feat(news): dynamic settings, IFinlyticLogger, live log streaming, and EF migration

This commit is contained in:
2026-08-15 21:30:01 +02:00
parent a1f2b888f6
commit a94c36a878
12 changed files with 800 additions and 555 deletions
+25 -1
View File
@@ -1,5 +1,8 @@
using FinlyticCore.Database;
using FinlyticCore.Entities.Settings;
using FinlyticNews.Entities; using FinlyticNews.Entities;
using Microsoft.EntityFrameworkCore; using Microsoft.EntityFrameworkCore;
using Microsoft.EntityFrameworkCore.Design;
namespace FinlyticNews.Database; namespace FinlyticNews.Database;
@@ -7,7 +10,7 @@ namespace FinlyticNews.Database;
/// Entity Framework Core database context for the news microservice, /// Entity Framework Core database context for the news microservice,
/// managing article sources, processed news, and matched assets. /// managing article sources, processed news, and matched assets.
/// </summary> /// </summary>
public class NewsDbContext : DbContext public class NewsDbContext : DbContext, ISettingsDbContext
{ {
/// <summary> /// <summary>
/// Initializes a new instance of the <see cref="NewsDbContext"/> class. /// Initializes a new instance of the <see cref="NewsDbContext"/> class.
@@ -17,6 +20,11 @@ public class NewsDbContext : DbContext
{ {
} }
/// <summary>
/// Gets or sets the database set for dynamic settings.
/// </summary>
public DbSet<SettingEntity> DynamicSettings => Set<SettingEntity>();
/// <summary> /// <summary>
/// Gets or sets the database set for configured article sources. /// Gets or sets the database set for configured article sources.
/// </summary> /// </summary>
@@ -45,6 +53,12 @@ public class NewsDbContext : DbContext
{ {
base.OnModelCreating(modelBuilder); base.OnModelCreating(modelBuilder);
modelBuilder.Entity<SettingEntity>(entity =>
{
entity.HasKey(e => e.Id);
entity.HasIndex(e => e.Key).IsUnique();
});
// Configure unique index on SourceUrl for deduplication check (idempotency) // Configure unique index on SourceUrl for deduplication check (idempotency)
modelBuilder.Entity<NewsArticleEntity>() modelBuilder.Entity<NewsArticleEntity>()
.HasIndex(a => a.SourceUrl) .HasIndex(a => a.SourceUrl)
@@ -68,3 +82,13 @@ public class NewsDbContext : DbContext
.OnDelete(DeleteBehavior.Cascade); .OnDelete(DeleteBehavior.Cascade);
} }
} }
public class NewsDbContextFactory : IDesignTimeDbContextFactory<NewsDbContext>
{
public NewsDbContext CreateDbContext(string[] args)
{
var optionsBuilder = new DbContextOptionsBuilder<NewsDbContext>();
optionsBuilder.UseNpgsql("Host=localhost;Database=news;Username=postgres;Password=postgres");
return new NewsDbContext(optionsBuilder.Options);
}
}
@@ -0,0 +1,205 @@
// <auto-generated />
using System;
using FinlyticNews.Database;
using Microsoft.EntityFrameworkCore;
using Microsoft.EntityFrameworkCore.Infrastructure;
using Microsoft.EntityFrameworkCore.Migrations;
using Microsoft.EntityFrameworkCore.Storage.ValueConversion;
using Npgsql.EntityFrameworkCore.PostgreSQL.Metadata;
#nullable disable
namespace FinlyticNews.Migrations
{
[DbContext(typeof(NewsDbContext))]
[Migration("20260815183946_AddDynamicSettings")]
partial class AddDynamicSettings
{
/// <inheritdoc />
protected override void BuildTargetModel(ModelBuilder modelBuilder)
{
#pragma warning disable 612, 618
modelBuilder
.HasAnnotation("ProductVersion", "10.0.9")
.HasAnnotation("Relational:MaxIdentifierLength", 63);
NpgsqlModelBuilderExtensions.UseIdentityByDefaultColumns(modelBuilder);
modelBuilder.Entity("FinlyticCore.Entities.Settings.SettingEntity", b =>
{
b.Property<Guid>("Id")
.ValueGeneratedOnAdd()
.HasColumnType("uuid");
b.Property<string>("Key")
.IsRequired()
.HasMaxLength(150)
.HasColumnType("character varying(150)");
b.Property<DateTime>("LastUpdatedUtc")
.HasColumnType("timestamp with time zone");
b.Property<string>("ServiceIdentifier")
.IsRequired()
.HasMaxLength(100)
.HasColumnType("character varying(100)");
b.Property<string>("ValueJson")
.IsRequired()
.HasColumnType("text");
b.HasKey("Id");
b.HasIndex("Key")
.IsUnique();
b.ToTable("DynamicSettings");
});
modelBuilder.Entity("FinlyticNews.Entities.ArticleSourceEntity", b =>
{
b.Property<Guid>("Id")
.ValueGeneratedOnAdd()
.HasColumnType("uuid");
b.Property<string>("Name")
.IsRequired()
.HasColumnType("text");
b.Property<string>("Source")
.IsRequired()
.HasColumnType("text");
b.Property<string>("Type")
.IsRequired()
.HasColumnType("text");
b.HasKey("Id");
b.ToTable("ArticleSources");
});
modelBuilder.Entity("FinlyticNews.Entities.MatchedAssetEntity", b =>
{
b.Property<Guid>("Id")
.ValueGeneratedOnAdd()
.HasColumnType("uuid");
b.Property<string>("Isin")
.IsRequired()
.HasColumnType("text");
b.Property<string>("Name")
.IsRequired()
.HasColumnType("text");
b.Property<Guid>("NewsArticleId")
.HasColumnType("uuid");
b.HasKey("Id");
b.HasIndex("Isin");
b.HasIndex("NewsArticleId");
b.ToTable("MatchedAssets");
});
modelBuilder.Entity("FinlyticNews.Entities.NewsArticleEntity", b =>
{
b.Property<Guid>("Id")
.ValueGeneratedOnAdd()
.HasColumnType("uuid");
b.Property<string>("Author")
.HasColumnType("text");
b.Property<string>("ContentRaw")
.IsRequired()
.HasColumnType("text");
b.Property<string>("Language")
.HasColumnType("text");
b.Property<DateTime>("PublishedAt")
.HasColumnType("timestamp with time zone");
b.Property<DateTime>("ScrapedAt")
.HasColumnType("timestamp with time zone");
b.Property<string>("SourceUrl")
.IsRequired()
.HasColumnType("text");
b.Property<string>("Status")
.IsRequired()
.HasColumnType("text");
b.Property<string>("Summary")
.HasColumnType("text");
b.Property<string>("Title")
.IsRequired()
.HasColumnType("text");
b.HasKey("Id");
b.HasIndex("SourceUrl")
.IsUnique();
b.HasIndex("Status");
b.HasIndex("PublishedAt", "ScrapedAt");
b.ToTable("NewsArticles");
});
modelBuilder.Entity("FinlyticNews.Entities.NewsSettingsEntity", b =>
{
b.Property<Guid>("Id")
.ValueGeneratedOnAdd()
.HasColumnType("uuid");
b.Property<int>("ArticleRetentionDays")
.HasColumnType("integer");
b.Property<int>("DefaultPageSize")
.HasColumnType("integer");
b.Property<string>("N8nWebhookUrl")
.IsRequired()
.HasColumnType("text");
b.Property<int>("PollingFrequencyMinutes")
.HasColumnType("integer");
b.Property<int>("ScrapingIntervalMinutes")
.HasColumnType("integer");
b.Property<DateTime>("UpdatedAt")
.HasColumnType("timestamp with time zone");
b.HasKey("Id");
b.ToTable("Settings");
});
modelBuilder.Entity("FinlyticNews.Entities.MatchedAssetEntity", b =>
{
b.HasOne("FinlyticNews.Entities.NewsArticleEntity", "NewsArticle")
.WithMany("MatchedAssets")
.HasForeignKey("NewsArticleId")
.OnDelete(DeleteBehavior.Cascade)
.IsRequired();
b.Navigation("NewsArticle");
});
modelBuilder.Entity("FinlyticNews.Entities.NewsArticleEntity", b =>
{
b.Navigation("MatchedAssets");
});
#pragma warning restore 612, 618
}
}
}
@@ -0,0 +1,43 @@
using System;
using Microsoft.EntityFrameworkCore.Migrations;
#nullable disable
namespace FinlyticNews.Migrations
{
/// <inheritdoc />
public partial class AddDynamicSettings : Migration
{
/// <inheritdoc />
protected override void Up(MigrationBuilder migrationBuilder)
{
migrationBuilder.CreateTable(
name: "DynamicSettings",
columns: table => new
{
Id = table.Column<Guid>(type: "uuid", nullable: false),
Key = table.Column<string>(type: "character varying(150)", maxLength: 150, nullable: false),
ValueJson = table.Column<string>(type: "text", nullable: false),
ServiceIdentifier = table.Column<string>(type: "character varying(100)", maxLength: 100, nullable: false),
LastUpdatedUtc = table.Column<DateTime>(type: "timestamp with time zone", nullable: false)
},
constraints: table =>
{
table.PrimaryKey("PK_DynamicSettings", x => x.Id);
});
migrationBuilder.CreateIndex(
name: "IX_DynamicSettings_Key",
table: "DynamicSettings",
column: "Key",
unique: true);
}
/// <inheritdoc />
protected override void Down(MigrationBuilder migrationBuilder)
{
migrationBuilder.DropTable(
name: "DynamicSettings");
}
}
}
@@ -22,6 +22,37 @@ namespace FinlyticNews.Migrations
NpgsqlModelBuilderExtensions.UseIdentityByDefaultColumns(modelBuilder); NpgsqlModelBuilderExtensions.UseIdentityByDefaultColumns(modelBuilder);
modelBuilder.Entity("FinlyticCore.Entities.Settings.SettingEntity", b =>
{
b.Property<Guid>("Id")
.ValueGeneratedOnAdd()
.HasColumnType("uuid");
b.Property<string>("Key")
.IsRequired()
.HasMaxLength(150)
.HasColumnType("character varying(150)");
b.Property<DateTime>("LastUpdatedUtc")
.HasColumnType("timestamp with time zone");
b.Property<string>("ServiceIdentifier")
.IsRequired()
.HasMaxLength(100)
.HasColumnType("character varying(100)");
b.Property<string>("ValueJson")
.IsRequired()
.HasColumnType("text");
b.HasKey("Id");
b.HasIndex("Key")
.IsUnique();
b.ToTable("DynamicSettings");
});
modelBuilder.Entity("FinlyticNews.Entities.ArticleSourceEntity", b => modelBuilder.Entity("FinlyticNews.Entities.ArticleSourceEntity", b =>
{ {
b.Property<Guid>("Id") b.Property<Guid>("Id")
+7
View File
@@ -1,3 +1,5 @@
using FinlyticCore.Database;
using FinlyticCore.Services;
using FinlyticNews.Database; using FinlyticNews.Database;
using FinlyticNews.Services; using FinlyticNews.Services;
using FinlyticNews.Util; using FinlyticNews.Util;
@@ -10,10 +12,15 @@ var builder = Host.CreateApplicationBuilder(args);
builder.Services.AddDbContext<NewsDbContext>(options => builder.Services.AddDbContext<NewsDbContext>(options =>
options.UseNpgsql(builder.Configuration.GetConnectionString("DefaultConnection"))); options.UseNpgsql(builder.Configuration.GetConnectionString("DefaultConnection")));
builder.Services.AddScoped<ISettingsDbContext>(sp => sp.GetRequiredService<NewsDbContext>());
// Register standard HttpClient // Register standard HttpClient
builder.Services.AddHttpClient(); builder.Services.AddHttpClient();
// Register Core Services
builder.Services.AddSingleton<ISettingsService, SettingsService>();
builder.Services.AddSingleton(typeof(IFinlyticLogger<>), typeof(FinlyticLogger<>));
// Register Discovery Adapters // Register Discovery Adapters
builder.Services.AddSingleton<ArticleDiscoveryAdapter, RssDiscoveryAdapter>(); builder.Services.AddSingleton<ArticleDiscoveryAdapter, RssDiscoveryAdapter>();
builder.Services.AddSingleton<ArticleDiscoveryAdapter, FinanznachrichtenRssDiscoveryAdapter>(); builder.Services.AddSingleton<ArticleDiscoveryAdapter, FinanznachrichtenRssDiscoveryAdapter>();
@@ -1,21 +1,18 @@
using System;
using System.Collections.Generic;
using System.Linq;
using System.Net.Http;
using System.Threading;
using System.Threading.Tasks;
using FinlyticCore.Dtos.News; using FinlyticCore.Dtos.News;
using FinlyticCore.Services;
using FinlyticNews.Adapters.Discovery; using FinlyticNews.Adapters.Discovery;
using Microsoft.Extensions.Logging; using FinlyticNews.Util;
namespace FinlyticNews.Services; namespace FinlyticNews.Services;
/// <summary>
/// Defines operations for discovering article links from various news feeds and web pages.
