fuzzy: added 'logo' as a known stop word to ignore during fuzzy name …

…matching
stojy · Sep 20, 2023 · 7b2ef9a · 7b2ef9a
1 parent fd8506c
commit 7b2ef9a
Show file tree

Hide file tree

Showing 3 changed files with 6 additions and 5 deletions.
diff --git a/ClrVpin.Tests/Shared/FuzzyTests.cs b/ClrVpin.Tests/Shared/FuzzyTests.cs
@@ -301,6 +301,7 @@ public void MatchTest(string gameName, string fileName, bool expectedSuccess)
     [TestCase("Mephisto (Cirsa 1987)", "jp mephisto.directb2s", 145, TestName = "author 'jp'")]
     [TestCase("Martian Queen (LTD do Brasil 1981)", "marqueen_w.directb2s", 149, TestName = "known word substitution: marqueen and trim single character 'w'")]
     [TestCase("Martian Queen (LTD do Brasil 1981)", "q_marqueen_w.directb2s", 149, TestName = "known word substitution: marqueen and trim single character 'q' and 'w'")]
+    [TestCase("O Gaucho (LTD do Brasil 1975)", "o gaucho_logo.png", 145, TestName = "drop word: logo")]
     public void MatchScoreTest(string databaseName, string fileOrFeedName, int expectedScore)
     {
         // exactly same as MatchTest.. with a score validation

diff --git a/ClrVpin/Shared/Fuzzy/Fuzzy.cs b/ClrVpin/Shared/Fuzzy/Fuzzy.cs
@@ -47,12 +47,12 @@ static Fuzzy()
         // - https://regex101.com/r/DoztL5/1
         Authors = new[] { "jps", "jp's", "jp", "sg1bson", "vpw", "starlion", "pinball58", "vp99", "balutito", "siggis", "uws" };
         string[] language = { "a", "and", "n'", "'n", "n", "the", "en" };
-        string[] vpx = {"vpx8", "vpx", "mod", "vp10", "4k", "b2s", "4player", "2021", "2022", "2023", "2024" }; // order is important, e.g. vpx8 to be stripped before vpx
+        string[] vpx = {"vpx8", "vpx", "mod", "vp10", "4k", "b2s", "4player", "2021", "2022", "2023", "2024", "logo" }; // order is important, e.g. vpx8 to be stripped before vpx
         string[] technologyTypes = { TableType.ElectroMagnetic.ToLower(), TableType.SolidState.ToLower(), TableType.PureMechanical.ToLower() };
         string[] descriptions = { "no leds", "upgrade", "premium" };
         string[] versions = { "beta1", "beta" }; // order is important, e.g. ensure beta1 is removed before beta
         pattern = string.Join('|', Authors.Concat(language).Concat(vpx).Concat(technologyTypes).Concat(descriptions).Concat(versions));
-        _wholeWordRegex = new Regex($"(?<=^|[^a-z^A-Z])({pattern})(?=$|[^a-zA-Z])", RegexOptions.Compiled);
+        _stopWholeWordRegex = new Regex($"(?<=^|[^a-z^A-Z])({pattern})(?=$|[^a-zA-Z])", RegexOptions.Compiled);
 
         // first pass single whitespace
         // - performed BEFORE other checks that aren't sensitive to these changes
@@ -195,7 +195,7 @@ public static string CleanPostSplit(string name, bool removeAllWhiteSpace)
         cleanName = cleanName.ToNullLowerAndTrim() ?? "";
 
         // trim (whole) words
-        cleanName = _wholeWordRegex.Replace(cleanName, "");
+        cleanName = _stopWholeWordRegex.Replace(cleanName, "");
 
         // trim pseudo white space, e.g. trailing '_' char caused by whole removal: blah_VPX8
         cleanName = cleanName.TrimPseudoWhitespace();
@@ -663,7 +663,7 @@ private class MatchDetail
     private static readonly Regex _fileNameInfoRegex;
     private static readonly Regex _trimSpecialAndNonAsciiCharRegex;
     private static readonly Regex _trimTrailingPeriodRegex;
-    private static readonly Regex _wholeWordRegex;
+    private static readonly Regex _stopWholeWordRegex;
     private static readonly Regex _addSpacingFirstPassRegex;
     private static readonly Regex _addSpacingSecondPassRegex;
     private static readonly Regex _versionRegex;

diff --git a/Utils/Extensions/StringExtensions.cs b/Utils/Extensions/StringExtensions.cs
@@ -99,5 +99,5 @@ public static string TrimSingleLetterWords(this string source)
         return source.TrimPseudoWhitespace();
     }
 
-    private static readonly char[] _pseudoWhiteSpaceChars = { ' ', '_', '-' };
+    private static readonly char[] _pseudoWhiteSpaceChars = { ' ', '_', '-' }; // _ = snake case, - = kebab case
 }