mirror of
https://github.com/yanyiwu/cppjieba.git
synced 2025-07-18 00:00:12 +08:00
更新Demo示例文件，demo只使用一个Application实例即可。
This commit is contained in:
yanyiwu
parent
e5d1ac7bc8
commit
8ce2af9706
27
README.md
27
README.md
@ -42,13 +42,13 @@ make
|
||||
./load_test
|
||||
```
|
||||
|
||||
## 演示
|
||||
## Demo
|
||||
|
||||
```
|
||||
./segment.demo
|
||||
./demo
|
||||
```
|
||||
|
||||
详细请看 `test/segment_demo.cpp`.
|
||||
详细请看 `test/demo.cpp`.
|
||||
|
||||
|
||||
## 服务使用
|
||||
@ -233,9 +233,7 @@ Query方法先使用Mix方法切词，对于切出来的较长的词再使用Ful
|
||||
|
||||
### 自定义用户词典
|
||||
|
||||
自定义词典示例请看`test/testdata/userdict.utf8`。
|
||||
|
||||
载入自定义词典示例请看`test/segment.cpp`，产生的可执行文件示例请见 `build/segment.demo`
|
||||
自定义词典示例请看`dict/user.dict.utf8`。
|
||||
|
||||
没有使用自定义用户词典时的结果:
|
||||
|
||||
@ -251,31 +249,22 @@ Query方法先使用Mix方法切词，对于切出来的较长的词再使用Ful
|
||||
|
||||
### 关键词抽取
|
||||
|
||||
```
|
||||
make && ./keyword.demo
|
||||
```
|
||||
|
||||
你将看到如下结果：
|
||||
|
||||
```
|
||||
我是拖拉机学院手扶拖拉机专业的。不用多久，我就会升职加薪，当上CEO，走上人生巅峰。
|
||||
["CEO:11.7392", "升职:10.8562", "加薪:10.6426", "手扶拖拉机:10.0089", "巅峰:9.49396"]
|
||||
```
|
||||
|
||||
详细请见 `test/keyword_demo.cpp`.
|
||||
详细请见 `test/demo.cpp`.
|
||||
|
||||
### 词性标注
|
||||
|
||||
```
|
||||
./tagging.demo
|
||||
```
|
||||
|
||||
详情请看 `test/tagging_demo.cpp`.
|
||||
|
||||
```
|
||||
我是蓝翔技工拖拉机学院手扶拖拉机专业的。不用多久，我就会升职加薪，当上总经理，出任CEO，迎娶白富美，走上人生巅峰。
|
||||
["我:r", "是:v", "蓝翔:x", "技工:n", "拖拉机:n", "学院:n", "手扶拖拉机:n", "专业:n", "的:uj", "。:x", "不用:v", "多久:m", "，:x", "我:r", "就:d", "会:v", "升职:v", "加薪:nr", "，:x", "当:t", "上:f", "总经理:n", "，:x", "出任:v", "CEO:eng", "，:x", "迎娶:v", "白富美:x", "，:x", "走上:v", "人生:n", "巅峰:n", "。:x"]
|
||||
```
|
||||
|
||||
详细请看 `test/demo.cpp`.