/// </summary>
public interface IArticleDiscoveryService public interface IArticleDiscoveryService
{ {
/// <summary>
/// Discovers article links from a given source URL using configured adapters.
/// </summary>
/// <param name="url">The URL of the source news page or feed.</param>
/// <param name="adapterType">The type identifier of the adapter to use (e.g. "rss", "html").</param>
/// <param name="ct">The token to monitor for cancellation requests.</param>
/// <returns>A list of discovered absolute article URLs (optionally carrying ISINs), or null if the URL was invalid.</returns>
Task<List<DiscoveredArticle>?> DiscoverLinksAsync(string url, string adapterType, CancellationToken ct = default); Task<List<DiscoveredArticle>?> DiscoverLinksAsync(string url, string adapterType, CancellationToken ct = default);
} }
@@ -23,22 +20,16 @@ public interface IArticleDiscoveryService
public class ArticleDiscoveryService : IArticleDiscoveryService public class ArticleDiscoveryService : IArticleDiscoveryService
{ {
private readonly IEnumerable<ArticleDiscoveryAdapter> _adapters; private readonly IEnumerable<ArticleDiscoveryAdapter> _adapters;
private readonly ILogger<ArticleDiscoveryService> _logger; private readonly IFinlyticLogger<ArticleDiscoveryService> _finlyticLogger;
private readonly HttpClient _httpClient; private readonly HttpClient _httpClient;
/// <summary>
/// Initializes a new instance of the <see cref="ArticleDiscoveryService"/> class.
/// </summary>
/// <param name="adapters">Registered list of specialized adapters.</param>
/// <param name="logger">The application logging channel.</param>
/// <param name="httpClient">The HTTP client to fetch feeds.</param>
public ArticleDiscoveryService( public ArticleDiscoveryService(
IEnumerable<ArticleDiscoveryAdapter> adapters, IEnumerable<ArticleDiscoveryAdapter> adapters,
ILogger<ArticleDiscoveryService> logger, IFinlyticLogger<ArticleDiscoveryService> finlyticLogger,
HttpClient httpClient) HttpClient httpClient)
{ {
_adapters = adapters; _adapters = adapters;
_logger = logger; _finlyticLogger = finlyticLogger;
_httpClient = httpClient; _httpClient = httpClient;
} }
@@ -47,37 +38,35 @@ public class ArticleDiscoveryService : IArticleDiscoveryService
{ {
if (!Uri.TryCreate(url, UriKind.Absolute, out var uri)) if (!Uri.TryCreate(url, UriKind.Absolute, out var uri))
{ {
_logger.LogWarning("[{Channel}] Invalid source URL passed for discovery: {Url}", "NewsChannel", url); await _finlyticLogger.LogWarningAsync(SettingKeys.NewsChannel, "[ArticleDiscoveryService] Invalid source URL passed for discovery: {Url}", url);
return null; return null;
} }
try try
{ {
_logger.LogDebug("Fetching content from discovery source: {Url}", uri); await _finlyticLogger.LogDebugAsync(SettingKeys.NewsChannel, "[ArticleDiscoveryService] Fetching content from discovery source: {Url}", uri);
var content = await _httpClient.GetStringAsync(uri, ct); var content = await _httpClient.GetStringAsync(uri, ct);
if (string.IsNullOrWhiteSpace(content)) return []; if (string.IsNullOrWhiteSpace(content)) return [];
var trimmedContent = content.TrimStart(); var trimmedContent = content.TrimStart();
// 1. Zuerst gezielt nach registriertem Adapter suchen (z. B. finanznachrichten_rss)
var adapter = _adapters.FirstOrDefault(a => var adapter = _adapters.FirstOrDefault(a =>
a.Name.Equals(adapterType, StringComparison.OrdinalIgnoreCase) || a.Name.Equals(adapterType, StringComparison.OrdinalIgnoreCase) ||
uri.Host.Contains(a.Name, StringComparison.OrdinalIgnoreCase)); uri.Host.Contains(a.Name, StringComparison.OrdinalIgnoreCase));
if (adapter != null) if (adapter != null)
{ {
_logger.LogInformation("[{Channel}] Using specialized adapter {AdapterName} for source: {Url}", "NewsChannel", adapter.Name, url); await _finlyticLogger.LogInfoAsync(SettingKeys.NewsChannel, "[ArticleDiscoveryService] Using specialized adapter {AdapterName} for source: {Url}", adapter.Name, url);
return adapter.ExtractUrls(content, url); return adapter.ExtractUrls(content, url);
} }
// 2. Fallback: Automatische Erkennung für generische RSS/Atom-Feeds
if (adapterType.Equals("rss", StringComparison.OrdinalIgnoreCase) || if (adapterType.Equals("rss", StringComparison.OrdinalIgnoreCase) ||
trimmedContent.StartsWith("<?xml", StringComparison.OrdinalIgnoreCase) || trimmedContent.StartsWith("<?xml", StringComparison.OrdinalIgnoreCase) ||
trimmedContent.StartsWith("<rss", StringComparison.OrdinalIgnoreCase) || trimmedContent.StartsWith("<rss", StringComparison.OrdinalIgnoreCase) ||
trimmedContent.StartsWith("<feed", StringComparison.OrdinalIgnoreCase)) trimmedContent.StartsWith("<feed", StringComparison.OrdinalIgnoreCase))
{ {
_logger.LogInformation("[{Channel}] Syndication (RSS) format detected for source: {Url}", "NewsChannel", url); await _finlyticLogger.LogInfoAsync(SettingKeys.NewsChannel, "[ArticleDiscoveryService] Syndication (RSS) format detected for source: {Url}", url);
var rssAdapter = _adapters.FirstOrDefault(a => a.Name.Equals("rss", StringComparison.OrdinalIgnoreCase)) var rssAdapter = _adapters.FirstOrDefault(a => a.Name.Equals("rss", StringComparison.OrdinalIgnoreCase))
?? new RssDiscoveryAdapter(); ?? new RssDiscoveryAdapter();
@@ -85,12 +74,12 @@ public class ArticleDiscoveryService : IArticleDiscoveryService
return rssAdapter.ExtractUrls(content, url); return rssAdapter.ExtractUrls(content, url);
} }
_logger.LogWarning("[{Channel}] No suitable discovery adapter found for type '{Type}' and URL: {Url}", "NewsChannel", adapterType, url); await _finlyticLogger.LogWarningAsync(SettingKeys.NewsChannel, "[ArticleDiscoveryService] No suitable discovery adapter found for type '{Type}' and URL: {Url}", adapterType, url);
return []; return [];
} }
catch (Exception ex) catch (Exception ex)
{ {
_logger.LogError(ex, "[{Channel}] Failed to run article discovery on URL: {Url}", "NewsChannel", url); await _finlyticLogger.LogErrorAsync(SettingKeys.NewsChannel, ex, "[ArticleDiscoveryService] Failed to run article discovery on URL: {Url}", url);
return []; return [];
} }
} }
+16 -20
View File
@@ -1,14 +1,18 @@
using System;
using System.Collections.Generic;
using System.Net.Http;
using System.Text.Json; using System.Text.Json;
using System.Text.Json.Serialization; using System.Threading;
using System.Threading.Tasks;
using FinlyticCore.Dtos.News;
using FinlyticCore.Services;
using FinlyticCore.Util; using FinlyticCore.Util;
using FinlyticNews.Util;
using Microsoft.Extensions.Configuration; using Microsoft.Extensions.Configuration;
using Microsoft.Extensions.DependencyInjection; using Microsoft.Extensions.DependencyInjection;
using Microsoft.Extensions.Logging;
namespace FinlyticNews.Services; namespace FinlyticNews.Services;
using FinlyticCore.Dtos.News;
/// <summary> /// <summary>
/// Defines integration operations with the external n8n AI workflow webhook. /// Defines integration operations with the external n8n AI workflow webhook.
/// </summary> /// </summary>
@@ -26,21 +30,18 @@ public class N8nService : IN8nService
private readonly HttpClient _httpClient; private readonly HttpClient _httpClient;
private readonly IServiceScopeFactory _scopeFactory; private readonly IServiceScopeFactory _scopeFactory;
private readonly IConfiguration _configuration; private readonly IConfiguration _configuration;
private readonly ILogger<N8nService> _logger; private readonly IFinlyticLogger<N8nService> _finlyticLogger;
/// <summary>
/// Initializes a new instance of the <see cref="N8nService"/> class.
