Refactor TItleMatchingService

README.md

@@ -2,10 +2,10 @@
 
 ## English description coming soon
 
-## 12.02.2024: Erste Testversion
+## Erste Testversion
 Wer möchte kann den UmlautAdaptarr jetzt gerne testen! Über Feedback würde ich mich sehr freuen!
 
-Es sollte mit allen *arrs funktionieren, hat aber nur bei Sonarr schon Auswirkungen (abgesehen vom Caching).
+Es sollte mit allen *arrs funktionieren, hat aber nur bei Sonarr und Lidarr schon Auswirkungen (abgesehen vom Caching).
 
 Momentan ist docker dafür nötig, wer kein Docker nutzt muss sich noch etwas gedulden. 
 
@@ -38,8 +38,9 @@ Einige Beispiele findet ihr unter Features.
 
 | Feature                                                           | Status        |
 |-------------------------------------------------------------------|---------------|
-| Sonarr & Prowlarr Support                                         | ✓             |
-| Lidarr Support                                                    | (✓) kein RSS Sync|
+| Prowlarr Support                                                    | ✓|
+| Sonarr Support                                         | ✓             |
+| Lidarr Support                                                    | ✓|
 | Releases mit deutschem Titel werden erkannt   | ✓             |
 | Releases mit TVDB-Alias Titel werden erkannt  | ✓             |
 | Korrekte Suche und Erkennung von Titel mit Umlauten                            | ✓             |
@@ -54,16 +55,18 @@ Einige Beispiele findet ihr unter Features.
 In den Klammern am Ende des Releasenamens (Bild 2 & 4) steht zu Anschauungszwecken der deutsche Titel der vorher nicht gefunden bzw. akzeptiert wurde. Das bleibt natürlich nicht so ;)
 
 **Vorher:** Release wird zwar gefunden, kann aber kann nicht zu geordnet werden.
-![Vorherige Suche ohne deutsche Titel](https://i.imgur.com/7pfRzgH.png)  
+![Vorherige Suche ohne deutsche Titel](https://github.com/PCJones/UmlautAdaptarr/assets/377223/1fce2909-a36c-4f1b-8497-85903357fee3)
 
 **Jetzt:** 2-3 weitere Releases werden gefunden, außerdem meckert Sonarr nicht mehr über den Namen und würde es bei einer automatischen Suche ohne Probleme importieren.
-![Jetzige Suche mit deutschen Titeln](https://i.imgur.com/k55YIN9.png)  
+![Jetzige Suche mit deutschen Titeln](https://github.com/PCJones/UmlautAdaptarr/assets/377223/0edf43ba-2beb-4f22-aaf4-30f9a619dbd6)
+
 
 **Vorher:** Es werden nur Releases mit dem englischen Titel der Serie gefunden
-![Vorherige Suche, englische Titel](https://i.imgur.com/pbRlOeX.png)  
+![Vorherige Suche, englische Titel](https://github.com/PCJones/UmlautAdaptarr/assets/377223/ed7ca0fa-ac36-4584-87ac-b29f32dd9ace)
 
 **Jetzt:**  Es werden auch Titel mit dem deutschen Namen gefunden :D (haben nicht alle Suchergebnisse auf den Screenshot gepasst)
-![Jetzige Suche, deutsche und englische Titel](https://i.imgur.com/eeq0Voj.png)  
+![Jetzige Suche, deutsche und englische Titel](https://github.com/PCJones/UmlautAdaptarr/assets/377223/1c2dbe1a-5943-4fc4-91ef-29708082900e)
+
 
 **Vorher:** Die deutsche Produktion `Alone - Überlebe die Wildnis` hat auf [TheTVDB](https://thetvdb.com/series/alone-uberlebe-die-wildnis) den Englischen Namen `Alone Germany`.
 

