New: Use languages from Torznab/Newznab attributes if given

(cherry picked from commit 9c5a07f62a6e32832c10c80813cd3b98c5859989)

Fixes #3701

src/NzbDrone.Core/Indexers/IndexerBase.cs

@@ -1,6 +1,7 @@
 using System;
 using System.Collections.Generic;
 using System.Linq;
+using System.Text.RegularExpressions;
 using FluentValidation.Results;
 using NLog;
 using NzbDrone.Common.Http;
@@ -15,6 +16,8 @@ namespace NzbDrone.Core.Indexers
     public abstract class IndexerBase<TSettings> : IIndexer
         where TSettings : IIndexerSettings, new()
     {
+        private static readonly Regex MultiRegex = new (@"\b(?<multi>multi)\b", RegexOptions.Compiled | RegexOptions.IgnoreCase);
+
         protected readonly IIndexerStatusService _indexerStatusService;
         protected readonly IConfigService _configService;
         protected readonly IParsingService _parsingService;
@@ -75,6 +78,7 @@ namespace NzbDrone.Core.Indexers
         protected virtual IList<ReleaseInfo> CleanupReleases(IEnumerable<ReleaseInfo> releases)
         {
             var result = releases.DistinctBy(v => v.Guid).ToList();
+            var settings = Definition.Settings as IIndexerSettings;
 
             result.ForEach(c =>
             {

src/NzbDrone.Core/Indexers/Newznab/NewznabRssParser.cs

@@ -4,6 +4,8 @@ using System.Linq;
 using System.Xml.Linq;
 using NzbDrone.Common.Extensions;
 using NzbDrone.Core.Indexers.Exceptions;
+using NzbDrone.Core.Languages;
+using NzbDrone.Core.Parser;
 using NzbDrone.Core.Parser.Model;
 
 namespace NzbDrone.Core.Indexers.Newznab
@@ -101,6 +103,30 @@ namespace NzbDrone.Core.Indexers.Newznab
             return ParseUrl(item.TryGetValue("comments"));
         }
 
+        protected override List<Language> GetLanguages(XElement item)
+        {
+            var languges = TryGetMultipleNewznabAttributes(item, "language");
+            var results = new List<Language>();
+
+            // Try to find <language> elements for some indexers that suck at following the rules.
+            if (languges.Count == 0)
+            {
+                languges = item.Elements("language").Select(e => e.Value).ToList();
+            }
+
+            foreach (var language in languges)
+            {
+                var mappedLanguage = IsoLanguages.FindByName(language)?.Language ?? null;
+
+                if (mappedLanguage != null)
+                {
+                    results.Add(mappedLanguage);
+                }
+            }
+
+            return results;
+        }
+
         protected override long GetSize(XElement item)
         {
             long size;
@@ -165,5 +191,22 @@ namespace NzbDrone.Core.Indexers.Newznab
 
             return defaultValue;
         }
+
+        protected List<string> TryGetMultipleNewznabAttributes(XElement item, string key)
+        {
+            var attrElements = item.Elements(ns + "attr").Where(e => e.Attribute("name").Value.Equals(key, StringComparison.OrdinalIgnoreCase));
+            var results = new List<string>();
+
+            foreach (var element in attrElements)
+            {
+                var attrValue = element.Attribute("value");
+                if (attrValue != null)
+                {
+                    results.Add(attrValue.Value);
+                }
+            }
+
+            return results;
+        }
     }
 }

src/NzbDrone.Core/Indexers/RssParser.cs

@@ -12,6 +12,7 @@ using NzbDrone.Common.Extensions;
 using NzbDrone.Common.Http;
 using NzbDrone.Common.Instrumentation;
 using NzbDrone.Core.Indexers.Exceptions;
+using NzbDrone.Core.Languages;
 using NzbDrone.Core.Parser.Model;
 
 namespace NzbDrone.Core.Indexers
@@ -158,6 +159,7 @@ namespace NzbDrone.Core.Indexers
             releaseInfo.DownloadUrl = GetDownloadUrl(item);
             releaseInfo.InfoUrl = GetInfoUrl(item);
             releaseInfo.CommentUrl = GetCommentUrl(item);
+            releaseInfo.Languages = GetLanguages(item);
 
             try
             {
@@ -224,6 +226,11 @@ namespace NzbDrone.Core.Indexers
             return ParseUrl((string)item.Element("comments"));
         }
 
+        protected virtual List<Language> GetLanguages(XElement item)
+        {
+            return new List<Language>();
+        }
+
         protected virtual long GetSize(XElement item)
         {
             if (UseEnclosureLength)

src/NzbDrone.Core/Indexers/Torznab/TorznabRssParser.cs

@@ -4,6 +4,8 @@ using System.Linq;
 using System.Xml.Linq;
 using NzbDrone.Common.Extensions;
 using NzbDrone.Core.Indexers.Exceptions;
+using NzbDrone.Core.Languages;
+using NzbDrone.Core.Parser;
 using NzbDrone.Core.Parser.Model;
 
 namespace NzbDrone.Core.Indexers.Torznab
@@ -82,6 +84,30 @@ namespace NzbDrone.Core.Indexers.Torznab
             return ParseUrl(item.TryGetValue("comments"));
         }
 
+        protected override List<Language> GetLanguages(XElement item)
+        {
+            var languges = TryGetMultipleTorznabAttributes(item, "language");
+            var results = new List<Language>();
+
+            // Try to find <language> elements for some indexers that suck at following the rules.
+            if (languges.Count == 0)
+            {
+                languges = item.Elements("language").Select(e => e.Value).ToList();
+            }
+
+            foreach (var language in languges)
+            {
+                var mappedLanguage = IsoLanguages.FindByName(language)?.Language ?? null;
+
+                if (mappedLanguage != null)
+                {
+                    results.Add(mappedLanguage);
+                }
+            }
+
+            return results;
+        }
+
         protected override long GetSize(XElement item)
         {
             long size;
@@ -170,5 +196,22 @@ namespace NzbDrone.Core.Indexers.Torznab
 
             return defaultValue;
         }
+
+        protected List<string> TryGetMultipleTorznabAttributes(XElement item, string key)
+        {
+            var attrElements = item.Elements(ns + "attr").Where(e => e.Attribute("name").Value.Equals(key, StringComparison.OrdinalIgnoreCase));
+            var results = new List<string>();
+
+            foreach (var element in attrElements)
+            {
+                var attrValue = element.Attribute("value");
+                if (attrValue != null)
+                {
+                    results.Add(attrValue.Value);
+                }
+            }
+
+            return results;
+        }
     }
 }

src/NzbDrone.Core/Parser/Model/ReleaseInfo.cs

@@ -1,13 +1,20 @@
 using System;
+using System.Collections.Generic;
 using System.Text;
 using Newtonsoft.Json;
 using NzbDrone.Core.Download.Pending;
 using NzbDrone.Core.Indexers;
+using NzbDrone.Core.Languages;
 
 namespace NzbDrone.Core.Parser.Model
 {
     public class ReleaseInfo
     {
+        public ReleaseInfo()
+        {
+            Languages = new List<Language>();
+        }
+
         public string Guid { get; set; }
         public string Title { get; set; }
         public long Size { get; set; }
@@ -28,6 +35,8 @@ namespace NzbDrone.Core.Parser.Model
         public string Codec { get; set; }
         public string Resolution { get; set; }
 
+        public List<Language> Languages { get; set; }
+
         // Used to track pending releases that are being reprocessed
         [JsonIgnore]
         public PendingReleaseReason? PendingReleaseReason { get; set; }