cppjieba_rb 0.2.1
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +7 -0
- data/.gitignore +18 -0
- data/.gitmodules +3 -0
- data/.travis.yml +26 -0
- data/Gemfile +3 -0
- data/LICENSE.txt +22 -0
- data/README.md +81 -0
- data/Rakefile +20 -0
- data/cppjieba_rb.gemspec +50 -0
- data/ext/cppjieba/.gitignore +17 -0
- data/ext/cppjieba/.travis.yml +22 -0
- data/ext/cppjieba/CMakeLists.txt +28 -0
- data/ext/cppjieba/ChangeLog.md +236 -0
- data/ext/cppjieba/README.md +285 -0
- data/ext/cppjieba/README_EN.md +111 -0
- data/ext/cppjieba/appveyor.yml +32 -0
- data/ext/cppjieba/deps/CMakeLists.txt +1 -0
- data/ext/cppjieba/deps/gtest/CMakeLists.txt +5 -0
- data/ext/cppjieba/deps/gtest/include/gtest/gtest-death-test.h +283 -0
- data/ext/cppjieba/deps/gtest/include/gtest/gtest-message.h +230 -0
- data/ext/cppjieba/deps/gtest/include/gtest/gtest-param-test.h +1421 -0
- data/ext/cppjieba/deps/gtest/include/gtest/gtest-param-test.h.pump +487 -0
- data/ext/cppjieba/deps/gtest/include/gtest/gtest-printers.h +796 -0
- data/ext/cppjieba/deps/gtest/include/gtest/gtest-spi.h +232 -0
- data/ext/cppjieba/deps/gtest/include/gtest/gtest-test-part.h +176 -0
- data/ext/cppjieba/deps/gtest/include/gtest/gtest-typed-test.h +259 -0
- data/ext/cppjieba/deps/gtest/include/gtest/gtest.h +2155 -0
- data/ext/cppjieba/deps/gtest/include/gtest/gtest_pred_impl.h +358 -0
- data/ext/cppjieba/deps/gtest/include/gtest/gtest_prod.h +58 -0
- data/ext/cppjieba/deps/gtest/include/gtest/internal/gtest-death-test-internal.h +308 -0
- data/ext/cppjieba/deps/gtest/include/gtest/internal/gtest-filepath.h +210 -0
- data/ext/cppjieba/deps/gtest/include/gtest/internal/gtest-internal.h +1226 -0
- data/ext/cppjieba/deps/gtest/include/gtest/internal/gtest-linked_ptr.h +233 -0
- data/ext/cppjieba/deps/gtest/include/gtest/internal/gtest-param-util-generated.h +4822 -0
- data/ext/cppjieba/deps/gtest/include/gtest/internal/gtest-param-util-generated.h.pump +301 -0
- data/ext/cppjieba/deps/gtest/include/gtest/internal/gtest-param-util.h +619 -0
- data/ext/cppjieba/deps/gtest/include/gtest/internal/gtest-port.h +1788 -0
- data/ext/cppjieba/deps/gtest/include/gtest/internal/gtest-string.h +350 -0
- data/ext/cppjieba/deps/gtest/include/gtest/internal/gtest-tuple.h +968 -0
- data/ext/cppjieba/deps/gtest/include/gtest/internal/gtest-tuple.h.pump +336 -0
- data/ext/cppjieba/deps/gtest/include/gtest/internal/gtest-type-util.h +3330 -0
- data/ext/cppjieba/deps/gtest/include/gtest/internal/gtest-type-util.h.pump +296 -0
- data/ext/cppjieba/deps/gtest/src/.deps/.dirstamp +0 -0
- data/ext/cppjieba/deps/gtest/src/.deps/gtest-all.Plo +681 -0
- data/ext/cppjieba/deps/gtest/src/.deps/gtest_main.Plo +509 -0
- data/ext/cppjieba/deps/gtest/src/.dirstamp +0 -0
- data/ext/cppjieba/deps/gtest/src/gtest-all.cc +48 -0
- data/ext/cppjieba/deps/gtest/src/gtest-death-test.cc +1234 -0
- data/ext/cppjieba/deps/gtest/src/gtest-filepath.cc +380 -0
- data/ext/cppjieba/deps/gtest/src/gtest-internal-inl.h +1038 -0
- data/ext/cppjieba/deps/gtest/src/gtest-port.cc +746 -0
- data/ext/cppjieba/deps/gtest/src/gtest-printers.cc +356 -0
- data/ext/cppjieba/deps/gtest/src/gtest-test-part.cc +110 -0
- data/ext/cppjieba/deps/gtest/src/gtest-typed-test.cc +110 -0
- data/ext/cppjieba/deps/gtest/src/gtest.cc +4898 -0
- data/ext/cppjieba/deps/gtest/src/gtest_main.cc +39 -0
- data/ext/cppjieba/deps/limonp/ArgvContext.hpp +70 -0
- data/ext/cppjieba/deps/limonp/BlockingQueue.hpp +49 -0
- data/ext/cppjieba/deps/limonp/BoundedBlockingQueue.hpp +67 -0
- data/ext/cppjieba/deps/limonp/BoundedQueue.hpp +65 -0
- data/ext/cppjieba/deps/limonp/Closure.hpp +206 -0
- data/ext/cppjieba/deps/limonp/Colors.hpp +31 -0
- data/ext/cppjieba/deps/limonp/Condition.hpp +38 -0
- data/ext/cppjieba/deps/limonp/Config.hpp +103 -0
