进行聚合。\n"
- "底层 BE 进行数据 Compaction 的阶段。该阶段，BE "
- "会对已导入的不同批次的数据进行进一步的聚合。\n"
+ "底层 BE 进行数据 Compaction 的阶段。该阶段，BE 会对已导入的不同批次的数据进行进一步的聚合。\n"
"数据查询阶段。在数据查询时，对于查询涉及到的数据，会进行对应的聚合。\n"
- "数据在不同时间，可能聚合的程度不一致。比如一批数据刚导入时，可能还未与之前已存在的数据"
- "进行聚合。但是对于用户而言，用户只能查询到聚合后的数据。即不同的聚合程度对于用户查询而"
- "言是透明的。用户需始终认为数据以最终的完成的聚合程度存在，而不应假设某些聚合还未发生。"
- "（可参阅聚合模型的局限性一节获得更多详情。）\n"
+ "数据在不同时间，可能聚合的程度不一致。比如一批数据刚导入时，可能还未与之前已存在的数据进行聚合。但是对于用户而言，用户只能查询到聚合后的数据。即不同的聚合程度对于用户查询而言是透明的。用户需始终认为数据以最终的完成的聚合程度存在，而不应假设某些聚合还未发生。（可参阅聚合模型的局限性一节获得更多详情。）\n"
"\n"
"Unique 模型\n"
- "在某些多维分析场景下，用户更关注的是如何保证 Key 的唯一性，即如何获得 Primary Key "
- "唯一性约束。因此，我们引入了 /；·90Unique "
- "数据模型。在1."
- "2版本之前，该模型本质上是聚合模型的一个特例，也是一种简化的表结构表示方式。由于聚合模"
- "型的实现方式是读时合并（merge on "
- "read)，因此在一些聚合查询上性能不佳（参考后续章节聚合模型的局限性的描述），在1."
- "2版本我们引入了Unique模型新的实现方式，写时合并（merge on "
- "write），通过在写入时做一些额外的工作，实现了最优的查询性能。写时合并将在未来替换读时"
- "合并成为Unique模型的默认实现方式，两者将会短暂的共存一段时间。下面将对两种实现方式分别"
- "举例进行说明。\n"
+ "在某些多维分析场景下，用户更关注的是如何保证 Key 的唯一性，即如何获得 Primary Key 唯一性约束。因此，我们引入了 /；·90Unique 数据模型。在1.2版本之前，该模型本质上是聚合模型的一个特例，也是一种简化的表结构表示方式。由于聚合模型的实现方式是读时合并（merge on read)，因此在一些聚合查询上性能不佳（参考后续章节聚合模型的局限性的描述），在1.2版本我们引入了Unique模型新的实现方式，写时合并（merge on write），通过在写入时做一些额外的工作，实现了最优的查询性能。写时合并将在未来替换读时合并成为Unique模型的默认实现方式，两者将会短暂的共存一段时间。下面将对两种实现方式分别举例进行说明。\n"
"\n"
"读时合并（与聚合模型相同的实现方式）\n"
"ColumnName Type IsKey Comment\n"
@@ -964,8 +883,7 @@ void testJiebaMatchHuge(CuTest* tc) {
"phone LARGEINT No 用户电话\n"
"address VARCHAR(500) No 用户住址\n"
"register_time DATETIME No 用户注册时间\n"
- "这是一个典型的用户基础信息表。这类数据没有聚合需求，只需保证主键唯一性。（这里的主键为"
- " user_id + username）。那么我们的建表语句如下：\n"
+ "这是一个典型的用户基础信息表。这类数据没有聚合需求，只需保证主键唯一性。（这里的主键为 user_id + username）。那么我们的建表语句如下：\n"
"\n"
"CREATE TABLE IF NOT EXISTS example_db.example_tbl\n"
"(\n"
@@ -1014,17 +932,13 @@ void testJiebaMatchHuge(CuTest* tc) {
"\"replication_allocation\" = \"tag.location.default: 1\"\n"
");\n"
"\n"
- "即Unique 模型的读时合并实现完全可以用聚合模型中的 REPLACE "
- "方式替代。其内部的实现方式和数据存储方式也完全一样。这里不再继续举例说明。\n"
+ "即Unique 模型的读时合并实现完全可以用聚合模型中的 REPLACE 方式替代。其内部的实现方式和数据存储方式也完全一样。这里不再继续举例说明。\n"
"\n"
"SinceVersion 1.2\n"
"写时合并\n"
- "Unqiue模型的写时合并实现，与聚合模型就是完全不同的两种模型了，查询性能更接近于duplicat"
- "e模型，在有主键约束需求的场景上相比聚合模型有较大的查询性能优势，尤其是在聚合查询以及"
- "需要用索引过滤大量数据的查询中。\n"
+ "Unqiue模型的写时合并实现，与聚合模型就是完全不同的两种模型了，查询性能更接近于duplicate模型，在有主键约束需求的场景上相比聚合模型有较大的查询性能优势，尤其是在聚合查询以及需要用索引过滤大量数据的查询中。\n"
"\n"
- "在 1.2.0 "
- "版本中，作为一个新的feature，写时合并默认关闭，用户可以通过添加下面的property来开启\n"
+ "在 1.2.0 版本中，作为一个新的feature，写时合并默认关闭，用户可以通过添加下面的property来开启\n"
"\n"
"\"enable_unique_key_merge_on_write\" = \"true\"\n"
"\n"
@@ -1059,23 +973,15 @@ void testJiebaMatchHuge(CuTest* tc) {
"phone LARGEINT NONE 用户电话\n"
"address VARCHAR(500) NONE 用户住址\n"
"register_time DATETIME NONE 用户注册时间\n"
- "在开启了写时合并选项的Unique表上，数据在导入阶段就会去将被覆盖和被更新的数据进行标记删"
- "除，同时将新的数据写入新的文件。在查询的时候，所有被标记删除的数据都会在文件级别被过滤"
- "掉，读取出来的数据就都是最新的数据，消除掉了读时合并中的数据聚合过程，并且能够在很多情"
- "况下支持多种谓词的下推。因此在许多场景都能带来比较大的性能提升，尤其是在有聚合查询的情"
- "况下。\n"
+ "在开启了写时合并选项的Unique表上，数据在导入阶段就会去将被覆盖和被更新的数据进行标记删除，同时将新的数据写入新的文件。在查询的时候，所有被标记删除的数据都会在文件级别被过滤掉，读取出来的数据就都是最新的数据，消除掉了读时合并中的数据聚合过程，并且能够在很多情况下支持多种谓词的下推。因此在许多场景都能带来比较大的性能提升，尤其是在有聚合查询的情况下。\n"
"\n"
"【注意】\n"
"\n"
"新的Merge-on-write实现默认关闭，且只能在建表时通过指定property的方式打开。\n"
- "旧的Merge-on-"
- "read的实现无法无缝升级到新版本的实现（数据组织方式完全不同），如果需要改为使用写时合并"
- "的实现版本，需要手动执行insert into unique-mow-table select * from source table.\n"
- "在Unique模型上独有的delete sign 和 sequence "
- "col，在写时合并的新版实现中仍可以正常使用，用法没有变化。\n"
+ "旧的Merge-on-read的实现无法无缝升级到新版本的实现（数据组织方式完全不同），如果需要改为使用写时合并的实现版本，需要手动执行insert into unique-mow-table select * from source table.\n"
+ "在Unique模型上独有的delete sign 和 sequence col，在写时合并的新版实现中仍可以正常使用，用法没有变化。\n"
"Duplicate 模型\n"
- "在某些多维分析场景下，数据既没有主键，也没有聚合需求。因此，我们引入 Duplicate "
- "数据模型来满足这类需求。举例说明。\n"
+ "在某些多维分析场景下，数据既没有主键，也没有聚合需求。因此，我们引入 Duplicate 数据模型来满足这类需求。举例说明。\n"
"\n"
"ColumnName Type SortKey Comment\n"
"timestamp DATETIME Yes 日志时间\n"
@@ -1101,22 +1007,14 @@ void testJiebaMatchHuge(CuTest* tc) {
"\"replication_allocation\" = \"tag.location.default: 1\"\n"
");\n"
"\n"
- "这种数据模型区别于 Aggregate 和 Unique "
- "模型。数据完全按照导入文件中的数据进行存储，不会有任何聚合。即使两行数据完全相同，也都"
- "会保留。 而在建表语句中指定的 DUPLICATE "
- "KEY，只是用来指明底层数据按照那些列进行排序。（更贴切的名称应该为 “Sorted "
- "Column”，这里取名 “DUPLICATE KEY” 只是用以明确表示所用的数据模型。关于 “Sorted "
- "Column”的更多解释，可以参阅前缀索引）。在 DUPLICATE KEY "
- "的选择上，我们建议适当的选择前 2-4 列就可以。\n"
+ "这种数据模型区别于 Aggregate 和 Unique 模型。数据完全按照导入文件中的数据进行存储，不会有任何聚合。即使两行数据完全相同，也都会保留。 而在建表语句中指定的 DUPLICATE KEY，只是用来指明底层数据按照那些列进行排序。（更贴切的名称应该为 “Sorted Column”，这里取名 “DUPLICATE KEY” 只是用以明确表示所用的数据模型。关于 “Sorted Column”的更多解释，可以参阅前缀索引）。在 DUPLICATE KEY 的选择上，我们建议适当的选择前 2-4 列就可以。\n"
"\n"
- "这种数据模型适用于既没有聚合需求，又没有主键唯一性约束的原始数据的存储。更多使用场景，"
- "可参阅聚合模型的局限性小节。\n"
+ "这种数据模型适用于既没有聚合需求，又没有主键唯一性约束的原始数据的存储。更多使用场景，可参阅聚合模型的局限性小节。\n"
"\n"
"聚合模型的局限性\n"
"这里我们针对 Aggregate 模型，来介绍下聚合模型的局限性。\n"
"\n"
- "在聚合模型中，模型对外展现的，是最终聚合后的数据。也就是说，任何还未聚合的数据（比如说"
- "两个不同导入批次的数据），必须通过某种方式，以保证对外展示的一致性。我们举例说明。\n"
+ "在聚合模型中，模型对外展现的，是最终聚合后的数据。也就是说，任何还未聚合的数据（比如说两个不同导入批次的数据），必须通过某种方式，以保证对外展示的一致性。我们举例说明。\n"
"\n"
"假设表结构如下：\n"
"\n"
@@ -1137,9 +1035,7 @@ void testJiebaMatchHuge(CuTest* tc) {
"10001 2017-11-20 1\n"
"10001 2017-11-21 5\n"
"10003 2017-11-22 22\n"
- "可以看到，用户 10001 "
- "分属在两个导入批次中的数据还没有聚合。但是为了保证用户只能查询到如下最终聚合后的数据："
- "\n"
+ "可以看到，用户 10001 分属在两个导入批次中的数据还没有聚合。但是为了保证用户只能查询到如下最终聚合后的数据：\n"
"\n"
"user_id date cost\n"
"10001 2017-11-20 51\n"
@@ -1148,8 +1044,7 @@ void testJiebaMatchHuge(CuTest* tc) {
"10003 2017-11-22 22\n"
"我们在查询引擎中加入了聚合算子，来保证数据对外的一致性。\n"
"\n"
- "另外，在聚合列（Value）上，执行与聚合类型不一致的聚合类查询时，要注意语意。比如我们在"
- "如上示例中执行如下查询：\n"
+ "另外，在聚合列（Value）上，执行与聚合类型不一致的聚合类查询时，要注意语意。比如我们在如上示例中执行如下查询：\n"
"\n"
"SELECT MIN(cost) FROM table;\n"
"\n"
@@ -1161,10 +1056,7 @@ void testJiebaMatchHuge(CuTest* tc) {
"\n"
"SELECT COUNT(*) FROM table;\n"
"\n"
- "在其他数据库中，这类查询都会很快的返回结果。因为在实现上，我们可以通过如“导入时对行进"
- "行计数，保存 count 的统计信息”，或者在查询时“仅扫描某一列数据，获得 count "
- "值”的方式，只需很小的开销，即可获得查询结果。但是在 Doris "
- "的聚合模型中，这种查询的开销非常大。\n"
+ "在其他数据库中，这类查询都会很快的返回结果。因为在实现上，我们可以通过如“导入时对行进行计数，保存 count 的统计信息”，或者在查询时“仅扫描某一列数据，获得 count 值”的方式，只需很小的开销，即可获得查询结果。但是在 Doris 的聚合模型中，这种查询的开销非常大。\n"
"\n"
"我们以刚才的数据为例：\n"
"\n"
@@ -1186,39 +1078,23 @@ void testJiebaMatchHuge(CuTest* tc) {
"10001 2017-11-21 5\n"
"10002 2017-11-21 39\n"
"10003 2017-11-22 22\n"
- "所以，select count(*) from table; 的正确结果应该为 4。但如果我们只扫描 user_id "
- "这一列，如果加上查询时聚合，最终得到的结果是 3（10001, 10002, "
- "10003）。而如果不加查询时聚合，则得到的结果是 "
- "5（两批次一共5行数据）。可见这两个结果都是不对的。\n"
+ "所以，select count(*) from table; 的正确结果应该为 4。但如果我们只扫描 user_id 这一列，如果加上查询时聚合，最终得到的结果是 3（10001, 10002, 10003）。而如果不加查询时聚合，则得到的结果是 5（两批次一共5行数据）。可见这两个结果都是不对的。\n"
"\n"
- "为了得到正确的结果，我们必须同时读取 user_id 和 date "
- "这两列的数据，再加上查询时聚合，才能返回 4 这个正确的结果。也就是说，在 count() "
- "查询中，Doris 必须扫描所有的 AGGREGATE KEY 列（这里就是 user_id 和 "
- "date），并且聚合后，才能得到语意正确的结果。当聚合列非常多时，count() "
- "查询需要扫描大量的数据。\n"
+ "为了得到正确的结果，我们必须同时读取 user_id 和 date 这两列的数据，再加上查询时聚合，才能返回 4 这个正确的结果。也就是说，在 count() 查询中，Doris 必须扫描所有的 AGGREGATE KEY 列（这里就是 user_id 和 date），并且聚合后，才能得到语意正确的结果。当聚合列非常多时，count() 查询需要扫描大量的数据。\n"
"\n"
- "因此，当业务上有频繁的 count() 查询时，我们建议用户通过增加一个值恒为 1 "
- "的，聚合类型为 SUM 的列来模拟 count()。如刚才的例子中的表结构，我们修改如下：\n"
+ "因此，当业务上有频繁的 count() 查询时，我们建议用户通过增加一个值恒为 1 的，聚合类型为 SUM 的列来模拟 count()。如刚才的例子中的表结构，我们修改如下：\n"
"\n"
"ColumnName Type AggregateType Comment\n"
"user_id BIGINT 用户id\n"
"date DATE 数据灌入日期\n"
"cost BIGINT SUM 用户总消费\n"
"count BIGINT SUM 用于计算count\n"
- "增加一个 count 列，并且导入数据中，该列值恒为 1。则 select count() from table; "
- "的结果等价于 select sum(count) from "
- "table;"
- "。而后者的查询效率将远高于前者。不过这种方式也有使用限制，就是用户需要自行保证，不会重"
- "复导入 AGGREGATE KEY 列都相同的行。否则，select sum(count) from table; "
- "只能表述原始导入的行数，而不是 select count() from table; 的语义。\n"
+ "增加一个 count 列，并且导入数据中，该列值恒为 1。则 select count() from table; 的结果等价于 select sum(count) from table;。而后者的查询效率将远高于前者。不过这种方式也有使用限制，就是用户需要自行保证，不会重复导入 AGGREGATE KEY 列都相同的行。否则，select sum(count) from table; 只能表述原始导入的行数，而不是 select count() from table; 的语义。\n"
"\n"
- "另一种方式，就是 将如上的 count 列的聚合类型改为 REPLACE，且依然值恒为 1。那么 select "