/// </summary>
public N8nService( public N8nService(
HttpClient httpClient, HttpClient httpClient,
IServiceScopeFactory scopeFactory, IServiceScopeFactory scopeFactory,
IConfiguration configuration, IConfiguration configuration,
ILogger<N8nService> logger) IFinlyticLogger<N8nService> finlyticLogger)
{ {
_httpClient = httpClient; _httpClient = httpClient;
_scopeFactory = scopeFactory; _scopeFactory = scopeFactory;
_configuration = configuration; _configuration = configuration;
_logger = logger; _finlyticLogger = finlyticLogger;
} }
/// <inheritdoc /> /// <inheritdoc />
@@ -48,7 +49,6 @@ public class N8nService : IN8nService
{ {
string? targetUrl = null; string? targetUrl = null;
// 1. Dynamic Settings Resolution (DB Scope -> AppSettings Fallback)
using (var scope = _scopeFactory.CreateScope()) using (var scope = _scopeFactory.CreateScope())
{ {
var settingsDb = scope.ServiceProvider.GetService<ISettingsDbService>(); var settingsDb = scope.ServiceProvider.GetService<ISettingsDbService>();
@@ -67,17 +67,16 @@ public class N8nService : IN8nService
if (string.IsNullOrWhiteSpace(targetUrl)) if (string.IsNullOrWhiteSpace(targetUrl))
{ {
_logger.LogError("[{Channel}] N8nWebhookUrl is not configured in DB or application settings.", "NewsChannel"); await _finlyticLogger.LogErrorAsync(SettingKeys.NewsChannel, "[N8nService] N8nWebhookUrl is not configured in DB or application settings.");
return null; return null;
} }
_logger.LogInformation("[{Channel}] Posting article to n8n webhook pipeline at: {Url}", "NewsChannel", targetUrl); await _finlyticLogger.LogInfoAsync(SettingKeys.NewsChannel, "[N8nService] Posting article to n8n webhook pipeline at: {Url}", targetUrl);
var payload = new N8nRequestPayload(content, filteredAssets ?? []); var payload = new N8nRequestPayload(content, filteredAssets ?? []);
try try
{ {
// Zero-Allocation / Source-Generated Request Serialization
var jsonContent = JsonSerializer.Serialize(payload, FinlyticJsonSerializerContext.Default.N8nRequestPayload); var jsonContent = JsonSerializer.Serialize(payload, FinlyticJsonSerializerContext.Default.N8nRequestPayload);
using var requestContent = new StringContent(jsonContent, System.Text.Encoding.UTF8, "application/json"); using var requestContent = new StringContent(jsonContent, System.Text.Encoding.UTF8, "application/json");
@@ -86,7 +85,7 @@ public class N8nService : IN8nService
if (!response.IsSuccessStatusCode) if (!response.IsSuccessStatusCode)
{ {
var errorMsg = await response.Content.ReadAsStringAsync(ct); var errorMsg = await response.Content.ReadAsStringAsync(ct);
_logger.LogError("[{Channel}] n8n webhook returned status code {StatusCode}. Error payload: {Error}", "NewsChannel", response.StatusCode, errorMsg); await _finlyticLogger.LogErrorAsync(SettingKeys.NewsChannel, "[N8nService] n8n webhook returned status code {StatusCode}. Error payload: {Error}", response.StatusCode, errorMsg);
return null; return null;
} }
@@ -95,18 +94,16 @@ public class N8nService : IN8nService
var root = doc.RootElement; var root = doc.RootElement;
// 2. Robust n8n Array-Unwrapping (Handles [{ "json": { ... } }])
if (root.ValueKind == JsonValueKind.Array) if (root.ValueKind == JsonValueKind.Array)
{ {
if (root.GetArrayLength() == 0) if (root.GetArrayLength() == 0)
{ {
_logger.LogWarning("[{Channel}] n8n webhook returned an empty array.", "NewsChannel"); await _finlyticLogger.LogWarningAsync(SettingKeys.NewsChannel, "[N8nService] n8n webhook returned an empty array.");
return null; return null;
} }
root = root[0]; root = root[0];
} }
// 3. Dynamic Node Wrapper Unwrapping ("json", "output", "data", "body")
if (root.ValueKind == JsonValueKind.Object) if (root.ValueKind == JsonValueKind.Object)
{ {
if (root.TryGetProperty("json", out var jsonChild) && jsonChild.ValueKind == JsonValueKind.Object) if (root.TryGetProperty("json", out var jsonChild) && jsonChild.ValueKind == JsonValueKind.Object)
@@ -119,13 +116,12 @@ public class N8nService : IN8nService
root = bodyChild; root = bodyChild;
} }
// 4. Source-Generated Deserialization directly from JsonElement
var result = root.Deserialize(FinlyticJsonSerializerContext.Default.N8nResponsePayload); var result = root.Deserialize(FinlyticJsonSerializerContext.Default.N8nResponsePayload);
return result; return result;
} }
catch (Exception ex) catch (Exception ex)
{ {
_logger.LogError(ex, "[{Channel}] Failed to communicate with or parse response from n8n webhook workflow.", "NewsChannel"); await _finlyticLogger.LogErrorAsync(SettingKeys.NewsChannel, ex, "[N8nService] Failed to communicate with or parse response from n8n webhook workflow.");
return null; return null;
} }
} }
+48 -116
View File
@@ -1,6 +1,12 @@
using System;
using System.Collections.Generic;
using System.Linq;
using System.Threading.Tasks;
using FinlyticCore.Dtos.News; using FinlyticCore.Dtos.News;
using FinlyticCore.Services;
using FinlyticNews.Database; using FinlyticNews.Database;
using FinlyticNews.Entities; using FinlyticNews.Entities;
using FinlyticNews.Util;
using Microsoft.EntityFrameworkCore; using Microsoft.EntityFrameworkCore;
namespace FinlyticNews.Services; namespace FinlyticNews.Services;
@@ -39,22 +45,11 @@ public interface INewsDbService
/// </summary> /// </summary>
Task<NewsArticleEntity?> SaveArticleClassificationAsync(Guid id, N8nResponsePayload payload, List<MatchedAssetEntity> matchedAssets); Task<NewsArticleEntity?> SaveArticleClassificationAsync(Guid id, N8nResponsePayload payload, List<MatchedAssetEntity> matchedAssets);
/// <summary> Task<List<NewsArticleEntity>> GetArticlesByStatusAsync(string status);
/// Retrieves articles ready for historical sync or sentiment processing.
/// </summary>
Task<List<NewsArticleEntity>> GetCompletedArticlesAsync(int limit, int offset, string? isin = null);
/// <summary>
/// Fetches articles matching specific lifecycle statuses (e.g. "Pending", "Scraping" for Phase 1 retry).
/// </summary>
Task<List<NewsArticleEntity>> GetArticlesByStatusAsync(params string[] statuses);
/// <summary>
/// Fetches public daily news for API endpoints, filtering out intermediate or failed lifecycle states by default.
/// </summary>
Task<List<NewsArticleEntity>> GetFilteredNewsAsync( Task<List<NewsArticleEntity>> GetFilteredNewsAsync(
int limit = 20, int limit,
int offset = 0, int offset,
string? isin = null, string? isin = null,
DateTime? date = null, DateTime? date = null,
string? status = null, string? status = null,
@@ -67,15 +62,12 @@ public interface INewsDbService
public class NewsDbService : INewsDbService public class NewsDbService : INewsDbService
{ {
private readonly NewsDbContext _context; private readonly NewsDbContext _context;
private readonly ILogger<NewsDbService> _logger; private readonly IFinlyticLogger<NewsDbService> _finlyticLogger;
/// <summary> public NewsDbService(NewsDbContext context, IFinlyticLogger<NewsDbService> finlyticLogger)
/// Initializes a new instance of the <see cref="NewsDbService"/> class.
/// </summary>
public NewsDbService(NewsDbContext context, ILogger<NewsDbService> logger)
{ {
_context = context; _context = context;
_logger = logger; _finlyticLogger = finlyticLogger;
} }
/// <inheritdoc /> /// <inheritdoc />
@@ -103,9 +95,6 @@ public class NewsDbService : INewsDbService
} }
/// <inheritdoc /> /// <inheritdoc />
/// <summary>
/// Lifecycle Step 1: Creates a new article in 'Pending' state as an immediate lock.
/// </summary>
public async Task<NewsArticleEntity> CreatePendingArticleAsync( public async Task<NewsArticleEntity> CreatePendingArticleAsync(
string url, string url,
List<string>? discoveredIsins = null, List<string>? discoveredIsins = null,
@@ -121,7 +110,7 @@ public class NewsDbService : INewsDbService
if (existingArticle != null) if (existingArticle != null)
{ {
_logger.LogDebug("[Lifecycle] Article URL already exists (Duplicate hit): {Url}", trimmedUrl); await _finlyticLogger.LogDebugAsync(SettingKeys.NewsChannel, "[Lifecycle] Article URL already exists (Duplicate hit): {Url}", trimmedUrl);
return existingArticle; return existingArticle;
} }
@@ -146,7 +135,7 @@ public class NewsDbService : INewsDbService
Language = language, Language = language,
ScrapedAt = DateTime.UtcNow, ScrapedAt = DateTime.UtcNow,
PublishedAt = finalPublishedAt, PublishedAt = finalPublishedAt,
Status = "Pending" // 1. Pending State Status = "Pending"
}; };
if (discoveredIsins != null && discoveredIsins.Count > 0) if (discoveredIsins != null && discoveredIsins.Count > 0)
@@ -167,11 +156,11 @@ public class NewsDbService : INewsDbService
{ {
_context.NewsArticles.Add(article); _context.NewsArticles.Add(article);
await _context.SaveChangesAsync(); await _context.SaveChangesAsync();
_logger.LogDebug("[Lifecycle] Registered new article with status 'Pending'. ID: {Id}, Url: {Url}", article.Id, trimmedUrl); await _finlyticLogger.LogDebugAsync(SettingKeys.NewsChannel, "[Lifecycle] Registered new article with status 'Pending'. ID: {Id}, Url: {Url}", article.Id, trimmedUrl);
} }
catch (DbUpdateConcurrencyException) catch (DbUpdateConcurrencyException)
{ {
_logger.LogWarning("[{Channel}] Concurrency hit during insert for URL: {Url}. Fetching existing fallback.", "NewsChannel", trimmedUrl); await _finlyticLogger.LogWarningAsync(SettingKeys.NewsChannel, "[NewsChannel] Concurrency hit during insert for URL: {Url}. Fetching existing fallback.", trimmedUrl);
return await _context.NewsArticles.FirstAsync(a => a.SourceUrl == trimmedUrl); return await _context.NewsArticles.FirstAsync(a => a.SourceUrl == trimmedUrl);
} }
@@ -179,9 +168,6 @@ public class NewsDbService : INewsDbService
} }
/// <inheritdoc /> /// <inheritdoc />
/// <summary>
/// Lifecycle Step 2 & 5: Updates state (e.g. Pending -> Processing -> Scraping / Failed / Analyzed).
/// </summary>
public async Task UpdateArticleStatusAsync(Guid id, string status) public async Task UpdateArticleStatusAsync(Guid id, string status)
{ {
var rowsAffected = await _context.NewsArticles var rowsAffected = await _context.NewsArticles
@@ -190,11 +176,11 @@ public class NewsDbService : INewsDbService
if (rowsAffected == 0) if (rowsAffected == 0)
{ {
_logger.LogWarning("[{Channel}] Attempted status transition for non-existing article. ID: {Id}", "NewsChannel", id); await _finlyticLogger.LogWarningAsync(SettingKeys.NewsChannel, "[NewsChannel] Attempted status transition for non-existing article. ID: {Id}", id);
} }
else else
{ {
_logger.LogDebug("[Lifecycle] Transitioned article {Id} to status '{Status}'", id, status); await _finlyticLogger.LogDebugAsync(SettingKeys.NewsChannel, "[Lifecycle] Transitioned article {Id} to status '{Status}'", id, status);
} }
} }
@@ -207,18 +193,15 @@ public class NewsDbService : INewsDbService
if (rowsAffected == 0) if (rowsAffected == 0)
{ {
_logger.LogWarning("[{Channel}] Attempted URL update for non-existing article. ID: {Id}", "NewsChannel", id); await _finlyticLogger.LogWarningAsync(SettingKeys.NewsChannel, "[NewsChannel] Attempted URL update for non-existing article. ID: {Id}", id);
} }
else else
{ {
_logger.LogDebug("[Lifecycle] Resolved redirect for article {Id} -> New URL: {Url}", id, resolvedUrl); await _finlyticLogger.LogDebugAsync(SettingKeys.NewsChannel, "[Lifecycle] Resolved redirect for article {Id} -> New URL: {Url}", id, resolvedUrl);
} }
} }
/// <inheritdoc /> /// <inheritdoc />
/// <summary>
/// Lifecycle Step 4: Persists n8n classification and transitions status to 'Completed'.
