# Iterative c++ solution in 0ms

• ``````vector<string> letterCombinations(string digits) {
vector<string> result;
if(digits.empty()) return vector<string>();
static const vector<string> v = {"", "", "abc", "def", "ghi", "jkl", "mno", "pqrs", "tuv", "wxyz"};
result.push_back("");   // add a seed for the initial case
for(int i = 0 ; i < digits.size(); ++i) {
int num = digits[i]-'0';
if(num < 0 || num > 9) break;
const string& candidate = v[num];
if(candidate.empty()) continue;
vector<string> tmp;
for(int j = 0 ; j < candidate.size() ; ++j) {
for(int k = 0 ; k < result.size() ; ++k) {
tmp.push_back(result[k] + candidate[j]);
}
}
result.swap(tmp);
}
return result;
}
``````

Simple and efficient iterative solution.

Explanation with sample input "123"

Initial state:

• result = {""}

Stage 1 for number "1":

• result has {""}
• candiate is "abc"
• generate three strings "" + "a", ""+"b", ""+"c" and put into tmp,
tmp = {"a", "b","c"}
• swap result and tmp (swap does not take memory copy)
• Now result has {"a", "b", "c"}

Stage 2 for number "2":

• result has {"a", "b", "c"}
• candidate is "def"
• generate nine strings and put into tmp,
"a" + "d", "a"+"e", "a"+"f",
"b" + "d", "b"+"e", "b"+"f",
"c" + "d", "c"+"e", "c"+"f"
• so tmp has {"ad", "ae", "af", "bd", "be", "bf", "cd", "ce", "cf" }
• swap result and tmp
• Now result has {"ad", "ae", "af", "bd", "be", "bf", "cd", "ce", "cf" }

Stage 3 for number "3":

• result has {"ad", "ae", "af", "bd", "be", "bf", "cd", "ce", "cf" }
• candidate is "ghi"
• generate 27 strings and put into tmp,
• add "g" for each of "ad", "ae", "af", "bd", "be", "bf", "cd", "ce", "cf"
• add "h" for each of "ad", "ae", "af", "bd", "be", "bf", "cd", "ce", "cf"
• add "h" for each of "ad", "ae", "af", "bd", "be", "bf", "cd", "ce", "cf"
• so, tmp has
{"adg", "aeg", "afg", "bdg", "beg", "bfg", "cdg", "ceg", "cfg"
"adh", "aeh", "afh", "bdh", "beh", "bfh", "cdh", "ceh", "cfh"
"adi", "aei", "afi", "bdi", "bei", "bfi", "cdi", "cei", "cfi" }
• swap result and tmp
• Now result has
{"adg", "aeg", "afg", "bdg", "beg", "bfg", "cdg", "ceg", "cfg"
"adh", "aeh", "afh", "bdh", "beh", "bfh", "cdh", "ceh", "cfh"
"adi", "aei", "afi", "bdi", "bei", "bfi", "cdi", "cei", "cfi" }

Finally, return result.

• It is very smart to use the swap() method.
Concise and high performance.

• AHaaaaaaaaaaaaaaa,swap() is so great!

• Really smart.I thought out almost the same method but I used std::move() instead.

• Like this one. `swap` is neat here.

• Sample input is supposed to be "234"?

• like Permutations by BFS , every iteration create a new buffer which substitute the last one .

• I learned a lot from this, great solution, thanks for sharing!

• Amazing swap() technique.

• What is the time complexity of that solution?

• Explanation with sample input "123" is supposed to be "234"?

• I'm confused about the swap() part. Why do we need the swap? What if we just use result = tmp; there? I did this and it said accepted.
I think every iteration we have a new tmp, I don't know why the swap is necessary.