[wip] bridge python and c# and bring in rank torrent name (#177)

* [wip] bridge python and c# and bring in rank torrent name

* Container restores package now

Includes two dev scripts to install the python packages locally for debugging purposes.

* Introduce slightly turned title matching scoring, by making it length aware

this should help with sequels such as Terminator 2, vs Terminator etc

* Version bump

Also fixes postgres healthcheck so that it utilises the user from the stack.env file

deployment/docker/docker-compose.yaml

@@ -17,12 +17,8 @@ services:
   ## All downloaded metadata is stored in this database.
   postgres:
     env_file: stack.env
-    environment:
-      PGUSER: ${POSTGRES_USER}
     healthcheck:
-      test:
-        - CMD-SHELL
-        - pg_isready
+      test: [ "CMD", "sh", "-c", "pg_isready -h localhost -U $$POSTGRES_USER" ]
       timeout: 10s
       interval: 10s
       retries: 3
@@ -44,9 +40,7 @@ services:
   redis:
     env_file: stack.env
     healthcheck:
-      test:
-        - CMD-SHELL
-        - redis-cli ping
+      test: ["CMD-SHELL", "redis-cli ping"]
       timeout: 10s
       interval: 10s
       retries: 3
@@ -66,9 +60,7 @@ services:
   rabbitmq:
     env_file: stack.env
     healthcheck:
-      test:
-        - CMD-SHELL
-        - rabbitmq-diagnostics -q ping
+      test: ["CMD-SHELL", "rabbitmq-diagnostics -q ping"]
       timeout: 10s
       interval: 10s
       retries: 3
@@ -91,22 +83,17 @@ services:
     depends_on:
       metadata:
         condition: service_completed_successfully
-        required: true
       migrator:
         condition: service_completed_successfully
-        required: true
       postgres:
         condition: service_healthy
-        required: true
       rabbitmq:
         condition: service_healthy
-        required: true
       redis:
         condition: service_healthy
-        required: true
     env_file: stack.env
     hostname: knightcrawler-addon
-    image: gabisonfire/knightcrawler-addon:2.0.7
+    image: gabisonfire/knightcrawler-addon:2.0.8
     labels:
       logging: promtail
     networks:
@@ -120,21 +107,16 @@ services:
     depends_on:
       metadata:
         condition: service_completed_successfully
-        required: true
       migrator:
         condition: service_completed_successfully
-        required: true
       postgres:
         condition: service_healthy
-        required: true
       rabbitmq:
         condition: service_healthy
-        required: true
       redis:
         condition: service_healthy
-        required: true
     env_file: stack.env
-    image: gabisonfire/knightcrawler-consumer:2.0.7
+    image: gabisonfire/knightcrawler-consumer:2.0.8
     labels:
       logging: promtail
     networks:
@@ -146,21 +128,16 @@ services:
     depends_on:
       metadata:
         condition: service_completed_successfully
-        required: true
       migrator:
         condition: service_completed_successfully
-        required: true
       postgres:
         condition: service_healthy
-        required: true
       rabbitmq:
         condition: service_healthy
-        required: true
       redis:
         condition: service_healthy
-        required: true
     env_file: stack.env
-    image: gabisonfire/knightcrawler-debrid-collector:2.0.7
+    image: gabisonfire/knightcrawler-debrid-collector:2.0.8
     labels:
       logging: promtail
     networks:
@@ -173,9 +150,8 @@ services:
     depends_on:
       migrator:
         condition: service_completed_successfully
-        required: true
     env_file: stack.env
-    image: gabisonfire/knightcrawler-metadata:2.0.7
+    image: gabisonfire/knightcrawler-metadata:2.0.8
     networks:
       - knightcrawler-network
     restart: "no"
@@ -185,9 +161,8 @@ services:
     depends_on:
       postgres:
         condition: service_healthy
-        required: true
     env_file: stack.env
-    image: gabisonfire/knightcrawler-migrator:2.0.7
+    image: gabisonfire/knightcrawler-migrator:2.0.8
     networks:
       - knightcrawler-network
     restart: "no"
@@ -197,21 +172,16 @@ services:
     depends_on:
       metadata:
         condition: service_completed_successfully
-        required: true
       migrator:
         condition: service_completed_successfully
-        required: true
       postgres:
         condition: service_healthy
-        required: true
       rabbitmq:
         condition: service_healthy
-        required: true
       redis:
         condition: service_healthy
-        required: true
     env_file: stack.env
-    image: gabisonfire/knightcrawler-producer:2.0.7
+    image: gabisonfire/knightcrawler-producer:2.0.8
     labels:
       logging: promtail
     networks:
@@ -223,11 +193,10 @@ services:
     depends_on:
       qbittorrent:
         condition: service_healthy
-        required: true
     deploy:
       replicas: ${QBIT_REPLICAS:-0}
     env_file: stack.env
-    image: gabisonfire/knightcrawler-qbit-collector:2.0.7
+    image: gabisonfire/knightcrawler-qbit-collector:2.0.8
     labels:
       logging: promtail
     networks:
@@ -246,9 +215,7 @@ services:
       TORRENTING_PORT: "6881"
       WEBUI_PORT: "8080"
     healthcheck:
-      test:
-        - CMD-SHELL
-        - curl --fail http://localhost:8080
+      test: ["CMD-SHELL", "curl --fail http://localhost:8080"]
       timeout: 10s
       interval: 10s
       retries: 3