- "sum(count) from table; 和 select count(*) from table; "
- "的结果将是一致的。并且这种方式，没有导入重复行的限制。\n"
+ "另一种方式，就是 将如上的 count 列的聚合类型改为 REPLACE，且依然值恒为 1。那么 select sum(count) from table; 和 select count(*) from table; 的结果将是一致的。并且这种方式，没有导入重复行的限制。\n"
"\n"
"Unique模型的写时合并实现\n"
- "Unique模型的写时合并实现没有聚合模型的局限性，还是以刚才的数据为例，写时合并为每次导入"
- "的rowset增加了对应的delete bitmap，来标记哪些数据被覆盖。第一批数据导入后状态如下\n"
+ "Unique模型的写时合并实现没有聚合模型的局限性，还是以刚才的数据为例，写时合并为每次导入的rowset增加了对应的delete bitmap，来标记哪些数据被覆盖。第一批数据导入后状态如下\n"
"\n"
"batch 1\n"
"\n"
@@ -1238,176 +1114,147 @@ void testJiebaMatchHuge(CuTest* tc) {
"10001 2017-11-20 1 false\n"
"10001 2017-11-21 5 false\n"
"10003 2017-11-22 22 false\n"
- "在查询时，所有在delete "
- "bitmap中被标记删除的数据都不会读出来，因此也无需进行做任何数据聚合，上述数据中有效的行"
- "数为4行，查询出的结果也应该是4行，也就可以采取开销最小的方式来获取结果，即前面提到的“"
- "仅扫描某一列数据，获得 count 值”的方式。\n"
+ "在查询时，所有在delete bitmap中被标记删除的数据都不会读出来，因此也无需进行做任何数据聚合，上述数据中有效的行数为4行，查询出的结果也应该是4行，也就可以采取开销最小的方式来获取结果，即前面提到的“仅扫描某一列数据，获得 count 值”的方式。\n"
"\n"
- "在测试环境中，count(*) "
- "查询在Unique模型的写时合并实现上的性能，相比聚合模型有10倍以上的提升。\n"
+ "在测试环境中，count(*) 查询在Unique模型的写时合并实现上的性能，相比聚合模型有10倍以上的提升。\n"
"\n"
"Duplicate 模型\n"
- "Duplicate 模型没有聚合模型的这个局限性。因为该模型不涉及聚合语意，在做 count(*) "
- "查询时，任意选择一列查询，即可得到语意正确的结果。\n"
+ "Duplicate 模型没有聚合模型的这个局限性。因为该模型不涉及聚合语意，在做 count(*) 查询时，任意选择一列查询，即可得到语意正确的结果。\n"
"\n"
"key 列\n"
- "Duplicate、Aggregate、Unique 模型，都会在建表指定 key "
- "列，然而实际上是有所区别的：对于 Duplicate 模型，表的key列，可以认为只是 "
- "“排序列”，并非起到唯一标识的作用。而 Aggregate、Unique 模型这种聚合类型的表，key "
- "列是兼顾 “排序列” 和 “唯一标识列”，是真正意义上的“ key 列”。\n"
+ "Duplicate、Aggregate、Unique 模型，都会在建表指定 key 列，然而实际上是有所区别的：对于 Duplicate 模型，表的key列，可以认为只是 “排序列”，并非起到唯一标识的作用。而 Aggregate、Unique 模型这种聚合类型的表，key 列是兼顾 “排序列” 和 “唯一标识列”，是真正意义上的“ key 列”。\n"
"\n"
"数据模型的选择建议\n"
"因为数据模型在建表时就已经确定，且无法修改。所以，选择一个合适的数据模型非常重要。\n"
"\n"
- "Aggregate "
- "模型可以通过预聚合，极大地降低聚合查询时所需扫描的数据量和查询的计算量，非常适合有固定"
- "模式的报表类查询场景。但是该模型对 count(*) 查询很不友好。同时因为固定了 Value "
- "列上的聚合方式，在进行其他类型的聚合查询时，需要考虑语意正确性。\n"
- "Unique 模型针对需要唯一主键约束的场景，可以保证主键唯一性约束。但是无法利用 ROLLUP "
- "等预聚合带来的查询优势。\n"
+ "Aggregate 模型可以通过预聚合，极大地降低聚合查询时所需扫描的数据量和查询的计算量，非常适合有固定模式的报表类查询场景。但是该模型对 count(*) 查询很不友好。同时因为固定了 Value 列上的聚合方式，在进行其他类型的聚合查询时，需要考虑语意正确性。\n"
+ "Unique 模型针对需要唯一主键约束的场景，可以保证主键唯一性约束。但是无法利用 ROLLUP 等预聚合带来的查询优势。\n"
"对于聚合查询有较高性能需求的用户，推荐使用自1.2版本加入的写时合并实现。\n"
- "Unique "
- "模型仅支持整行更新，如果用户既需要唯一主键约束，又需要更新部分列（例如将多张源表导入到"
- "一张 doris 表的情形），则可以考虑使用 Aggregate 模型，同时将非主键列的聚合类型设置为 "
- "REPLACE_IF_NOT_NULL。具体的用法可以参考语法手册\n"
- "Duplicate 适合任意维度的 Ad-hoc "
- "查询。虽然同样无法利用预聚合的特性，但是不受聚合模型的约束，可以发挥列存模型的优势（只"
- "读取相关列，而不需要读取所有 Key 列）。";
- try {
- auto stringReader = std::make_unique<lucene::util::SStringReader<char>>(
- field_value_data, strlen(field_value_data), false);
- auto* stream = analyzer->tokenStream(field->name(), stringReader.get());
- field->setValue(stream);
- w.addDocument(&doc);
-
- w.close();
- doc.clear();
- _CLDELETE(stream)
- } catch (CLuceneError& r) {
- printf("clucene error in testJiebaMatchHuge: %s\n", r.what());
- }
+ "Unique 模型仅支持整行更新，如果用户既需要唯一主键约束，又需要更新部分列（例如将多张源表导入到一张 doris 表的情形），则可以考虑使用 Aggregate 模型，同时将非主键列的聚合类型设置为 REPLACE_IF_NOT_NULL。具体的用法可以参考语法手册\n"
+ "Duplicate 适合任意维度的 Ad-hoc 查询。虽然同样无法利用预聚合的特性，但是不受聚合模型的约束，可以发挥列存模型的优势（只读取相关列，而不需要读取所有 Key 列）。";
+ auto stringReader =
+ _CLNEW lucene::util::SStringReader<char>(field_value_data, strlen(field_value_data), false);
+ field->setValue(stringReader);
+ w.addDocument(&doc);
+
+ w.close();
IndexSearcher searcher(&dir);
- std::vector<std::string> analyse_result;
+ lucene::util::Reader* reader = nullptr;
+ std::vector<std::wstring> analyse_result;
const char* value = "相关";
- auto analyzer1 = std::make_unique<lucene::analysis::LanguageBasedAnalyzer>(L"chinese", false);
- auto reader = std::make_unique<lucene::util::SStringReader<char>>(value, strlen(value), false);
+ analyzer = _CLNEW lucene::analysis::LanguageBasedAnalyzer(L"chinese", false);
+ reader = _CLNEW lucene::util::SStringReader<char>(value, strlen(value), false);
- lucene::analysis::TokenStream* token_stream = analyzer1->tokenStream(field_name, reader.get());
+ lucene::analysis::TokenStream* token_stream = analyzer->tokenStream(field_name, reader);
lucene::analysis::Token token;
while (token_stream->next(&token)) {
- if (token.termLength<char>() != 0) {
- analyse_result.emplace_back(token.termBuffer<char>(), token.termLength<char>());
+ if(token.termLength<TCHAR>() != 0) {
+ analyse_result.emplace_back(token.termBuffer<TCHAR>(), token.termLength<TCHAR>());
}
}
if (token_stream != nullptr) {
token_stream->close();
}
- _CLDELETE(token_stream)
- auto query = std::make_unique<lucene::search::BooleanQuery>();
+
+ lucene::search::Query* query = _CLNEW lucene::search::BooleanQuery();
for (const auto& t : analyse_result) {
- std::wstring token_ws = StringUtil::string_to_wstring(t);
- auto* term = _CLNEW lucene::index::Term(field_name, token_ws.c_str());
- dynamic_cast<lucene::search::BooleanQuery*>(query.get())
+ //std::wstring token_ws = std::wstring(token.begin(), token.end());
+ auto* term =
+ _CLNEW lucene::index::Term(field_name, t.c_str());
+ dynamic_cast<lucene::search::BooleanQuery*>(query)
->add(_CLNEW lucene::search::TermQuery(term), true,
lucene::search::BooleanClause::SHOULD);
_CLDECDELETE(term);
}
- Hits* hits1 = searcher.search(query.get());
+ Hits *hits1 = searcher.search(query);
CLUCENE_ASSERT(1 == hits1->length());
- _CLDELETE(hits1)
- _CLDELETE_ARRAY(field_name)
+
+ doc.clear();
}
void testChineseMatch(CuTest* tc) {
RAMDirectory dir;
- auto* field_name = lucene::util::Misc::_charToWide("chinese");
- auto analyzer = std::make_unique<lucene::analysis::LanguageBasedAnalyzer>();
+
+ auto analyzer = _CLNEW lucene::analysis::LanguageBasedAnalyzer();
analyzer->setLanguage(L"cjk");
- try {
- IndexWriter w(&dir, analyzer.get(), true);
- w.setUseCompoundFile(false);
-
- Document doc;
- auto field = _CLNEW Field(field_name, Field::INDEX_TOKENIZED | Field::STORE_NO);
- doc.add(*field);
-
- const char* field_value_data = "人民可以得到更多实惠";
- auto stringReader = std::make_unique<lucene::util::SStringReader<char>>(
- field_value_data, strlen(field_value_data), false);
- auto* stream = analyzer->tokenStream(field->name(), stringReader.get());
- field->setValue(stream);
- w.addDocument(&doc);
-
- const char* field_value_data1 = "中国人民银行";
- auto stringReader1 = std::make_unique<lucene::util::SStringReader<char>>(
- field_value_data1, strlen(field_value_data1), false);
- auto* stream1 = analyzer->tokenStream(field->name(), stringReader1.get());
- field->setValue(stream1);
- w.addDocument(&doc);
-
- const char* field_value_data2 = "洛杉矶人，洛杉矶居民";
- auto stringReader2 = std::make_unique<lucene::util::SStringReader<char>>(
- field_value_data2, strlen(field_value_data2), false);
- auto* stream2 = analyzer->tokenStream(field->name(), stringReader2.get());
- field->setValue(stream2);
- w.addDocument(&doc);
-
- const char* field_value_data3 = "民族，人民";
- auto stringReader3 = std::make_unique<lucene::util::SStringReader<char>>(
- field_value_data3, strlen(field_value_data3), false);
- auto* stream3 = analyzer->tokenStream(field->name(), stringReader3.get());
- field->setValue(stream3);
- w.addDocument(&doc);
-
- w.close();
- doc.clear();
- _CLDELETE(stream)
- _CLDELETE(stream1)
- _CLDELETE(stream2)
- _CLDELETE(stream3)
- } catch (const CLuceneError& e) {
- std::cout << "clucene error in testChineseMatch:" << e.what();
- }
+
+ IndexWriter w(&dir, analyzer, true);
+ auto field_name = lucene::util::Misc::_charToWide("chinese");
+
+ Document doc;
+ auto field = _CLNEW Field(field_name, Field::INDEX_TOKENIZED | Field::STORE_NO);
+ doc.add(*field);
+
+
+ const char* field_value_data = "人民可以得到更多实惠";
+ auto stringReader =
+ _CLNEW lucene::util::SimpleInputStreamReader(new lucene::util::AStringReader(field_value_data), lucene::util::SimpleInputStreamReader::UTF8);
+ field->setValue(stringReader);