UmlautAdaptarr/Models/SearchItem.cs

@@ -54,6 +54,7 @@ namespace UmlautAdaptarr.Models
             else
             {
                 TitleSearchVariations = GenerateVariations(germanTitle, mediaType).ToArray();
+
                 var allTitleVariations = new List<string>(TitleSearchVariations);
 
                 // If aliases are not null, generate variations for each and add them to the list
@@ -66,8 +67,22 @@ namespace UmlautAdaptarr.Models
                     }
                 }
 
-                TitleMatchVariations = allTitleVariations.Distinct().ToArray();
                 AuthorMatchVariations = [];
+
+                // if a german title ends with (DE) also add a search string that replaces (DE) with GERMAN
+                // also add a matching title without (DE)
+                if (germanTitle?.EndsWith("(DE)") ?? false)
+                {
+                    TitleSearchVariations = [.. TitleSearchVariations, .. 
+                        GenerateVariations(
+                            germanTitle.Replace("(DE)", " GERMAN").RemoveExtraWhitespaces(),
+                        mediaType)];
+
+                    allTitleVariations.AddRange(GenerateVariations(germanTitle.Replace("(DE)", "").Trim(), mediaType));
+
+                }
+
+                TitleMatchVariations = allTitleVariations.Distinct(StringComparer.InvariantCultureIgnoreCase).ToArray();
             }
         }
 
@@ -111,7 +126,7 @@ namespace UmlautAdaptarr.Models
             {
                 var cleanTitleWithoutArticle = germanTitle[3..].Trim();
                 baseVariations.AddRange(GenerateVariations(cleanTitleWithoutArticle, mediaType));
-            }
+            }        
 
             // Remove multiple spaces
             var cleanedVariations = baseVariations.Select(variation => variation.RemoveExtraWhitespaces());

UmlautAdaptarr/Providers/SonarrClient.cs

@@ -39,6 +39,7 @@ namespace UmlautAdaptarr.Providers
                             logger.LogWarning($"Sonarr Show {show.id} doesn't have a tvdbId.");
                             continue;
                         }
+
                         (var germanTitle, var aliases) = await titleService.FetchGermanTitleAndAliasesByExternalIdAsync(_mediaType, tvdbId);
                         var searchItem = new SearchItem
                         (

UmlautAdaptarr/Services/CacheService.cs

@@ -84,7 +84,7 @@ namespace UmlautAdaptarr.Services
                             continue;
                         }
                         // After finding a potential item, compare normalizedTitle with each German title variation
-                        foreach (var variation in item?.TitleSearchVariations ?? [])
+                        foreach (var variation in item?.TitleMatchVariations ?? [])
                         {
                             var normalizedVariation = variation.RemoveAccentButKeepGermanUmlauts().ToLower();
                             if (normalizedTitle.StartsWith(variation, StringComparison.OrdinalIgnoreCase))

UmlautAdaptarr/Services/TitleMatchingService.cs

@@ -1,4 +1,5 @@
-using System.Text.RegularExpressions;
+using Microsoft.Extensions.FileSystemGlobbing.Internal;
+using System.Text.RegularExpressions;
 using System.Xml.Linq;
 using UmlautAdaptarr.Models;
 using UmlautAdaptarr.Utilities;
@@ -19,7 +20,7 @@ namespace UmlautAdaptarr.Services
                 if (titleElement != null)
                 {
                     var originalTitle = titleElement.Value;
-                    var normalizedOriginalTitle = NormalizeTitle(originalTitle);
+                    var cleanTitleSeperatedBySpace = ReplaceSeperatorsWithSpace(originalTitle.RemoveAccentButKeepGermanUmlauts());
 
                     var categoryElement = item.Element("category");
                     var category = categoryElement?.Value;
@@ -33,7 +34,7 @@ namespace UmlautAdaptarr.Services
                     if (useCacheService)
                     {
                         // Use CacheService to find a matching SearchItem by title
-                        searchItem = cacheService.SearchItemByTitle(mediaType, normalizedOriginalTitle);
+                        searchItem = cacheService.SearchItemByTitle(mediaType, cleanTitleSeperatedBySpace);
                     }
 