- data/ext/cppjieba/deps/limonp/FileLock.hpp +74 -0
- data/ext/cppjieba/deps/limonp/ForcePublic.hpp +7 -0
- data/ext/cppjieba/deps/limonp/LocalVector.hpp +139 -0
- data/ext/cppjieba/deps/limonp/Logging.hpp +76 -0
- data/ext/cppjieba/deps/limonp/Md5.hpp +411 -0
- data/ext/cppjieba/deps/limonp/MutexLock.hpp +51 -0
- data/ext/cppjieba/deps/limonp/NonCopyable.hpp +21 -0
- data/ext/cppjieba/deps/limonp/StdExtension.hpp +159 -0
- data/ext/cppjieba/deps/limonp/StringUtil.hpp +365 -0
- data/ext/cppjieba/deps/limonp/Thread.hpp +44 -0
- data/ext/cppjieba/deps/limonp/ThreadPool.hpp +86 -0
- data/ext/cppjieba/dict/README.md +31 -0
- data/ext/cppjieba/dict/hmm_model.utf8 +34 -0
- data/ext/cppjieba/dict/idf.utf8 +258826 -0
- data/ext/cppjieba/dict/jieba.dict.utf8 +348982 -0
- data/ext/cppjieba/dict/pos_dict/char_state_tab.utf8 +6653 -0
- data/ext/cppjieba/dict/pos_dict/prob_emit.utf8 +166 -0
- data/ext/cppjieba/dict/pos_dict/prob_start.utf8 +259 -0
- data/ext/cppjieba/dict/pos_dict/prob_trans.utf8 +5222 -0
- data/ext/cppjieba/dict/stop_words.utf8 +1534 -0
- data/ext/cppjieba/dict/user.dict.utf8 +4 -0
- data/ext/cppjieba/include/cppjieba/DictTrie.hpp +227 -0
- data/ext/cppjieba/include/cppjieba/FullSegment.hpp +93 -0
- data/ext/cppjieba/include/cppjieba/HMMModel.hpp +129 -0
- data/ext/cppjieba/include/cppjieba/HMMSegment.hpp +190 -0
- data/ext/cppjieba/include/cppjieba/Jieba.hpp +108 -0
- data/ext/cppjieba/include/cppjieba/KeywordExtractor.hpp +153 -0
- data/ext/cppjieba/include/cppjieba/MPSegment.hpp +137 -0
- data/ext/cppjieba/include/cppjieba/MixSegment.hpp +109 -0
- data/ext/cppjieba/include/cppjieba/PosTagger.hpp +77 -0
- data/ext/cppjieba/include/cppjieba/PreFilter.hpp +54 -0
- data/ext/cppjieba/include/cppjieba/QuerySegment.hpp +90 -0
- data/ext/cppjieba/include/cppjieba/SegmentBase.hpp +46 -0
- data/ext/cppjieba/include/cppjieba/SegmentTagged.hpp +23 -0
- data/ext/cppjieba/include/cppjieba/TextRankExtractor.hpp +190 -0
- data/ext/cppjieba/include/cppjieba/Trie.hpp +174 -0
- data/ext/cppjieba/include/cppjieba/Unicode.hpp +215 -0
- data/ext/cppjieba/test/CMakeLists.txt +5 -0
- data/ext/cppjieba/test/demo.cpp +80 -0
- data/ext/cppjieba/test/load_test.cpp +54 -0
- data/ext/cppjieba/test/testdata/curl.res +1 -0
- data/ext/cppjieba/test/testdata/extra_dict/jieba.dict.small.utf8 +109750 -0
- data/ext/cppjieba/test/testdata/gbk_dict/hmm_model.gbk +34 -0
- data/ext/cppjieba/test/testdata/gbk_dict/jieba.dict.gbk +348982 -0
- data/ext/cppjieba/test/testdata/jieba.dict.0.1.utf8 +93 -0
- data/ext/cppjieba/test/testdata/jieba.dict.0.utf8 +93 -0
- data/ext/cppjieba/test/testdata/jieba.dict.1.utf8 +67 -0
- data/ext/cppjieba/test/testdata/jieba.dict.2.utf8 +64 -0
- data/ext/cppjieba/test/testdata/load_test.urls +2 -0
- data/ext/cppjieba/test/testdata/review.100 +100 -0
- data/ext/cppjieba/test/testdata/review.100.res +200 -0
- data/ext/cppjieba/test/testdata/server.conf +19 -0
- data/ext/cppjieba/test/testdata/testlines.gbk +9 -0
- data/ext/cppjieba/test/testdata/testlines.utf8 +8 -0
- data/ext/cppjieba/test/testdata/userdict.2.utf8 +1 -0
- data/ext/cppjieba/test/testdata/userdict.english +2 -0
- data/ext/cppjieba/test/testdata/userdict.utf8 +8 -0
- data/ext/cppjieba/test/testdata/weicheng.utf8 +247 -0
- data/ext/cppjieba/test/unittest/CMakeLists.txt +24 -0
- data/ext/cppjieba/test/unittest/gtest_main.cpp +39 -0
- data/ext/cppjieba/test/unittest/jieba_test.cpp +133 -0
- data/ext/cppjieba/test/unittest/keyword_extractor_test.cpp +79 -0
- data/ext/cppjieba/test/unittest/pos_tagger_test.cpp +41 -0
- data/ext/cppjieba/test/unittest/pre_filter_test.cpp +43 -0
- data/ext/cppjieba/test/unittest/segments_test.cpp +256 -0
- data/ext/cppjieba/test/unittest/textrank_test.cpp +86 -0
- data/ext/cppjieba/test/unittest/trie_test.cpp +177 -0
- data/ext/cppjieba/test/unittest/unicode_test.cpp +43 -0