/// </summary>
public async Task<NewsArticleEntity?> SaveArticleClassificationAsync(Guid id, N8nResponsePayload payload, List<MatchedAssetEntity> matchedAssets) public async Task<NewsArticleEntity?> SaveArticleClassificationAsync(Guid id, N8nResponsePayload payload, List<MatchedAssetEntity> matchedAssets)
{ {
var article = await _context.NewsArticles var article = await _context.NewsArticles
@@ -226,7 +209,7 @@ public class NewsDbService : INewsDbService
if (article == null) if (article == null)
{ {
_logger.LogWarning("[{Channel}] Article with ID {Id} not found for classification update.", "NewsChannel", id); await _finlyticLogger.LogWarningAsync(SettingKeys.NewsChannel, "[NewsChannel] Article with ID {Id} not found for classification update.", id);
return null; return null;
} }
@@ -235,8 +218,6 @@ public class NewsDbService : INewsDbService
article.Summary = payload.Summary; article.Summary = payload.Summary;
article.ContentRaw = payload.ContentRaw; article.ContentRaw = payload.ContentRaw;
article.Language = payload.Language; article.Language = payload.Language;
// 🎯 Step 4: Classification finished -> Transition to 'Completed' (triggers MQTT broadcast)
article.Status = "Completed"; article.Status = "Completed";
if (DateTime.TryParse(payload.PublishedAt, out var publishedDate)) if (DateTime.TryParse(payload.PublishedAt, out var publishedDate))
@@ -251,7 +232,6 @@ public class NewsDbService : INewsDbService
article.ScrapedAt = scrapedDate.ToUniversalTime(); article.ScrapedAt = scrapedDate.ToUniversalTime();
} }
// Clean up previous temporary assets
await _context.MatchedAssets.Where(m => m.NewsArticleId == id).ExecuteDeleteAsync(); await _context.MatchedAssets.Where(m => m.NewsArticleId == id).ExecuteDeleteAsync();
article.MatchedAssets = new List<MatchedAssetEntity>(); article.MatchedAssets = new List<MatchedAssetEntity>();
@@ -267,113 +247,65 @@ public class NewsDbService : INewsDbService
} }
await _context.SaveChangesAsync(); await _context.SaveChangesAsync();
_logger.LogInformation("[Lifecycle] Article {Id} successfully classified and marked 'Completed'. Title: '{Title}'", article.Id, article.Title); await _finlyticLogger.LogInfoAsync(SettingKeys.NewsChannel, "[Lifecycle] Article {Id} successfully classified and marked 'Completed'. Title: '{Title}'", article.Id, article.Title);
return article; return article;
} }
/// <inheritdoc /> /// <inheritdoc />
public async Task<List<NewsArticleEntity>> GetCompletedArticlesAsync(int limit, int offset, string? isin = null) public async Task<List<NewsArticleEntity>> GetArticlesByStatusAsync(string status)
{ {
var query = _context.NewsArticles return await _context.NewsArticles
.Include(a => a.MatchedAssets) .Include(a => a.MatchedAssets)
.Where(a => a.Status == "Completed" || a.Status == "Analyzed"); .Where(a => a.Status == status)
if (!string.IsNullOrWhiteSpace(isin))
{
var cleanIsin = isin.Trim();
query = query.Where(a => a.MatchedAssets.Any(m => m.Isin == cleanIsin || m.Name == cleanIsin));
}
return await query
.OrderByDescending(a => a.PublishedAt) .OrderByDescending(a => a.PublishedAt)
.ThenByDescending(a => a.ScrapedAt)
.ThenByDescending(a => a.Id)
.Skip(offset)
.Take(limit)
.AsNoTracking() .AsNoTracking()
.ToListAsync(); .ToListAsync();
} }
/// <inheritdoc /> /// <inheritdoc />
/// <summary>
/// Lifecycle Helper: Retrieves articles by target lifecycle status (e.g. "Pending", "Scraping" for Phase 1 Retry).
/// </summary>
public async Task<List<NewsArticleEntity>> GetArticlesByStatusAsync(params string[] statuses)
{
IQueryable<NewsArticleEntity> query = _context.NewsArticles
.Include(a => a.MatchedAssets)
.AsNoTracking();
if (statuses != null && statuses.Length > 0)
{
var cleanStatuses = statuses.Select(s => s.Trim()).ToList();
query = query.Where(a => cleanStatuses.Contains(a.Status));
}
else
{
query = query.Where(a => a.Status != "Failed" && a.Status != "Duplicate");
}
return await query.ToListAsync();
}
/// <inheritdoc />
/// <summary>
/// API Gateway Helper: Retrieves articles for UI rendering, excluding intermediate/failed states by default.
/// </summary>
public async Task<List<NewsArticleEntity>> GetFilteredNewsAsync( public async Task<List<NewsArticleEntity>> GetFilteredNewsAsync(
int limit = 20, int limit,
int offset = 0, int offset,
string? isin = null, string? isin = null,
DateTime? date = null, DateTime? date = null,
string? status = null, string? status = null,
string? searchQuery = null) string? searchQuery = null)
{ {
IQueryable<NewsArticleEntity> query = _context.NewsArticles var query = _context.NewsArticles
.Include(a => a.MatchedAssets) .Include(a => a.MatchedAssets)
.AsNoTracking(); .AsNoTracking()
.AsQueryable();
// 1. Status Filter
if (!string.IsNullOrWhiteSpace(status))
{
var targetStatus = status.Trim();
query = query.Where(a => a.Status == targetStatus);
}
else
{
// By default, only show fully processed articles to the API/UI
query = query.Where(a => a.Status == "Completed" || a.Status == "Analyzed");
}
// 2. Date Filter
if (date.HasValue)
{
// Erstelle ein exaktes UTC-Datum von 00:00:00 Uhr am gebuchten Tag
var targetDate = new DateTime(date.Value.Year, date.Value.Month, date.Value.Day, 0, 0, 0, DateTimeKind.Utc);
var nextDate = targetDate.AddDays(1);
query = query.Where(a => a.PublishedAt >= targetDate && a.PublishedAt < nextDate);
}
// 3. ISIN / Symbol Filter
if (!string.IsNullOrWhiteSpace(isin)) if (!string.IsNullOrWhiteSpace(isin))
{ {
var cleanIsin = isin.Trim(); var cleanIsin = isin.Trim();
query = query.Where(a => a.MatchedAssets.Any(m => m.Isin == cleanIsin || m.Name == cleanIsin)); query = query.Where(a => a.MatchedAssets.Any(m => m.Isin == cleanIsin));
}
if (date.HasValue)
{
var startUtc = date.Value.Date.ToUniversalTime();
var endUtc = startUtc.AddDays(1);
query = query.Where(a => a.PublishedAt >= startUtc && a.PublishedAt < endUtc);
}
if (!string.IsNullOrWhiteSpace(status))
{
query = query.Where(a => a.Status == status);
} }
// 4. Search Term
if (!string.IsNullOrWhiteSpace(searchQuery)) if (!string.IsNullOrWhiteSpace(searchQuery))
{ {
var q = searchQuery.Trim(); var cleanSearch = searchQuery.Trim().ToLower();
query = query.Where(a => EF.Functions.ILike(a.Title, $"%{q}%") || (a.Summary != null && EF.Functions.ILike(a.Summary, $"%{q}%"))); query = query.Where(a =>
a.Title.ToLower().Contains(cleanSearch) ||
a.Summary.ToLower().Contains(cleanSearch) ||
a.MatchedAssets.Any(m => m.Name.ToLower().Contains(cleanSearch)));
} }
return await query return await query
.OrderByDescending(a => a.PublishedAt) .OrderByDescending(a => a.PublishedAt)
.ThenByDescending(a => a.ScrapedAt)
.ThenByDescending(a => a.Id)
.Skip(offset) .Skip(offset)
.Take(limit) .Take(limit)
.ToListAsync(); .ToListAsync();
@@ -1,18 +1,22 @@
using System;
using System.Collections.Concurrent; using System.Collections.Concurrent;
using System.Collections.Generic;
using System.IO; using System.IO;
using System.Linq;
using System.Text.Json; using System.Text.Json;
using System.Text.Json.Serialization;
using System.Text.RegularExpressions; using System.Text.RegularExpressions;
using System.Threading;
using System.Threading.Tasks;
using FinlyticAssets.Models; using FinlyticAssets.Models;
using FinlyticAssets.Util; using FinlyticAssets.Util;
using FinlyticCore.Dtos.News; using FinlyticCore.Dtos.News;
using FinlyticCore.Services;
using FinlyticCore.Util; using FinlyticCore.Util;
using FinlyticNews.Entities; using FinlyticNews.Entities;
using FinlyticNews.Util; using FinlyticNews.Util;
using Microsoft.Extensions.Configuration; using Microsoft.Extensions.Configuration;
using Microsoft.Extensions.DependencyInjection; using Microsoft.Extensions.DependencyInjection;
using Microsoft.Extensions.Hosting; using Microsoft.Extensions.Hosting;
using Microsoft.Extensions.Logging;
namespace FinlyticNews.Services; namespace FinlyticNews.Services;
@@ -22,9 +26,6 @@ namespace FinlyticNews.Services;
/// </summary> /// </summary>
public class NewsScraperBackgroundService : BackgroundService public class NewsScraperBackgroundService : BackgroundService
{ {
/// <summary>
/// Internal wrapper to associate compiled regex patterns with the unmodified AssetIndex record.
