third_codes/cppjieba
1
0
Fork 0
mirror of https://github.com/yanyiwu/cppjieba.git synced 2025-07-18 00:00:12 +08:00
Code Issues Packages Projects Releases Wiki Activity

rename dicts -> dict

Browse Source
This commit is contained in:
wyy 2013-12-24 02:32:00 -08:00
parent 657aee0fda
commit 418b18db55
18 changed files with 23 additions and 32 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
CMakeLists.txt
View File

@ -9,7 +9,7 @@ IF (DEFINED ENC)
ENDIF() ENDIF()
#ADD_DEFINITIONS(-DNO_FILTER) #ADD_DEFINITIONS(-DNO_FILTER)
ADD_SUBDIRECTORY(src) ADD_SUBDIRECTORY(src)
ADD_SUBDIRECTORY(dicts) ADD_SUBDIRECTORY(dict)
ADD_SUBDIRECTORY(scripts) ADD_SUBDIRECTORY(scripts)
ADD_SUBDIRECTORY(conf) ADD_SUBDIRECTORY(conf)
ADD_SUBDIRECTORY(test) ADD_SUBDIRECTORY(test)

4
conf/server.conf
View File

@ -10,10 +10,10 @@ thread_num=4
daemonize=true daemonize=true
#dict path #dict path
dict_path=/usr/share/CppJieba/dicts/jieba.dict.utf8 dict_path=/usr/share/CppJieba/dict/jieba.dict.utf8
#model path #model path
model_path=/usr/share/CppJieba/dicts/hmm_model.utf8 model_path=/usr/share/CppJieba/dict/hmm_model.utf8
#pid file #pid file
pid_file=/var/run/CppJieba/cjserver.pid pid_file=/var/run/CppJieba/cjserver.pid

0
dicts/CMakeLists.txt → dict/CMakeLists.txt
View File

0
dicts/hmm_model.gbk → dict/hmm_model.gbk
View File

0
dicts/hmm_model.utf8 → dict/hmm_model.utf8
View File

0
dicts/jieba.dict.gbk → dict/jieba.dict.gbk
View File

0
dicts/jieba.dict.utf8 → dict/jieba.dict.utf8
View File

2
scripts/cjseg.sh
View File

@ -2,4 +2,4 @@ if [ $# -lt 1 ]; then
echo "usage: $0 <file>" echo "usage: $0 <file>"
exit 1 exit 1
fi fi
cjsegment --dictpath /usr/share/CppJieba/dicts/jieba.dict.utf8 --modelpath /usr/share/CppJieba/dicts/hmm_model.utf8 $1 cjsegment --dictpath /usr/share/CppJieba/dict/jieba.dict.utf8 --modelpath /usr/share/CppJieba/dict/hmm_model.utf8 $1

10
src/segment.cpp
View File

@ -45,11 +45,11 @@ int main(int argc, char ** argv)
<<"\t--modelpath\tsee example\n" <<"\t--modelpath\tsee example\n"
<<"\t--maxlen\tspecify the granularity of cut used in cutQuery. \n\t\t\tIf not specified, the default is 3\n" <<"\t--maxlen\tspecify the granularity of cut used in cutQuery. \n\t\t\tIf not specified, the default is 3\n"
<<"example:\n" <<"example:\n"
<<"\t"<<argv[0]<<" ../test/testdata/testlines.utf8 --dictpath ../dicts/jieba.dict.utf8 --algorithm cutDAG\n" <<"\t"<<argv[0]<<" ../test/testdata/testlines.utf8 --dictpath ../dict/jieba.dict.utf8 --algorithm cutDAG\n"
<<"\t"<<argv[0]<<" ../test/testdata/testlines.utf8 --dictpath ../dicts/jieba.dict.utf8 --algorithm cutFull\n" <<"\t"<<argv[0]<<" ../test/testdata/testlines.utf8 --dictpath ../dict/jieba.dict.utf8 --algorithm cutFull\n"
<<"\t"<<argv[0]<<" ../test/testdata/testlines.utf8 --modelpath ../dicts/hmm_model.utf8 --algorithm cutHMM\n" <<"\t"<<argv[0]<<" ../test/testdata/testlines.utf8 --modelpath ../dict/hmm_model.utf8 --algorithm cutHMM\n"
<<"\t"<<argv[0]<<" ../test/testdata/testlines.utf8 --dictpath ../dicts/jieba.dict.utf8 --modelpath ../dicts/hmm_model.utf8 --algorithm cutMix\n" <<"\t"<<argv[0]<<" ../test/testdata/testlines.utf8 --dictpath ../dict/jieba.dict.utf8 --modelpath ../dict/hmm_model.utf8 --algorithm cutMix\n"
<<"\t"<<argv[0]<<" ../test/testdata/testlines.utf8 --dictpath ../dicts/jieba.dict.utf8 --modelpath ../dicts/hmm_model.utf8 --algorithm cutQuery --maxlen 3\n" <<"\t"<<argv[0]<<" ../test/testdata/testlines.utf8 --dictpath ../dict/jieba.dict.utf8 --modelpath ../dict/hmm_model.utf8 --algorithm cutQuery --maxlen 3\n"
<<endl; <<endl;
return EXIT_FAILURE; return EXIT_FAILURE;