- data/ext/cppjieba_rb/cppjieba_rb.c +10 -0
- data/ext/cppjieba_rb/extconf.rb +26 -0
- data/ext/cppjieba_rb/internal.cc +148 -0
- data/lib/cppjieba_rb/segment.rb +20 -0
- data/lib/cppjieba_rb/version.rb +3 -0
- data/lib/cppjieba_rb.rb +34 -0
- data/test/test_keyword.rb +17 -0
- data/test/test_segment.rb +24 -0
- data/test/test_tagging.rb +19 -0
- metadata +244 -0
@@ -0,0 +1,86 @@
|
|
1
|
+
#include "cppjieba/TextRankExtractor.hpp"
|
2
|
+
#include "gtest/gtest.h"
|
3
|
+
|
4
|
+
using namespace cppjieba;
|
5
|
+
|
6
|
+
TEST(TextRankExtractorTest, Test1) {
|
7
|
+
TextRankExtractor Extractor(
|
8
|
+
"../test/testdata/extra_dict/jieba.dict.small.utf8",
|
9
|
+
"../dict/hmm_model.utf8",
|
10
|
+
"../dict/stop_words.utf8");
|
11
|
+
{
|
12
|
+
string s("你好世界世界而且而且");
|
13
|
+
string res;
|
14
|
+
size_t topN = 5;
|
15
|
+
|
16
|
+
{
|
17
|
+
vector<string> words;
|
18
|
+
Extractor.Extract(s, words, topN);
|
19
|
+
res << words;
|
20
|
+
ASSERT_EQ(res, "[\"世界\", \"你好\"]");
|
21
|
+
}
|
22
|
+
|
23
|
+
{
|
24
|
+
vector<pair<string, double> > words;
|
25
|
+
Extractor.Extract(s, words, topN);
|
26
|
+
res << words;
|
27
|
+
ASSERT_EQ(res, "[世界:1, 你好:0.519787]");
|
28
|
+
}
|
29
|
+
|
30
|
+
{
|
31
|
+
vector<TextRankExtractor::Word> words;
|
32
|
+
Extractor.Extract(s, words, topN);
|
33
|
+
res << words;
|
34
|
+
ASSERT_EQ(res, "[{\"word\": \"世界\", \"offset\": [6, 12], \"weight\": 1}, {\"word\": \"你好\", \"offset\": [0], \"weight\": 0.519787}]");
|
35
|
+
}
|
36
|
+
}
|
37
|
+
|
38
|
+
{
|
39
|
+
string s("\xe6\x88\x91\xe6\x98\xaf\xe6\x8b\x96\xe6\x8b\x89\xe6\x9c\xba\xe5\xad\xa6\xe9\x99\xa2\xe6\x89\x8b\xe6\x89\xb6\xe6\x8b\x96\xe6\x8b\x89\xe6\x9c\xba\xe4\xb8\x93\xe4\xb8\x9a\xe7\x9a\x84\xe3\x80\x82\xe4\xb8\x8d\xe7\x94\xa8\xe5\xa4\x9a\xe4\xb9\x85\xef\xbc\x8c\xe6\x88\x91\xe5\xb0\xb1\xe4\xbc\x9a\xe5\x8d\x87\xe8\x81\x8c\xe5\x8a\xa0\xe8\x96\xaa\xef\xbc\x8c\xe5\xbd\x93\xe4\xb8\x8a CEO\xef\xbc\x8c\xe8\xb5\xb0\xe4\xb8\x8a\xe4\xba\xba\xe7\x94\x9f\xe5\xb7\x85\xe5\xb3\xb0");
|
40
|
+
string res;
|
41
|
+
vector<TextRankExtractor::Word> wordweights;
|
42
|
+
size_t topN = 5;
|
43
|
+
Extractor.Extract(s, wordweights, topN);
|
44
|
+
res << wordweights;
|
45
|
+
ASSERT_EQ(res, "[{\"word\": \"当上\", \"offset\": [87], \"weight\": 1}, {\"word\": \"不用\", \"offset\": [48], \"weight\": 0.989848}, {\"word\": \"多久\", \"offset\": [54], \"weight\": 0.985126}, {\"word\": \"加薪\", \"offset\": [78], \"weight\": 0.983046}, {\"word\": \"升职\", \"offset\": [72], \"weight\": 0.980278}]");
|
46
|
+
//ASSERT_EQ(res, "[{\"word\": \"专业\", \"offset\": [36], \"weight\": 1}, {\"word\": \"CEO\", \"offset\": [94], \"weight\": 0.95375}, {\"word\": \"手扶拖拉机\", \"offset\": [21], \"weight\": 0.801701}, {\"word\": \"当上\", \"offset\": [87], \"weight\": 0.798968}, {\"word\": \"走上\", \"offset\": [100], \"weight\": 0.775505}]");
|
47
|
+
}
|
48
|
+
|
49
|
+
{
|
50
|
+
string s("一部iPhone6");
|
51
|
+
string res;
|
52
|
+
vector<TextRankExtractor::Word> wordweights;
|
53
|
+
size_t topN = 5;
|
54
|
+
Extractor.Extract(s, wordweights, topN);
|
55
|
+
res << wordweights;
|
56
|
+
ASSERT_EQ(res, "[{\"word\": \"一部\", \"offset\": [0], \"weight\": 1}, {\"word\": \"iPhone6\", \"offset\": [6], \"weight\": 0.996126}]");
|
57
|
+
}
|
58
|
+
}
|
59
|
+
|
60
|
+
TEST(TextRankExtractorTest, Test2) {
|
61
|
+
TextRankExtractor Extractor(
|
62
|
+
"../test/testdata/extra_dict/jieba.dict.small.utf8",
|
63
|
+
"../dict/hmm_model.utf8",
|
64
|
+
"../dict/stop_words.utf8",
|
65
|
+
"../test/testdata/userdict.utf8");
|
66
|
+
|
67
|
+
{
|
68
|
+
string s("\xe8\x93\x9d\xe7\xbf\x94\xe4\xbc\x98\xe7\xa7\x80\xe6\xaf\x95\xe4\xb8\x9a\xe7\x94\x9f");
|
69
|
+
string res;
|
70
|
+
vector<TextRankExtractor::Word> wordweights;
|
71
|
+
size_t topN = 5;
|
72
|
+