/// </summary>
private record CompiledAssetMatcher( private record CompiledAssetMatcher(
AssetIndex Asset, AssetIndex Asset,
string CoreName, string CoreName,
@@ -33,63 +34,63 @@ public class NewsScraperBackgroundService : BackgroundService
); );
private readonly IServiceScopeFactory _scopeFactory; private readonly IServiceScopeFactory _scopeFactory;
private readonly ILogger<NewsScraperBackgroundService> _logger; private readonly IFinlyticLogger<NewsScraperBackgroundService> _finlyticLogger;
private readonly NewsMqttClient _mqttClient; private readonly NewsMqttClient _mqttClient;
private readonly int _intervalMinutes;
private readonly string _indexPath; private readonly string _indexPath;
// In-Memory Cache for compiled asset matchers to prevent re-reading & re-compiling Regex
private List<CompiledAssetMatcher>? _cachedAssetMatchers; private List<CompiledAssetMatcher>? _cachedAssetMatchers;
private DateTime _lastIndexLoadTime = DateTime.MinValue; private DateTime _lastIndexLoadTime = DateTime.MinValue;
public NewsScraperBackgroundService( public NewsScraperBackgroundService(
IServiceScopeFactory scopeFactory, IServiceScopeFactory scopeFactory,
ILogger<NewsScraperBackgroundService> logger, IFinlyticLogger<NewsScraperBackgroundService> finlyticLogger,
NewsMqttClient mqttClient, NewsMqttClient mqttClient,
IConfiguration configuration) IConfiguration configuration)
{ {
_scopeFactory = scopeFactory; _scopeFactory = scopeFactory;
_logger = logger; _finlyticLogger = finlyticLogger;
_mqttClient = mqttClient; _mqttClient = mqttClient;
_intervalMinutes = configuration.GetValue<int>("ScrapingSettings:IntervalMinutes", 15);
_indexPath = Path.Combine(Volumes.IndexRelativePath, "index.json"); _indexPath = Path.Combine(Volumes.IndexRelativePath, "index.json");
} }
protected override async Task ExecuteAsync(CancellationToken stoppingToken) protected override async Task ExecuteAsync(CancellationToken stoppingToken)
{ {
_logger.LogInformation("[{Channel}] NewsScraperBackgroundService started. Interval: {Minutes} minutes.", "NewsChannel", _intervalMinutes); await _finlyticLogger.LogInfoAsync(SettingKeys.NewsChannel, "[NewsScraperBackgroundService] NewsScraperBackgroundService started.");
while (!stoppingToken.IsCancellationRequested) while (!stoppingToken.IsCancellationRequested)
{ {
try try
{ {
await RunScrapingCycleAsync(stoppingToken); using var scope = _scopeFactory.CreateScope();
var settings = scope.ServiceProvider.GetRequiredService<ISettingsService>();
bool enabled = await settings.GetSettingAsync(SettingKeys.EnableAutoScraping, stoppingToken);
if (enabled)
{
await RunScrapingCycleAsync(stoppingToken);
}
else
{
await _finlyticLogger.LogInfoAsync(SettingKeys.NewsChannel, "[NewsScraperBackgroundService] Auto-scraping is disabled via settings.");
}
} }
catch (Exception ex) when (ex is not OperationCanceledException) catch (Exception ex) when (ex is not OperationCanceledException)
{ {
_logger.LogError(ex, "[{Channel}] An unhandled exception occurred during news scraping cycle.", "NewsChannel"); await _finlyticLogger.LogErrorAsync(SettingKeys.NewsChannel, ex, "[NewsScraperBackgroundService] An unhandled exception occurred during news scraping cycle.");
} }
int intervalMinutes = _intervalMinutes; int intervalMinutes = 15;
try try
{ {
using var scope = _scopeFactory.CreateScope(); using var scope = _scopeFactory.CreateScope();
var settingsDb = scope.ServiceProvider.GetService<ISettingsDbService>(); var settings = scope.ServiceProvider.GetRequiredService<ISettingsService>();
if (settingsDb != null) intervalMinutes = await settings.GetSettingAsync(SettingKeys.ScrapeIntervalMinutes, stoppingToken);
{
var settings = await settingsDb.GetSettingsAsync();
if (settings?.ScrapingIntervalMinutes > 0)
{
intervalMinutes = settings.ScrapingIntervalMinutes;
}
}
} }
catch { /* Ignore settings DB lookup failures */ } catch { }
var jitterSeconds = Random.Shared.Next(0, 300); var jitterSeconds = Random.Shared.Next(0, 60);
var nextRunDelay = TimeSpan.FromMinutes(intervalMinutes) + TimeSpan.FromSeconds(jitterSeconds); var nextRunDelay = TimeSpan.FromMinutes(intervalMinutes) + TimeSpan.FromSeconds(jitterSeconds);
_logger.LogInformation("[{Channel}] Scraping cycle completed. Next cycle in {Delay} (interval: {Minutes}m).", "NewsChannel", nextRunDelay, intervalMinutes); await _finlyticLogger.LogInfoAsync(SettingKeys.NewsChannel, "[NewsScraperBackgroundService] Scraping cycle completed. Next cycle in {Delay} (interval: {Minutes}m).", nextRunDelay, intervalMinutes);
try try
{ {
@@ -101,7 +102,7 @@ public class NewsScraperBackgroundService : BackgroundService
} }
} }
_logger.LogInformation("[{Channel}] NewsScraperBackgroundService stopping.", "NewsChannel"); await _finlyticLogger.LogInfoAsync(SettingKeys.NewsChannel, "[NewsScraperBackgroundService] NewsScraperBackgroundService stopping.");
} }
private async Task RunScrapingCycleAsync(CancellationToken stoppingToken) private async Task RunScrapingCycleAsync(CancellationToken stoppingToken)
@@ -111,28 +112,28 @@ public class NewsScraperBackgroundService : BackgroundService
var discoveryService = scope.ServiceProvider.GetRequiredService<IArticleDiscoveryService>(); var discoveryService = scope.ServiceProvider.GetRequiredService<IArticleDiscoveryService>();
var scraperService = scope.ServiceProvider.GetRequiredService<IPlaywrightScraperService>(); var scraperService = scope.ServiceProvider.GetRequiredService<IPlaywrightScraperService>();
var n8nService = scope.ServiceProvider.GetRequiredService<IN8nService>(); var n8nService = scope.ServiceProvider.GetRequiredService<IN8nService>();
var settings = scope.ServiceProvider.GetRequiredService<ISettingsService>();
var maxArticlesPerFeed = await settings.GetSettingAsync(SettingKeys.MaxArticlesPerFeed, stoppingToken);
// Load pre-compiled asset index matchers for zero-latency pre-filtering
var assetMatchers = await GetOrLoadAssetMatchersAsync(); var assetMatchers = await GetOrLoadAssetMatchersAsync();
_logger.LogInformation("[{Channel}] Loaded {Count} asset index items for text pre-filtering.", "NewsChannel", assetMatchers.Count); await _finlyticLogger.LogInfoAsync(SettingKeys.NewsChannel, "[NewsScraperBackgroundService] Loaded {Count} asset index items for text pre-filtering.", assetMatchers.Count);
// 1. Scraping Retry Phase: query articles in status "Scraping" (failed Playwright runs)
var failedArticles = await dbService.GetArticlesByStatusAsync("Scraping"); var failedArticles = await dbService.GetArticlesByStatusAsync("Scraping");
if (failedArticles.Count > 0) if (failedArticles.Count > 0)
{ {
_logger.LogInformation("[{Channel}] Found {Count} articles in status 'Scraping' that failed to scrape previously. Retrying...", "NewsChannel", failedArticles.Count); await _finlyticLogger.LogInfoAsync(SettingKeys.NewsChannel, "[NewsScraperBackgroundService] Found {Count} articles in status 'Scraping' that failed to scrape previously. Retrying...", failedArticles.Count);
foreach (var article in failedArticles) foreach (var article in failedArticles)
{ {
if (stoppingToken.IsCancellationRequested) break; if (stoppingToken.IsCancellationRequested) return;
await ProcessSingleArticleAsync(article, dbService, scraperService, n8nService, assetMatchers, stoppingToken); await ProcessSingleArticleAsync(article, scraperService, n8nService, dbService, assetMatchers, stoppingToken);
} }
} }
// 2. Link Discovery Phase: query RSS feeds and listing pages
var sources = await dbService.GetSourcesAsync(); var sources = await dbService.GetSourcesAsync();
if (sources.Count == 0) if (sources.Count == 0)
{ {
_logger.LogWarning("[{Channel}] No article sources configured in database. Skipping cycle.", "NewsChannel"); await _finlyticLogger.LogWarningAsync(SettingKeys.NewsChannel, "[NewsScraperBackgroundService] No article sources configured in database. Skipping cycle.");
return; return;
} }
@@ -140,31 +141,30 @@ public class NewsScraperBackgroundService : BackgroundService
{ {
if (stoppingToken.IsCancellationRequested) break; if (stoppingToken.IsCancellationRequested) break;
_logger.LogInformation("[{Channel}] Starting article link discovery for source: {SourceName} ({Url})", "NewsChannel", source.Name, source.Source); await _finlyticLogger.LogInfoAsync(SettingKeys.NewsChannel, "[NewsScraperBackgroundService] Starting article link discovery for source: {SourceName} ({Url})", source.Name, source.Source);
var discoveredArticles = await discoveryService.DiscoverLinksAsync(source.Source, source.Type, stoppingToken); var discoveredArticles = await discoveryService.DiscoverLinksAsync(source.Source, source.Type, stoppingToken);
if (discoveredArticles == null || discoveredArticles.Count == 0) if (discoveredArticles == null || discoveredArticles.Count == 0)
{ {
_logger.LogDebug("No links discovered from source: {SourceName}", source.Name); await _finlyticLogger.LogDebugAsync(SettingKeys.NewsChannel, "[NewsScraperBackgroundService] No links discovered from source: {SourceName}", source.Name);
continue; continue;
} }
_logger.LogInformation("[{Channel}] Discovered {Count} potential article links from {SourceName}.", "NewsChannel", discoveredArticles.Count, source.Name); await _finlyticLogger.LogInfoAsync(SettingKeys.NewsChannel, "[NewsScraperBackgroundService] Discovered {Count} potential article links from {SourceName}.", discoveredArticles.Count, source.Name);
foreach (var discovered in discoveredArticles) var toProcess = discoveredArticles.Take(maxArticlesPerFeed > 0 ? maxArticlesPerFeed : 20);
foreach (var discovered in toProcess)
{ {
if (stoppingToken.IsCancellationRequested) break; if (stoppingToken.IsCancellationRequested) break;
// Idempotency Check & Deduplication if (await dbService.IsUrlDuplicateAsync(discovered.Url))
var isDuplicate = await dbService.IsUrlDuplicateAsync(discovered.Url);
if (isDuplicate)
{ {
_logger.LogDebug("Skipping duplicate article URL: {Url}", discovered.Url); await _finlyticLogger.LogDebugAsync(SettingKeys.NewsChannel, "[NewsScraperBackgroundService] Skipping duplicate article URL: {Url}", discovered.Url);
continue; continue;
} }
// Register Initial Lock State in the database ("Pending") NewsArticleEntity article;
NewsArticleEntity? article;
try try
{ {
article = await dbService.CreatePendingArticleAsync( article = await dbService.CreatePendingArticleAsync(
@@ -173,53 +173,48 @@ public class NewsScraperBackgroundService : BackgroundService
discovered.Title, discovered.Title,
discovered.Summary, discovered.Summary,
discovered.PublishedAt, discovered.PublishedAt,
discovered.Language); discovered.Language
);
} }
catch (Exception ex) catch (Exception ex)
{ {
_logger.LogError(ex, "[{Channel}] Failed to register initial pending state for URL: {Url}. Skipping.", "NewsChannel", discovered.Url); await _finlyticLogger.LogErrorAsync(SettingKeys.NewsChannel, ex, "[NewsScraperBackgroundService] Failed to register initial pending state for URL: {Url}. Skipping.", discovered.Url);
continue; continue;
} }
if (article == null || article.Id == Guid.Empty) if (article.Id == Guid.Empty)
{ {
_logger.LogWarning("[{Channel}] Created pending article has invalid/empty ID for URL: {Url}. Skipping.", "NewsChannel", discovered.Url); await _finlyticLogger.LogWarningAsync(SettingKeys.NewsChannel, "[NewsScraperBackgroundService] Created pending article has invalid/empty ID for URL: {Url}. Skipping.", discovered.Url);
continue; continue;
} }
// Process single article pipeline await ProcessSingleArticleAsync(article, scraperService, n8nService, dbService, assetMatchers, stoppingToken);
await ProcessSingleArticleAsync(article, dbService, scraperService, n8nService, assetMatchers, stoppingToken);
} }
} }
} }
private async Task ProcessSingleArticleAsync( private async Task ProcessSingleArticleAsync(
NewsArticleEntity article, NewsArticleEntity article,
INewsDbService dbService,
IPlaywrightScraperService scraperService, IPlaywrightScraperService scraperService,
IN8nService n8nService, IN8nService n8nService,
INewsDbService dbService,
List<CompiledAssetMatcher> assetMatchers, List<CompiledAssetMatcher> assetMatchers,
CancellationToken stoppingToken) CancellationToken stoppingToken)
{ {
try try
{ {
// 3. Extraction with Headless Browser (Transitions to "Processing")
await dbService.UpdateArticleStatusAsync(article.Id, "Processing"); await dbService.UpdateArticleStatusAsync(article.Id, "Processing");
var (resolvedUrl, rawText) = await scraperService.ScrapeArticleAsync(article.SourceUrl);
if (string.IsNullOrWhiteSpace(rawText)) var (resolvedUrl, rawContent) = await scraperService.ScrapeArticleAsync(article.SourceUrl);
{
throw new InvalidOperationException("Scraping returned empty text body content.");
}
