Added (some) support for case insensitive expand operations

skybrud · Apr 12, 2021 · e5858ac · e5858ac
1 parent 27118d6
commit e5858ac
Show file tree

Hide file tree

Showing 2 changed files with 46 additions and 2 deletions.
diff --git a/src/Skybrud.TextAnalysis/Hunspell/Expand/HunspellExpandOptions.cs b/src/Skybrud.TextAnalysis/Hunspell/Expand/HunspellExpandOptions.cs
@@ -22,21 +22,30 @@ public class HunspellExpandOptions {
         /// If specified, suggested words with a distance highter than the value of this property will be ignored.</remarks>
         public int MaxDistance { get; set; }
 
+        /// <summary>
+        /// Gets or sets whether the expand should be case insensitive.
+        /// </summary>
+        /// <remarks>This option may not be fully implemented throughout the expand implementation.</remarks>
+        public bool CaseInsentive { get; set; }
+
         #endregion
 
         #region Constructors
 
         /// <summary>
         /// Initializes a new instance with default options.
         /// </summary>
-        public HunspellExpandOptions() { }
+        public HunspellExpandOptions() {
+            CaseInsentive = true;
+        }
 
         /// <summary>
         /// Initializes a new instance based on the specified <paramref name="text"/>.
         /// </summary>
         /// <param name="text">Gets or sets the text. This will typically be the raw string the user enters into the search field.</param>
         public HunspellExpandOptions(string text) {
             Text = text;
+            CaseInsentive = true;
         }
 
         #endregion

diff --git a/src/Skybrud.TextAnalysis/Hunspell/HunspellTextAnalyzer.cs b/src/Skybrud.TextAnalysis/Hunspell/HunspellTextAnalyzer.cs
@@ -360,9 +360,44 @@ public virtual HunspellExpandResult Expand(HunspellExpandOptions options) {
 
                     }
 
+                    // Get the suggestions for "piece"
+                    string[] suggestions = Suggest(piece);
+
+                    // If case insensitivity is enabled, we're looking for a suggestion that is spelled the same way,
+                    // but with different casing. If a match is found, all other suggestions are ignored as they may
+                    // lead to a higher amount of "unrelated" search results.
+                    //
+                    // This will typically be the case if the user enters a name in lowercase (eg. "ole" instead of "Ole"),
+                    // which then technically is a spelling error. The same may be the case for abbreviations.
+                    if (options.CaseInsentive) {
+
+                        string insensitiveMatch = suggestions.FirstOrDefault(x => string.Equals(x, piece, StringComparison.InvariantCultureIgnoreCase));
+
+                        if (insensitiveMatch != null) {
+
+                            // Append the match with the correct casing
+                            or.Append(insensitiveMatch);
+
+                            // Iterate over the stem(s) of "suggestion"
+                            foreach (HunspellStemResult stem in Stem(insensitiveMatch)) {
+
+                                // Append each variant/morph to the list
+                                foreach (string variant in Morph(stem)) {
+                                    or.Append(variant);
+                                }
+
+                            }
+
+                            continue;
+
+                        }
+
+                    }
+
+                    // Append the word as entered by the user
                     or.Append(piece);
 
-                    foreach (string suggestion in Suggest(piece)) {
+                    foreach (string suggestion in suggestions) {
 
                         // Calculate the Levenshtein distance
                         int distance = Levenshtein(piece, suggestion);