/// <summary>Find suggested spellings for a given input word.</summary>
/// <param name="input">The word being spell checked.</param>
/// <param name="verbosity">The value controlling the quantity/closeness of the retuned suggestions.</param>
/// <param name="maxEditDistance">The maximum edit distance between input and suggested words.</param>
/// <returns>A List of SuggestItem object representing suggested correct spellings for the input word,
/// sorted by edit distance, and secondarily by count frequency.</returns>
public override List <string> Lookup(string inpt)
{
Verbosity verbosity = SymSpell.Verbosity.Closest;
int maxEditDistance = 2;
//verbosity=Top: the suggestion with the highest term frequency of the suggestions of smallest edit distance found
//verbosity=Closest: all suggestions of smallest edit distance found, the suggestions are ordered by term frequency
//verbosity=All: all suggestions <= maxEditDistance, the suggestions are ordered by edit distance, then by term frequency (slower, no early termination)
// maxEditDistance used in Lookup can't be bigger than the maxDictionaryEditDistance
// used to construct the underlying dictionary structure.
if (maxEditDistance > MaxDictionaryEditDistance)
{
throw new ArgumentOutOfRangeException(maxEditDistance.ToString());
}
List <SuggestItem> suggestions = new List <SuggestItem>();
List <string> Namzatlar = new List <string>();
string newinput = inpt.ToLower().Replace(Uyghur.Sozghuch, "");
int inputLen = newinput.Length;
// early exit - word is too big to possibly match any words
if (inputLen - maxEditDistance > maxLength)
{
return(Namzatlar);
}
// deletes we've considered already
HashSet <string> hashset1 = new HashSet <string>();
// suggestions we've considered already
HashSet <string> hashset2 = new HashSet <string>();
// quick look for exact match
long suggestionCount = 0;
if (words.TryGetValue(newinput, out suggestionCount))
{
suggestions.Add(new SuggestItem(newinput, 0, suggestionCount));
}
hashset2.Add(newinput); // we considered the input already in the word.TryGetValue above
int maxEditDistance2 = maxEditDistance;
int candidatePointer = 0;
var singleSuggestion = new string[1] {
string.Empty
};
List <string> candidates = new List <string>();
//add original prefix
int inputPrefixLen = inputLen;
if (inputPrefixLen > prefixLength)
{
inputPrefixLen = prefixLength;
candidates.Add(newinput.Substring(0, inputPrefixLen));
}
else
{
candidates.Add(newinput);
}
var distanceComparer = new EditDistance(newinput, this.distanceAlgorithm);
while (candidatePointer < candidates.Count)
{
string candidate = candidates[candidatePointer++];
int candidateLen = candidate.Length;
int lengthDiff = inputPrefixLen - candidateLen;
//save some time - early termination
//if canddate distance is already higher than suggestion distance, than there are no better suggestions to be expected
if (lengthDiff > maxEditDistance2)
{
// skip to next candidate if Verbosity.All, look no further if Verbosity.Top or Closest
// (candidates are ordered by delete distance, so none are closer than current)
if (verbosity == Verbosity.All)
{
continue;
}
break;
}
//read candidate entry from dictionary
string[] dictSuggestions;
if (deletes.TryGetValue(GetStringHash(candidate), out dictSuggestions))
{
//iterate through suggestions (to other correct dictionary items) of delete item and add them to suggestion list
for (int i = 0; i < dictSuggestions.Length; i++)
{
var suggestion = dictSuggestions[i];
int suggestionLen = suggestion.Length;
if (suggestion == newinput)
{
continue;
}
if ((Math.Abs(suggestionLen - inputLen) > maxEditDistance2) || // input and sugg lengths diff > allowed/current best distance
(suggestionLen < candidateLen) || // sugg must be for a different delete string, in same bin only because of hash collision
(suggestionLen == candidateLen && suggestion != candidate)) // if sugg len = delete len, then it either equals delete or is in same bin only because of hash collision
{
continue;
}
var suggPrefixLen = Math.Min(suggestionLen, prefixLength);
if (suggPrefixLen > inputPrefixLen && (suggPrefixLen - candidateLen) > maxEditDistance2)
{
continue;
}
//True Damerau-Levenshtein Edit Distance: adjust distance, if both distances>0
//We allow simultaneous edits (deletes) of maxEditDistance on on both the dictionary and the input term.
//For replaces and adjacent transposes the resulting edit distance stays <= maxEditDistance.