17
test/load_test.cpp
View File

@ -22,28 +22,19 @@ void cut(const ISegment * seg, const char * const filePath, size_t times = 10)
for(uint i = 0; i < times; i ++) for(uint i = 0; i < times; i ++)
{ {
LogInfo("times[%u]", i); LogInfo("times[%u]", i);
//ifile.seekg(0); res.clear();
//while(getline(ifile, line)) seg->cut(doc, res);
//{
// if(!line.empty())
// {
res.clear();
seg->cut(doc, res);
//print(res);
//cout<<join(res.begin(), res.end(),"/")<<endl;
// }
//}
} }
} }
int main(int argc, char ** argv) int main(int argc, char ** argv)
{ {
{ {
MixSegment seg("../dicts/jieba.dict.utf8", "../dicts/hmm_model.utf8"); MixSegment seg("../dict/jieba.dict.utf8", "../dict/hmm_model.utf8");
if(!seg) if(!seg)
{ {
cout<<"seg init failed."<<endl; cout<<"seg init failed."<<endl;
return false; return EXIT_FAILURE;
} }
cut(&seg, "../test/testdata/weicheng.utf8"); cut(&seg, "../test/testdata/weicheng.utf8");
} }

4
test/segment.cpp
View File

@ -23,8 +23,8 @@ void cut(const ISegment * seg, const char * const filePath)
} }
const char * const TEST_FILE = "../test/testdata/testlines.utf8"; const char * const TEST_FILE = "../test/testdata/testlines.utf8";
const char * const JIEBA_DICT_FILE = "../dicts/jieba.dict.utf8"; const char * const JIEBA_DICT_FILE = "../dict/jieba.dict.utf8";
const char * const HMM_DICT_FILE = "../dicts/hmm_model.utf8"; const char * const HMM_DICT_FILE = "../dict/hmm_model.utf8";
int main(int argc, char ** argv) int main(int argc, char ** argv)
{ {

4
test/server.cpp
View File

@ -8,8 +8,8 @@
using namespace Husky; using namespace Husky;
using namespace CppJieba; using namespace CppJieba;
const char * const DEFAULT_DICTPATH = "../dicts/jieba.dict.utf8"; const char * const DEFAULT_DICTPATH = "../dict/jieba.dict.utf8";
const char * const DEFAULT_MODELPATH = "../dicts/hmm_model.utf8"; const char * const DEFAULT_MODELPATH = "../dict/hmm_model.utf8";
class ServerDemo: public IRequestHandler class ServerDemo: public IRequestHandler
{ {

2
test/unittest/TFullSegment.cpp
View File

@ -5,7 +5,7 @@ using namespace CppJieba;
TEST(FullSegment, Test1) TEST(FullSegment, Test1)
{ {
FullSegment segment("../dicts/jieba.dict.utf8"); FullSegment segment("../dict/jieba.dict.utf8");
const char* str = "我来自北京邮电大学。。。 学号 123456"; const char* str = "我来自北京邮电大学。。。 学号 123456";
const char* res[] = {"我", "来自", "北京", "北京邮电", "北京邮电大学", "邮电", "邮电大学", "电大", "大学", "。", "。", "。", " ", "学号", " 123456"}; const char* res[] = {"我", "来自", "北京", "北京邮电", "北京邮电大学", "邮电", "邮电大学", "电大", "大学", "。", "。", "。", " ", "学号", " 123456"};
vector<string> words; vector<string> words;