+ w.addDocument(&doc);
+
+ const char* field_value_data1 = "中国人民银行";
+ auto stringReader1 =
+ _CLNEW lucene::util::SimpleInputStreamReader(new lucene::util::AStringReader(field_value_data1), lucene::util::SimpleInputStreamReader::UTF8);
+ field->setValue(stringReader1);
+ w.addDocument(&doc);
+
+ const char* field_value_data2 = "洛杉矶人，洛杉矶居民";
+ auto stringReader2 =
+ _CLNEW lucene::util::SimpleInputStreamReader(new lucene::util::AStringReader(field_value_data2), lucene::util::SimpleInputStreamReader::UTF8);
+ field->setValue(stringReader2);
+ w.addDocument(&doc);
+
+ const char* field_value_data3 = "民族，人民";
+ auto stringReader3 =
+ _CLNEW lucene::util::SimpleInputStreamReader(new lucene::util::AStringReader(field_value_data3), lucene::util::SimpleInputStreamReader::UTF8);
+ field->setValue(stringReader3);
+ w.addDocument(&doc);
+
+ w.close();
+
IndexSearcher searcher(&dir);
- std::vector<std::string> analyse_result;
+ lucene::util::Reader* reader = nullptr;
+ std::vector<std::wstring> analyse_result;
const char* value = "民族";
- auto analyzer1 = std::make_unique<lucene::analysis::LanguageBasedAnalyzer>(L"cjk", false);
- auto reader = std::make_unique<lucene::util::SStringReader<char>>(value, strlen(value), false);
- lucene::analysis::TokenStream* token_stream = analyzer1->tokenStream(field_name, reader.get());
+ analyzer = _CLNEW lucene::analysis::LanguageBasedAnalyzer(L"cjk", false);
+ reader = _CLNEW lucene::util::SimpleInputStreamReader(new lucene::util::AStringReader(value), lucene::util::SimpleInputStreamReader::UTF8);
+
+ lucene::analysis::TokenStream* token_stream = analyzer->tokenStream(field_name, reader);
lucene::analysis::Token token;
while (token_stream->next(&token)) {
- if (token.termLength<char>() != 0) {
- analyse_result.emplace_back(token.termBuffer<char>(), token.termLength<char>());
+ if(token.termLength<TCHAR>() != 0) {
+ analyse_result.emplace_back(token.termBuffer<TCHAR>(), token.termLength<TCHAR>());
}
}
if (token_stream != nullptr) {
token_stream->close();
}
- _CLDELETE(token_stream)
- auto query = std::make_unique<lucene::search::BooleanQuery>();
+
+ lucene::search::Query* query = _CLNEW lucene::search::BooleanQuery();
for (const auto& t : analyse_result) {
- std::wstring token_ws = StringUtil::string_to_wstring(t);
- auto* term = _CLNEW lucene::index::Term(field_name, token_ws.c_str());
- dynamic_cast<lucene::search::BooleanQuery*>(query.get())
+ //std::wstring token_ws = std::wstring(token.begin(), token.end());
+ auto* term =
+ _CLNEW lucene::index::Term(field_name, t.c_str());
+ dynamic_cast<lucene::search::BooleanQuery*>(query)
->add(_CLNEW lucene::search::TermQuery(term), true,
lucene::search::BooleanClause::SHOULD);
_CLDECDELETE(term);
}
- Hits* hits1 = searcher.search(query.get());
+ Hits *hits1 = searcher.search(query);
CLUCENE_ASSERT(1 == hits1->length());
- _CLDELETE(hits1)
- _CLDELETE_ARRAY(field_name)
+
+ doc.clear();
}
void testLanguageBasedAnalyzer(CuTest* tc) {
@@ -1443,8 +1290,8 @@ void testLanguageBasedAnalyzer(CuTest* tc) {
_CLDELETE(ts);
}
-CuSuite* testchinese(void) {
- CuSuite* suite = CuSuiteNew(_T("CLucene chinese tokenizer Test"));
+CuSuite *testchinese(void) {
+ CuSuite *suite = CuSuiteNew(_T("CLucene chinese tokenizer Test"));
SUITE_ADD_TEST(suite, testFile);
SUITE_ADD_TEST(suite, testCJK);
diff --git a/src/test/document/TestDocument.cpp b/src/test/document/TestDocument.cpp
index 7dc44a6864..93d52e8e15 100644
--- a/src/test/document/TestDocument.cpp
+++ b/src/test/document/TestDocument.cpp
@@ -58,123 +58,95 @@ public:
void TestReaderValueField(CuTest *tc) {
RAMDirectory dir;
- SimpleAnalyzer<char> analyzer;
+ SimpleAnalyzer<TCHAR> analyzer;
IndexWriter w(&dir, &analyzer, true);
- w.setUseCompoundFile(false);
auto field_name = lucene::util::Misc::_charToWide("f3");
Document doc;
auto field = _CLNEW Field(field_name, Field::INDEX_TOKENIZED | Field::STORE_NO);
+ auto value1 = lucene::util::Misc::_charToWide("value1");
- auto char_string_reader = std::make_unique<lucene::util::SStringReader<char>>();
- char_string_reader->init("value1", 6, true);
- auto stream = analyzer.tokenStream(field->name(), char_string_reader.get());
- field->setValue(stream);
+ auto stringReader = _CLNEW StringReader(value1, wcslen(value1), false);
+ field->setValue(stringReader);
doc.add(*field);
w.addDocument(&doc);
w.close();
IndexSearcher searcher(&dir);
- auto t1 = std::make_unique<Term>(_T("f3"), _T("value1"));
- auto query1 =std::make_unique<TermQuery>(t1.get());
- Hits *hits1 = searcher.search(query1.get());
+ Term *t1 = _CLNEW Term(_T("f3"), _T("value1"));
+ auto *query1 = _CLNEW TermQuery(t1);
+ Hits *hits1 = searcher.search(query1);
CLUCENE_ASSERT(1 == hits1->length());
- _CLDELETE(stream)
- _CLDELETE(hits1)
- _CLDELETE_ARRAY(field_name)
}
void TestMultiSetValueField(CuTest *tc) {
RAMDirectory dir;
- SimpleAnalyzer<char> analyzer;
+ SimpleAnalyzer<TCHAR> analyzer;
IndexWriter w(&dir, &analyzer, true);
- w.setUseCompoundFile(false);
auto field_name = lucene::util::Misc::_charToWide("f3");
Document doc;
auto field = _CLNEW Field(field_name, Field::INDEX_TOKENIZED | Field::STORE_NO);
- auto char_string_reader = std::make_unique<lucene::util::SStringReader<char>>();
- char_string_reader->init("value1", 6, false);
- auto stream = analyzer.tokenStream(field->name(), char_string_reader.get());
- field->setValue(stream);
- char_string_reader->init("value2", 6, false);
- auto stream2 = analyzer.tokenStream(field->name(), char_string_reader.get());
- field->setValue(stream2);
+ auto value1 = lucene::util::Misc::_charToWide("value1");
+ field->setValue(value1, false);
+ auto value2 = lucene::util::Misc::_charToWide("value2");
+ field->setValue(value2, false);
doc.add(*field);
w.addDocument(&doc);
w.close();
IndexSearcher searcher(&dir);
- auto t1 = std::make_unique<Term>(_T("f3"), _T("value1"));
- auto query1 = std::make_unique<TermQuery>(t1.get());
- Hits *hits1 = searcher.search(query1.get());
+ Term *t1 = _CLNEW Term(_T("f3"), _T("value1"));
+ auto *query1 = _CLNEW TermQuery(t1);
+ Hits *hits1 = searcher.search(query1);
CLUCENE_ASSERT(0 == hits1->length());
- auto t2 = std::make_unique<Term>(_T("f3"), _T("value2"));
- auto query2 = std::make_unique<TermQuery>(t2.get());
- Hits *hits2 = searcher.search(query2.get());
+ Term *t2 = _CLNEW Term(_T("f3"), _T("value2"));
+ auto *query2 = _CLNEW TermQuery(t2);
+ Hits *hits2 = searcher.search(query2);
CLUCENE_ASSERT(1 == hits2->length());
doc.clear();
- _CLDELETE(stream)
- _CLDELETE(stream2)
- _CLDELETE(hits1)
- _CLDELETE(hits2)
- _CLDELETE_ARRAY(field_name)
+ //_CLDELETE(field)
}
void TestMultiAddValueField(CuTest *tc) {
RAMDirectory dir;
auto field_name = lucene::util::Misc::_charToWide("f3");
- SimpleAnalyzer<char> analyzer;
+ SimpleAnalyzer<TCHAR> analyzer;
IndexWriter w(&dir, &analyzer, true);
- w.setUseCompoundFile(false);
Document doc;
- auto field1 = _CLNEW Field(field_name, Field::INDEX_TOKENIZED | Field::STORE_NO);
- auto char_string_reader = std::make_unique<lucene::util::SStringReader<char>>();
- char_string_reader->init("value1", 6, false);
- auto stream = analyzer.tokenStream(field1->name(), char_string_reader.get());
- field1->setValue(stream);
- doc.add(*field1);
-
- auto field2 = _CLNEW Field(field_name, Field::INDEX_TOKENIZED | Field::STORE_NO);
- auto char_string_reader2 = std::make_unique<lucene::util::SStringReader<char>>();
- char_string_reader2->init("value2", 6, false);
- auto stream2 = analyzer.tokenStream(field2->name(), char_string_reader2.get());
- field2->setValue(stream2);
- doc.add(*field2);
-
- try {
- w.addDocument(&doc);
- w.close();
- } catch (CLuceneError& ae) {
- std::cout <<ae.what() << std::endl;
- throw ae;
- }
+ doc.add(*_CLNEW Field(field_name, _T("value1"), Field::INDEX_TOKENIZED | Field::STORE_NO));