deployment/docker/src/components/infrastructure.yaml

@@ -13,7 +13,7 @@ x-redishealth: &redis-health
   <<: *base-health
 
 x-postgreshealth: &postgresdb-health
-  test: pg_isready
+  test: [ "CMD", "sh", "-c", "pg_isready -h localhost -U $$POSTGRES_USER" ]
   <<: *base-health
 
 x-qbit: &qbit-health

deployment/docker/src/components/knightcrawler.yaml

@@ -20,7 +20,7 @@ x-depends: &knightcrawler-app-depends
 
 services:
   metadata:
-    image: gabisonfire/knightcrawler-metadata:2.0.7
+    image: gabisonfire/knightcrawler-metadata:2.0.8
     env_file: ../../.env
     networks:
       - knightcrawler-network
@@ -30,7 +30,7 @@ services:
         condition: service_completed_successfully
 
   migrator:
-    image: gabisonfire/knightcrawler-migrator:2.0.7
+    image: gabisonfire/knightcrawler-migrator:2.0.8
     env_file: ../../.env
     networks:
       - knightcrawler-network
@@ -40,7 +40,7 @@ services:
         condition: service_healthy
 
   addon:
-    image: gabisonfire/knightcrawler-addon:2.0.7
+    image: gabisonfire/knightcrawler-addon:2.0.8
     <<: [*knightcrawler-app, *knightcrawler-app-depends]
     restart: unless-stopped
     hostname: knightcrawler-addon
@@ -48,22 +48,22 @@ services:
       - "7000:7000"
 
   consumer:
-    image: gabisonfire/knightcrawler-consumer:2.0.7
+    image: gabisonfire/knightcrawler-consumer:2.0.8
     <<: [*knightcrawler-app, *knightcrawler-app-depends]
     restart: unless-stopped
 
   debridcollector:
-    image: gabisonfire/knightcrawler-debrid-collector:2.0.7
+    image: gabisonfire/knightcrawler-debrid-collector:2.0.8
     <<: [*knightcrawler-app, *knightcrawler-app-depends]
     restart: unless-stopped
         
   producer:
-    image: gabisonfire/knightcrawler-producer:2.0.7
+    image: gabisonfire/knightcrawler-producer:2.0.8
     <<: [*knightcrawler-app, *knightcrawler-app-depends]
     restart: unless-stopped
 
   qbitcollector:
-    image: gabisonfire/knightcrawler-qbit-collector:2.0.7
+    image: gabisonfire/knightcrawler-qbit-collector:2.0.8
     <<: [*knightcrawler-app, *knightcrawler-app-depends]
     restart: unless-stopped
     depends_on:

src/producer/.dockerignore

@@ -0,0 +1,2 @@
+**/python/
+.idea/

src/producer/Producer.sln

@@ -6,6 +6,12 @@ Project("{FAE04EC0-301F-11D3-BF4B-00C04F79EFBC}") = "SharedContracts", "..\share
 EndProject
 Project("{2150E333-8FDC-42A3-9474-1A3956D46DE8}") = "shared", "shared", "{FF5CA857-51E8-4446-8840-2A1D24ED3952}"
 EndProject
+Project("{2150E333-8FDC-42A3-9474-1A3956D46DE8}") = "eng", "eng", "{1AE7F597-24C4-4575-B59F-67A625D95C1E}"
+	ProjectSection(SolutionItems) = preProject
+		eng\install-python-reqs.ps1 = eng\install-python-reqs.ps1
+		eng\install-python-reqs.sh = eng\install-python-reqs.sh
+	EndProjectSection
+EndProject
 Global
 	GlobalSection(SolutionConfigurationPlatforms) = preSolution
 		Debug|Any CPU = Debug|Any CPU

src/producer/eng/install-python-reqs.ps1

@@ -0,0 +1,2 @@
+mkdir -p ../src/python
+pip install --force-reinstall rank-torrent-name==0.1.6 -t ../src/python/

src/producer/eng/install-python-reqs.sh

@@ -0,0 +1,4 @@
+#!/bin/bash
+
+mkdir -p ../src/python
+pip install --force-reinstall rank-torrent-name==0.1.6 -t ../src/python/

src/producer/src/.dockerignore

@@ -0,0 +1,2 @@
+**/python/
+.idea/

src/producer/src/Dockerfile

@@ -8,13 +8,21 @@ WORKDIR /src/producer/src
 RUN dotnet restore -a $TARGETARCH
 RUN dotnet publish -c Release --no-restore -o /src/out -a $TARGETARCH
 
-
-FROM mcr.microsoft.com/dotnet/aspnet:8.0-alpine
+FROM mcr.microsoft.com/dotnet/aspnet:8.0-alpine3.19
 
 WORKDIR /app
+
+ENV PYTHONUNBUFFERED=1
+RUN apk add --update --no-cache python3=~3.11.8-r0 py3-pip && ln -sf python3 /usr/bin/python
+
 COPY --from=build /src/out .
+RUN rm -rf /app/python && mkdir -p /app/python
+RUN pip3 install --force-reinstall rank-torrent-name==0.1.6 -t /app/python
 RUN addgroup -S producer && adduser -S -G producer producer
 USER producer
 HEALTHCHECK --interval=30s --timeout=30s --start-period=5s --retries=3 \
   CMD pgrep -f dotnet || exit 1
+
+ENV PYTHONNET_PYDLL=/usr/lib/libpython3.11.so.1.0
+
 ENTRYPOINT ["dotnet", "Producer.dll"]

src/producer/src/Features/Crawlers/Dmm/DebridMediaManagerCrawler.cs

@@ -1,3 +1,5 @@
+using Microsoft.VisualBasic;
+
 namespace Producer.Features.Crawlers.Dmm;
 
 public partial class DebridMediaManagerCrawler(
@@ -5,11 +7,12 @@ public partial class DebridMediaManagerCrawler(
     ILogger<DebridMediaManagerCrawler> logger,
     IDataStorage storage,
     GithubConfiguration githubConfiguration,
-    IParseTorrentTitle parseTorrentTitle,
+    IRankTorrentName rankTorrentName,
     IDistributedCache cache) : BaseCrawler(logger, storage)
 {
     [GeneratedRegex("""<iframe src="https:\/\/debridmediamanager.com\/hashlist#(.*)"></iframe>""")]
     private static partial Regex HashCollectionMatcher();
+    private LengthAwareRatioScorer _lengthAwareRatioScorer = new();
 
     private const string DownloadBaseUrl = "https://raw.githubusercontent.com/debridmediamanager/hashlists/main";
     protected override IReadOnlyDictionary<string, string> Mappings => new Dictionary<string, string>();
@@ -108,73 +111,64 @@ public partial class DebridMediaManagerCrawler(
             return null;
         }
         