|
||||
|
||||
支持自定义词性。
|
||||
比如在(`dict/user.dict.utf8`)增加一行
|
||||
|
||||
|
||||
@ -1,7 +1,5 @@
|
||||
SET(EXECUTABLE_OUTPUT_PATH ${PROJECT_BINARY_DIR})
|
||||
|
||||
ADD_EXECUTABLE(segment.demo segment_demo.cpp)
|
||||
ADD_EXECUTABLE(keyword.demo keyword_demo.cpp)
|
||||
ADD_EXECUTABLE(tagging.demo tagging_demo.cpp)
|
||||
ADD_EXECUTABLE(demo demo.cpp)
|
||||
ADD_EXECUTABLE(load_test load_test.cpp)
|
||||
ADD_SUBDIRECTORY(unittest)
|
||||
|
||||
72
test/demo.cpp
Normal file
72
test/demo.cpp
Normal file
@ -0,0 +1,72 @@
|
||||
#include "../src/Application.hpp"
|
||||
|
||||
using namespace CppJieba;
|
||||
|
||||
void LoadSentences(const string& filepath, vector<string>& sentences) {
|
||||
ifstream ifile(filepath.c_str());
|
||||
if(!ifile.is_open()) {
|
||||
LogFatal("open %s failed.", filepath.c_str());
|
||||
}
|
||||
string line;
|
||||
while(getline(ifile, line)) {
|
||||
if(!line.empty()) {
|
||||
sentences.push_back(line);
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
int main(int argc, char** argv) {
|
||||
CppJieba::Application app("../dict/");
|
||||
vector<string> words;
|
||||
string result;
|
||||
string s;
|
||||
|
||||
vector<string> sentences;
|
||||
LoadSentences("../test/testdata/testlines.utf8", sentences);
|
||||
|
||||
cout << "\n\e[32m" << "[demo] METHOD_MP" << "\e[0m\n"; // colorful
|
||||
for (size_t i = 0; i < sentences.size(); i++) {
|
||||
app.cut(sentences[i], words, METHOD_MP);
|
||||
cout << join(words.begin(), words.end(), "/") << endl;
|
||||
}
|
||||
|
||||
cout << "\n\e[32m" << "[demo] METHOD_HMM" << "\e[0m\n"; // colorful
|
||||
for (size_t i = 0; i < sentences.size(); i++) {
|
||||
app.cut(sentences[i], words, METHOD_HMM);
|
||||
cout << join(words.begin(), words.end(), "/") << endl;
|
||||
}
|
||||
|
||||
cout << "\n\e[32m" << "[demo] METHOD_MIX" << "\e[0m\n"; // colorful
|
||||
for (size_t i = 0; i < sentences.size(); i++) {
|
||||
app.cut(sentences[i], words, METHOD_MIX);
|
||||
cout << join(words.begin(), words.end(), "/") << endl;
|
||||
}
|
||||
|
||||
cout << "\n\e[32m" << "[demo] METHOD_FULL" << "\e[0m\n"; // colorful
|
||||
for (size_t i = 0; i < sentences.size(); i++) {
|
||||
app.cut(sentences[i], words, METHOD_FULL);
|
||||
cout << join(words.begin(), words.end(), "/") << endl;
|
||||
}
|
||||
|
||||
cout << "\n\e[32m" << "[demo] METHOD_QUERY" << "\e[0m\n"; // colorful
|
||||
for (size_t i = 0; i < sentences.size(); i++) {
|
||||
app.cut(sentences[i], words, METHOD_QUERY);
|
||||
cout << join(words.begin(), words.end(), "/") << endl;
|
||||
}
|
||||
|
||||
cout << "\n\e[32m" << "[demo] TAGGING" << "\e[0m\n"; // colorful
|
||||
vector<pair<string, string> > tagres;
|
||||
s = "我是蓝翔技工拖拉机学院手扶拖拉机专业的。不用多久，我就会升职加薪，当上总经理，出任CEO，迎娶白富美，走上人生巅峰。";
|
||||
app.tag(s, tagres);
|
||||
cout << s << endl;
|
||||
cout << tagres << endl;;
|
||||
|
||||
cout << "\n\e[32m" << "[demo] KEYWORD" << "\e[0m\n"; // colorful
|
||||
vector<pair<string, double> > keywordres;
|
||||
s = "我是拖拉机学院手扶拖拉机专业的。不用多久，我就会升职加薪，当上CEO，走上人生巅峰。";
|
||||
app.extract(s, keywordres, 5);
|
||||
cout << s << endl;
|
||||
cout << keywordres << endl;
|
||||
|