+ doc.add(*_CLNEW Field(field_name, _T("value2"), Field::INDEX_TOKENIZED | Field::STORE_NO));
+
+ w.addDocument(&doc);
+ w.close();
- auto t1 = std::make_unique<Term>(_T("f3"), _T("value1"));
- auto query1 = std::make_unique<TermQuery>(t1.get());
+ Term *t1 = _CLNEW Term(_T("f3"), _T("value1"));
+ auto *query1 = _CLNEW TermQuery(t1);
IndexSearcher searcher(&dir);
- Hits *hits1 = searcher.search(query1.get());
+ Hits *hits1 = searcher.search(query1);
CLUCENE_ASSERT(1 == hits1->length());
- auto t2 = std::make_unique<Term>(_T("f3"), _T("value2"));
- auto query2 = std::make_unique<TermQuery>(t2.get());
- Hits *hits2 = searcher.search(query2.get());
+ Term *t2 = _CLNEW Term(_T("f3"), _T("value2"));
+ auto *query2 = _CLNEW TermQuery(t2);
+ Hits *hits2 = searcher.search(query2);
CLUCENE_ASSERT(1 == hits2->length());
doc.removeFields(_T("f3"));
CLUCENE_ASSERT(doc.getFields()->size() == 0);
+
+ _CLDELETE(query1);
+ _CLDELETE(query2);
+ _CLDELETE(t1);
+ _CLDELETE(t2);
_CLDELETE(hits1);
_CLDELETE(hits2);
- _CLDELETE(stream);
- _CLDELETE(stream2);
- _CLDELETE_ARRAY(field_name)
}
void TestFields(CuTest *tc) {
@@ -454,42 +426,38 @@ const int32_t MAX_FIELD_LEN = 0x7FFFFFFFL;
const int32_t MAX_BUFFER_DOCS = 100000000;
const int32_t MERGE_FACTOR = 100000000;
void TestAddDocument(CuTest *tc) {
- try {
- RAMDirectory dir;
- SimpleAnalyzer<char> sanalyzer;
- IndexWriter w(&dir, &sanalyzer, true);
- w.setUseCompoundFile(false);
- w.setMaxBufferedDocs(MAX_BUFFER_DOCS);
- w.setRAMBufferSizeMB(256);
- w.setMaxFieldLength(MAX_FIELD_LEN);
- w.setMergeFactor(MERGE_FACTOR);
- w.setDocumentWriter(_CLNEW SDocumentsWriter<char>(w.getDirectory(), &w));
- Document doc;
- auto field_name = lucene::util::Misc::_charToWide("f3");
- auto field = _CLNEW Field(field_name, Field::INDEX_TOKENIZED | Field::STORE_NO);
- doc.add(*field);
+ RAMDirectory dir;
+ SimpleAnalyzer<char> sanalyzer;
+ IndexWriter w(&dir, NULL, true);
+ w.setUseCompoundFile(false);
+ w.setMaxBufferedDocs(MAX_BUFFER_DOCS);
+ w.setRAMBufferSizeMB(256);
+ w.setMaxFieldLength(MAX_FIELD_LEN);
+ w.setMergeFactor(MERGE_FACTOR);
+ w.setDocumentWriter(_CLNEW SDocumentsWriter<char>(w.getDirectory(), &w));
+ Document doc;
+ auto field_name = lucene::util::Misc::_charToWide("f3");
+ auto field = _CLNEW Field(field_name, Field::INDEX_TOKENIZED | Field::STORE_NO);
+ doc.add(*field);
- for (int i = 0; i <= 2000000; i++) {
- std::string value1 = "value1";
- if (i > 0) value1 = generateRandomString(2000);
- auto stringReader = _CLNEW lucene::util::SStringReader<char>(
- value1.c_str(), strlen(value1.c_str()), false);
- auto stream = sanalyzer.reusableTokenStream(field_name, stringReader);
+ for (int i = 0; i <= 2000000; i++) {
+ std::string value1 = "value1";
+ if (i > 0)
+ value1 = generateRandomString(2000);
+ auto stringReader = _CLNEW lucene::util::SStringReader<char>(
+ value1.c_str(), strlen(value1.c_str()), false);
+ auto stream = sanalyzer.reusableTokenStream(field_name, stringReader);
- field->setValue(stream);
- w.addDocument(&doc, &sanalyzer);
- }
- IndexSearcher searcher(&dir);
- Term* t2 = _CLNEW Term(_T("f3"), _T("value1"));
- auto* query2 = _CLNEW TermQuery(t2);
- Hits* hits2 = searcher.search(query2);
- CLUCENE_ASSERT(1 == hits2->length());
- doc.clear();
- w.close();
- } catch (CLuceneError& ae) {
- std::cout <<ae.what() << std::endl;
- throw ae;
+ field->setValue(stream);
+ w.addDocument(&doc, &sanalyzer);
}
+ IndexSearcher searcher(&dir);
+ Term *t2 = _CLNEW Term(_T("f3"), _T("value1"));
+ auto *query2 = _CLNEW TermQuery(t2);
+ Hits *hits2 = searcher.search(query2);
+ CLUCENE_ASSERT(1 == hits2->length());
+ doc.clear();
+ w.close();
}
void TestNewFieldBench(CuTest *tc) {
@@ -527,10 +495,18 @@ void TestNewFieldBench(CuTest *tc) {
CuSuite *testdocument(void) {
CuSuite *suite = CuSuiteNew(_T("CLucene Document Test"));
+ //SUITE_ADD_TEST(suite, TestCompressedDocument);
+ //SUITE_ADD_TEST(suite, TestBinaryDocument);
+ //SUITE_ADD_TEST(suite, TestLazyCompressedDocument);
+ //SUITE_ADD_TEST(suite, TestLazyBinaryDocument);
+ // SUITE_ADD_TEST(suite, TestFieldSelectors);
SUITE_ADD_TEST(suite, TestFields);
SUITE_ADD_TEST(suite, TestMultiSetValueField);
SUITE_ADD_TEST(suite, TestMultiAddValueField);
+ //SUITE_ADD_TEST(suite, TestSetFieldBench);
+ //SUITE_ADD_TEST(suite, TestNewFieldBench);
SUITE_ADD_TEST(suite, TestReaderValueField);
- //SUITE_ADD_TEST(suite, TestAddDocument);
+ SUITE_ADD_TEST(suite, TestAddDocument);
+ //SUITE_ADD_TEST(suite, TestDateTools);
return suite;
}
diff --git a/src/test/test.h b/src/test/test.h
index cb92953882..a6d72732e7 100644
--- a/src/test/test.h
+++ b/src/test/test.h
@@ -155,7 +155,7 @@ void TestAssertIndexReaderEquals(CuTest *tc, IndexReader* reader1, IndexReader*
extern unittest tests[];
-#define CLUCENE_DATA_LOCATION1 "./data/"
+#define CLUCENE_DATA_LOCATION1 "../../src/test/data/"
#define CLUCENE_DATA_LOCATION2 "../src/test/data/"
#define CLUCENE_DATA_LOCATION3 "../../../src/test/data/"
#define CLUCENE_DATA_LOCATIONENV "srcdir"
diff --git a/src/test/tests.cpp b/src/test/tests.cpp
index 3a193a3c38..282794fe43 100644
--- a/src/test/tests.cpp
+++ b/src/test/tests.cpp
@@ -7,13 +7,45 @@
#include "test.h"
unittest tests[] = {
+// {"threads", testatomicupdates},
+// {"indexreader", testindexreader},
+// {"indexsearcher", testIndexSearcher},
+// {"reuters", testreuters},
{"analysis", testanalysis},
{"analyzers", testanalyzers},
{"analysis", teststandard95},
{"document", testdocument},
{"field", testField},
+// {"numbertools", testNumberTools},
+// {"debug", testdebug},
+// {"ramdirectory", testRAMDirectory},
+// {"indexwriter", testindexwriter},
+// {"indexmodifier", testIndexModifier},
+// {"addIndexesNoOptimize", testAddIndexesNoOptimize},
+// {"highfreq", testhighfreq},
+// {"priorityqueue", testpriorityqueue},
+// {"datetools", testDateTools},
+// {"queryparser", testQueryParser},
+// {"mfqueryparser", testMultiFieldQueryParser},
+// {"boolean", testBoolean},
+// {"search", testsearch},
+// {"rangefilter", testRangeFilter},
+// {"queries", testqueries},
+// {"csrqueries", testConstantScoreQueries},
+// {"termvector", testtermvector},
+// {"sort", testsort},
+// {"duplicates", testduplicates},
+// {"datefilter", testdatefilter},
+// {"wildcard", testwildcard},
+// {"store", teststore},
+// {"utf8", testutf8},
+// {"bitset", testBitSet},
{"bkd", testBKD},
{"MSBRadixSorter",testMSBRadixSorter},
+// {"extractterms", testExtractTerms},
+// {"spanqueries", testSpanQueries},
+// {"stringbuffer", testStringBuffer},
+// {"termvectorsreader", testTermVectorsReader},
{"strconvert", testStrConvert},
{"searchRange", testSearchRange},
{"MultiPhraseQuery", testMultiPhraseQuery},
diff --git a/src/test/util/TestBKD.cpp b/src/test/util/TestBKD.cpp
index 330f72f739..ec6c8f9900 100644
--- a/src/test/util/TestBKD.cpp
+++ b/src/test/util/TestBKD.cpp
@@ -36,30 +36,25 @@ void TestVisitor1::visit(int docID) {
}
}
-int TestVisitor1::matches(uint8_t* packedValue) {
+bool TestVisitor1::matches(uint8_t *packedValue) {
std::vector<uint8_t> result(4);
std::copy(packedValue, packedValue + 4, result.begin());
int x = NumericUtils::sortableBytesToInt(result, 0);
if (x >= queryMin && x <= queryMax) {
- return 0;
- }
- if (x < queryMin) {
- return -1;
- }
- if (x > queryMax) {
- return 1;
+ return true;
}
+ return false;
}
void TestVisitor1::visit(roaring::Roaring *docID, std::vector<uint8_t> &packedValue) {
- if (matches(packedValue.data()) != 0) {
+ if (!matches(packedValue.data())) {
return;
}
visit(*docID);
}
void TestVisitor1::visit(bkd::bkd_docid_set_iterator *iter, std::vector<uint8_t> &packedValue) {
- if (matches(packedValue.data()) != 0) {
+ if (!matches(packedValue.data())) {
return;
}
