feat(News): refactor news service and add sentiment tracking

FinlyticNews/Adapters/Discovery/ArticleDiscoveryAdapter.cs

@@ -0,0 +1,22 @@
+using FinlyticCore.Dtos.News;
+
+namespace FinlyticNews.Adapters.Discovery;
+
+/// <summary>
+/// Abstract base class representing an adapter capable of extracting article URLs from web content.
+/// </summary>
+public abstract class ArticleDiscoveryAdapter
+{
+    /// <summary>
+    /// Gets the unique adapter keyword name used for matching (e.g. "rss", "html").
+    /// </summary>
+    public abstract string Name { get; }
+
+    /// <summary>
+    /// Extracts a list of absolute article URLs from the retrieved web/feed page content.
+    /// </summary>
+    /// <param name="content">The raw text or XML content loaded from the source.</param>
+    /// <param name="sourceUrl">The original URL of the source page (used to resolve relative links).</param>
+    /// <returns>A list of resolved absolute URLs.</returns>
+    public abstract List<DiscoveredArticle> ExtractUrls(string content, string sourceUrl);
+}

FinlyticNews/Adapters/Discovery/FinanznachrichtenRssDiscoveryAdapter.cs

@@ -0,0 +1,93 @@
+using System.Text.RegularExpressions;
+using System.Xml.Linq;
+using FinlyticCore.Dtos.News;
+
+namespace FinlyticNews.Adapters.Discovery;
+
+/// <summary>
+/// Specialized RSS discovery adapter for Finanznachrichten.de, capable of parsing custom fn:isin elements and standard RSS metadata.
+/// </summary>
+public partial class FinanznachrichtenRssDiscoveryAdapter : ArticleDiscoveryAdapter
+{
+    private static readonly XNamespace FnNamespace = "http://www.finanznachrichten.de/service/rss";
+
+    /// <inheritdoc />
+    public override string Name => "finanznachrichten_rss";
+
+    /// <inheritdoc />
+    public override List<DiscoveredArticle> ExtractUrls(string content, string sourceUrl)
+    {
+        if (string.IsNullOrWhiteSpace(content)) return [];
+
+        var articles = new List<DiscoveredArticle>();
+        if (!Uri.TryCreate(sourceUrl, UriKind.Absolute, out var baseUri)) return articles;
+
+        try
+        {
+            var doc = XDocument.Parse(content);
+            var channel = doc.Root?.Element("channel");
+            var channelTitle = channel?.Element("title")?.Value?.Trim();
+            var channelLanguage = channel?.Element("language")?.Value?.Trim();
+
+            var items = doc.Descendants("item");
+
+            foreach (var item in items)
+            {
+                var link = item.Element("link")?.Value?.Trim() ?? item.Element("guid")?.Value?.Trim();
+                if (string.IsNullOrWhiteSpace(link)) continue;
+
+                // ISIN Extraktion
+                var isins = new HashSet<string>(StringComparer.OrdinalIgnoreCase);
+                foreach (var isinEl in item.Elements(FnNamespace + "isin"))
+                {
+                    if (string.IsNullOrWhiteSpace(isinEl.Value)) continue;
+
+                    var parts = isinEl.Value.Split(new[] { ',', ';', ' ' }, StringSplitOptions.RemoveEmptyEntries);
+                    foreach (var part in parts)
+                    {
+                        var trimmed = part.Trim().ToUpperInvariant();
+                        if (IsValidIsin(trimmed))
+                        {
+                            isins.Add(trimmed);
+                        }
+                    }
+                }
+
+                var title = item.Element("title")?.Value?.Trim();
+                var description = item.Element("description")?.Value?.Trim();
+
+                DateTime? pubDate = null;
+                var pubDateValue = item.Element("pubDate")?.Value;
+                if (!string.IsNullOrWhiteSpace(pubDateValue) && DateTime.TryParse(pubDateValue, out var parsedDate))
+                {
+                    pubDate = parsedDate.ToUniversalTime();
+                }
+
+                if (Uri.TryCreate(baseUri, link, out var absoluteUri))
+                {
+                    articles.Add(new DiscoveredArticle(
+                        Url: absoluteUri.ToString(),
+                        Isins: isins.Count > 0 ? isins.ToList() : null,
+                        Title: title,
+                        Summary: description,
+                        PublishedAt: pubDate,
+                        Language: channelLanguage,
+                        SourceName: channelTitle
+                    ));
+                }
+            }
+        }
+        catch
+        {
+            // XML Malformed / Parsing Errors ignorieren
+        }
+
+        // Deduplizierung nach URL
+        return articles.GroupBy(a => a.Url, StringComparer.OrdinalIgnoreCase)
+                       .Select(g => g.First())
+                       .ToList();
+    }
+
+    private static bool IsValidIsin(string input) 
+        => input.Length == 12 && char.IsLetter(input[0]) && char.IsLetter(input[1]);
+}

FinlyticNews/Adapters/Discovery/RssDiscoveryAdapter.cs

@@ -0,0 +1,73 @@
+using System.ServiceModel.Syndication;
+using System.Xml;
+using FinlyticCore.Dtos.News;
+
+namespace FinlyticNews.Adapters.Discovery;
+
+/// <summary>
+/// Discovery adapter targeting standard RSS/Atom syndication feeds.
+/// </summary>
+public class RssDiscoveryAdapter : ArticleDiscoveryAdapter
+{
+    /// <inheritdoc />
+    public override string Name => "rss";
+
+    /// <inheritdoc />
+    public override List<DiscoveredArticle> ExtractUrls(string content, string sourceUrl)
+    {
+        if (string.IsNullOrWhiteSpace(content)) return [];
+
+        var articles = new List<DiscoveredArticle>();
+        if (!Uri.TryCreate(sourceUrl, UriKind.Absolute, out var baseUri)) return articles;
+
+        try
+        {
+            using var stringReader = new StringReader(content);
+            using var xmlReader = XmlReader.Create(stringReader, new XmlReaderSettings 
+            { 
+                DtdProcessing = DtdProcessing.Ignore,
+                IgnoreWhitespace = true 
+            });
+
+            var feed = SyndicationFeed.Load(xmlReader);
+            var feedLanguage = feed?.Language;
+            var feedTitle = feed?.Title?.Text;
+
+            if (feed?.Items != null)
+            {
+                foreach (var item in feed.Items)
+                {
+                    var link = item.Links.FirstOrDefault()?.Uri?.ToString() ?? item.Id;
+
+                    if (!string.IsNullOrWhiteSpace(link) && Uri.TryCreate(baseUri, link, out var absoluteUri))
+                    {
+                        var title = item.Title?.Text?.Trim();
+                        var summary = item.Summary?.Text?.Trim();
+                        
+                        DateTime? pubDate = item.PublishDate != DateTimeOffset.MinValue 
+                            ? item.PublishDate.UtcDateTime 
+                            : (item.LastUpdatedTime != DateTimeOffset.MinValue ? item.LastUpdatedTime.UtcDateTime : null);
+
+                        articles.Add(new DiscoveredArticle(
+                            Url: absoluteUri.ToString(),
+                            Isins: null,
+                            Title: title,
+                            Summary: summary,
+                            PublishedAt: pubDate,
+                            Language: feedLanguage,
+                            SourceName: feedTitle
+                        ));
+                    }
+                }
+            }
+        }
+        catch
+        {
+            // Graceful handling fehlerhafter Feeds
+        }
+
+        return articles.GroupBy(u => u.Url, StringComparer.OrdinalIgnoreCase)
+                       .Select(g => g.First())
+                       .ToList();
+    }
+}