||||
return EXIT_SUCCESS;
|
||||
}
|
||||
@ -1,16 +0,0 @@
|
||||
#include "../src/KeywordExtractor.hpp"
|
||||
using namespace CppJieba;
|
||||
|
||||
int main(int argc, char ** argv) {
|
||||
KeywordExtractor extractor("../dict/jieba.dict.utf8", "../dict/hmm_model.utf8", "../dict/idf.utf8", "../dict/stop_words.utf8");
|
||||
//KeywordExtractor extractor("../dict/jieba.dict.utf8", "../dict/hmm_model.utf8", "../dict/idf.utf8", "../dict/stop_words.utf8", "../dict/user.dict.utf8");
|
||||
string s("我是拖拉机学院手扶拖拉机专业的。不用多久，我就会升职加薪，当上CEO，走上人生巅峰。");
|
||||
vector<pair<string, double> > wordweights;
|
||||
vector<string> words;
|
||||
size_t topN = 5;
|
||||
extractor.extract(s, wordweights, topN);
|
||||
cout<< s << '\n' << wordweights << endl;
|
||||
extractor.extract(s, words, topN);
|
||||
cout<< s << '\n' << words << endl;
|
||||
return EXIT_SUCCESS;
|
||||
}
|
||||
@ -1,55 +0,0 @@
|
||||
#include <iostream>
|
||||
#include <fstream>
|
||||
|
||||
#define LOGGER_LEVEL LL_WARN
|
||||
|
||||
#include "../src/MPSegment.hpp"
|
||||
#include "../src/HMMSegment.hpp"
|
||||
#include "../src/MixSegment.hpp"
|
||||
|
||||
using namespace CppJieba;
|
||||
|
||||
const char * const TEST_FILE = "../test/testdata/testlines.utf8";
|
||||
const char * const JIEBA_DICT_FILE = "../dict/jieba.dict.utf8";
|
||||
const char * const HMM_DICT_FILE = "../dict/hmm_model.utf8";
|
||||
const char * const USER_DICT_FILE = "../dict/user.dict.utf8";
|
||||
|
||||
void cut(const ISegment& seg, const char * const filePath) {
|
||||
ifstream ifile(filePath);
|
||||
vector<string> words;
|
||||
string line;
|
||||
string res;
|
||||
while(getline(ifile, line)) {
|
||||
if(!line.empty()) {
|
||||
words.clear();
|
||||
seg.cut(line, words);
|
||||
join(words.begin(), words.end(), res, "/");
|
||||
cout<< res <<endl;
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
|
||||
int main(int argc, char ** argv) {
|
||||
{
|
||||
printf("\e[32m%s\e[0m\n", "[demo] MPSegment"); // colorful
|
||||
MPSegment seg(JIEBA_DICT_FILE);
|
||||
cut(seg, TEST_FILE);
|
||||
}
|
||||
{
|
||||
printf("\e[32m%s\e[0m\n", "[demo] HMMSegment"); // colorful
|
||||
HMMSegment seg(HMM_DICT_FILE);
|
||||
cut(seg, TEST_FILE);
|
||||
}
|
||||
{
|
||||
printf("\e[32m%s\e[0m\n", "[demo] MixSegment"); // colorful
|
||||
MixSegment seg(JIEBA_DICT_FILE, HMM_DICT_FILE);
|
||||
cut(seg, TEST_FILE);
|
||||
}
|
||||
{
|
||||
printf("\e[32m%s\e[0m\n", "[demo] MixSegment with UserDict"); // colorful
|
||||
MixSegment seg(JIEBA_DICT_FILE, HMM_DICT_FILE, USER_DICT_FILE);
|
||||
cut(seg, TEST_FILE);
|
||||
}
|
||||
return EXIT_SUCCESS;
|
||||
}
|
||||
@ -1,11 +0,0 @@
|
||||
#include "../src/PosTagger.hpp"
|
||||
using namespace CppJieba;
|
||||
|
||||
int main(int argc, char ** argv) {
|
||||
PosTagger tagger("../dict/jieba.dict.utf8", "../dict/hmm_model.utf8", "../dict/user.dict.utf8");
|
||||
string s("我是蓝翔技工拖拉机学院手扶拖拉机专业的。不用多久，我就会升职加薪，当上总经理，出任CEO，迎娶白富美，走上人生巅峰。");
|
||||
vector<pair<string, string> > res;
|
||||
tagger.tag(s, res);
|
||||
cout << res << endl;
|
||||
return EXIT_SUCCESS;
|
||||
}
|
||||
Loading…
x
Reference in New Issue
Block a user