int32_t docID = iter->docid_set->nextDoc();
@@ -69,7 +64,8 @@ void TestVisitor1::visit(bkd::bkd_docid_set_iterator *iter, std::vector<uint8_t>
}
}
-int TestVisitor1::visit(int docID, std::vector<uint8_t>& packedValue) {
+void TestVisitor1::visit(
+ int docID, std::vector<uint8_t> &packedValue) {
int x = NumericUtils::sortableBytesToInt(packedValue, 0);
if (0) {
wcout << L"visit docID=" << docID << L" x=" << x << endl;
@@ -77,29 +73,17 @@ int TestVisitor1::visit(int docID, std::vector<uint8_t>& packedValue) {
if (x >= queryMin && x <= queryMax) {
//wcout << L"visit docID=" << docID << L" x=" << x << endl;
hits->set(docID);
- return 0;
- }
- if (x < queryMin) {
- return -1;
}
- if (x > queryMax) {
- return 1;
- }
- return 0;
}
-lucene::util::bkd::relation TestVisitor1::compare_prefix(std::vector<uint8_t>& prefix) {
- return lucene::util::bkd::relation::CELL_CROSSES_QUERY;
-}
-
-lucene::util::bkd::relation TestVisitor1::compare(std::vector<uint8_t>& minPacked,
- std::vector<uint8_t>& maxPacked) {
+lucene::util::bkd::relation TestVisitor1::compare(
+ std::vector<uint8_t> &minPacked, std::vector<uint8_t> &maxPacked) {
int min = NumericUtils::sortableBytesToInt(minPacked, 0);
int max = NumericUtils::sortableBytesToInt(maxPacked, 0);
assert(max >= min);
if (0) {
- wcout << L"compare: min=" << min << L" max=" << max << L" vs queryMin=" << queryMin
- << L" queryMax=" << queryMax << endl;
+ wcout << L"compare: min=" << min << L" max=" << max << L" vs queryMin="
+ << queryMin << L" queryMax=" << queryMax << endl;
}
if (max < queryMin || min > queryMax) {
@@ -111,262 +95,104 @@ lucene::util::bkd::relation TestVisitor1::compare(std::vector<uint8_t>& minPacke
}
}
-template <predicate QT>
-TestVisitor<QT>::TestVisitor(const uint8_t* qMin, const uint8_t* qMax, BitSet* h) {
+TestVisitor::TestVisitor(const uint8_t *qMin, const uint8_t *qMax,
+ BitSet *h, predicate p) {
queryMin = qMin;
queryMax = qMax;
hits = h;
+ pred = p;
}
-template <predicate QT>
-int TestVisitor<QT>::matches(uint8_t* packed_value) {
- bool all_greater_than_max = true;
- bool all_within_range = true;
-
+bool TestVisitor::matches(uint8_t *packedValue) {
for (int dim = 0; dim < reader->num_data_dims_; dim++) {
int offset = dim * reader->bytes_per_dim_;
-
- auto result_max = lucene::util::FutureArrays::CompareUnsigned(
- packed_value, offset, offset + reader->bytes_per_dim_, queryMax, offset,
- offset + reader->bytes_per_dim_);
-
- auto result_min = lucene::util::FutureArrays::CompareUnsigned(
- packed_value, offset, offset + reader->bytes_per_dim_, queryMin, offset,
- offset + reader->bytes_per_dim_);
-
- all_greater_than_max &= (result_max > 0);
- all_within_range &= (result_min > 0 && result_max < 0);
-
- if (!all_greater_than_max && !all_within_range) {
- return -1;
- }
- }
-
- if (all_greater_than_max) {
- return 1;
- } else if (all_within_range) {
- return 0;
- } else {
- return -1;
- }
-}
-
-template <>
-int TestVisitor<predicate::EQ>::matches(uint8_t* packed_value) {
- // if query type is equal, query_min == query_max
- if (reader->num_data_dims_ == 1) {
- return std::memcmp(packed_value, queryMin, reader->bytes_per_dim_);
- } else {
- // if all dim value > matched value, then return > 0, otherwise return < 0
- int return_result = 0;
- for (int dim = 0; dim < reader->num_data_dims_; dim++) {
- int offset = dim * reader->bytes_per_dim_;
- auto result = lucene::util::FutureArrays::CompareUnsigned(
- packed_value, offset, offset + reader->bytes_per_dim_, queryMin, offset,
- offset + reader->bytes_per_dim_);
- if (result < 0) {
- return -1;
- } else if (result > 0) {
- return_result = 1;
+ if (pred == L) {
+ if (lucene::util::FutureArrays::CompareUnsigned(
+ packedValue, offset, offset + reader->bytes_per_dim_, queryMax, offset,
+ offset + reader->bytes_per_dim_) >= 0) {
+ // Doc's value is too high, in this dimension
+ return false;
}
- }
- return return_result;
- }
-}
-
-template <>
-int TestVisitor<predicate::L>::matches(uint8_t* packed_value) {
- if (reader->num_data_dims_ == 1) {
- auto result = std::memcmp(packed_value, queryMax, reader->bytes_per_dim_);
- if (result >= 0) {
- return 1;
- }
- return 0;
- } else {
- bool all_greater_or_equal = true;
- bool all_lesser = true;
-
- for (int dim = 0; dim < reader->num_data_dims_; dim++) {
- int offset = dim * reader->bytes_per_dim_;
- auto result = lucene::util::FutureArrays::CompareUnsigned(
- packed_value, offset, offset + reader->bytes_per_dim_, queryMax, offset,
- offset + reader->bytes_per_dim_);
-
- all_greater_or_equal &=
- (result >= 0); // Remains true only if all results are greater or equal
- all_lesser &= (result < 0); // Remains true only if all results are lesser
- }
-
- // Return 1 if all values are greater or equal, 0 if all are lesser, otherwise -1
- return all_greater_or_equal ? 1 : (all_lesser ? 0 : -1);
- }
-}
-
-template <>
-int TestVisitor<predicate::LE>::matches(uint8_t* packed_value) {
- if (reader->num_data_dims_ == 1) {
- auto result = std::memcmp(packed_value, queryMax, reader->bytes_per_dim_);
- if (result > 0) {
- return 1;
- }
- return 0;
- } else {
- bool all_greater = true;
- bool all_lesser_or_equal = true;
-
- for (int dim = 0; dim < reader->num_data_dims_; dim++) {
- int offset = dim * reader->bytes_per_dim_;
- auto result = lucene::util::FutureArrays::CompareUnsigned(
- packed_value, offset, offset + reader->bytes_per_dim_, queryMax, offset,
- offset + reader->bytes_per_dim_);
-
- all_greater &= (result > 0); // Remains true only if all results are greater
- all_lesser_or_equal &=
- (result <= 0); // Remains true only if all results are lesser or equal
- }
-
- // Return 1 if all values are greater or equal, 0 if all are lesser, otherwise -1
- return all_greater ? 1 : (all_lesser_or_equal ? 0 : -1);
- }
-}
-
-template <>
-int TestVisitor<predicate::G>::matches(uint8_t* packed_value) {
- if (reader->num_data_dims_ == 1) {
- auto result = std::memcmp(packed_value, queryMin, reader->bytes_per_dim_);
- if (result <= 0) {
- return -1;
- }
- return 0;
- } else {
- for (int dim = 0; dim < reader->num_data_dims_; dim++) {
- int offset = dim * reader->bytes_per_dim_;
- auto result = lucene::util::FutureArrays::CompareUnsigned(
- packed_value, offset, offset + reader->bytes_per_dim_, queryMin, offset,
- offset + reader->bytes_per_dim_);
- if (result <= 0) {
- return -1;
+ } else if (pred == G) {
+ if (lucene::util::FutureArrays::CompareUnsigned(
+ packedValue, offset, offset + reader->bytes_per_dim_, queryMin, offset,
+ offset + reader->bytes_per_dim_) <= 0) {
+ // Doc's value is too high, in this dimension
+ return false;
}
- }
- return 0;
- }
-}
-
-template <>
-int TestVisitor<predicate::GE>::matches(uint8_t* packed_value) {
- if (reader->num_data_dims_ == 1) {
- auto result = std::memcmp(packed_value, queryMin, reader->bytes_per_dim_);
- if (result < 0) {
- return -1;
- }
- return 0;
- } else {
- for (int dim = 0; dim < reader->num_data_dims_; dim++) {
- int offset = dim * reader->bytes_per_dim_;
- auto result = lucene::util::FutureArrays::CompareUnsigned(
- packed_value, offset, offset + reader->bytes_per_dim_, queryMin, offset,
- offset + reader->bytes_per_dim_);
- if (result < 0) {
- return -1;
+ } else {
+ if (lucene::util::FutureArrays::CompareUnsigned(
+ packedValue, offset, offset + reader->bytes_per_dim_, queryMin, offset,
+ offset + reader->bytes_per_dim_) < 0) {
+ // Doc's value is too low, in this dimension
+ return false;
+ }
+ if (lucene::util::FutureArrays::CompareUnsigned(
+ packedValue, offset, offset + reader->bytes_per_dim_, queryMax, offset,
+ offset + reader->bytes_per_dim_) > 0) {
+ // Doc's value is too high, in this dimension
+ return false;
}
}
- return 0;
}
+ return true;
}
-template <predicate QT>
-void TestVisitor<QT>::visit(int rowID) {