Extractor.Extract(s, wordweights, topN);
|
73
|
+
res << wordweights;
|
74
|
+
ASSERT_EQ(res, "[{\"word\": \"蓝翔\", \"offset\": [0], \"weight\": 1}, {\"word\": \"毕业生\", \"offset\": [12], \"weight\": 0.996685}, {\"word\": \"优秀\", \"offset\": [6], \"weight\": 0.992994}]");
|
75
|
+
}
|
76
|
+
|
77
|
+
{
|
78
|
+
string s("一部iPhone6");
|
79
|
+
string res;
|
80
|
+
vector<TextRankExtractor::Word> wordweights;
|
81
|
+
size_t topN = 5;
|
82
|
+
Extractor.Extract(s, wordweights, topN);
|
83
|
+
res << wordweights;
|
84
|
+
ASSERT_EQ(res, "[{\"word\": \"一部\", \"offset\": [0], \"weight\": 1}, {\"word\": \"iPhone6\", \"offset\": [6], \"weight\": 0.996126}]");
|
85
|
+
}
|
86
|
+
}
|
@@ -0,0 +1,177 @@
|
|
1
|
+
#include "cppjieba/DictTrie.hpp"
|
2
|
+
#include "cppjieba/MPSegment.hpp"
|
3
|
+
#include "gtest/gtest.h"
|
4
|
+
|
5
|
+
using namespace cppjieba;
|
6
|
+
|
7
|
+
static const char* const DICT_FILE = "../test/testdata/extra_dict/jieba.dict.small.utf8";
|
8
|
+
|
9
|
+
TEST(TrieTest, Empty) {
|
10
|
+
vector<Unicode> keys;
|
11
|
+
vector<const DictUnit*> values;
|
12
|
+
Trie trie(keys, values);
|
13
|
+
}
|
14
|
+
|
15
|
+
TEST(TrieTest, Construct) {
|
16
|
+
vector<Unicode> keys;
|
17
|
+
vector<const DictUnit*> values;
|
18
|
+
keys.push_back(DecodeRunesInString("你"));
|
19
|
+
values.push_back((const DictUnit*)(NULL));
|
20
|
+
Trie trie(keys, values);
|
21
|
+
}
|
22
|
+
|
23
|
+
TEST(DictTrieTest, NewAndDelete) {
|
24
|
+
DictTrie * trie;
|
25
|
+
trie = new DictTrie(DICT_FILE);
|
26
|
+
delete trie;
|
27
|
+
}
|
28
|
+
|
29
|
+
TEST(DictTrieTest, Test1) {
|
30
|
+
string s1, s2;
|
31
|
+
DictTrie trie(DICT_FILE);
|
32
|
+
ASSERT_LT(trie.GetMinWeight() + 15.6479, 0.001);
|
33
|
+
string word("来到");
|
34
|
+
cppjieba::RuneStrArray uni;
|
35
|
+
ASSERT_TRUE(DecodeRunesInString(word, uni));
|
36
|
+
//DictUnit nodeInfo;
|
37
|
+
//nodeInfo.word = uni;
|
38
|
+
//nodeInfo.tag = "v";
|
39
|
+
//nodeInfo.weight = -8.87033;
|
40
|
+
//s1 << nodeInfo;
|
41
|
+
//s2 << (*trie.Find(uni.begin(), uni.end()));
|
42
|
+
const DictUnit* du = trie.Find(uni.begin(), uni.end());
|
43
|
+
ASSERT_TRUE(du != NULL);
|
44
|
+
ASSERT_EQ(2u, du->word.size());
|
45
|
+
ASSERT_EQ(26469u, du->word[0]);
|
46
|
+
ASSERT_EQ(21040u, du->word[1]);
|
47
|
+
ASSERT_EQ("v", du->tag);
|
48
|
+
ASSERT_NEAR(-8.870, du->weight, 0.001);
|
49
|
+
|
50
|
+
//EXPECT_EQ("[\"26469\", \"21040\"] v -8.870", s2);
|
51
|
+
word = "清华大学";
|
52
|
+
LocalVector<pair<size_t, const DictUnit*> > res;
|
53
|
+
const char * words[] = {"清", "清华", "清华大学"};
|
54
|
+
for (size_t i = 0; i < sizeof(words)/sizeof(words[0]); i++) {
|
55
|
+
ASSERT_TRUE(DecodeRunesInString(words[i], uni));
|
56
|
+
res.push_back(make_pair(uni.size() - 1, trie.Find(uni.begin(), uni.end())));
|
57
|
+
//resMap[uni.size() - 1] = trie.Find(uni.begin(), uni.end());
|
58
|
+
}
|
59
|
+
vector<pair<size_t, const DictUnit*> > vec;
|
60
|
+
vector<struct Dag> dags;
|
61
|
+
ASSERT_TRUE(DecodeRunesInString(word, uni));
|
62
|
+
trie.Find(uni.begin(), uni.end(), dags);
|
63
|
+
ASSERT_EQ(dags.size(), uni.size());
|
64
|
+
ASSERT_NE(dags.size(), 0u);
|
65
|
+
s1 << res;
|
66
|
+
s2 << dags[0].nexts;
|
67
|
+
ASSERT_EQ(s1, s2);
|
68
|
+
|
69
|
+
}
|
70
|
+
|
71
|
+
TEST(DictTrieTest, UserDict) {
|
72
|
+
DictTrie trie(DICT_FILE, "../test/testdata/userdict.utf8");
|
73
|
+
string word = "云计算";
|
74
|
+
cppjieba::RuneStrArray unicode;
|
75
|
+
ASSERT_TRUE(DecodeRunesInString(word, unicode));
|
76
|
+
const DictUnit * unit = trie.Find(unicode.begin(), unicode.end());
|
77
|
+
ASSERT_TRUE(unit != NULL);
|
78
|
+
ASSERT_NEAR(unit->weight, -14.100, 0.001);
|
79
|
+
|
80
|
+
word = "蓝翔";
|
81
|
+
ASSERT_TRUE(DecodeRunesInString(word, unicode));
|
82
|
+
unit = trie.Find(unicode.begin(), unicode.end());
|
83
|
+
ASSERT_TRUE(unit != NULL);