2
test/unittest/THMMSegment.cpp
View File

@ -5,7 +5,7 @@ using namespace CppJieba;
TEST(HMMSegmentTest, Test1) TEST(HMMSegmentTest, Test1)
{ {
HMMSegment segment("../dicts/hmm_model.utf8");; HMMSegment segment("../dict/hmm_model.utf8");;
const char* str = "我来自北京邮电大学。。。 学号 123456"; const char* str = "我来自北京邮电大学。。。 学号 123456";
const char* res[] = {"我来", "自北京", "邮电大学", "。", "。", "。", " ", "学号", " 123456"}; const char* res[] = {"我来", "自北京", "邮电大学", "。", "。", "。", " ", "学号", " 123456"};
//string s; //string s;

2
test/unittest/TMPSegment.cpp
View File

@ -5,7 +5,7 @@ using namespace CppJieba;
TEST(MPSegmentTest, Test1) TEST(MPSegmentTest, Test1)
{ {
MPSegment segment("../dicts/jieba.dict.utf8");; MPSegment segment("../dict/jieba.dict.utf8");;
const char* str = "我来自北京邮电大学。。。 学号 123456"; const char* str = "我来自北京邮电大学。。。 学号 123456";
const char* res[] = {"我", "来自", "北京邮电大学", "。","。","。"," ","学","号", " 123456"}; const char* res[] = {"我", "来自", "北京邮电大学", "。","。","。"," ","学","号", " 123456"};
vector<string> words; vector<string> words;

2
test/unittest/TMixSegment.cpp
View File

@ -5,7 +5,7 @@ using namespace CppJieba;
TEST(MixSegmentTest, Test1) TEST(MixSegmentTest, Test1)
{ {
MixSegment segment("../dicts/jieba.dict.utf8", "../dicts/hmm_model.utf8");; MixSegment segment("../dict/jieba.dict.utf8", "../dict/hmm_model.utf8");;
const char* str = "我来自北京邮电大学。。。 学号 123456"; const char* str = "我来自北京邮电大学。。。 学号 123456";
const char* res[] = {"我", "来自", "北京邮电大学", "。","。","。"," ","学号", " 123456"}; const char* res[] = {"我", "来自", "北京邮电大学", "。","。","。"," ","学号", " 123456"};
vector<string> words; vector<string> words;

2
test/unittest/TQuerySegment.cpp
View File

@ -5,7 +5,7 @@ using namespace CppJieba;
TEST(QuerySegment, Test1) TEST(QuerySegment, Test1)
{ {
QuerySegment segment("../dicts/jieba.dict.utf8", "../dicts/hmm_model.utf8", 3); QuerySegment segment("../dict/jieba.dict.utf8", "../dict/hmm_model.utf8", 3);
const char* str = "小明硕士毕业于中国科学院计算所，后在日本京都大学深造"; const char* str = "小明硕士毕业于中国科学院计算所，后在日本京都大学深造";
const char* res[] = {"小明", "硕士", "毕业", "于", "中国", "中国科学院", "科学", "科学院", "学院", "计算所", "，", "后", "在", "日本", "日本京都大学", "京都", "京都大学", "大学", "深造"}; const char* res[] = {"小明", "硕士", "毕业", "于", "中国", "中国科学院", "科学", "科学院", "学院", "计算所", "，", "后", "在", "日本", "日本京都大学", "京都", "京都大学", "大学", "深造"};
vector<string> words; vector<string> words;

2
test/unittest/TTrie.cpp
View File

@ -3,7 +3,7 @@
using namespace CppJieba; using namespace CppJieba;
static const char* const DICT_FILE = "../dicts/jieba.dict.utf8"; static const char* const DICT_FILE = "../dict/jieba.dict.utf8";
TEST(TrieTest, Test1) TEST(TrieTest, Test1)
{ {