+void TestVisitor::visit(int rowID) {
hits->set(rowID);
if (0) {
std::wcout << L"visit docID=" << rowID << std::endl;
}
}
-template <predicate QT>
-int TestVisitor<QT>::visit(int rowID, std::vector<uint8_t>& packedValue) {
+void TestVisitor::visit(int rowID, std::vector<uint8_t> &packedValue) {
if (0) {
int x = lucene::util::NumericUtils::sortableBytesToLong(packedValue, 0);
std::wcout << L"visit docID=" << rowID << L" x=" << x << std::endl;
}
- auto result = matches(packedValue.data());
- if (result != 0) {
- return result;
+ if (matches(packedValue.data())) {
+ hits->set(rowID);
}
- hits->set(rowID);
- return 0;
}
-template <>
-lucene::util::bkd::relation TestVisitor<predicate::L>::compare(std::vector<uint8_t>& min_packed,
- std::vector<uint8_t>& max_packed) {
+lucene::util::bkd::relation TestVisitor::compare(std::vector<uint8_t> &minPacked,
+ std::vector<uint8_t> &maxPacked) {
bool crosses = false;
- for (int dim = 0; dim < reader->num_data_dims_; dim++) {
- int offset = dim * reader->bytes_per_dim_;
- if (lucene::util::FutureArrays::CompareUnsigned(
- min_packed.data(), offset, offset + reader->bytes_per_dim_, queryMax, offset,
- offset + reader->bytes_per_dim_) >= 0) {
- return lucene::util::bkd::relation::CELL_OUTSIDE_QUERY;
- }
- crosses |= lucene::util::FutureArrays::CompareUnsigned(
- min_packed.data(), offset, offset + reader->bytes_per_dim_, queryMin,
- offset, offset + reader->bytes_per_dim_) <= 0 ||
- lucene::util::FutureArrays::CompareUnsigned(
- max_packed.data(), offset, offset + reader->bytes_per_dim_, queryMax,
- offset, offset + reader->bytes_per_dim_) >= 0;
- }
- if (crosses) {
- return lucene::util::bkd::relation::CELL_CROSSES_QUERY;
- } else {
- return lucene::util::bkd::relation::CELL_INSIDE_QUERY;
- }
-}
-template <>
-lucene::util::bkd::relation TestVisitor<predicate::G>::compare(std::vector<uint8_t>& min_packed,
- std::vector<uint8_t>& max_packed) {
- bool crosses = false;
for (int dim = 0; dim < reader->num_data_dims_; dim++) {
int offset = dim * reader->bytes_per_dim_;
- if (lucene::util::FutureArrays::CompareUnsigned(
- max_packed.data(), offset, offset + reader->bytes_per_dim_, queryMin, offset,
- offset + reader->bytes_per_dim_) <= 0) {
- return lucene::util::bkd::relation::CELL_OUTSIDE_QUERY;
+
+ if (pred == L) {
+ if (lucene::util::FutureArrays::CompareUnsigned(
+ minPacked.data(), offset, offset + reader->bytes_per_dim_, queryMax, offset,
+ offset + reader->bytes_per_dim_) >= 0) {
+ return lucene::util::bkd::relation::CELL_OUTSIDE_QUERY;
+ }
+ } else if (pred == G) {
+ if (lucene::util::FutureArrays::CompareUnsigned(
+ maxPacked.data(), offset, offset + reader->bytes_per_dim_, queryMin, offset,
+ offset + reader->bytes_per_dim_) <= 0) {
+ return lucene::util::bkd::relation::CELL_OUTSIDE_QUERY;
+ }
+ } else {
+ if (lucene::util::FutureArrays::CompareUnsigned(
+ minPacked.data(), offset, offset + reader->bytes_per_dim_, queryMax, offset,
+ offset + reader->bytes_per_dim_) > 0 ||
+ lucene::util::FutureArrays::CompareUnsigned(
+ maxPacked.data(), offset, offset + reader->bytes_per_dim_, queryMin, offset,
+ offset + reader->bytes_per_dim_) < 0) {
+ return lucene::util::bkd::relation::CELL_OUTSIDE_QUERY;
+ }
}
+
crosses |= lucene::util::FutureArrays::CompareUnsigned(
- min_packed.data(), offset, offset + reader->bytes_per_dim_, queryMin,
+ minPacked.data(), offset, offset + reader->bytes_per_dim_, queryMin,
offset, offset + reader->bytes_per_dim_) <= 0 ||
lucene::util::FutureArrays::CompareUnsigned(
- max_packed.data(), offset, offset + reader->bytes_per_dim_, queryMax,
+ maxPacked.data(), offset, offset + reader->bytes_per_dim_, queryMax,
offset, offset + reader->bytes_per_dim_) >= 0;
}
- if (crosses) {
- return lucene::util::bkd::relation::CELL_CROSSES_QUERY;
- } else {
- return lucene::util::bkd::relation::CELL_INSIDE_QUERY;
- }
-}
-template <predicate QT>
-lucene::util::bkd::relation TestVisitor<QT>::compare(std::vector<uint8_t>& min_packed,
- std::vector<uint8_t>& max_packed) {
- bool crosses = false;
- for (int dim = 0; dim < reader->num_data_dims_; dim++) {
- int offset = dim * reader->bytes_per_dim_;
- if (lucene::util::FutureArrays::CompareUnsigned(
- min_packed.data(), offset, offset + reader->bytes_per_dim_, queryMax, offset,
- offset + reader->bytes_per_dim_) > 0 ||
- lucene::util::FutureArrays::CompareUnsigned(
- max_packed.data(), offset, offset + reader->bytes_per_dim_, queryMin, offset,
- offset + reader->bytes_per_dim_) < 0) {
- return lucene::util::bkd::relation::CELL_OUTSIDE_QUERY;
- }
- crosses |= lucene::util::FutureArrays::CompareUnsigned(
- min_packed.data(), offset, offset + reader->bytes_per_dim_, queryMin,
- offset, offset + reader->bytes_per_dim_) < 0 ||
- lucene::util::FutureArrays::CompareUnsigned(
- max_packed.data(), offset, offset + reader->bytes_per_dim_, queryMax,
- offset, offset + reader->bytes_per_dim_) > 0;
- }
if (crosses) {
return lucene::util::bkd::relation::CELL_CROSSES_QUERY;
} else {
@@ -374,23 +200,6 @@ lucene::util::bkd::relation TestVisitor<QT>::compare(std::vector<uint8_t>& min_p
}
}
-template <predicate QT>
-lucene::util::bkd::relation TestVisitor<QT>::compare_prefix(std::vector<uint8_t>& prefix) {
- if (lucene::util::FutureArrays::CompareUnsigned(prefix.data(), 0, prefix.size(), queryMax, 0,
- prefix.size()) > 0 ||
- lucene::util::FutureArrays::CompareUnsigned(prefix.data(), 0, prefix.size(), queryMin, 0,
- prefix.size()) < 0) {
- return lucene::util::bkd::relation::CELL_OUTSIDE_QUERY;
- }
- if (lucene::util::FutureArrays::CompareUnsigned(prefix.data(), 0, prefix.size(), queryMin, 0,
- prefix.size()) > 0 &&
- lucene::util::FutureArrays::CompareUnsigned(prefix.data(), 0, prefix.size(), queryMax, 0,
- prefix.size()) < 0) {
- return lucene::util::bkd::relation::CELL_INSIDE_QUERY;
- }
- return lucene::util::bkd::relation::CELL_CROSSES_QUERY;
-}
-
Directory *getDirectory(int numPoints) {
Directory *dir;
if (numPoints > 100000) {
@@ -470,7 +279,7 @@ void testSameInts1DRead(CuTest *tc) {
r->intersect(v.get());
} catch (CLuceneError &r) {
//printf("something wrong in read\n");
- printf("clucene error in testSameInts1DRead: %s\n", r.what());
+ printf("clucene error: %s\n", r.what());
}
for (int docID = 0; docID < N; docID++) {
bool expected = docID >= queryMin && docID <= queryMax;
@@ -502,7 +311,7 @@ void testSameInts1DRead(CuTest *tc) {
void testBug1Write(CuTest *tc) {
const int N = 8;
- Directory *dir(FSDirectory::getDirectory("testBug1"));
+ Directory *dir(FSDirectory::getDirectory("TestBKDTree"));
shared_ptr<bkd::bkd_writer> w =
make_shared<bkd::bkd_writer>(N, 1, 1, 4, 4, 100.0f, N, true);
w->docs_seen_ = N;
@@ -522,9 +331,9 @@ void testBug1Write(CuTest *tc) {
int64_t indexFP;
{
- std::unique_ptr<IndexOutput> out(dir->createOutput("bkd"));
- std::unique_ptr<IndexOutput> meta_out(dir->createOutput("bkd_meta"));
- std::unique_ptr<IndexOutput> index_out(dir->createOutput("bkd_index"));
+ std::unique_ptr<IndexOutput> out(dir->createOutput("bkd3"));
+ std::unique_ptr<IndexOutput> meta_out(dir->createOutput("bkd3_meta"));
+ std::unique_ptr<IndexOutput> index_out(dir->createOutput("bkd3_index"));
try {
indexFP = w->finish(out.get(), index_out.get());
w->meta_finish(meta_out.get(), indexFP, 0);
@@ -539,15 +348,13 @@ void testBug1Write(CuTest *tc) {
void testBug1Read(CuTest *tc) {
uint64_t str = Misc::currentTimeMillis();
- auto *dir = FSDirectory::getDirectory("testBug1");
+ Directory *dir(FSDirectory::getDirectory("TestBKDTree"));
{
- auto closeDirectory = true;
- auto bkd_reader =
- std::make_shared<lucene::util::bkd::bkd_reader>(dir, closeDirectory);
- if (!bkd_reader->open()) {
- printf("can not open bkd file\n");
- exit(1);
- }
+ IndexInput *in_(dir->openInput("bkd3"));