|
84
|
+
ASSERT_EQ(unit->tag, "nz");
|
85
|
+
ASSERT_NEAR(unit->weight, -14.100, 0.001);
|
86
|
+
|
87
|
+
word = "区块链";
|
88
|
+
ASSERT_TRUE(DecodeRunesInString(word, unicode));
|
89
|
+
unit = trie.Find(unicode.begin(), unicode.end());
|
90
|
+
ASSERT_TRUE(unit != NULL);
|
91
|
+
ASSERT_EQ(unit->tag, "nz");
|
92
|
+
ASSERT_NEAR(unit->weight, -15.6478, 0.001);
|
93
|
+
}
|
94
|
+
|
95
|
+
TEST(DictTrieTest, UserDictWithMaxWeight) {
|
96
|
+
DictTrie trie(DICT_FILE, "../test/testdata/userdict.utf8", DictTrie::WordWeightMax);
|
97
|
+
string word = "云计算";
|
98
|
+
cppjieba::RuneStrArray unicode;
|
99
|
+
ASSERT_TRUE(DecodeRunesInString(word, unicode));
|
100
|
+
const DictUnit * unit = trie.Find(unicode.begin(), unicode.end());
|
101
|
+
ASSERT_TRUE(unit);
|
102
|
+
ASSERT_NEAR(unit->weight, -2.975, 0.001);
|
103
|
+
}
|
104
|
+
|
105
|
+
TEST(DictTrieTest, Dag) {
|
106
|
+
DictTrie trie(DICT_FILE, "../test/testdata/userdict.utf8");
|
107
|
+
|
108
|
+
{
|
109
|
+
string word = "清华大学";
|
110
|
+
cppjieba::RuneStrArray unicode;
|
111
|
+
ASSERT_TRUE(DecodeRunesInString(word, unicode));
|
112
|
+
vector<struct Dag> res;
|
113
|
+
trie.Find(unicode.begin(), unicode.end(), res);
|
114
|
+
|
115
|
+
size_t nexts_sizes[] = {3, 2, 2, 1};
|
116
|
+
ASSERT_EQ(res.size(), sizeof(nexts_sizes)/sizeof(nexts_sizes[0]));
|
117
|
+
for (size_t i = 0; i < res.size(); i++) {
|
118
|
+
ASSERT_EQ(res[i].nexts.size(), nexts_sizes[i]);
|
119
|
+
}
|
120
|
+
}
|
121
|
+
|
122
|
+
{
|
123
|
+
string word = "北京邮电大学";
|
124
|
+
cppjieba::RuneStrArray unicode;
|
125
|
+
ASSERT_TRUE(DecodeRunesInString(word, unicode));
|
126
|
+
vector<struct Dag> res;
|
127
|
+
trie.Find(unicode.begin(), unicode.end(), res);
|
128
|
+
|
129
|
+
size_t nexts_sizes[] = {3, 1, 2, 2, 2, 1};
|
130
|
+
ASSERT_EQ(res.size(), sizeof(nexts_sizes)/sizeof(nexts_sizes[0]));
|
131
|
+
for (size_t i = 0; i < res.size(); i++) {
|
132
|
+
ASSERT_EQ(res[i].nexts.size(), nexts_sizes[i]);
|
133
|
+
}
|
134
|
+
}
|
135
|
+
|
136
|
+
{
|
137
|
+
string word = "长江大桥";
|
138
|
+
cppjieba::RuneStrArray unicode;
|
139
|
+
ASSERT_TRUE(DecodeRunesInString(word, unicode));
|
140
|
+
vector<struct Dag> res;
|
141
|
+
trie.Find(unicode.begin(), unicode.end(), res);
|
142
|
+
|
143
|
+
size_t nexts_sizes[] = {3, 1, 2, 1};
|
144
|
+
ASSERT_EQ(res.size(), sizeof(nexts_sizes)/sizeof(nexts_sizes[0]));
|
145
|
+
for (size_t i = 0; i < res.size(); i++) {
|
146
|
+
ASSERT_EQ(res[i].nexts.size(), nexts_sizes[i]);
|
147
|
+
}
|
148
|
+
}
|
149
|
+
|
150
|
+
{
|
151
|
+
string word = "长江大桥";
|
152
|
+
cppjieba::RuneStrArray unicode;
|
153
|
+
ASSERT_TRUE(DecodeRunesInString(word, unicode));
|
154
|
+
vector<struct Dag> res;
|
155
|
+
trie.Find(unicode.begin(), unicode.end(), res, 3);
|
156
|
+
|
157
|
+
size_t nexts_sizes[] = {2, 1, 2, 1};
|
158
|
+
ASSERT_EQ(res.size(), sizeof(nexts_sizes)/sizeof(nexts_sizes[0]));
|
159
|
+
for (size_t i = 0; i < res.size(); i++) {
|
160
|
+
ASSERT_EQ(res[i].nexts.size(), nexts_sizes[i]);
|
161
|
+
}
|
162
|
+
}
|
163
|
+
|
164
|
+
{
|
165
|
+
string word = "长江大桥";
|
166
|
+
cppjieba::RuneStrArray unicode;
|
167
|
+
ASSERT_TRUE(DecodeRunesInString(word, unicode));
|
168
|
+
vector<struct Dag> res;
|
169
|
+
trie.Find(unicode.begin(), unicode.end(), res, 4);
|
170
|
+
|
171
|
+
size_t nexts_sizes[] = {3, 1, 2, 1};
|
172
|
+
ASSERT_EQ(res.size(), sizeof(nexts_sizes)/sizeof(nexts_sizes[0]));
|
173
|
+
for (size_t i = 0; i < res.size(); i++) {
|
174
|
+
ASSERT_EQ(res[i].nexts.size(), nexts_sizes[i]);
|
175
|
+
}
|
176
|
+
}
|
177
|
+
}
|
@@ -0,0 +1,43 @@
|
|
1
|
+
#include "cppjieba/Unicode.hpp"
|
2
|
+
#include "limonp/StdExtension.hpp"
|
3
|
+
#include "gtest/gtest.h"
|
4
|
+
|
5
|
+
using namespace cppjieba;
|
6
|
+
using namespace std;
|
7
|
+
|
8
|
+