-        var parsedTorrent = parseTorrentTitle.Parse(torrentTitle.CleanTorrentTitleForImdb());
+        var parsedTorrent = rankTorrentName.Parse(torrentTitle.CleanTorrentTitleForImdb());
         
-        var (cached, cachedResult) = await CheckIfInCacheAndReturn(parsedTorrent.Title);
-        
-        if (cached)
-        {
-            logger.LogInformation("[{ImdbId}] Found cached imdb result for {Title}", cachedResult.ImdbId, parsedTorrent.Title);
-            return new()
-            {
-                Source = Source,
-                Name = cachedResult.Title,
-                Imdb = cachedResult.ImdbId,
-                Size = bytesElement.GetInt64().ToString(),
-                InfoHash = hashElement.ToString(),
-                Seeders = 0,
-                Leechers = 0,
-                Category = parsedTorrent.TorrentType switch
-                {
-                    TorrentType.Movie => "movies",
-                    TorrentType.Tv => "tv",
-                    _ => "unknown",
-                },
-            };
-        }
-        
-        var imdbEntry = await Storage.FindImdbMetadata(parsedTorrent.Title, parsedTorrent.TorrentType, parsedTorrent.Year);
-
-        if (imdbEntry.Count == 0)
+        if (!parsedTorrent.Success)
         {
             return null;
         }
         
-        var scoredTitles = await ScoreTitles(parsedTorrent, imdbEntry);
+        var (cached, cachedResult) = await CheckIfInCacheAndReturn(parsedTorrent.ParsedTitle);
+        
+        if (cached)
+        {
+            logger.LogInformation("[{ImdbId}] Found cached imdb result for {Title}", cachedResult.ImdbId, parsedTorrent.ParsedTitle);
+            return MapToTorrent(cachedResult, bytesElement, hashElement, parsedTorrent);
+        }
+
+        var year = parsedTorrent.Year != 0 ? parsedTorrent.Year.ToString() : null;
+        var imdbEntries = await Storage.FindImdbMetadata(parsedTorrent.ParsedTitle, parsedTorrent.IsMovie ? "movies" : "tv", year);
+
+        if (imdbEntries.Count == 0)
+        {
+            return null;
+        }
+        
+        var scoredTitles = await ScoreTitles(parsedTorrent, imdbEntries);
         
         if (!scoredTitles.Success)
         {
             return null;
         }
         
-        logger.LogInformation("[{ImdbId}] Found best match for {Title}: {BestMatch} with score {Score}", scoredTitles.BestMatch.Value.ImdbId, parsedTorrent.Title, scoredTitles.BestMatch.Value.Title, scoredTitles.BestMatch.Score);
+        logger.LogInformation("[{ImdbId}] Found best match for {Title}: {BestMatch} with score {Score}", scoredTitles.BestMatch.Value.ImdbId, parsedTorrent.ParsedTitle, scoredTitles.BestMatch.Value.Title, scoredTitles.BestMatch.Score);
 
-        var torrent = new IngestedTorrent
+        return MapToTorrent(scoredTitles.BestMatch.Value, bytesElement, hashElement, parsedTorrent);
+    }
+
+    private IngestedTorrent MapToTorrent(ImdbEntry result, JsonElement bytesElement, JsonElement hashElement, ParseTorrentTitleResponse parsedTorrent) =>
+        new()
         {
             Source = Source,
-            Name = scoredTitles.BestMatch.Value.Title,
-            Imdb = scoredTitles.BestMatch.Value.ImdbId,
+            Name = result.Title,
+            Imdb = result.ImdbId,
             Size = bytesElement.GetInt64().ToString(),
             InfoHash = hashElement.ToString(),
             Seeders = 0,
             Leechers = 0,
-            Category = parsedTorrent.TorrentType switch
+            Category = parsedTorrent.IsMovie switch
             {
-                TorrentType.Movie => "movies",
-                TorrentType.Tv => "tv",
-                _ => "unknown",
+                true => "movies",
+                false => "tv",
             },
         };
 
