// Suggestion checks the spelling of a word
func (m *Model) Suggestion(word string, editDistanceMax int) []string {
// and sort based on likelyness
// change suggestions to a *Term will have to change slicing operations
m.m.Lock()
wordRune := []rune(word)
if len(wordRune)-editDistanceMax > m.Max {
return []string{}
}
hashset1 := util.NewHash()
hashset2 := util.NewHash()
candidates := []string{}
candidates = append(candidates, word)
//hashset1.Add(word)
suggestions := []string{}
//fmt.Println(len(candidates))
for len(candidates) > 0 {
// this does not change the slice the undlerying array is still the same
//candidate, candidates := candidates[0], candidates[1:]
// can do this https://play.golang.org/p/pyum8Afrin
candidate := candidates[0]
candidates = append(candidates[:0], candidates[1:]...)
candidateRune := []rune(candidate)
if temp, ok := m.Data[candidate]; ok {
if temp.Count > 0 && hashset2.Add(candidate) {
//fmt.Println("ok added to suggestions")
suggestions = append(suggestions, temp.Suggestions[:]...)
}
for _, suggestion := range suggestions {
suggestionRune := []rune(suggestion)
if hashset2.Add(suggestion) {
// check the beginning and back of the suggestion
prefix := 0
suffix := 0
for (prefix < len(suggestionRune)) && (prefix < len(wordRune)) &&
(suggestionRune[prefix] == wordRune[prefix]) {
prefix++
}
for (suffix < len(suggestionRune)-prefix) && (suffix < len(wordRune)-prefix) &&
(suggestionRune[len(suggestionRune)-suffix-prefix-1] == wordRune[len(wordRune)-suffix-prefix-1]) {
suffix++
}
distance := 0
if suggestion != word {
if len(suggestionRune) == len(wordRune) {
if (prefix > 0) || (suffix > 0) {
distance = len(wordRune) - len(candidateRune)
}
} else if len(word) == len(candidate) {
if (prefix > 0) || (suffix > 0) {
distance = len(suggestionRune) - len(candidateRune)
}
} else {
if (prefix > 0) || (suffix > 0) {
// can used optimized one here from edit.go
distance = DamerauLevenshteinDistance(
suggestionRune[prefix:len(suggestionRune)-suffix],
wordRune[prefix:len(wordRune)-suffix])
} else {
if (prefix > 0) || (suffix > 0) {
distance = DamerauLevenshteinDistance(suggestionRune, wordRune)
}
}
}
}
// remove for verbose missing here
if distance <= editDistanceMax {
if val, ok := m.Data[suggestion]; ok {
suggestions = append(suggestions, val.Suggestions[:]...)
}
}
}
}
}
if len(wordRune)-len(candidateRune) < editDistanceMax {
// verbose stuff here missing
// this does the same thing as Edits()
for i := 0; i < len(candidateRune); i++ {
delete := string(wordRune[:i]) + string(wordRune[i+1:])
if hashset1.Add(delete) {
candidates = append(candidates, delete)
}
}
}
}
m.m.Unlock()
temp := []string{}
a := make(map[string]struct{})
for _, val := range suggestions {
a[val] = struct{}{}
}
for k := range a {
temp = append(temp, k)
}
return temp
}
func generateWords(password string, m spell.Speller) []Word {
var words []Word
var wordsCollection []Word
// collect best edit distance
bestFoundDistance := 9999
preanalysisRules := []string{":", "r"}
//preanalysisRules := []string{":", "r", "}", "{"}
if !*bruteRules {
preanalysisRules = preanalysisRules[:1]
}
var prePassword string
for _, preRule := range preanalysisRules {
prePassword = rules.ApplyRules([]string{preRule}, password)
var suggestions []string
if len(*wordDebug) > 0 {
suggestions = []string{*wordDebug}
} else if *simpleWords {
suggestions = generateSimpleWords(prePassword, m)
} else {
suggestions = generateAdvancedWords(prePassword, m)
}
hashset1 := util.NewHash()
for _, val := range suggestions {
hashset1.Add(val)
}
// rules for each of the suggestions
for _, suggestion := range suggestions {
suggestion = strings.Replace(suggestion, " ", "", -1)
suggestion = strings.Replace(suggestion, "-", "", -1)
if !hashset1.Exists(suggestion) {
suggestions = append(suggestions, suggestion)
hashset1.Add(suggestion)
}
}
/*
// TODO what is the point of this??
// debugging??
if len(suggestions) != hashset1.Len() {
// make these sorted
//fmt.Println(suggestions)
//fmt.Println(hashset1)
}
*/
for _, suggestion := range suggestions {
distance := Levenshtein(suggestion, prePassword)
temp := Word{
suggestion: suggestion,
distance: distance,
password: prePassword,
preRule: preRule,
bestRuleLength: 9999,
}
words = append(words, temp)
}
}
sort.Sort(Words(words))
for _, word := range words {
if !*moreWords {
if word.distance < bestFoundDistance {
bestFoundDistance = word.distance
}
} else if word.distance > bestFoundDistance {
if *debug {
log.Println("best found distance suboptimal")
}
break
}
// filter words with large edit distance
if word.distance <= *maxWordDist {
wordsCollection = append(wordsCollection, word)
} else {
if *debug {
log.Println("max distance exceeded")
}
}
}
if *maxWords > 0 {
if *maxWords > len(wordsCollection) {
wordsCollection = wordsCollection[:]
} else {
wordsCollection = wordsCollection[:*maxWords]
}
}
return wordsCollection
}