TEST(UnicodeTest, Test1) {
|
9
|
+
string s = "你好世界";
|
10
|
+
RuneStrArray runes;
|
11
|
+
ASSERT_TRUE(DecodeRunesInString(s, runes));
|
12
|
+
string actual;
|
13
|
+
string expected = "[\"{\"rune\": \"20320\", \"offset\": 0, \"len\": 3}\", \"{\"rune\": \"22909\", \"offset\": 3, \"len\": 3}\", \"{\"rune\": \"19990\", \"offset\": 6, \"len\": 3}\", \"{\"rune\": \"30028\", \"offset\": 9, \"len\": 3}\"]";
|
14
|
+
actual << runes;
|
15
|
+
ASSERT_EQ(expected, actual);
|
16
|
+
}
|
17
|
+
|
18
|
+
TEST(UnicodeTest, Illegal) {
|
19
|
+
string s = "123\x80";
|
20
|
+
RuneStrArray runes;
|
21
|
+
ASSERT_FALSE(DecodeRunesInString(s, runes));
|
22
|
+
string actual;
|
23
|
+
string expected = "[]";
|
24
|
+
actual << runes;
|
25
|
+
ASSERT_EQ(expected, actual);
|
26
|
+
}
|
27
|
+
|
28
|
+
TEST(UnicodeTest, Rand) {
|
29
|
+
const size_t ITERATION = 1024;
|
30
|
+
const size_t MAX_LEN = 256;
|
31
|
+
string s;
|
32
|
+
srand(time(NULL));
|
33
|
+
|
34
|
+
for (size_t i = 0; i < ITERATION; i++) {
|
35
|
+
size_t len = rand() % MAX_LEN;
|
36
|
+
s.resize(len);
|
37
|
+
for (size_t j = 0; j < len; j++) {
|
38
|
+
s[rand() % len] = rand();
|
39
|
+
}
|
40
|
+
RuneStrArray runes;
|
41
|
+
DecodeRunesInString(s, runes);
|
42
|
+
}
|
43
|
+
}
|
@@ -0,0 +1,26 @@
|
|
1
|
+
require "mkmf"
|
2
|
+
abs = File.expand_path File.dirname(__FILE__)
|
3
|
+
|
4
|
+
LIBDIR = RbConfig::CONFIG['libdir']
|
5
|
+
INCLUDEDIR = RbConfig::CONFIG['includedir']
|
6
|
+
|
7
|
+
HEADER_DIRS = [
|
8
|
+
INCLUDEDIR,
|
9
|
+
"#{abs}/../cppjieba/include",
|
10
|
+
"#{abs}/../cppjieba/deps"
|
11
|
+
]
|
12
|
+
|
13
|
+
LIB_DIRS = [
|
14
|
+
LIBDIR
|
15
|
+
]
|
16
|
+
|
17
|
+
dir_config('cppjieba_src', HEADER_DIRS, LIB_DIRS)
|
18
|
+
|
19
|
+
CONFIG["CXXFLAGS"] += " -std=c++11 -O3"
|
20
|
+
$CXXFLAGS = "#{$CXXFLAGS} -std=c++11 -O3"
|
21
|
+
create_makefile 'cppjieba_rb'
|
22
|
+
# respect header changes
|
23
|
+
headers = Dir.glob('*.{hpp,h}').join ' '
|
24
|
+
File.open 'Makefile', 'a' do |f|
|
25
|
+
f.puts "\n$(OBJS): #{headers}"
|
26
|
+
end
|
@@ -0,0 +1,148 @@
|
|
1
|
+
#include <ruby.h>
|
2
|
+
#include <ruby/encoding.h>
|
3
|
+
#include "cppjieba/Jieba.hpp"
|
4
|
+
|
5
|
+
#define GET_CPPJIEBA(_data) jieba_cpp_data* _data; \
|
6
|
+
TypedData_Get_Struct(self, jieba_cpp_data, &jieba_cpp_type, _data)
|
7
|
+
|
8
|
+
typedef struct {
|
9
|
+
cppjieba::Jieba* jieba;
|
10
|
+
} jieba_cpp_data;
|
11
|
+
|
12
|
+
// make compiler happy
|
13
|
+
typedef VALUE (ruby_method)(...);
|
14
|
+
|
15
|
+
static ID rb_sMp;
|
16
|
+
static ID rb_sMix;
|
17
|
+
static ID rb_sHmm;
|
18
|
+
static ID rb_sQuery;
|
19
|
+
static ID rb_sFull;
|
20
|
+
static rb_encoding* u8_enc;
|
21
|
+
VALUE rb_cCppjiebaRb_Internal;
|
22
|
+
extern "C" VALUE rb_mCppjiebaRb;
|
23
|
+
|
24
|
+
static void jieba_cpp_free(void* _this)
|
25
|
+
{
|
26
|
+
jieba_cpp_data* data = static_cast<jieba_cpp_data*>(_this);
|
27
|
+
delete data->jieba;
|
28
|
+
data->jieba = nullptr;
|
29
|
+
}
|
30
|
+
|
31
|
+
static size_t jieba_cpp_memsize(const void* _)
|
32
|
+
{
|
33
|
+
return sizeof(jieba_cpp_data);
|
34
|
+
}
|
35
|
+
|
36
|
+
static const rb_data_type_t jieba_cpp_type = {
|
37
|
+
"jieba/internal",
|
38
|
+
{NULL, jieba_cpp_free, jieba_cpp_memsize,},
|
39
|
+
0, 0,
|
40
|
+
RUBY_TYPED_FREE_IMMEDIATELY,
|
41
|
+
};
|
42
|
+
|
43
|
+
extern "C" {
|
44
|
+
|
45
|
+
VALUE internal_alloc(VALUE self)
|
46
|
+
{
|
47
|
+
jieba_cpp_data* data;
|
48
|
+
return TypedData_Make_Struct(self, jieba_cpp_data, &jieba_cpp_type, data);
|
49
|
+
}
|
50
|
+
|
51
|
+
VALUE internal_initialize(VALUE self,
|
52
|
+
VALUE dict_path,
|
53
|
+
VALUE model_path,
|
54
|
+
VALUE user_dict_path,
|
55
|
+
VALUE idf_path,
|
56
|
+
VALUE stop_word_path)
|
57
|
+
{
|
58
|
+
GET_CPPJIEBA(data);
|
59
|
+
data->jieba = new cppjieba::Jieba(StringValueCStr(dict_path),
