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

add Jieba::ResetSeparators api and unittest

Browse Source
This commit is contained in:
yanyiwu 2016-05-09 22:49:51 +08:00
parent 6e3ecec599
commit c425bcc49f
3 changed files with 25 additions and 21 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
include/cppjieba/Jieba.hpp
View File

@ -67,6 +67,15 @@ class Jieba {
return dict_trie_.InsertUserWord(word, tag); return dict_trie_.InsertUserWord(word, tag);
} }
void ResetSeparators(const string& s) {
//TODO
mp_seg_.ResetSeparators(s);
hmm_seg_.ResetSeparators(s);
mix_seg_.ResetSeparators(s);
full_seg_.ResetSeparators(s);
query_seg_.ResetSeparators(s);
}
const DictTrie* GetDictTrie() const { const DictTrie* GetDictTrie() const {
return &dict_trie_; return &dict_trie_;
} }

26
include/cppjieba/SegmentBase.hpp
View File

@ -8,20 +8,20 @@
namespace cppjieba { namespace cppjieba {
const char* const SPECIAL_RUNES = " \t\n，。"; const char* const SPECIAL_SEPARATORS = " \t\n，。";
using namespace limonp; using namespace limonp;
class SegmentBase { class SegmentBase {
public: public:
SegmentBase() { SegmentBase() {
XCHECK(Insert(SPECIAL_RUNES)); XCHECK(ResetSeparators(SPECIAL_SEPARATORS));
} }
~SegmentBase() { ~SegmentBase() {
} }
protected: bool ResetSeparators(const string& s) {
bool Insert(const string& s) { symbols_.clear();
RuneStrArray runes; RuneStrArray runes;
if (!DecodeRunesInString(s, runes)) { if (!DecodeRunesInString(s, runes)) {
XLOG(ERROR) << "decode " << s << " failed"; XLOG(ERROR) << "decode " << s << " failed";
@ -35,23 +35,7 @@ class SegmentBase {
} }
return true; return true;
} }
//bool Remove(const string& s) { protected:
// RuneStrArray runes;
// if (!DecodeRunesInString(s, runes)) {
// XLOG(ERROR) << "decode " << s << " failed";
// return false;
// }
// for (size_t i = 0; i < runes.size(); i++) {
// unordered_set<Rune>::iterator iter = symbols_.find(runes[i].rune);
// if (iter == symbols_.end()) {
// XLOG(ERROR) << s.substr(runes[i].offset, runes[i].len) << " not found";
// return false;
// }
// symbols_.erase(iter);
// }
// return true;
//}
unordered_set<Rune> symbols_; unordered_set<Rune> symbols_;
}; // class SegmentBase }; // class SegmentBase

11
test/unittest/jieba_test.cpp
View File

@ -103,4 +103,15 @@ TEST(JiebaTest, InsertUserWord) {
result << words; result << words;
ASSERT_EQ(result, StringFormat("[\"%s\"]", newWord.c_str())); ASSERT_EQ(result, StringFormat("[\"%s\"]", newWord.c_str()));
} }
ASSERT_TRUE(jieba.InsertUserWord("同一个世界，同一个梦想"));
jieba.Cut("同一个世界，同一个梦想", words);
result = Join(words.begin(), words.end(), "/");
ASSERT_EQ(result, "同一个/世界/，/同一个/梦想");
jieba.ResetSeparators("");
jieba.Cut("同一个世界，同一个梦想", words);
result = Join(words.begin(), words.end(), "/");
ASSERT_EQ(result, "同一个世界，同一个梦想");
} }