-        return torrent;
-    }
-
-    private async Task<(bool Success, ExtractedResult<ImdbEntry>? BestMatch)> ScoreTitles(TorrentMetadata parsedTorrent, List<ImdbEntry> imdbEntries)
+    private async Task<(bool Success, ExtractedResult<ImdbEntry>? BestMatch)> ScoreTitles(ParseTorrentTitleResponse parsedTorrent, List<ImdbEntry> imdbEntries)
     {
-        var lowerCaseTitle = parsedTorrent.Title.ToLowerInvariant();
+        var lowerCaseTitle = parsedTorrent.ParsedTitle.ToLowerInvariant();
        
         // Scoring directly operates on the List<ImdbEntry>, no need for lookup table.
-        var scoredResults = Process.ExtractAll(new(){Title = lowerCaseTitle}, imdbEntries, x => x.Title?.ToLowerInvariant(), scorer: new DefaultRatioScorer(), cutoff: 90);
+        var scoredResults = Process.ExtractAll(new(){Title = lowerCaseTitle}, imdbEntries, x => x.Title?.ToLowerInvariant(), scorer: _lengthAwareRatioScorer, cutoff: 90);
 
         var best = scoredResults.MaxBy(x => x.Score);
 
@@ -192,7 +186,7 @@ public partial class DebridMediaManagerCrawler(
     {
         var cacheOptions = new DistributedCacheEntryOptions
         {
-            AbsoluteExpirationRelativeToNow = TimeSpan.FromMinutes(15),
+            AbsoluteExpirationRelativeToNow = TimeSpan.FromDays(1),
         };
         
         return cache.SetStringAsync(lowerCaseTitle, JsonSerializer.Serialize(best.Value), cacheOptions);

src/producer/src/Features/DataProcessing/LengthAwareRatioScorer.cs

@@ -0,0 +1,24 @@
+namespace Producer.Features.DataProcessing
+{
+    public class LengthAwareRatioScorer : IRatioScorer
+    {
+        private readonly IRatioScorer _defaultScorer = new DefaultRatioScorer();
+
+        public int Score(string input1, string input2)
+        {
+            var score = _defaultScorer.Score(input1, input2);
+            var lengthRatio = (double)Math.Min(input1.Length, input2.Length) / Math.Max(input1.Length, input2.Length);
+            var result = (int)(score * lengthRatio);
+            return result > 100 ? 100 : result;
+        }
+
+        public int Score(string input1, string input2, PreprocessMode preprocessMode)
+        {
+            var score = _defaultScorer.Score(input1, input2, preprocessMode);
+            var lengthRatio = (double)Math.Min(input1.Length, input2.Length) / Math.Max(input1.Length, input2.Length);
+            var result = (int)(score * lengthRatio);
+
+            return result > 100 ? 100 : result;
+        }
+    }
+}

src/producer/src/Features/DataProcessing/ServiceCollectionExtensions.cs

@@ -9,7 +9,8 @@ internal static class ServiceCollectionExtensions
        
         services.AddTransient<IDataStorage, DapperDataStorage>();
         services.AddTransient<IMessagePublisher, TorrentPublisher>();
-        services.AddSingleton<IParseTorrentTitle, ParseTorrentTitle>();
+        services.RegisterPythonEngine();
+        services.AddSingleton<IRankTorrentName, RankTorrentName>();
         services.AddStackExchangeRedisCache(options =>
         {
             options.Configuration = redisConfiguration.ConnectionString;