|
60
|
+
StringValueCStr(model_path),
|
61
|
+
StringValueCStr(user_dict_path),
|
62
|
+
StringValueCStr(idf_path),
|
63
|
+
StringValueCStr(stop_word_path));
|
64
|
+
}
|
65
|
+
|
66
|
+
VALUE internal_extract_keyword(VALUE self, VALUE text_rbs, VALUE topN)
|
67
|
+
{
|
68
|
+
std::string text = StringValueCStr(text_rbs);
|
69
|
+
int top_n = NUM2INT(topN);
|
70
|
+
GET_CPPJIEBA(data);
|
71
|
+
|
72
|
+
std::vector<std::pair<std::string, double> > top_words;
|
73
|
+
|
74
|
+
data->jieba->extractor.Extract(text, top_words, top_n);
|
75
|
+
VALUE arr = rb_ary_new2(top_words.size());
|
76
|
+
for (auto iter = top_words.begin(); iter != top_words.end(); iter++) {
|
77
|
+
VALUE inner_arr = rb_ary_new2(2);
|
78
|
+
rb_ary_push(inner_arr, rb_enc_str_new(iter->first.c_str(), iter->first.length(), u8_enc));
|
79
|
+
rb_ary_push(inner_arr, rb_float_new(iter->second));
|
80
|
+
rb_ary_push(arr, inner_arr);
|
81
|
+
}
|
82
|
+
return arr;
|
83
|
+
}
|
84
|
+
|
85
|
+
static VALUE internal_segment(VALUE self, VALUE text_rbs, VALUE mode, VALUE max_length_, VALUE hmm_)
|
86
|
+
{
|
87
|
+
std::string text = StringValueCStr(text_rbs);
|
88
|
+
size_t max_length = NUM2UINT(max_length_);
|
89
|
+
int hmm = (hmm_ == Qtrue ? 1 : 0);
|
90
|
+
GET_CPPJIEBA(data);
|
91
|
+
ID rb_sMode = SYM2ID(mode);
|
92
|
+
std::vector<std::string> words;
|
93
|
+
|
94
|
+
if (rb_sMode == rb_sMp) {
|
95
|
+
data->jieba->CutSmall(text, words, max_length);
|
96
|
+
} else if (rb_sMode == rb_sMix) {
|
97
|
+
data->jieba->Cut(text, words, hmm);
|
98
|
+
} else if (rb_sMode == rb_sHmm) {
|
99
|
+
data->jieba->CutHMM(text, words);
|
100
|
+
} else if (rb_sMode == rb_sQuery) {
|
101
|
+
data->jieba->CutForSearch(text, words, hmm);
|
102
|
+
} else if (rb_sMode == rb_sFull) {
|
103
|
+
data->jieba->CutAll(text, words);
|
104
|
+
}
|
105
|
+
|
106
|
+
VALUE arr = rb_ary_new2(words.size());
|
107
|
+
for (auto iter = words.begin(); iter != words.end(); iter++) {
|
108
|
+
rb_ary_push(arr, rb_enc_str_new(iter->c_str(), iter->length(), u8_enc));
|
109
|
+
}
|
110
|
+
return arr;
|
111
|
+
}
|
112
|
+
|
113
|
+
static VALUE internal_segment_tag(VALUE self, VALUE text_rbs)
|
114
|
+
{
|
115
|
+
std::string text = StringValueCStr(text_rbs);
|
116
|
+
GET_CPPJIEBA(data);
|
117
|
+
|
118
|
+
std::vector<std::pair<std::string, std::string>> words;
|
119
|
+
data->jieba->Tag(text, words);
|
120
|
+
|
121
|
+
VALUE result = rb_hash_new();
|
122
|
+
for (auto iter = words.begin(); iter != words.end(); iter++) {
|
123
|
+
rb_hash_aset(result,
|
124
|
+
rb_enc_str_new(iter->first.c_str(), iter->first.length(), u8_enc),
|
125
|
+
rb_enc_str_new(iter->second.c_str(), iter->second.length(), u8_enc));
|
126
|
+
}
|
127
|
+
return result;
|
128
|
+
}
|
129
|
+
|
130
|
+
|
131
|
+
void Init_internal()
|
132
|
+
{
|
133
|
+
rb_sMp = rb_intern("mp");
|
134
|
+
rb_sMix = rb_intern("mix");
|
135
|
+
rb_sHmm = rb_intern("hmm");
|
136
|
+
rb_sQuery = rb_intern("query");
|
137
|
+
rb_sFull = rb_intern("full");
|
138
|
+
u8_enc = rb_utf8_encoding();
|
139
|
+
|
140
|
+
rb_cCppjiebaRb_Internal = rb_define_class_under(rb_mCppjiebaRb, "Internal", rb_cData);
|
141
|
+
rb_define_alloc_func(rb_cCppjiebaRb_Internal, internal_alloc);
|
142
|
+
rb_define_method(rb_cCppjiebaRb_Internal, "initialize", (ruby_method*) &internal_initialize, 5);
|
143
|
+
rb_define_method(rb_cCppjiebaRb_Internal, "extract_keyword", (ruby_method*) &internal_extract_keyword, 2);
|
144
|
+
rb_define_method(rb_cCppjiebaRb_Internal, "segment", (ruby_method*) &internal_segment, 4);
|
145
|
+
rb_define_method(rb_cCppjiebaRb_Internal, "segment_tag", (ruby_method*) &internal_segment_tag, 1);
|
146
|
+
}
|
147
|
+
|
148
|
+
}
|
@@ -0,0 +1,20 @@
|
|
1
|
+
module CppjiebaRb
|
2
|
+
class Segment
|
3
|
+