//For inserts and deletes the resulting edit distance might exceed maxEditDistance.
//To prevent suggestions of a higher edit distance, we need to calculate the resulting edit distance, if there are simultaneous edits on both sides.
//Example: (bank==bnak and bank==bink, but bank!=kanb and bank!=xban and bank!=baxn for maxEditDistance=1)
//Two deletes on each side of a pair makes them all equal, but the first two pairs have edit distance=1, the others edit distance=2.
int distance = 0;
int min = 0;
if (candidateLen == 0)
{
//suggestions which have no common chars with input (inputLen<=maxEditDistance && suggestionLen<=maxEditDistance)
distance = Math.Max(inputLen, suggestionLen);
if (distance > maxEditDistance2 || !hashset2.Add(suggestion))
{
continue;
}
}
else if (suggestionLen == 1)
{
if (newinput.IndexOf(suggestion[0]) < 0)
{
distance = inputLen;
}
else
{
distance = inputLen - 1;
}
if (distance > maxEditDistance2 || !hashset2.Add(suggestion))
{
continue;
}
}
else
//number of edits in prefix ==maxediddistance AND no identic suffix
//, then editdistance>maxEditDistance and no need for Levenshtein calculation
// (inputLen >= prefixLength) && (suggestionLen >= prefixLength)
if ((prefixLength - maxEditDistance == candidateLen) &&
(((min = Math.Min(inputLen, suggestionLen) - prefixLength) > 1) &&
(newinput.Substring(inputLen + 1 - min) != suggestion.Substring(suggestionLen + 1 - min))) ||
((min > 0) && (newinput[inputLen - min] != suggestion[suggestionLen - min]) &&
((newinput[inputLen - min - 1] != suggestion[suggestionLen - min]) ||
(newinput[inputLen - min] != suggestion[suggestionLen - min - 1]))))
{
continue;
}
else
{
// DeleteInSuggestionPrefix is somewhat expensive, and only pays off when verbosity is Top or Closest.
if ((verbosity != Verbosity.All && !DeleteInSuggestionPrefix(candidate, candidateLen, suggestion, suggestionLen)) ||
!hashset2.Add(suggestion))
{
continue;
}
distance = distanceComparer.Compare(suggestion, maxEditDistance2);
if (distance < 0)
{
continue;
}
}
//save some time
//do not process higher distances than those already found, if verbosity<All (note: maxEditDistance2 will always equal maxEditDistance when Verbosity.All)
if (distance <= maxEditDistance2)
{
suggestionCount = words[suggestion];
SuggestItem si = new SuggestItem(suggestion, distance, suggestionCount);
if (suggestions.Count > 0)
{
switch (verbosity)
{
case Verbosity.Closest:
{
//we will calculate DamLev distance only to the smallest found distance so far
if (distance < maxEditDistance2)
{
suggestions.Clear();
}
break;
}
case Verbosity.Top:
{
if (distance < maxEditDistance2 || suggestionCount > suggestions[0].count)
{
maxEditDistance2 = distance;
suggestions[0] = si;
}
continue;
}
}
}
if (verbosity != Verbosity.All)
{
maxEditDistance2 = distance;
}
suggestions.Add(si);
}
} //end foreach
} //end if
//add edits
//derive edits (deletes) from candidate (input) and add them to candidates list
//this is a recursive process until the maximum edit distance has been reached
if ((lengthDiff < maxEditDistance) && (candidateLen <= prefixLength))
{
//save some time
//do not create edits with edit distance smaller than suggestions already found
if (verbosity != Verbosity.All && lengthDiff >= maxEditDistance2)
{
continue;
}
for (int i = 0; i < candidateLen; i++)
{
string delete = candidate.Remove(i, 1);
if (hashset1.Add(delete))
{
candidates.Add(delete);
}
}
}
} //end while
//sort by ascending edit distance, then by descending word frequency
if (suggestions.Count > 1)
{
suggestions.Sort();
}
foreach (SuggestItem item in suggestions)
{
Namzatlar.Add(item.term);
}
if (gYeziq == Uyghur.YEZIQ.ULY)
{
newinput = newinput.Replace('o', 'ö').Replace('u', 'ü').Replace('é', 'e');
if (IsListed(newinput))
{
Namzatlar.Insert(0, newinput);
}
}
return(Namzatlar);
} //end if
public List <SuggestItem> Lookup(string input, string language, int editDistanceMax, int verbose)
{
// editDistanceMax used in Lookup can't be bigger than the editDistanceMax use to construct
// the underlying dictionary structure.