src/producer/src/GlobalUsings.cs

@@ -7,6 +7,8 @@ global using System.Text.RegularExpressions;
 global using System.Xml.Linq;
 global using FuzzySharp;
 global using FuzzySharp.Extractor;
+global using FuzzySharp.PreProcess;
+global using FuzzySharp.SimilarityRatio.Scorer;
 global using FuzzySharp.SimilarityRatio.Scorer.StrategySensitive;
 global using LZStringCSharp;
 global using MassTransit;
@@ -23,11 +25,10 @@ global using Producer.Features.Crawlers.Torrentio;
 global using Producer.Features.CrawlerSupport;
 global using Producer.Features.DataProcessing;
 global using Producer.Features.JobSupport;
-global using PromKnight.ParseTorrentTitle;
-global using Serilog;
 global using SharedContracts.Configuration;
 global using SharedContracts.Dapper;
 global using SharedContracts.Extensions;
 global using SharedContracts.Models;
+global using SharedContracts.Python;
+global using SharedContracts.Python.RTN;
 global using SharedContracts.Requests;
-global using StackExchange.Redis;

src/producer/src/Producer.csproj

@@ -19,6 +19,7 @@
       <PackageReference Include="Microsoft.Extensions.Hosting" Version="8.0.0" />
       <PackageReference Include="Microsoft.Extensions.Http" Version="8.0.0" />
       <PackageReference Include="Polly" Version="8.3.0" />
+      <PackageReference Include="pythonnet" Version="3.0.3" />
       <PackageReference Include="Quartz.Extensions.DependencyInjection" Version="3.8.0" />
       <PackageReference Include="Quartz.Extensions.Hosting" Version="3.8.0" />
       <PackageReference Include="Serilog" Version="3.1.1" />
@@ -34,9 +35,9 @@
       </None>
     </ItemGroup>
 
-    <ItemGroup>
-        <Content Remove="Data\**" />
-        <None Include="Data\**">
+    <ItemGroup Condition="'$(Configuration)' == 'Debug'">
+        <Content Remove="python\**" />
+        <None Include="python\**">
             <CopyToOutputDirectory>Always</CopyToOutputDirectory>
         </None>
     </ItemGroup>

src/shared/Dapper/DapperDataStorage.cs

@@ -115,10 +115,10 @@ public class DapperDataStorage(PostgresConfiguration configuration, RabbitMqConf
             return result.ToList();
         }, "Error getting imdb metadata.", cancellationToken);
 
-    public async Task<List<ImdbEntry>> FindImdbMetadata(string? parsedTorrentTitle, TorrentType torrentType, string? year, CancellationToken cancellationToken = default) =>
+    public async Task<List<ImdbEntry>> FindImdbMetadata(string? parsedTorrentTitle, string torrentType, string? year, CancellationToken cancellationToken = default) =>
         await ExecuteCommandAsync(async connection =>
         {
-            var query = $"select \"imdb_id\" as \"ImdbId\", \"title\" as \"Title\", \"year\" as \"Year\" from search_imdb_meta('{parsedTorrentTitle.Replace("'", "").Replace("\"", "")}', '{(torrentType == TorrentType.Movie ? "movie" : "tvSeries")}'";
+            var query = $"select \"imdb_id\" as \"ImdbId\", \"title\" as \"Title\", \"year\" as \"Year\" from search_imdb_meta('{parsedTorrentTitle.Replace("'", "").Replace("\"", "")}', '{(torrentType.Equals("movie", StringComparison.OrdinalIgnoreCase) ? "movie" : "tvSeries")}'";
             query += year is not null ? $", '{year}'" : ", NULL";
             query += ", 15)";
             

src/shared/Dapper/IDataStorage.cs

@@ -9,7 +9,7 @@ public interface IDataStorage
     Task<DapperResult<PageIngestedResult, PageIngestedResult>> MarkPageAsIngested(string pageId, CancellationToken cancellationToken = default);
     Task<DapperResult<int, int>> GetRowCountImdbMetadata(CancellationToken cancellationToken = default);
     Task<List<ImdbEntry>> GetImdbEntriesForRequests(int year, int batchSize, string? stateLastProcessedImdbId, CancellationToken cancellationToken = default);
-    Task<List<ImdbEntry>> FindImdbMetadata(string? parsedTorrentTitle, TorrentType parsedTorrentTorrentType, string? parsedTorrentYear, CancellationToken cancellationToken = default);
+    Task<List<ImdbEntry>> FindImdbMetadata(string? parsedTorrentTitle, string parsedTorrentTorrentType, string? parsedTorrentYear, CancellationToken cancellationToken = default);
     Task InsertTorrent(Torrent torrent, CancellationToken cancellationToken = default);
     Task InsertFiles(IEnumerable<TorrentFile> files, CancellationToken cancellationToken = default);
     Task InsertSubtitles(IEnumerable<SubtitleFile> subtitles, CancellationToken cancellationToken = default);