VALID_MODES = %i[mix hmm mp query full].freeze
|
4
|
+
|
5
|
+
def initialize(opts = nil)
|
6
|
+
opts ||= {}
|
7
|
+
unless opts[:mode].nil? || VALID_MODES.include?(opts[:mode])
|
8
|
+
raise ArgumentError, "The mode is #{opts[:mode]}. It should be one of :mix :hmm :mp"
|
9
|
+
end
|
10
|
+
|
11
|
+
@mode = opts[:mode] || :mix
|
12
|
+
@max_word_length = opts[:max_word_length] || 8
|
13
|
+
@hmm = opts[:hmm] || true
|
14
|
+
end
|
15
|
+
|
16
|
+
def segment(str)
|
17
|
+
CppjiebaRb.internal.segment(str, @mode, @max_word_length, @hmm)
|
18
|
+
end
|
19
|
+
end
|
20
|
+
end
|
data/lib/cppjieba_rb.rb
ADDED
@@ -0,0 +1,34 @@
|
|
1
|
+
require 'cppjieba_rb/cppjieba_rb'
|
2
|
+
require 'cppjieba_rb/version'
|
3
|
+
require 'cppjieba_rb/segment'
|
4
|
+
|
5
|
+
module CppjiebaRb
|
6
|
+
EXT_BASE = File.join(File.dirname(__FILE__), '..', 'ext', 'cppjieba', 'dict')
|
7
|
+
DICT_PATH = File.join(EXT_BASE, 'jieba.dict.utf8')
|
8
|
+
HMM_DICT_PATH = File.join(EXT_BASE, 'hmm_model.utf8')
|
9
|
+
USER_DICT = File.join(EXT_BASE, 'user.dict.utf8')
|
10
|
+
IDF_PATH = File.join(EXT_BASE, 'idf.utf8')
|
11
|
+
STOP_WORD_PATH = File.join(EXT_BASE, 'stop_words.utf8')
|
12
|
+
|
13
|
+
def self.extract_keyword(str, top_n)
|
14
|
+
internal.extract_keyword(str, top_n)
|
15
|
+
end
|
16
|
+
|
17
|
+
def self.segment(str, opts = nil)
|
18
|
+
CppjiebaRb::Segment.new(opts).segment(str)
|
19
|
+
end
|
20
|
+
|
21
|
+
def self.segment_tag(str)
|
22
|
+
internal.segment_tag(str)
|
23
|
+
end
|
24
|
+
|
25
|
+
class << self
|
26
|
+
def internal
|
27
|
+
@backend ||= CppjiebaRb::Internal.new(DICT_PATH,
|
28
|
+
HMM_DICT_PATH,
|
29
|
+
USER_DICT,
|
30
|
+
IDF_PATH,
|
31
|
+
STOP_WORD_PATH)
|
32
|
+
end
|
33
|
+
end
|
34
|
+
end
|
@@ -0,0 +1,17 @@
|
|
1
|
+
# coding: utf-8
|
2
|
+
require 'minitest/autorun'
|
3
|
+
require 'cppjieba_rb'
|
4
|
+
|
5
|
+
class JiebaTest < Minitest::Test
|
6
|
+
def test_keywords
|
7
|
+
results = CppjiebaRb.extract_keyword "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。", 5
|
8
|
+
|
9
|
+
assert_equal [["CEO",
|
10
|
+
11.739204307083542],
|
11
|
+
["升职", 10.8561552143],
|
12
|
+
["加薪", 10.642581114],
|
13
|
+
["手扶拖拉机", 10.0088573539],
|
14
|
+
["巅峰", 9.49395840471]], results
|
15
|
+
|
16
|
+
end
|
17
|
+
end
|
@@ -0,0 +1,24 @@
|
|
1
|
+
# coding: utf-8
|
2
|
+
require 'minitest/autorun'
|
3
|
+
require 'cppjieba_rb'
|
4
|
+
|
5
|
+
class JiebaTest < Minitest::Test
|
6
|
+
def test_mix_segment
|
7
|
+
words = CppjiebaRb.segment "我来到南京市长江大桥"
|
8
|
+
assert_equal %w(我 来到 南京市 长江大桥), words
|
9
|
+
|
10
|
+
words = CppjiebaRb.segment "令狐冲是云计算行业的专家"
|
11
|
+
assert_equal %w(令狐冲 是 云计算 行业 的 专家), words
|
12
|
+
end
|
13
|
+
|
14
|
+
def test_hmm_segment
|
15
|
+
words = CppjiebaRb.segment "令狐冲是云计算行业的专家", mode: :hmm
|
16
|
+
assert_equal %w(令狐冲 是 云计算 行业 的 专家), words
|
17
|
+
end
|
18
|
+
|
19
|
+
def test_max_prob_segment
|
20
|
+
words = CppjiebaRb.segment "令狐冲是云计算行业的专家", mode: :mp
|
21
|
+
assert_equal %w(令狐冲 是 云计算 行业 的 专家), words
|
22
|
+
end
|
23
|
+
|
24
|
+
end
|
@@ -0,0 +1,19 @@
|
|
1
|
+
# coding: utf-8
|
2
|
+
require 'minitest/autorun'
|
3
|
+
require 'cppjieba_rb'
|
4
|
+
class JiebaTest < Minitest::Test
|
5
|
+
def test_tagging
|
6
|
+
pairs = CppjiebaRb.segment_tag '我是蓝翔技工拖拉机学院手扶拖拉机专业的。'
|
7
|
+
assert_equal({ '我' => 'r', '是' => 'v', '蓝翔' => 'nz', '技工' => 'n',
|
8
|
+
'拖拉机' => 'n', '学院' => 'n', '手扶拖拉机' => 'n', '专业' => 'n',
|
9
|
+
'的' => 'uj', '。' => 'x' }, pairs)
|
10
|
+
end
|
11
|
+
|
12
|
+
def test_tagging_with_user_dict
|
13
|
+
pairs = CppjiebaRb.segment_tag '我是蓝翔技工拖拉机学院手扶拖拉机专业的。'
|
14
|
+
assert_equal({ '我' => 'r', '是' => 'v', '蓝翔' => 'nz', '技工' => 'n',
|
15
|
+
'拖拉机' => 'n', '学院' => 'n', '手扶拖拉机' => 'n', '专业' => 'n',
|
16
|
+
'的' => 'uj', '。' => 'x' }, pairs)
|
17
|
+
end
|
18
|
+
|
19
|
+
end
|