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

fix bug about optional argument hmm

Browse Source
This commit is contained in:
yanyiwu 2015-09-13 18:06:44 +08:00
parent f517601c29
commit 8eef9a13a8
2 changed files with 39 additions and 11 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
src/MixSegment.hpp
View File

@ -34,12 +34,13 @@ class MixSegment: public SegmentBase {
} }
void cut(Unicode::const_iterator begin, Unicode::const_iterator end, vector<Unicode>& res, bool hmm) const { void cut(Unicode::const_iterator begin, Unicode::const_iterator end, vector<Unicode>& res, bool hmm) const {
if (!hmm) {
mpSeg_.cut(begin, end, res);
return;
}
vector<Unicode> words; vector<Unicode> words;
words.reserve(end - begin); words.reserve(end - begin);
mpSeg_.cut(begin, end, words); mpSeg_.cut(begin, end, words);
if (!hmm) {
return;
}
vector<Unicode> hmmRes; vector<Unicode> hmmRes;
hmmRes.reserve(end - begin); hmmRes.reserve(end - begin);

43
test/unittest/TSegments.cpp
View File

@ -12,15 +12,42 @@ using namespace CppJieba;
TEST(MixSegmentTest, Test1) { TEST(MixSegmentTest, Test1) {
MixSegment segment("../dict/jieba.dict.utf8", "../dict/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，用AK47"; string sentence;
const char* res[] = {"我", "来自", "北京邮电大学", "。","。","。", "学号", "123456","，","用","AK47"};
const char* str2 = "B超 T恤";
const char* res2[] = {"B超"," ", "T恤"};
vector<string> words; vector<string> words;
segment.cut(str, words); string actual;
ASSERT_EQ(words, vector<string>(res, res + sizeof(res)/sizeof(res[0]))); string expected;
segment.cut(str2, words);
ASSERT_EQ(words, vector<string>(res2, res2 + sizeof(res2)/sizeof(res2[0]))); {
sentence = "我来自北京邮电大学。。。学号123456，用AK47";
expected = "我/来自/北京邮电大学/。/。/。/学号/123456/，/用/AK47";
segment.cut(sentence, words);
actual = join(words.begin(), words.end(), "/");
ASSERT_EQ(actual, expected);
}
{
sentence = "B超 T恤";
expected = "B超/ /T恤";
segment.cut(sentence, words);
actual = join(words.begin(), words.end(), "/");
ASSERT_EQ(actual, expected);
}
{
sentence = "他来到了网易杭研大厦";
expected = "他/来到/了/网易/杭/研/大厦";
segment.cut(sentence, words, false);
actual = join(words.begin(), words.end(), "/");
ASSERT_EQ(actual, expected);
}
{
sentence = "他来到了网易杭研大厦";
expected = "他/来到/了/网易/杭研/大厦";
segment.cut(sentence, words);
actual = join(words.begin(), words.end(), "/");
ASSERT_EQ(actual, expected);
}
} }
TEST(MixSegmentTest, NoUserDict) { TEST(MixSegmentTest, NoUserDict) {