//if (editDistanceMax > this.editDistanceMax) throw new ArgumentOutOfRangeException();
//save some time
if (input.Length - editDistanceMax > maxLength)
{
return(new List <SuggestItem>());
}
List <string> candidates = new List <string>();
HashSet <string> hashset1 = new HashSet <string>();
List <SuggestItem> suggestions = new List <SuggestItem>();
HashSet <string> hashset2 = new HashSet <string>();
int editDistanceMax2 = editDistanceMax;
int candidatePointer = 0;
//add original term
candidates.Add(input);
//add original prefix
if (input.Length > lp)
{
candidates.Add(input.Substring(0, lp));
}
var distanceComparer = new EditDistance(EditDistance.DistanceAlgorithm.DamerauOSA);
while (candidatePointer < candidates.Count)
{
string candidate = candidates[candidatePointer++];
int lengthDiff = Math.Min(input.Length, lp) - candidate.Length;
//save some time
//early termination
//suggestion distance=candidate.distance... candidate.distance+editDistanceMax
//if canddate distance is already higher than suggestion distance, than there are no better suggestions to be expected
if ((verbose < 2) && (suggestions.Count > 0) && (lengthDiff > suggestions[0].distance))
{
goto sort;
}
//read candidate entry from dictionary
if (dictionary.TryGetValue(language + candidate, out int valueo))
{
DictionaryItem value = new DictionaryItem();
if (valueo >= 0)
{
value.suggestions.Add((Int32)valueo);
}
else
{
value = itemlist[-valueo - 1];
}
//if count>0 then candidate entry is correct dictionary term, not only delete item
if (value.count > 0)
{
int distance = input.Length - candidate.Length;
//save some time
//do not process higher distances than those already found, if verbose<2
if ((distance <= editDistanceMax) &&
((verbose == 2) || (suggestions.Count == 0) || (distance <= suggestions[0].distance)) &&
(hashset2.Add(candidate)))
{
//Fix: previously not allways all suggestons within editdistance (verbose=1) or the best suggestion (verbose=0) were returned : e.g. elove did not return love
//suggestions.Clear() was not executed in this branch, if a suggestion with lower edit distance was added here (for verbose<2).
//Then possibly suggestions with higher edit distance remained on top, the suggestion with lower edit distance were added to the end.
//All of them where deleted later once a suggestion with a lower distance than the first item in the list was later added in the other branch.
//Therefore returned suggestions were not always complete for verbose<2.
//remove all existing suggestions of higher distance, if verbose<2
if ((verbose < 2) && (suggestions.Count > 0) && (suggestions[0].distance > distance))
{
suggestions.Clear();
}
//add correct dictionary term term to suggestion list
SuggestItem si = new SuggestItem()
{
term = candidate,
count = value.count,
distance = distance
};
suggestions.Add(si);
//early termination
if ((verbose < 2) && (distance == 0))
{
goto sort;
}
}
}
//iterate through suggestions (to other correct dictionary items) of delete item and add them to suggestion list
foreach (int suggestionint in value.suggestions)
{
//save some time
//skipping double items early: different deletes of the input term can lead to the same suggestion
//index2word
string suggestion = wordlist[suggestionint];
//True Damerau-Levenshtein Edit Distance: adjust distance, if both distances>0
//We allow simultaneous edits (deletes) of editDistanceMax on on both the dictionary and the input term.
//For replaces and adjacent transposes the resulting edit distance stays <= editDistanceMax.
//For inserts and deletes the resulting edit distance might exceed editDistanceMax.
//To prevent suggestions of a higher edit distance, we need to calculate the resulting edit distance, if there are simultaneous edits on both sides.
//Example: (bank==bnak and bank==bink, but bank!=kanb and bank!=xban and bank!=baxn for editDistanceMaxe=1)
//Two deletes on each side of a pair makes them all equal, but the first two pairs have edit distance=1, the others edit distance=2.