+ IndexInput *meta_in_(dir->openInput("bkd3_meta"));
+ IndexInput *index_in_(dir->openInput("bkd3_index"));
+
+ shared_ptr<bkd::bkd_reader> r = make_shared<bkd::bkd_reader>(in_);
// Simple 1D range query:
int value = 0;
auto result = std::make_unique<BitSet>(10);
@@ -557,24 +364,27 @@ void testBug1Read(CuTest *tc) {
const auto *max = reinterpret_cast<const uint8_t *>(value_bytes.data());
const auto *min = reinterpret_cast<const uint8_t *>(value_bytes.data());
- auto v = std::make_unique<TestVisitor<EQ>>(min, max, result.get());
+ auto v = std::make_unique<TestVisitor>(min, max, result.get(), EQ);
try {
- v->setReader(bkd_reader);
- bkd_reader->intersect(v.get());
+ v->setReader(r);
+ r->read_meta(meta_in_);
+ //auto type = r->read_type();
+ CuAssertEquals(tc, 0, r->type);
+ r->read_index(index_in_);
+ r->intersect(v.get());
} catch (CLuceneError &r) {
//printf("something wrong in read\n");
- printf("clucene error in testBug1Read: %s\n", r.what());
+ printf("clucene error: %s\n", r.what());
}
//printf("hits count=%d\n", result->count());
CuAssertEquals(tc, result->count(), 6);
//printf("\nFirst search time taken: %d ms\n\n", (int32_t) (Misc::currentTimeMillis() - str));
}
- _CLLDECDELETE(dir)
}
void testLowCardinalInts1DWrite(CuTest *tc) {
const int N = 1024 * 1024;
- Directory *dir(FSDirectory::getDirectory("testLowCardinalInts1D"));
+ Directory *dir(FSDirectory::getDirectory("TestBKDTree"));
shared_ptr<bkd::bkd_writer> w =
make_shared<bkd::bkd_writer>(N, 1, 1, 4, 512, 100.0f, N, true);
w->docs_seen_ = N;
@@ -592,9 +402,9 @@ void testLowCardinalInts1DWrite(CuTest *tc) {
// equivalent: ORIGINAL LINE: try (org.apache.lucene.store.IndexOutput out =
// dir.createOutput("bkd", org.apache.lucene.store.IOContext.DEFAULT))
{
- std::unique_ptr<IndexOutput> out(dir->createOutput("bkd"));
- std::unique_ptr<IndexOutput> meta_out(dir->createOutput("bkd_meta"));
- std::unique_ptr<IndexOutput> index_out(dir->createOutput("bkd_index"));
+ std::unique_ptr<IndexOutput> out(dir->createOutput("bkd2"));
+ std::unique_ptr<IndexOutput> meta_out(dir->createOutput("bkd2_meta"));
+ std::unique_ptr<IndexOutput> index_out(dir->createOutput("bkd2_index"));
//auto metaOffset = w->MetaInit(out.get());
try {
@@ -612,68 +422,68 @@ void testLowCardinalInts1DWrite(CuTest *tc) {
void testLowCardinalInts1DRead2(CuTest *tc) {
uint64_t str = Misc::currentTimeMillis();
const int N = 1024 * 1024;
- Directory *dir = FSDirectory::getDirectory("testLowCardinalInts1D");
+ Directory *dir(FSDirectory::getDirectory("TestBKDTree"));
{
+ IndexInput *in_(dir->openInput("bkd2"));
+ IndexInput *meta_in_(dir->openInput("bkd2_meta"));
+ IndexInput *index_in_(dir->openInput("bkd2_index"));
- auto closeDirectory = true;
- auto bkd_reader =
- std::make_shared<lucene::util::bkd::bkd_reader>(dir, closeDirectory);
- if (!bkd_reader->open()) {
- printf("can not open bkd file\n");
- exit(1);
- }
+ shared_ptr<bkd::bkd_reader> r = make_shared<bkd::bkd_reader>(in_);
// Simple 1D range query:
constexpr int queryMin = 0; //std::numeric_limits<int>::min();
constexpr int queryMax = 100;//std::numeric_limits<int>::max();
auto hits = std::make_shared<BitSet>(N);
auto v = std::make_unique<TestVisitor1>(queryMin, queryMax, hits);
try {
- bkd_reader->intersect(v.get());
+ r->read_meta(meta_in_);
+ //auto type = r->read_type();
+ CuAssertEquals(tc, 0, r->type);
+ r->read_index(index_in_);
+ r->intersect(v.get());
} catch (CLuceneError &r) {
//printf("something wrong in read\n");
- printf("clucene error in testLowCardinalInts1DRead2: %s\n", r.what());
+ printf("clucene error: %s\n", r.what());
}
//printf("hits count=%d\n", hits->count());
CuAssertEquals(tc, hits->count(), 12928);
//printf("\nFirst search time taken: %d ms\n\n", (int32_t) (Misc::currentTimeMillis() - str));
- _CLLDECDELETE(dir)
}
}
void testLowCardinalInts1DRead(CuTest *tc) {
uint64_t str = Misc::currentTimeMillis();
const int N = 1024 * 1024;
- Directory *dir = FSDirectory::getDirectory("testLowCardinalInts1D");
+ Directory *dir(FSDirectory::getDirectory("TestBKDTree"));
{
- auto closeDirectory = true;
- auto bkd_reader =
- std::make_shared<lucene::util::bkd::bkd_reader>(dir, closeDirectory);
- if (!bkd_reader->open()) {
- printf("can not open bkd file\n");
- exit(1);
- }
+ IndexInput *in_(dir->openInput("bkd2"));
+ IndexInput *meta_in_(dir->openInput("bkd2_meta"));
+ IndexInput *index_in_(dir->openInput("bkd2_index"));
+ shared_ptr<bkd::bkd_reader> r = make_shared<bkd::bkd_reader>(in_);
// Simple 1D range query:
constexpr int queryMin = 0;//std::numeric_limits<int>::min();
constexpr int queryMax = 1;//std::numeric_limits<int>::max();
auto hits = std::make_shared<BitSet>(N);
auto v = std::make_unique<TestVisitor1>(queryMin, queryMax, hits);
try {
- bkd_reader->intersect(v.get());
+ r->read_meta(meta_in_);
+ //auto type = r->read_type();
+ CuAssertEquals(tc, 0, r->type);
+ r->read_index(index_in_);
+ r->intersect(v.get());
} catch (CLuceneError &r) {
//printf("something wrong in read\n");
- printf("clucene error in testLowCardinalInts1DRead: %s\n", r.what());
+ printf("clucene error: %s\n", r.what());
}
//printf("hits count=%d\n", hits->count());
CuAssertEquals(tc, hits->count(), 256);
//printf("\nFirst search time taken: %d ms\n\n", (int32_t) (Misc::currentTimeMillis() - str));
- _CLLDECDELETE(dir)
}
}
void testBasicsInts1DWrite(CuTest *tc) {
const int N = 1024 * 1024;
- Directory *dir(FSDirectory::getDirectory("testBasicsInts1D"));
+ Directory *dir(FSDirectory::getDirectory("TestBKDTree"));
shared_ptr<bkd::bkd_writer> w =
make_shared<bkd::bkd_writer>(N, 1, 1, 4, 512, 100.0f, N, true);
w->docs_seen_ = N;
@@ -710,26 +520,26 @@ void testBasicsInts1DWrite(CuTest *tc) {
void testBasicsInts1DRead(CuTest *tc) {
uint64_t str = Misc::currentTimeMillis();
const int N = 1024 * 1024;
- Directory *dir = FSDirectory::getDirectory("testBasicsInts1D");
+ Directory *dir(FSDirectory::getDirectory("TestBKDTree"));
{
- auto closeDirectory = true;
- auto bkd_reader =
- std::make_shared<lucene::util::bkd::bkd_reader>(dir, closeDirectory);
- if (!bkd_reader->open()) {
- printf("can not open bkd file\n");
- exit(1);
- }
-
+ IndexInput *in_(dir->openInput("bkd"));
+ IndexInput *meta_in_(dir->openInput("bkd_meta"));
+ IndexInput *index_in_(dir->openInput("bkd_index"));
+ shared_ptr<bkd::bkd_reader> r = make_shared<bkd::bkd_reader>(in_);
// Simple 1D range query:
constexpr int queryMin = 1024;
constexpr int queryMax = std::numeric_limits<int>::max();
auto hits = std::make_shared<BitSet>(N);
auto v = std::make_unique<TestVisitor1>(queryMin, queryMax, hits);
try {
- bkd_reader->intersect(v.get());
+ r->read_meta(meta_in_);
+ //auto type = r->read_type();
+ CuAssertEquals(tc, 0, r->type);
+ r->read_index(index_in_);
+ r->intersect(v.get());
} catch (CLuceneError &r) {
//printf("something wrong in read\n");
- printf("clucene error in testBasicsInts1DRead: %s\n", r.what());
+ printf("clucene error: %s\n", r.what());
}
for (int docID = 0; docID < N; docID++) {
bool expected = docID >= queryMin && docID <= queryMax;
@@ -746,7 +556,7 @@ void testBasicsInts1DRead(CuTest *tc) {
auto v1 = std::make_unique<TestVisitor1>(queryMin, queryMax, hits1);
str = Misc::currentTimeMillis();
- bkd_reader->intersect(v1.get());
+ r->intersect(v1.get());
for (int docID = 0; docID < N; docID++) {
bool expected = docID >= queryMin && docID <= queryMax;
bool actual = hits1->get(N - docID - 1);
@@ -758,7 +568,7 @@ void testBasicsInts1DRead(CuTest *tc) {
}
//printf("\nSecond search time taken: %d ms\n\n", (int32_t) (Misc::currentTimeMillis() - str));
}
- //dir->close();
+ dir->close();
_CLDECDELETE(dir);
}
@@ -785,7 +595,7 @@ void testHttplogsRead(CuTest *tc) {
const auto *max = reinterpret_cast<const uint8_t *>(scratch2.data());