                     if (searchItem == null)
@@ -45,10 +46,10 @@ namespace UmlautAdaptarr.Services
                     switch (mediaType)
                     {
                         case "tv":
-                            FindAndReplaceForMoviesAndTV(logger, searchItem, titleElement, originalTitle, normalizedOriginalTitle!);
+                            FindAndReplaceForMoviesAndTV(logger, searchItem, titleElement, originalTitle, cleanTitleSeperatedBySpace!);
                             break;
                         case "movie":
-                            FindAndReplaceForMoviesAndTV(logger, searchItem, titleElement, originalTitle, normalizedOriginalTitle!);
+                            FindAndReplaceForMoviesAndTV(logger, searchItem, titleElement, originalTitle, cleanTitleSeperatedBySpace!);
                             break;
                         case "audio":
                             FindAndReplaceForAudio(searchItem, titleElement, originalTitle!);
@@ -67,23 +68,22 @@ namespace UmlautAdaptarr.Services
             var authorMatch = FindBestMatch(searchItem.AuthorMatchVariations, originalTitle.NormalizeForComparison(), originalTitle);
             var titleMatch = FindBestMatch(searchItem.TitleMatchVariations, originalTitle.NormalizeForComparison(), originalTitle);
 
-            if (authorMatch.Item1 && titleMatch.Item1)
+            if (authorMatch.foundMatch && titleMatch.foundMatch)
             {
-                int matchEndPositionInOriginal = Math.Max(authorMatch.Item3, titleMatch.Item3);
+                int matchEndPositionInOriginal = Math.Max(authorMatch.bestEndInOriginal, titleMatch.bestEndInOriginal);
 
-                var test = originalTitle[matchEndPositionInOriginal];
                 // Check and adjust for immediate following delimiter
-                if (matchEndPositionInOriginal < originalTitle.Length && new char[] { ' ', '-', '_', '.' }.Contains(originalTitle[matchEndPositionInOriginal]))
+                char[] delimiters = [' ', '-', '_', '.'];
+                if (matchEndPositionInOriginal < originalTitle.Length && delimiters.Contains(originalTitle[matchEndPositionInOriginal]))
                 {
                     matchEndPositionInOriginal++; // Skip the delimiter if it's immediately after the match
                 }
 
                 // Ensure we trim any leading delimiters from the suffix
                 string suffix = originalTitle[matchEndPositionInOriginal..].TrimStart([' ', '-', '_', '.']).Trim();
-                suffix = suffix.Replace("-", ".");
 
                 // Concatenate the expected title with the remaining suffix
-                var updatedTitle = $"{searchItem.ExpectedAuthor} - {searchItem.ExpectedTitle}-{suffix}";
+                var updatedTitle = $"{searchItem.ExpectedAuthor} - {searchItem.ExpectedTitle}-[{suffix}]";
 
                 // Update the title element
                 titleElement.Value = updatedTitle;
@@ -96,7 +96,7 @@ namespace UmlautAdaptarr.Services
         }
 
 
-        private Tuple<bool, int, int> FindBestMatch(string[] variations, string normalizedOriginal, string originalTitle)
+        private (bool foundMatch, int bestStart, int bestEndInOriginal) FindBestMatch(string[] variations, string normalizedOriginal, string originalTitle)
         {
             bool found = false;
             int bestStart = int.MaxValue;
@@ -119,8 +119,8 @@ namespace UmlautAdaptarr.Services
                 }
             }
 
-            if (!found) return Tuple.Create(false, 0, 0);
-            return Tuple.Create(found, bestStart, bestEndInOriginal);
+            if (!found) return (false, 0, 0);
+            return (found, bestStart, bestEndInOriginal);
         }
 
