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

update cppcommon

Browse Source
This commit is contained in:
gwdwyy 2013-07-18 11:23:29 +08:00
parent 46ecb1382a
commit 7aadcb2990
2 changed files with 16 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

19
src/cppcommon/str_functs.cpp
View File

@ -293,11 +293,6 @@ namespace CPPCOMMON
{ {
for(uint i = 0; i < uniLen; i++) for(uint i = 0; i < uniLen; i++)
{ {
//char c = 0;
//c = ((pUni[i]>>8) & 0x00ff);
//res += c;
//c = (pUni[i] & 0x00ff);
//res += c;
pair<char, char> char2= uint16ToChar2(pUni[i]); pair<char, char> char2= uint16ToChar2(pUni[i]);
res += char2.first; res += char2.first;
res += char2.second; res += char2.second;
@ -368,6 +363,19 @@ namespace CPPCOMMON
return res; return res;
} }
size_t getUtf8WordLen(const string& utf)
{
string uni = utf8ToUnicode(utf);
if(uni.empty()||uni.size()%2)
{
return 0;
}
else
{
return uni.size()/2;
}
}
} }
#ifdef TEST_STR_FUNCTS #ifdef TEST_STR_FUNCTS
@ -434,6 +442,7 @@ int main()
{ {
cout<<line<<endl; cout<<line<<endl;
string s = gbkToUtf8(line); string s = gbkToUtf8(line);
cout<<getUtf8WordLen(s)<<endl;
s = utf8ToGbk(s); s = utf8ToGbk(s);
cout<<s<<endl; cout<<s<<endl;
} }

2
src/cppcommon/str_functs.h
View File

@ -41,6 +41,8 @@ namespace CPPCOMMON
string gbkToUtf8(const string& gbk); string gbkToUtf8(const string& gbk);
string utf8ToGbk(const string& utf); string utf8ToGbk(const string& utf);
size_t getUtf8WordLen(const string& utf);
inline uint16_t twocharToUint16(char high, char low) inline uint16_t twocharToUint16(char high, char low)
{ {
return (((uint16_t(high) & 0x00ff ) << 8) | (uint16_t(low) & 0x00ff)); return (((uint16_t(high) & 0x00ff ) << 8) | (uint16_t(low) & 0x00ff));