// Update resolved URL if redirect occurred if (!string.IsNullOrWhiteSpace(resolvedUrl) &&
if (!string.Equals(resolvedUrl, article.SourceUrl, StringComparison.OrdinalIgnoreCase)) !resolvedUrl.Equals(article.SourceUrl, StringComparison.OrdinalIgnoreCase))
{ {
_logger.LogInformation("[{Channel}] Redirect detected. Initial: {OldUrl} -> Resolved: {NewUrl}", "NewsChannel", article.SourceUrl, resolvedUrl); await _finlyticLogger.LogInfoAsync(SettingKeys.NewsChannel, "[NewsScraperBackgroundService] Redirect detected. Initial: {OldUrl} -> Resolved: {NewUrl}", article.SourceUrl, resolvedUrl);
if (await dbService.IsUrlDuplicateAsync(resolvedUrl)) if (await dbService.IsUrlDuplicateAsync(resolvedUrl))
{ {
_logger.LogInformation("[{Channel}] Redirected URL {ResolvedUrl} is a duplicate. Terminating processing.", "NewsChannel", resolvedUrl); await _finlyticLogger.LogInfoAsync(SettingKeys.NewsChannel, "[NewsScraperBackgroundService] Redirected URL {ResolvedUrl} is a duplicate. Terminating processing.", resolvedUrl);
await dbService.UpdateArticleStatusAsync(article.Id, "Duplicate"); await dbService.UpdateArticleStatusAsync(article.Id, "Failed");
return; return;
} }
@@ -227,108 +222,85 @@ public class NewsScraperBackgroundService : BackgroundService
article.SourceUrl = resolvedUrl; article.SourceUrl = resolvedUrl;
} }
// 4. Pre-filtering Assets (Optimized with Pre-Compiled Regex Patterns) if (string.IsNullOrWhiteSpace(rawContent) || rawContent.Length < 60)
var title = article.Title ?? string.Empty;
var summary = article.Summary ?? string.Empty;
var preFilteredAssets = assetMatchers.Where(matcher =>
{ {
var asset = matcher.Asset;
// ISIN direct match
if (rawText.Contains(asset.Isin, StringComparison.OrdinalIgnoreCase) ||
title.Contains(asset.Isin, StringComparison.OrdinalIgnoreCase) ||
summary.Contains(asset.Isin, StringComparison.OrdinalIgnoreCase))
{
return true;
}
// Fast regex word boundary check on Full Name
if (matcher.WordRegex != null && (matcher.WordRegex.IsMatch(rawText) || matcher.WordRegex.IsMatch(title)))
{
return true;
}
// Fast regex word boundary check on Core Name
if (matcher.CoreName.Length >= 3 && matcher.CoreWordRegex != null &&
(matcher.CoreWordRegex.IsMatch(rawText) || matcher.CoreWordRegex.IsMatch(title)))
{
return true;
}
return false;
})
.Select(m => new FilteredAssetPayload(m.Asset.Name, m.Asset.Isin))
.ToList();
if (preFilteredAssets.Count == 0)
{
_logger.LogInformation("[{Channel}] Pre-filtering: Article {Id} does not reference any known assets. Terminating pipeline.", "NewsChannel", article.Id);
await dbService.UpdateArticleStatusAsync(article.Id, "Failed"); await dbService.UpdateArticleStatusAsync(article.Id, "Failed");
return; return;
} }
_logger.LogInformation("[{Channel}] Pre-filtering matched {Count} assets for article {Id}.", "NewsChannel", preFilteredAssets.Count, article.Id); var discoveredIsins = article.MatchedAssets.Select(m => m.Isin).Where(i => !string.IsNullOrEmpty(i)).ToList();
var preFilteredAssets = PreFilterAssets(rawContent, article.Title, assetMatchers, discoveredIsins);
// 5. Send to n8n Webhook Pipeline if (preFilteredAssets.Count == 0)
var n8nResponse = await n8nService.AnalyzeArticleAsync(rawText, preFilteredAssets, stoppingToken);
if (n8nResponse == null)
{ {
throw new InvalidOperationException("n8n AI webhook execution returned null or failed."); await _finlyticLogger.LogInfoAsync(SettingKeys.NewsChannel, "[NewsScraperBackgroundService] Pre-filtering: Article {Id} does not reference any known assets. Terminating pipeline.", article.Id);
await dbService.UpdateArticleStatusAsync(article.Id, "Failed");
return;
} }
// 6. Map and Save Completed Classification await _finlyticLogger.LogInfoAsync(SettingKeys.NewsChannel, "[NewsScraperBackgroundService] Pre-filtering matched {Count} assets for article {Id}.", preFilteredAssets.Count, article.Id);
var matchedEntities = new List<MatchedAssetEntity>();
foreach (var n8nAsset in n8nResponse.MatchedAssets) var n8nResponse = await n8nService.AnalyzeArticleAsync(rawContent, preFilteredAssets, stoppingToken);
if (n8nResponse == null)
{ {
var n8nCoreName = ExtractCoreAssetName(n8nAsset.Name); await dbService.UpdateArticleStatusAsync(article.Id, "Failed");
return;
}
var matchedIsin = preFilteredAssets.FirstOrDefault(fa => var matchedEntities = new List<MatchedAssetEntity>();
fa.Name.Equals(n8nAsset.Name, StringComparison.OrdinalIgnoreCase) || if (n8nResponse.MatchedAssets != null && n8nResponse.MatchedAssets.Count > 0)
n8nAsset.Name.Contains(fa.Name, StringComparison.OrdinalIgnoreCase) || {
(n8nCoreName.Length >= 3 && ExtractCoreAssetName(fa.Name).Equals(n8nCoreName, StringComparison.OrdinalIgnoreCase)))?.Isin; foreach (var asset in n8nResponse.MatchedAssets)
if (string.IsNullOrWhiteSpace(matchedIsin))
{ {
matchedIsin = assetMatchers.FirstOrDefault(m => var preMatch = preFilteredAssets.FirstOrDefault(p => p.Name.Equals(asset.Name, StringComparison.OrdinalIgnoreCase));
m.Asset.Name.Equals(n8nAsset.Name, StringComparison.OrdinalIgnoreCase) || var isin = preMatch?.Isin ?? asset.Ticker ?? "";
n8nAsset.Name.Contains(m.Asset.Name, StringComparison.OrdinalIgnoreCase) || if (string.IsNullOrWhiteSpace(isin)) continue;
(n8nCoreName.Length >= 3 && m.CoreName.Equals(n8nCoreName, StringComparison.OrdinalIgnoreCase)))?.Asset.Isin;
}
if (!string.IsNullOrWhiteSpace(matchedIsin))
{
matchedEntities.Add(new MatchedAssetEntity matchedEntities.Add(new MatchedAssetEntity
{ {
Id = Guid.NewGuid(),
NewsArticleId = article.Id, NewsArticleId = article.Id,
Name = n8nAsset.Name, Isin = isin.Trim().ToUpperInvariant(),
Isin = matchedIsin Name = !string.IsNullOrWhiteSpace(asset.Name) ? asset.Name.Trim() : isin.Trim().ToUpperInvariant()
}); });
} }
} }
var completedArticle = await dbService.SaveArticleClassificationAsync(article.Id, n8nResponse, matchedEntities); if (matchedEntities.Count == 0)
{
if (completedArticle != null) foreach (var preMatch in preFilteredAssets)
{
matchedEntities.Add(new MatchedAssetEntity
{
Id = Guid.NewGuid(),
NewsArticleId = article.Id,
Isin = preMatch.Isin,
Name = preMatch.Name
});
}
}
var updatedArticle = await dbService.SaveArticleClassificationAsync(article.Id, n8nResponse, matchedEntities);
if (updatedArticle != null)
{ {
// 7. MQTT Broadcast (Sends completed article to downstream services)
var dto = new NewsArticleDto var dto = new NewsArticleDto
{ {
Id = completedArticle.Id, Id = updatedArticle.Id,
Title = completedArticle.Title, Title = updatedArticle.Title,
Author = completedArticle.Author, Author = updatedArticle.Author,
Summary = completedArticle.Summary, Summary = updatedArticle.Summary,
ContentRaw = completedArticle.ContentRaw, ContentRaw = updatedArticle.ContentRaw,
Language = completedArticle.Language, Language = updatedArticle.Language,
SourceUrl = completedArticle.SourceUrl, SourceUrl = updatedArticle.SourceUrl,
ScrapedAt = completedArticle.ScrapedAt, ScrapedAt = updatedArticle.ScrapedAt,
PublishedAt = completedArticle.PublishedAt, PublishedAt = updatedArticle.PublishedAt,
MatchedAssets = completedArticle.MatchedAssets.Select(m => new MatchedAssetDto Status = updatedArticle.Status,
MatchedAssets = updatedArticle.MatchedAssets.Select(m => new MatchedAssetDto
{ {
Name = m.Name, Name = m.Name,
Isin = m.Isin Isin = m.Isin
}).ToList(), }).ToList()
Status = completedArticle.Status
}; };
await _mqttClient.BroadcastArticleAsync(dto); await _mqttClient.BroadcastArticleAsync(dto);
@@ -336,103 +308,120 @@ public class NewsScraperBackgroundService : BackgroundService
} }
catch (Exception ex) catch (Exception ex)
{ {
_logger.LogError(ex, "[{Channel}] Failed to complete processing pipeline for article: {Url}. Transitioning to 'Scraping' for next cycle retry.", "NewsChannel", article.SourceUrl); await _finlyticLogger.LogErrorAsync(SettingKeys.NewsChannel, ex, "[NewsScraperBackgroundService] Failed to complete processing pipeline for article: {Url}. Transitioning to 'Scraping' for next cycle retry.", article.SourceUrl);
try await dbService.UpdateArticleStatusAsync(article.Id, "Scraping");
{
await dbService.UpdateArticleStatusAsync(article.Id, "Scraping");
}
catch { /* Suppress database secondary errors */ }
} }
} }
/// <summary> private List<FilteredAssetPayload> PreFilterAssets(
/// Returns cached compiled asset matchers or parses the index file from disk if stale/missing. string content,
/// </summary> string? title,
List<CompiledAssetMatcher> assetMatchers,
List<string>? priorityIsins = null)
{
if (assetMatchers.Count == 0) return new List<FilteredAssetPayload>();
var fullText = (title != null ? title + " " + content : content);
var matched = new Dictionary<string, FilteredAssetPayload>(StringComparer.OrdinalIgnoreCase);
if (priorityIsins != null && priorityIsins.Count > 0)
{
foreach (var isin in priorityIsins)
{
var match = assetMatchers.FirstOrDefault(m => string.Equals(m.Asset.Isin, isin, StringComparison.OrdinalIgnoreCase));
if (match != null && !matched.ContainsKey(match.Asset.Isin))
{
matched[match.Asset.Isin] = new FilteredAssetPayload(match.Asset.Name, match.Asset.Isin);
}
}
}
foreach (var m in assetMatchers)
{
if (matched.ContainsKey(m.Asset.Isin)) continue;
if (fullText.Contains(m.Asset.Isin, StringComparison.OrdinalIgnoreCase))
{
matched[m.Asset.Isin] = new FilteredAssetPayload(m.Asset.Name, m.Asset.Isin);
continue;
}
if (m.WordRegex != null && m.WordRegex.IsMatch(fullText))
{
matched[m.Asset.Isin] = new FilteredAssetPayload(m.Asset.Name, m.Asset.Isin);
continue;
}
if (m.CoreWordRegex != null && m.CoreWordRegex.IsMatch(fullText))
{
matched[m.Asset.Isin] = new FilteredAssetPayload(m.Asset.Name, m.Asset.Isin);
}
}
return matched.Values.ToList();
}
private async Task<List<CompiledAssetMatcher>> GetOrLoadAssetMatchersAsync() private async Task<List<CompiledAssetMatcher>> GetOrLoadAssetMatchersAsync()
{ {
if (_cachedAssetMatchers != null && (DateTime.UtcNow - _lastIndexLoadTime).TotalMinutes < 30) if (_cachedAssetMatchers != null && (DateTime.UtcNow - _lastIndexLoadTime).TotalMinutes < 60)
{ {
return _cachedAssetMatchers; return _cachedAssetMatchers;
} }
if (!File.Exists(_indexPath)) if (!File.Exists(_indexPath))
{ {
_logger.LogWarning("[{Channel}] Asset index file not found at: {Path}. Pre-filtering will match 0 assets.", "NewsChannel", _indexPath); await _finlyticLogger.LogWarningAsync(SettingKeys.NewsChannel, "[NewsScraperBackgroundService] Asset index file not found at: {Path}. Pre-filtering will match 0 assets.", _indexPath);
return []; return new List<CompiledAssetMatcher>();
} }
try try
{ {
await using var stream = File.OpenRead(_indexPath); using var stream = File.OpenRead(_indexPath);
var indexList = await JsonSerializer.DeserializeAsync<List<AssetIndex>>(stream);
// Standard Deserialization for AssetIndex list if (indexList == null || indexList.Count == 0)
var rawList = await JsonSerializer.DeserializeAsync<List<AssetIndex>>(stream);
if (rawList != null)
{ {
_cachedAssetMatchers = rawList.Select(asset => return new List<CompiledAssetMatcher>();
{
var coreName = ExtractCoreAssetName(asset.Name);
return new CompiledAssetMatcher(
Asset: asset,
CoreName: coreName,
WordRegex: BuildWordRegex(asset.Name),
CoreWordRegex: BuildWordRegex(coreName)
);
}).ToList();
_lastIndexLoadTime = DateTime.UtcNow;
return _cachedAssetMatchers;
} }
var compiled = new List<CompiledAssetMatcher>(indexList.Count);
foreach (var asset in indexList)
{
if (string.IsNullOrWhiteSpace(asset.Name) || string.IsNullOrWhiteSpace(asset.Isin))
continue;
var rawName = asset.Name.Trim();
var coreName = ExtractCoreName(rawName);
Regex? wordRegex = null;
if (rawName.Length >= 4)
{
wordRegex = new Regex($@"\b{Regex.Escape(rawName)}\b", RegexOptions.IgnoreCase | RegexOptions.CultureInvariant | RegexOptions.Compiled);
}
Regex? coreWordRegex = null;
if (!string.IsNullOrWhiteSpace(coreName) && coreName.Length >= 4 && !coreName.Equals(rawName, StringComparison.OrdinalIgnoreCase))
{
coreWordRegex = new Regex($@"\b{Regex.Escape(coreName)}\b", RegexOptions.IgnoreCase | RegexOptions.CultureInvariant | RegexOptions.Compiled);
}
compiled.Add(new CompiledAssetMatcher(asset, coreName, wordRegex, coreWordRegex));
}
_cachedAssetMatchers = compiled;
_lastIndexLoadTime = DateTime.UtcNow;
return _cachedAssetMatchers;
} }
catch (Exception ex) catch (Exception ex)
{ {
_logger.LogError(ex, "[{Channel}] Failed to read or parse asset index file from {Path}.", "NewsChannel", _indexPath); await _finlyticLogger.LogErrorAsync(SettingKeys.NewsChannel, ex, "[NewsScraperBackgroundService] Failed to read or parse asset index file from {Path}.", _indexPath);
} return _cachedAssetMatchers ?? new List<CompiledAssetMatcher>();
return _cachedAssetMatchers ?? [];
}
/// <summary>
/// Helper to pre-compile Word Boundary Regex for an asset name.