int distance = 0;// editDistanceMax+1;
if (suggestion != input)
{
int min = 0;
if (Math.Abs(suggestion.Length - input.Length) > editDistanceMax2)
{
continue;
}
else if (candidate.Length == 0)
{
//suggestions which have no common chars with input (input.length<=editDistanceMax && suggestion.length<=editDistanceMax)
if (!hashset2.Add(suggestion))
{
continue;
}
distance = Math.Max(input.Length, suggestion.Length);
}
else
//number of edits in prefix ==maxediddistance AND no identic suffix, then editdistance>editdistancemax and no need for Levenshtein calculation
// (input.Length >= lp) && (suggestion.Length >= lp)
if ((lp - editDistanceMax == candidate.Length) && (((min = Math.Min(input.Length, suggestion.Length) - lp) > 1) && (input.Substring(input.Length + 1 - min) != suggestion.Substring(suggestion.Length + 1 - min))) || ((min > 0) && (input[input.Length - min] != suggestion[suggestion.Length - min]) && ((input[input.Length - min - 1] != suggestion[suggestion.Length - min]) || (input[input.Length - min] != suggestion[suggestion.Length - min - 1]))))
{
continue;
}
else
//edit distance of remaining string (after prefix)
if ((suggestion.Length == candidate.Length) && (input.Length <= lp))
{
if (!hashset2.Add(suggestion))
{
continue;
}
distance = input.Length - candidate.Length;
}
else if ((input.Length == candidate.Length) && (suggestion.Length <= lp))
{
if (!hashset2.Add(suggestion))
{
continue;
}
distance = suggestion.Length - candidate.Length;
}
else if (hashset2.Add(suggestion))
{
distance = distanceComparer.Compare(input, suggestion, editDistanceMax2);
if (distance < 0)
{
distance = editDistanceMax + 1;
}
}
else
{
continue;
}
}
else if (!hashset2.Add(suggestion))
{
continue;
}
//save some time
//do not process higher distances than those already found, if verbose<2
if ((verbose < 2) && (suggestions.Count > 0) && (distance > suggestions[0].distance))
{
continue;
}
if (distance <= editDistanceMax)
{
if (dictionary.TryGetValue(language + suggestion, out int value2))
{
SuggestItem si = new SuggestItem()
{
term = suggestion,
count = itemlist[-value2 - 1].count,
distance = distance
};
//we will calculate DamLev distance only to the smallest found distance sof far
if (verbose < 2)
{
editDistanceMax2 = distance;
}
//remove all existing suggestions of higher distance, if verbose<2
if ((verbose < 2) && (suggestions.Count > 0) && (suggestions[0].distance > distance))
{
suggestions.Clear();
}
suggestions.Add(si);
}
}
} //end foreach
} //end if
//add edits
//derive edits (deletes) from candidate (input) and add them to candidates list
//this is a recursive process until the maximum edit distance has been reached
if ((lengthDiff < editDistanceMax) && (candidate.Length <= lp))
{
//save some time
//do not create edits with edit distance smaller than suggestions already found
if ((verbose < 2) && (suggestions.Count > 0) && (lengthDiff >= suggestions[0].distance))
{
continue;
}
for (int i = 0; i < candidate.Length; i++)
{
string delete = candidate.Remove(i, 1);
if (hashset1.Add(delete))
{
candidates.Add(delete);
}
}
}
}//end while
//sort by ascending edit distance, then by descending word frequency
sort : if (verbose < 2)
{
suggestions.Sort((x, y) => - x.count.CompareTo(y.count));
}
else
{
suggestions.Sort((x, y) => 2 * x.distance.CompareTo(y.distance) - x.count.CompareTo(y.count));
}
if ((verbose == 0) && (suggestions.Count > 1))
{
return(suggestions.GetRange(0, 1));
}
else
{
return(suggestions);
}
}
public List <SuggestItem> LookupCompound(string input, int editDistanceMax)
{
//parse input string into single terms
string[] termList1 = ParseWords(input);
List <SuggestItem> suggestionsPreviousTerm; //suggestions for a single term
List <SuggestItem> suggestions = new List <SuggestItem>(); //suggestions for a single term
List <SuggestItem> suggestionParts = new List <SuggestItem>(); //1 line with separate parts
//translate every term to its best suggestion, otherwise it remains unchanged
bool lastCombi = false;
for (int i = 0; i < termList1.Length; i++)
{
suggestionsPreviousTerm = new List <SuggestItem>(suggestions.Count); for (int k = 0; k < suggestions.Count; k++)