src/shared/Extensions/ConfigurationExtensions.cs

@@ -1,4 +1,3 @@
-using Microsoft.Extensions.DependencyInjection;
 using Microsoft.Extensions.DependencyInjection.Extensions;
 
 namespace SharedContracts.Extensions;

src/shared/GlobalUsings.cs

@@ -6,10 +6,11 @@ global using MassTransit;
 global using Microsoft.AspNetCore.Builder;
 global using Microsoft.AspNetCore.Hosting;
 global using Microsoft.Extensions.Configuration;
+global using Microsoft.Extensions.DependencyInjection;
 global using Microsoft.Extensions.Hosting;
 global using Microsoft.Extensions.Logging;
 global using Npgsql;
-global using PromKnight.ParseTorrentTitle;
+global using Python.Runtime;
 global using Serilog;
 global using SharedContracts.Configuration;
 global using SharedContracts.Extensions;

src/shared/Python/PythonEngineService.cs

@@ -0,0 +1,49 @@
+namespace SharedContracts.Python;
+
+public class PythonEngineService(ILogger<PythonEngineService> logger) : IHostedService
+{
+    private IntPtr _mainThreadState;
+    private bool _isInitialized;
+    
+    public Task StartAsync(CancellationToken cancellationToken)
+    {
+        if (_isInitialized)
+        {
+            return Task.CompletedTask;
+        }
+        
+        try
+        {
+            var pythonDllEnv = Environment.GetEnvironmentVariable("PYTHONNET_PYDLL");
+            
+            if (string.IsNullOrWhiteSpace(pythonDllEnv))
+            {
+                logger.LogWarning("PYTHONNET_PYDLL env is not set. Exiting Application");
+                Environment.Exit(1);
+                return Task.CompletedTask;
+            }
+
+            Runtime.PythonDLL = pythonDllEnv;
+            PythonEngine.Initialize();
+            _mainThreadState = PythonEngine.BeginAllowThreads();
+            
+            _isInitialized = true;
+            logger.LogInformation("Python engine initialized");
+        }
+        catch (Exception e)
+        {
+            logger.LogWarning(e, "Failed to initialize Python engine");
+            Environment.Exit(1);
+        }
+        
+        return Task.CompletedTask;
+    }
+
+    public Task StopAsync(CancellationToken cancellationToken)
+    {
+        PythonEngine.EndAllowThreads(_mainThreadState);
+        PythonEngine.Shutdown();
+        
+        return Task.CompletedTask;
+    }
+}

src/shared/Python/RTN/IRankTorrentName.cs

@@ -0,0 +1,8 @@
+namespace SharedContracts.Python.RTN;
+
+public interface IRankTorrentName
+{
+    ParseTorrentTitleResponse Parse(string title);
+    bool IsTrash(string title);
+    bool TitleMatch(string title, string checkTitle);
+}

src/shared/Python/RTN/ParseTorrentTitleResponse.cs

@@ -0,0 +1,6 @@
+namespace SharedContracts.Python.RTN;
+
+public record ParseTorrentTitleResponse(bool Success, string ParsedTitle, int Year, int[]? Season = null, int[]? Episode = null)
+{
+    public bool IsMovie => Season == null && Episode == null;
+}

