Result = new IndexedString(rightindex, rightstring, false) Result = new IndexedString(entryindex, entrystring, true) Result = new IndexedString(leftindex, leftstring, false) String rightstring = (indexofentry + Entry.Length) // Generate substrings // string leftstring = ( 0, indexofentry) Int rightindex = Source.Index + indexofentry + Entry.Length Int entryindex = Source.Index + indexofentry // Compute realitve indexes // int leftindex = Source.Index IndexedString result = new IndexedString Private IndexedString splitByEntry(IndexedString Source, string Entry) X => x.Text.Length).Sum() > currentScore.Select( Var currentScore = result.Where(x => x.Word = true) // Comparing current score and temporary score // var temporaryScore = temporarylist.Where(x => x.Word = true) Temporarylist.AddRange(recursiveWordSearch(right, newdictionary)) Temporarylist.AddRange(recursiveWordSearch(left, newdictionary)) // Calling the method on the other two parts recursively // IndexedString middle = devidedBytheEntry

IndexedString devidedBytheEntry = splitByEntry(Stringtosplit, entry) // Deviding the string to 3 parts: the entry, left and right // // Trivial case // if (newdictionary.Length temporarylist = new List() String newdictionary = Dictionary.Where(x => (x)).ToArray() // Checking the buffer // int length = // Private Methods // private List recursiveWordSearch( Var sortedlist = recursiveWordSearch(stringtosplit, Dictionary).OrderBy(x => x.Index) IndexedString stringtosplit = new IndexedString( 0, Input, false) Public string SplitToWords( string Input) This.Dictionary = Dictionary.OrderByDescending(x=>x.Length).ToArray() The nested IndexedString class implements these features. I needed a class that makes strings trackable by storing their relative index to the initial input string, plusĪ variable to mark if a string has sense. The implementation consists of a main class and a nested one. This optimization boosts the performance by magnitudes. The recursive method then starts by checking the corresponding entry in the generated substring list and if it isn't empty, returns its value. This method works as intended, but it could be very slow for long strings, because theĪlgorithm computes the same substrings many times. If it is, then we keep it as the best solution (so far). Meaningful characters: implementing the 1. criteria). Combine all the results and check if it's better than the combined results from the previous entries (by comparing the numbers of the.Call our recursive method on the right string.Call our recursive method on the left string.Mark the entry as meaningful and add it to the possible resultlist.For each entry divide the string into 3 parts:.Iterate through all the entries in the sorted dictionary.If the sorted dictionary is empty then return the input string and mark it as nonsense.

Sort the dictionary descending by length (Implementing the 2.Make a sorted dictionary of the possible substrings.For this I used recursion in my code.įollowing are the steps in the recursive method: The next issue is to compute all the possible permutations. As it turned out, these conditions provide the correct solution most of the times. The most simple approach is to prioritize solutions with the longest words. It's easy to see that short (2-3 character long) words can often come up by chance.

Configure spacing between words in word