{
suggestionsPreviousTerm.Add(suggestions[k].ShallowCopy());
}
suggestions = Lookup(termList1[i], Verbosity.Top, editDistanceMax);
//combi check, always before split
if ((i > 0) && !lastCombi)
{
List <SuggestItem> suggestionsCombi = Lookup(termList1[i - 1] + termList1[i], Verbosity.Top, editDistanceMax);
if (suggestionsCombi.Count > 0)
{
SuggestItem best1 = suggestionParts[suggestionParts.Count - 1];
SuggestItem best2 = new SuggestItem();
if (suggestions.Count > 0)
{
best2 = suggestions[0];
}
else
{
best2.term = termList1[i];
best2.distance = editDistanceMax + 1;
best2.count = 0;
}
//if (suggestionsCombi[0].distance + 1 < DamerauLevenshteinDistance(termList1[i - 1] + " " + termList1[i], best1.term + " " + best2.term))
var distanceComparer1 = new EditDistance(termList1[i - 1] + " " + termList1[i], this.distanceAlgorithm); //new
int distance1 = distanceComparer1.Compare(best1.term + " " + best2.term, editDistanceMax);
if ((distance1 >= 0) && (suggestionsCombi[0].distance + 1 < distance1))
{
suggestionsCombi[0].distance++;
suggestionParts[suggestionParts.Count - 1] = suggestionsCombi[0];
lastCombi = true;
goto nextTerm;
}
}
}
lastCombi = false;
//alway split terms without suggestion / never split terms with suggestion ed=0 / never split single char terms
if ((suggestions.Count > 0) && ((suggestions[0].distance == 0) || (termList1[i].Length == 1)))
{
//choose best suggestion
suggestionParts.Add(suggestions[0]);
}
else
{
//if no perfect suggestion, split word into pairs
List <SuggestItem> suggestionsSplit = new List <SuggestItem>();
//add original term
if (suggestions.Count > 0)
{
suggestionsSplit.Add(suggestions[0]);
}
if (termList1[i].Length > 1)
{
for (int j = 1; j < termList1[i].Length; j++)
{
string part1 = termList1[i].Substring(0, j);
string part2 = termList1[i].Substring(j);
SuggestItem suggestionSplit = new SuggestItem();
List <SuggestItem> suggestions1 = Lookup(part1, Verbosity.Top, editDistanceMax);
if (suggestions1.Count > 0)
{
if ((suggestions.Count > 0) && (suggestions[0].term == suggestions1[0].term))
{
break; //if split correction1 == einzelwort correction
}
List <SuggestItem> suggestions2 = Lookup(part2, Verbosity.Top, editDistanceMax);
if (suggestions2.Count > 0)
{
if ((suggestions.Count > 0) && (suggestions[0].term == suggestions2[0].term))
{
break; //if split correction1 == einzelwort correction
}
//select best suggestion for split pair
suggestionSplit.term = suggestions1[0].term + " " + suggestions2[0].term;
var distanceComparer2 = new EditDistance(termList1[i], this.distanceAlgorithm); //new
int distance2 = distanceComparer2.Compare(suggestions1[0].term + " " + suggestions2[0].term, editDistanceMax);
if (distance2 < 0)
{
distance2 = editDistanceMax + 1;
}
suggestionSplit.distance = distance2;
suggestionSplit.count = Math.Min(suggestions1[0].count, suggestions2[0].count);
suggestionsSplit.Add(suggestionSplit);
//early termination of split
if (suggestionSplit.distance == 1)
{
break;
}
}
}
}
if (suggestionsSplit.Count > 0)
{
//select best suggestion for split pair
suggestionsSplit.Sort((x, y) => 2 * x.distance.CompareTo(y.distance) - x.count.CompareTo(y.count));
suggestionParts.Add(suggestionsSplit[0]);
}
else
{
SuggestItem si = new SuggestItem();
si.term = termList1[i];
si.count = 0;
si.distance = editDistanceMax + 1;
suggestionParts.Add(si);
}
}
else
{
SuggestItem si = new SuggestItem();
si.term = termList1[i];
si.count = 0;
si.distance = editDistanceMax + 1;
suggestionParts.Add(si);
}
}
nextTerm :;
}
SuggestItem suggestion = new SuggestItem();
suggestion.count = Int64.MaxValue;
string s = ""; foreach (SuggestItem si in suggestionParts)
{
s += si.term + " "; suggestion.count = Math.Min(suggestion.count, si.count);
} //Console.WriteLine(s);
suggestion.term = s.TrimEnd();
var distanceComparer3 = new EditDistance(suggestion.term, this.distanceAlgorithm); //new
suggestion.distance = distanceComparer3.Compare(input, int.MaxValue);
List <SuggestItem> suggestionsLine = new List <SuggestItem>();
suggestionsLine.Add(suggestion);
return(suggestionsLine);
}