src/shared/Python/RTN/RankTorrentName.cs

@@ -0,0 +1,118 @@
+namespace SharedContracts.Python.RTN;
+
+public class RankTorrentName : IRankTorrentName
+{
+    private const string SysModuleName = "sys";
+    private const string RtnModuleName = "RTN";
+
+    private readonly ILogger<RankTorrentName> _logger;
+    private dynamic? _sys;
+    private dynamic? _rtn;
+
+    public RankTorrentName(ILogger<RankTorrentName> logger)
+    {
+        _logger = logger;
+        InitModules();
+    }
+
+   
+    public ParseTorrentTitleResponse Parse(string title)
+    {
+        try
+        {
+            using var py = Py.GIL();
+            var result = _rtn?.parse(title);
+
+            if (result == null)
+            {
+                return new(false, string.Empty, 0);
+            }
+
+            return ParseResult(result);
+        }
+        catch (Exception e)
+        {
+            _logger.LogError(e, "Failed to parse title");
+            return new(false, string.Empty, 0);
+        }
+    }
+
+    public bool IsTrash(string title)
+    {
+        try
+        {
+            using var py = Py.GIL();
+            var result = _rtn?.check_trash(title);
+
+            if (result == null)
+            {
+                return false;
+            }
+
+            var response = result.As<bool>() ?? false;
+            
+            return response;
+        }
+        catch (Exception e)
+        {
+            _logger.LogError(e, "Failed to parse title");
+            return false;
+        }
+    }
+    
+    public bool TitleMatch(string title, string checkTitle)
+    {
+        try
+        {
+            using var py = Py.GIL();
+            var result = _rtn?.title_match(title, checkTitle);
+
+            if (result == null)
+            {
+                return false;
+            }
+
+            var response = result.As<bool>() ?? false;
+            
+            return response;
+        }
+        catch (Exception e)
+        {
+            _logger.LogError(e, "Failed to parse title");
+            return false;
+        }
+    }
+    
+    
+    private static ParseTorrentTitleResponse ParseResult(dynamic result)
+    {
+        var parsedTitle = result.GetAttr("parsed_title")?.As<string>() ?? string.Empty;
+        var year = result.GetAttr("year")?.As<int>() ?? 0;
+        var seasonList = result.GetAttr("season")?.As<PyList>();
+        var episodeList = result.GetAttr("episode")?.As<PyList>();
+        int[]? seasons = seasonList?.Length() > 0 ? seasonList.As<int[]>() : null;
+        int[]? episodes = episodeList?.Length() > 0 ? episodeList.As<int[]>() : null;
+
+        return new ParseTorrentTitleResponse(true, parsedTitle, year, seasons, episodes);
+    }
+
+    private void InitModules()
+    {
+        using var py = Py.GIL();
+        _sys = Py.Import(SysModuleName);
+
+        if (_sys == null)
+        {
+            _logger.LogError($"Failed to import Python module: {SysModuleName}");
+            return;
+        }
+
+        _sys.path.append(Path.Combine(AppContext.BaseDirectory, "python"));
+
+        _rtn = Py.Import(RtnModuleName);
+        if (_rtn == null)
+        {
+            _logger.LogError($"Failed to import Python module: {RtnModuleName}");
+        }
+    }
+}

src/shared/Python/ServiceCollectionExtensions.cs

@@ -0,0 +1,13 @@
+namespace SharedContracts.Python;
+
+public static class ServiceCollectionExtensions
+{
+    public static IServiceCollection RegisterPythonEngine(this IServiceCollection services)
+    {
+        services.AddSingleton<PythonEngineService>();
+        
+        services.AddHostedService(p => p.GetRequiredService<PythonEngineService>());
+
+        return services;
+    }
+}

src/shared/SharedContracts.csproj

@@ -16,7 +16,7 @@
   <PackageReference Include="MassTransit.Abstractions" Version="8.2.0" />
   <PackageReference Include="MassTransit.RabbitMQ" Version="8.2.0" />
   <PackageReference Include="Npgsql" Version="8.0.2" />
-  <PackageReference Include="PromKnight.ParseTorrentTitle" Version="1.0.4" />
+  <PackageReference Include="pythonnet" Version="3.0.3" />
   <PackageReference Include="Serilog" Version="3.1.1" />
   <PackageReference Include="Serilog.Extensions.Hosting" Version="8.0.0" />
   <PackageReference Include="Serilog.Settings.Configuration" Version="8.0.0" />