         // Maps an index from the normalized string back to a corresponding index in the original string
@@ -159,6 +159,7 @@ namespace UmlautAdaptarr.Services
             var titleMatchVariations = searchItem.TitleMatchVariations;
             var expectedTitle = searchItem.ExpectedTitle;
             var variationsOrderedByLength = titleMatchVariations!.OrderByDescending(variation => variation.Length);
+
             // Attempt to find a variation that matches the start of the original title
             foreach (var variation in variationsOrderedByLength)
             {
@@ -174,12 +175,6 @@ namespace UmlautAdaptarr.Services
                 // Check if the originalTitle starts with the variation (ignoring case and separators)
                 if (Regex.IsMatch(normalizedOriginalTitle, variationMatchPattern, RegexOptions.IgnoreCase))
                 {
-                    // Workaround for the rare case of e.g. "Frieren: Beyond Journey's End" that also has the alias "Frieren"
-                    if (expectedTitle!.StartsWith(variation, StringComparison.OrdinalIgnoreCase))
-                    {
-                        logger.LogWarning($"TitleMatchingService - Didn't rename: '{originalTitle}' because the expected title '{expectedTitle}' starts with the variation '{variation}'");
-                        continue;
-                    }
                     var originalTitleMatchPattern = "^" + Regex.Escape(variation).Replace("\\ ", "[._ ]");
 
                     // Find the first separator used in the original title for consistent replacement
@@ -191,8 +186,21 @@ namespace UmlautAdaptarr.Services
                     var variationLength = variation.Length;
                     var suffix = originalTitle[Math.Min(variationLength, originalTitle.Length)..];
 
-                    // Clean up any leading separators from the suffix
-                    suffix = Regex.Replace(suffix, "^[._ ]+", "");
+                    // Workaround for the rare case of e.g. "Frieren: Beyond Journey's End" that also has the alias "Frieren"
+                    if (expectedTitle!.StartsWith(variation, StringComparison.OrdinalIgnoreCase))
+                    {
+                        // See if we already matched the whole title by checking if S01E01 pattern is coming next to avoid false positives
+                        // - that won't help with movies but with tv shows
+                        var seasonMatchingPattern = $"^{separator}S\\d{{1,2}}E\\d{{1,2}}";
+                        if (!Regex.IsMatch(suffix, seasonMatchingPattern))
+                        {
+                            logger.LogWarning($"TitleMatchingService - Didn't rename: '{originalTitle}' because the expected title '{expectedTitle}' starts with the variation '{variation}'");
+                            continue;
+                        }
+                    }
+
+                    // Clean up any leading separator from the suffix
+                    suffix = Regex.Replace(suffix, "^ +", "");
 
                     // TODO EVALUTE! definitely make this optional - this adds GERMAN to the title is the title is german to make sure it's recognized as german
                     // can lead to problems with shows such as "dark" that have international dubs
@@ -218,9 +226,8 @@ namespace UmlautAdaptarr.Services
             }
         }
 
-        private static string NormalizeTitle(string title)
+        private static string ReplaceSeperatorsWithSpace(string title)
         {
-            title = title.RemoveAccentButKeepGermanUmlauts();
             // Replace all known separators with space for normalization
             return WordSeperationCharRegex().Replace(title, " ".ToString());
         }

UmlautAdaptarr/Utilities/UrlUtilities.cs

@@ -5,7 +5,7 @@ namespace UmlautAdaptarr.Utilities
 {
     public partial class UrlUtilities
     {
-        [GeneratedRegex(@"^(?!http:\/\/)([a-zA-Z0-9]+(\.[a-zA-Z0-9]+)+.*)$")]
+        [GeneratedRegex(@"^(?!http:\/\/)([a-zA-Z0-9-]+(\.[a-zA-Z0-9-]+)+.*)$")]
         private static partial Regex UrlMatchingRegex();
         public static bool IsValidDomain(string domain)
         {

build_and_push_docker_image.bat

@@ -0,0 +1,18 @@
+@echo off
+SET IMAGE_NAME=pcjones/umlautadaptarr
+
+echo Enter the version number for the Docker image:
+set /p VERSION="Version: "
+
+echo Building Docker image with version %VERSION%...
+docker build -t %IMAGE_NAME%:%VERSION% .
+docker tag %IMAGE_NAME%:%VERSION% %IMAGE_NAME%:latest
+
+echo Pushing Docker image with version %VERSION%...
+docker push %IMAGE_NAME%:%VERSION%
+
+echo Pushing Docker image with tag latest...
+docker push %IMAGE_NAME%:latest
+
+echo Done.
+pause