inline bool operator==(const std::tr1::unordered_set<size_t> &a, const std::tr1::unordered_set<size_t> &b) {
if (a.size() != b.size()) return false;
std::tr1::unordered_set<size_t>::const_iterator cit = a.begin();
while (a.end() != cit) {
if (b.end() == b.find(*cit))
return false;
cit++;
}
return true;
}
vector<bool> wordBreak1(string s, std::tr1::unordered_set<string> &dict) {
vector<bool> wordB(s.length()+1,false); //wordB stands for the string from 0 to i can be segument
wordB[0]=true;
for(int i=1;i<s.length()+1;i++)
{
for(int j=i-1;j>=0;j--)
{
if(wordB[j]&&dict.find(s.substr(j,i-j))!=dict.end())
{
wordB[i]=true;
break;
}
}
}
return wordB;
}
void dfs(int start,int n,string s,string cur,std::tr1::unordered_set<string> &dict,vector<bool> wordB)
{
if(start==n)
{
result.push_back(cur);
return ;
}
for(int i=start;i<n;i++)
{
string t=s.substr(start,i+1-start);
if(wordB[i+1]&&dict.find(t)!=dict.end())
{
int c=cur.size();//used to record the current element ,will be used to keep the size like the pop operation,all is caused by " "signal
cur+=t;
if(i<n-1)
cur.push_back(' ');//the last element do not need to add a space symbol
dfs(i+1,n,s,cur,dict,wordB);
cur.resize(c);//being used to back to the previous condition
}
}
}