/// </summary>
private static Regex? BuildWordRegex(string name)
{
if (string.IsNullOrWhiteSpace(name)) return null;
try
{
return new Regex($@"\b{Regex.Escape(name)}\b", RegexOptions.IgnoreCase | RegexOptions.CultureInvariant | RegexOptions.Compiled);
}
catch
{
return null;
} }
} }
/// <summary> private static string ExtractCoreName(string rawName)
/// Extracts the core name of an asset by removing parenthetical metadata and corporate suffixes.
/// </summary>
private static string ExtractCoreAssetName(string name)
{ {
if (string.IsNullOrWhiteSpace(name)) return string.Empty; var cleaned = Regex.Replace(rawName, @"\b(AG|SE|SA|NV|PLC|INC|CORP|LLC|GMBH|CO|KG|HOLDING|GROUP|CLASS\s+[A-Z])\b", "", RegexOptions.IgnoreCase);
return cleaned.Trim(' ', '.', ',', '-');
int parenIndex = name.IndexOf('(');
if (parenIndex >= 0)
{
name = name[..parenIndex];
}
name = name.Trim();
var suffixes = new[] { "Inc.", "Inc", "AG", "SE", "Co.", "Co", "Corp.", "Corp", "Ltd.", "Ltd", "plc", "GmbH", "SA", "NV", "Group" };
foreach (var suffix in suffixes)
{
if (name.EndsWith(" " + suffix, StringComparison.OrdinalIgnoreCase))
{
name = name[..^suffix.Length].Trim();
}
}
return name;
} }
} }
@@ -3,8 +3,9 @@ using System.Collections.Generic;
using System.Linq; using System.Linq;
using System.Threading; using System.Threading;
using System.Threading.Tasks; using System.Threading.Tasks;
using FinlyticCore.Services;
using FinlyticNews.Adapters.Scraping; using FinlyticNews.Adapters.Scraping;
using Microsoft.Extensions.Logging; using FinlyticNews.Util;
using Microsoft.Playwright; using Microsoft.Playwright;
namespace FinlyticNews.Services; namespace FinlyticNews.Services;
@@ -25,7 +26,7 @@ public interface IPlaywrightScraperService
/// <inheritdoc /> /// <inheritdoc />
public class PlaywrightScraperService : IPlaywrightScraperService, IAsyncDisposable public class PlaywrightScraperService : IPlaywrightScraperService, IAsyncDisposable
{ {
private readonly ILogger<PlaywrightScraperService> _logger; private readonly IFinlyticLogger<PlaywrightScraperService> _finlyticLogger;
private readonly IEnumerable<ArticleScraperAdapter> _scraperAdapters; private readonly IEnumerable<ArticleScraperAdapter> _scraperAdapters;
private IPlaywright? _playwright; private IPlaywright? _playwright;
@@ -36,21 +37,20 @@ public class PlaywrightScraperService : IPlaywrightScraperService, IAsyncDisposa
/// Initializes a new instance of the <see cref="PlaywrightScraperService"/> class. /// Initializes a new instance of the <see cref="PlaywrightScraperService"/> class.
/// </summary> /// </summary>
public PlaywrightScraperService( public PlaywrightScraperService(
ILogger<PlaywrightScraperService> logger, IFinlyticLogger<PlaywrightScraperService> finlyticLogger,
IEnumerable<ArticleScraperAdapter> scraperAdapters) IEnumerable<ArticleScraperAdapter> scraperAdapters)
{ {
_logger = logger; _finlyticLogger = finlyticLogger;
_scraperAdapters = scraperAdapters; _scraperAdapters = scraperAdapters;
} }
/// <inheritdoc /> /// <inheritdoc />
public async Task<(string ResolvedUrl, string Content)> ScrapeArticleAsync(string url) public async Task<(string ResolvedUrl, string Content)> ScrapeArticleAsync(string url)
{ {
_logger.LogInformation("[{Channel}] Launching browser context to scrape article: {Url}", "NewsChannel", url); await _finlyticLogger.LogInfoAsync(SettingKeys.NewsChannel, "[PlaywrightScraperService] Launching browser context to scrape article: {Url}", url);
var browser = await GetOrInitBrowserAsync(); var browser = await GetOrInitBrowserAsync();
// Fast, isolated browser context (incognito tab environment) per article
await using var context = await browser.NewContextAsync(new BrowserNewContextOptions await using var context = await browser.NewContextAsync(new BrowserNewContextOptions
{ {
UserAgent = "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36", UserAgent = "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/120.0.0.0 Safari/537.36",
@@ -61,7 +61,6 @@ public class PlaywrightScraperService : IPlaywrightScraperService, IAsyncDisposa
try try
{ {
// 1. Initial Navigation with DOMContentLoaded wait
var response = await page.GotoAsync(url, new PageGotoOptions var response = await page.GotoAsync(url, new PageGotoOptions
{ {
WaitUntil = WaitUntilState.DOMContentLoaded, WaitUntil = WaitUntilState.DOMContentLoaded,
@@ -74,115 +73,86 @@ public class PlaywrightScraperService : IPlaywrightScraperService, IAsyncDisposa
} }
var finalUrl = page.Url; var finalUrl = page.Url;
_logger.LogDebug("[{Channel}] Navigation completed. Initial final URL: {Url}", "NewsChannel", finalUrl); await _finlyticLogger.LogDebugAsync(SettingKeys.NewsChannel, "[PlaywrightScraperService] Navigation completed. Initial final URL: {Url}", finalUrl);
// 2. Resolve Host Specific Scraper Adapter var host = new Uri(finalUrl).Host;
var uri = new Uri(url); var adapter = _scraperAdapters.FirstOrDefault(a =>
var host = uri.Host; host.EndsWith(a.Hostname, StringComparison.OrdinalIgnoreCase) ||
var adapter = _scraperAdapters.FirstOrDefault(a => host.Contains(a.Hostname, StringComparison.OrdinalIgnoreCase)); a.Hostname.EndsWith(host, StringComparison.OrdinalIgnoreCase));
IPage targetPage = page;
if (adapter != null) if (adapter != null)
{ {
_logger.LogInformation("[{Channel}] Executing adapter redirect check for host: {Host}", "NewsChannel", adapter.Hostname); await _finlyticLogger.LogInfoAsync(SettingKeys.NewsChannel, "[PlaywrightScraperService] Executing adapter redirect check for host: {Host}", adapter.Hostname);
try try
{ {
var resolvedRedirectUrl = await adapter.TryResolveRedirectUrlAsync(page); var resolvedRedirectUrl = await adapter.TryResolveRedirectUrlAsync(page);
// Loop Protection: Navigate only if redirect target is a new URL
if (!string.IsNullOrWhiteSpace(resolvedRedirectUrl) && if (!string.IsNullOrWhiteSpace(resolvedRedirectUrl) &&
!string.Equals(page.Url, resolvedRedirectUrl, StringComparison.OrdinalIgnoreCase)) !resolvedRedirectUrl.Equals(finalUrl, StringComparison.OrdinalIgnoreCase))
{ {
_logger.LogInformation("[{Channel}] Redirect resolved to target URL: {Url}", "NewsChannel", resolvedRedirectUrl); await _finlyticLogger.LogInfoAsync(SettingKeys.NewsChannel, "[PlaywrightScraperService] Redirect resolved to target URL: {Url}", resolvedRedirectUrl);
var refererUrl = page.Url;
finalUrl = resolvedRedirectUrl;
var redirectResponse = await page.GotoAsync(resolvedRedirectUrl, new PageGotoOptions var redirectResponse = await page.GotoAsync(resolvedRedirectUrl, new PageGotoOptions
{ {
WaitUntil = WaitUntilState.DOMContentLoaded, WaitUntil = WaitUntilState.DOMContentLoaded,
Timeout = 30000, Timeout = 30000
Referer = refererUrl
}); });
if (redirectResponse == null) if (redirectResponse == null)
{ {
_logger.LogWarning("[{Channel}] Failed to load response for redirect URL: {Url}", "NewsChannel", resolvedRedirectUrl); await _finlyticLogger.LogWarningAsync(SettingKeys.NewsChannel, "[PlaywrightScraperService] Failed to load response for redirect URL: {Url}", resolvedRedirectUrl);
}
else
{
finalUrl = page.Url;
} }
// Check if redirect opened a new tab/popup finalUrl = page.Url;
var matchedPage = page.Context.Pages.FirstOrDefault(p => p.Url == resolvedRedirectUrl); host = new Uri(finalUrl).Host;
if (matchedPage != null)
{ adapter = _scraperAdapters.FirstOrDefault(a =>
targetPage = matchedPage; host.EndsWith(a.Hostname, StringComparison.OrdinalIgnoreCase) ||
} a.Hostname.EndsWith(host, StringComparison.OrdinalIgnoreCase));
} }
} }
catch (Exception ex) catch (Exception ex)
{ {
_logger.LogWarning(ex, "[{Channel}] Failed to resolve redirect through adapter for host: {Host}. Continuing with current page.", "NewsChannel", adapter.Hostname); await _finlyticLogger.LogWarningAsync(SettingKeys.NewsChannel, ex, "[PlaywrightScraperService] Failed to resolve redirect through adapter for host: {Host}. Continuing with current page.", adapter.Hostname);
} }
} }
// 3. Re-resolve detail page adapter for final target page string content;
var targetUri = new Uri(targetPage.Url); if (adapter != null)
var targetHost = targetUri.Host;
var targetAdapter = _scraperAdapters.FirstOrDefault(a => targetHost.Contains(a.Hostname, StringComparison.OrdinalIgnoreCase));
// Wait a brief moment for dynamic scripts / DOM settling
await targetPage.WaitForTimeoutAsync(1000);
// 4. Extract Content via Mozilla Readability (or Adapter Fallback)
string extractedText = string.Empty;
if (targetAdapter != null)
{ {
var readabilityResult = await targetAdapter.ExtractArticleContentAsync(targetPage); var result = await adapter.ExtractArticleContentAsync(page);
if (readabilityResult != null && !string.IsNullOrWhiteSpace(readabilityResult.TextContent)) content = result?.TextContent ?? await FallbackExtractContentAsync(page);
{
extractedText = readabilityResult.TextContent;
}
} }
else
// Standard Fallback: Body Text / Selector Extraction
if (string.IsNullOrWhiteSpace(extractedText))
{ {