const auto *min = reinterpret_cast<const uint8_t *>(scratch.data());
- auto v = std::make_unique<TestVisitor<G>>(min, max, result.get());
+ auto v = std::make_unique<TestVisitor>(min, max, result.get(), G);
v->setReader(r);
try {
str = Misc::currentTimeMillis();
@@ -798,7 +608,7 @@ void testHttplogsRead(CuTest *tc) {
//printf("\nsearch time taken: %d ms\n\n", (int32_t) (Misc::currentTimeMillis() - str));
} catch (CLuceneError &r) {
//printf("something wrong in read\n");
- printf("clucene error in testHttplogsRead: %s\n", r.what());
+ printf("clucene error: %s\n", r.what());
}
//printf("result size = %d\n", result->count());
CuAssertEquals(tc, result->count(), 8445);
@@ -886,7 +696,7 @@ void testSame(CuTest *tc) {
{
//std::shared_ptr<Directory> dir{getDirectory(10001)};
const int N = 1024 * 1024;
- Directory *dir(FSDirectory::getDirectory("testSame"));
+ Directory *dir(FSDirectory::getDirectory("TestBKDTree"));
shared_ptr<bkd::bkd_writer> w =
make_shared<bkd::bkd_writer>(N, 1, 1, 4, 512, 100.0f, N, true);
@@ -914,13 +724,11 @@ void testSame(CuTest *tc) {
// equivalent: ORIGINAL LINE: try (org.apache.lucene.store.IndexInput in =
// dir.openInput("bkd", org.apache.lucene.store.IOContext.DEFAULT))
{
- auto closeDirectory = true;
- auto bkd_reader =
- std::make_shared<lucene::util::bkd::bkd_reader>(dir, closeDirectory);
- if (!bkd_reader->open()) {
- printf("can not open bkd file\n");
- exit(1);
- }
+ IndexInput *in_(dir->openInput("bkd"));
+ IndexInput *meta_in_(dir->openInput("bkd_meta"));
+ IndexInput *index_in_(dir->openInput("bkd_index"));
+ //in_->seek(indexFP);
+ shared_ptr<bkd::bkd_reader> r = make_shared<bkd::bkd_reader>(in_);
// Simple 1D range query:
constexpr int queryMin = 100;
@@ -929,7 +737,11 @@ void testSame(CuTest *tc) {
//std::shared_ptr<BitSet> hits;
auto hits = std::make_shared<BitSet>(N);
auto v = std::make_unique<TestVisitor1>(queryMin, queryMax, hits);
- bkd_reader->intersect(v.get());
+ r->read_meta(meta_in_);
+ //auto type = r->read_type();
+ CuAssertEquals(tc, 0, r->type);
+ r->read_index(index_in_);
+ r->intersect(v.get());
for (int docID = 0; docID < N; docID++) {
bool expected = (100 >= queryMin && 100 <= queryMax);
@@ -941,7 +753,7 @@ void testSame(CuTest *tc) {
//assertEquals(L"docID=" + to_wstring(docID), expected, actual);
}
}
- //dir->close();
+ dir->close();
_CLDECDELETE(dir);
}
}
@@ -953,7 +765,7 @@ void equal_predicate(std::shared_ptr<lucene::util::bkd::bkd_reader> r) {
const auto *max = reinterpret_cast<const uint8_t *>(&value);
const auto *min = reinterpret_cast<const uint8_t *>(&value);
- auto v = std::make_unique<TestVisitor<EQ>>(min, max, result.get());
+ auto v = std::make_unique<TestVisitor>(min, max, result.get(), EQ);
v->setReader(r);
r->intersect(v.get());
printf("count: %d\n", result->count());
@@ -976,7 +788,7 @@ void less_equal_predicate(std::shared_ptr<lucene::util::bkd::bkd_reader> r) {
}
const auto *max = reinterpret_cast<const uint8_t *>(&value);
- auto v = std::make_unique<TestVisitor<LE>>(min.data(), max, result.get());
+ auto v = std::make_unique<TestVisitor>(min.data(), max, result.get(), LE);
v->setReader(r);
r->intersect(v.get());
printf("\ncount: %d\n", result->count());
@@ -1002,7 +814,7 @@ void less_predicate(std::shared_ptr<lucene::util::bkd::bkd_reader> r) {
}
const auto *max = reinterpret_cast<const uint8_t *>(&value);
- auto v = std::make_unique<TestVisitor<L>>(min.data(), max, result.get());
+ auto v = std::make_unique<TestVisitor>(min.data(), max, result.get(), L);
v->setReader(r);
r->intersect(v.get());
printf("count: %d\n", result->count());
@@ -1024,7 +836,7 @@ void greater_equal_predicate(std::shared_ptr<lucene::util::bkd::bkd_reader> r) {
}
const auto *min = reinterpret_cast<const uint8_t *>(&value);
- auto v = std::make_unique<TestVisitor<GE>>(min, max.data(), result.get());
+ auto v = std::make_unique<TestVisitor>(min, max.data(), result.get(), GE);
v->setReader(r);
r->intersect(v.get());
printf("count: %d\n", result->count());
@@ -1046,7 +858,7 @@ void greater_predicate(std::shared_ptr<lucene::util::bkd::bkd_reader> r) {
}
const auto *min = reinterpret_cast<const uint8_t *>(&value);
- auto v = std::make_unique<TestVisitor<G>>(min, max.data(), result.get());
+ auto v = std::make_unique<TestVisitor>(min, max.data(), result.get(), G);
v->setReader(r);
r->intersect(v.get());
printf("count: %d\n", result->count());
diff --git a/src/test/util/TestBKD.h b/src/test/util/TestBKD.h
index 327c70fe82..a66f252ad4 100644
--- a/src/test/util/TestBKD.h
+++ b/src/test/util/TestBKD.h
@@ -26,20 +26,19 @@ public:
}
}
void visit(std::vector<char>& docID, std::vector<uint8_t> &packedValue) override {
- if (matches(packedValue.data()) != 0) {
+ if (!matches(packedValue.data())) {
return;
}
visit(roaring::Roaring::read(docID.data(), false));
}
void visit(roaring::Roaring *docID, std::vector<uint8_t> &packedValue) override;
+ void visit(int docID, std::vector<uint8_t> &packedValue) override;
void visit(lucene::util::bkd::bkd_docid_set_iterator *iter, std::vector<uint8_t> &packedValue) override;
- int visit(int docid, std::vector<uint8_t> &packedValue) override;
- int matches(uint8_t *packedValue);
+ bool matches(uint8_t *packedValue);
lucene::util::bkd::relation compare(std::vector<uint8_t> &minPacked,
std::vector<uint8_t> &maxPacked) override;
- lucene::util::bkd::relation compare_prefix(std::vector<uint8_t> &prefix) override;
};
enum predicate {
@@ -50,17 +49,20 @@ enum predicate {
EQ
};
-template <predicate QT>
class TestVisitor : public lucene::util::bkd::bkd_reader::intersect_visitor {
private:
const uint8_t *queryMin;
const uint8_t *queryMax;
+ //int queryMin = 0;
+ //int queryMax = 0;
lucene::util::BitSet *hits;
+ //std::shared_ptr<lucene::util::BitSet> hits;
std::shared_ptr<lucene::util::bkd::bkd_reader> reader;
+ predicate pred;
public:
- TestVisitor(const uint8_t *queryMin, const uint8_t *queryMax, lucene::util::BitSet *hits);
- ~TestVisitor() override = default;
+ TestVisitor(const uint8_t *queryMin, const uint8_t *queryMax, lucene::util::BitSet *hits, predicate p);
+ virtual ~TestVisitor() = default;
void setReader(std::shared_ptr<lucene::util::bkd::bkd_reader> &r) { reader = r; };
@@ -99,10 +101,9 @@ public:
docID = iter->docid_set->nextDoc();
}
};
- int matches(uint8_t *packedValue);
- lucene::util::bkd::relation compare_prefix(std::vector<uint8_t> &prefix) override;
+ bool matches(uint8_t *packedValue);
- int visit(int rowID, std::vector<uint8_t> &packedValue) override;
+ void visit(int rowID, std::vector<uint8_t> &packedValue) override;
lucene::util::bkd::relation compare(std::vector<uint8_t> &minPacked,
std::vector<uint8_t> &maxPacked) override;
diff --git a/src/test/util/TestMSBRadixSorter.cpp b/src/test/util/TestMSBRadixSorter.cpp
index 347ee43e8c..d817e11833 100644
--- a/src/test/util/TestMSBRadixSorter.cpp
+++ b/src/test/util/TestMSBRadixSorter.cpp
@@ -89,21 +89,23 @@ void TestMSBRadixSorter::testOneValue()
void TestMSBRadixSorter::testNValues()
{
const int n = 1000;
+ std::vector<uint8_t> scratch(4);
auto y = std::vector<BytesRef>();
auto z = std::vector<BytesRef>();
for (int docID = 0; docID < n; docID++) {
- std::vector<uint8_t> scratch(4);
NumericUtils::intToSortableBytes(docID, scratch, 0);
- y.emplace_back(scratch);
+ BytesRef x1(scratch);
+ y.emplace_back(x1);
}
+ //for (int docID = 0; docID <n; docID++) {
for (int docID = n-1; docID >= 0; docID--) {
- std::vector<uint8_t> scratch(4);
NumericUtils::intToSortableBytes(docID, scratch, 0);
- z.emplace_back(scratch);
+ BytesRef x1(scratch);
+ z.emplace_back(x1);
}
- test(y, z, n);
+ test(y,z, n);
}
void testSorter(CuTest *tc) {
---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@doris.apache.org
For additional commands, e-mail: commits-help@doris.apache.org