var bodySelector = targetAdapter?.ArticleBodySelector ?? "body"; content = await FallbackExtractContentAsync(page);
var locator = targetPage.Locator(bodySelector);
if (await locator.CountAsync() > 0)
{
extractedText = await locator.First.InnerTextAsync();
}
if (string.IsNullOrWhiteSpace(extractedText))
{
extractedText = await targetPage.EvaluateAsync<string>(
$"() => document.querySelector('{bodySelector}')?.innerText ?? ''");
}
} }
return (finalUrl, extractedText?.Trim() ?? string.Empty); return (finalUrl, content);
} }
catch (Exception ex) catch (Exception ex)
{ {
_logger.LogError(ex, "[{Channel}] Failed to scrape page content from URL: {Url}", "NewsChannel", url); await _finlyticLogger.LogErrorAsync(SettingKeys.NewsChannel, ex, "[PlaywrightScraperService] Failed to scrape page content from URL: {Url}", url);
throw; throw;
} }
finally finally
{ {
await context.CloseAsync(); await page.CloseAsync();
} }
} }
/// <summary> private async Task<string> FallbackExtractContentAsync(IPage page)
/// Thread-safe singleton initialization of the Chromium browser instance. {
/// </summary> var innerText = await page.EvaluateAsync<string>(@"() => {
const scripts = document.querySelectorAll('script, style, noscript, nav, header, footer, iframe, svg');
scripts.forEach(s => s.remove());
const main = document.querySelector('article, main, .article-content, #content, .story-body') || document.body;
return main ? main.innerText : document.body.innerText;
}");
return innerText?.Trim() ?? string.Empty;
}
private async Task<IBrowser> GetOrInitBrowserAsync() private async Task<IBrowser> GetOrInitBrowserAsync()
{ {
if (_browser != null && _browser.IsConnected) if (_browser != null && _browser.IsConnected)
@@ -202,10 +172,16 @@ public class PlaywrightScraperService : IPlaywrightScraperService, IAsyncDisposa
_browser = await _playwright.Chromium.LaunchAsync(new BrowserTypeLaunchOptions _browser = await _playwright.Chromium.LaunchAsync(new BrowserTypeLaunchOptions
{ {
Headless = true, Headless = true,
Args = ["--no-sandbox", "--disable-setuid-sandbox", "--disable-dev-shm-usage"] Args = new[]
{
"--no-sandbox",
"--disable-setuid-sandbox",
"--disable-dev-shm-usage",
"--disable-gpu"
}
}); });
_logger.LogInformation("[{Channel}] Initialized shared Chromium browser instance.", "NewsChannel"); await _finlyticLogger.LogInfoAsync(SettingKeys.NewsChannel, "[PlaywrightScraperService] Initialized shared Chromium browser instance.");
return _browser; return _browser;
} }
finally finally
@@ -214,9 +190,6 @@ public class PlaywrightScraperService : IPlaywrightScraperService, IAsyncDisposa
} }
} }
/// <summary>
/// Disposes the Playwright and Browser instances cleanly during service shutdown.
/// </summary>
public async ValueTask DisposeAsync() public async ValueTask DisposeAsync()
{ {
if (_browser != null) if (_browser != null)
+122 -90
View File
@@ -1,10 +1,18 @@
using System;
using System.Collections.Generic;
using System.IO; using System.IO;
using System.Linq;
using System.Text.Json; using System.Text.Json;
using System.Threading;
using System.Threading.Tasks;
using FinlyticCore.Dtos; using FinlyticCore.Dtos;
using FinlyticCore.Dtos.News; using FinlyticCore.Dtos.News;
using FinlyticCore.Dtos.Sentiment; using FinlyticCore.Dtos.Sentiment;
using FinlyticCore.Dtos.Settings;
using FinlyticCore.Models; using FinlyticCore.Models;
using FinlyticCore.Services;
using FinlyticCore.Util; using FinlyticCore.Util;
using FinlyticNews.Database;
using FinlyticNews.Entities; using FinlyticNews.Entities;
using FinlyticNews.Services; using FinlyticNews.Services;
using Microsoft.Extensions.Configuration; using Microsoft.Extensions.Configuration;
@@ -40,12 +48,10 @@ public class NewsMqttClient : ManagedMqttClient, IHostedService
{ {
Host = _configuration["MQTT:Host"] ?? _configuration["MQTT__Host"] ?? "localhost", Host = _configuration["MQTT:Host"] ?? _configuration["MQTT__Host"] ?? "localhost",
Port = Convert.ToInt32(_configuration["MQTT:Port"] ?? _configuration["MQTT__Port"] ?? "1883"), Port = Convert.ToInt32(_configuration["MQTT:Port"] ?? _configuration["MQTT__Port"] ?? "1883"),
ClientId = ClientId = $"{(_configuration["MQTT:ClientId"] ?? _configuration["MQTT__ClientId"] ?? "FinlyticNews")}_{Guid.NewGuid()}"
$"{(_configuration["MQTT:ClientId"] ?? _configuration["MQTT__ClientId"] ?? "FinlyticNews")}_{Guid.NewGuid()}"
}; };
_logger.LogInformation("Starting News MQTT client. Host: {Host}, ClientId: {ClientId}", config.Host, _logger.LogInformation("Starting News MQTT client. Host: {Host}, ClientId: {ClientId}", config.Host, config.ClientId);
config.ClientId);
await ConnectAsync(config); await ConnectAsync(config);
} }
@@ -61,13 +67,22 @@ public class NewsMqttClient : ManagedMqttClient, IHostedService
{ {
_logger.LogInformation("News MQTT client connected. Subscribing to RPC topics..."); _logger.LogInformation("News MQTT client connected. Subscribing to RPC topics...");
// ZUSAMMENGELEGT: Unified News Fetching (news_Get deckt news_GetDaily mit ab)
await SubscribeAsync("services/request/news_Get/#"); await SubscribeAsync("services/request/news_Get/#");
await SubscribeAsync("services/request/news_GetById/#"); await SubscribeAsync("services/request/news_GetById/#");
await SubscribeAsync("services/request/news_GetPending/#"); await SubscribeAsync("services/request/news_GetPending/#");
await SubscribeAsync("services/request/news_UpdateStatus/#"); await SubscribeAsync("services/request/news_UpdateStatus/#");
await SubscribeAsync("services/request/news_settings_GetAll/#");
await SubscribeAsync("services/request/news_settings_Update/#");
await SubscribeAsync("services/request/health_Ping/#"); await SubscribeAsync("services/request/health_Ping/#");
await SubscribeAsync("services/config/updated/#"); await SubscribeAsync("services/config/updated/#");
FinlyticCore.Services.FinlyticLogBroadcaster.OnLogPublished = async (logDto) =>
{
if (IsConnected && string.Equals(logDto.ServiceName, "FinlyticNews", StringComparison.OrdinalIgnoreCase))
{
await PublishAsync("finlytic/logs/FinlyticNews", logDto);
}
};
} }
/// <summary> /// <summary>
@@ -75,12 +90,10 @@ public class NewsMqttClient : ManagedMqttClient, IHostedService
/// </summary> /// </summary>
public async Task BroadcastArticleAsync(NewsArticleDto article) public async Task BroadcastArticleAsync(NewsArticleDto article)
{ {
// 1. Primärer System-Broadcast
const string topic = "services/news/completed"; const string topic = "services/news/completed";
_logger.LogInformation("Broadcasting completed article to MQTT topic: {Topic}. ID: {Id}", topic, article.Id); _logger.LogInformation("Broadcasting completed article to MQTT topic: {Topic}. ID: {Id}", topic, article.Id);
await PublishAsync(topic, article); await PublishAsync(topic, article);
// 2. Zielgerichteter ISIN-Stream für Echtzeit-Frontend-Feeds
var firstIsin = article.MatchedAssets.FirstOrDefault()?.Isin; var firstIsin = article.MatchedAssets.FirstOrDefault()?.Isin;
if (!string.IsNullOrWhiteSpace(firstIsin)) if (!string.IsNullOrWhiteSpace(firstIsin))
{ {
@@ -94,14 +107,12 @@ public class NewsMqttClient : ManagedMqttClient, IHostedService
{ {
if (string.IsNullOrWhiteSpace(topic)) return; if (string.IsNullOrWhiteSpace(topic)) return;
// 1. System Config Updates
if (topic.StartsWith("services/config/updated", StringComparison.OrdinalIgnoreCase)) if (topic.StartsWith("services/config/updated", StringComparison.OrdinalIgnoreCase))
{ {
if (topic.EndsWith("FinlyticNews", StringComparison.OrdinalIgnoreCase)) if (topic.EndsWith("FinlyticNews", StringComparison.OrdinalIgnoreCase))
{ {
await OnConfigUpdatedAsync(payload); await OnConfigUpdatedAsync(payload);
} }
return; return;
} }
@@ -111,7 +122,6 @@ public class NewsMqttClient : ManagedMqttClient, IHostedService
var channel = segments[2]; var channel = segments[2];
var correlationId = segments[^1]; var correlationId = segments[^1];
// 2. Unified RPC Dispatching via Switch
switch (channel) switch (channel)
{ {
case "news_Get": case "news_Get":
@@ -130,6 +140,14 @@ public class NewsMqttClient : ManagedMqttClient, IHostedService
await OnUpdateNewsStatusAsync(payload, correlationId); await OnUpdateNewsStatusAsync(payload, correlationId);
break; break;
case "news_settings_GetAll":
await OnSettingsGetAllAsync(correlationId);
break;
case "news_settings_Update":
await OnSettingsUpdateAsync(payload, correlationId);
break;
case "health_Ping": case "health_Ping":
await OnHealthPingAsync(segments, correlationId); await OnHealthPingAsync(segments, correlationId);
break; break;
@@ -142,7 +160,10 @@ public class NewsMqttClient : ManagedMqttClient, IHostedService
private async Task OnGetNewsAsync(string payload, string correlationId) private async Task OnGetNewsAsync(string payload, string correlationId)
{ {
_logger.LogInformation("Received RPC news_Get request. Correlation: {CorrelationId}", correlationId); using var scope = _scopeFactory.CreateScope();
var finlyticLogger = scope.ServiceProvider.GetRequiredService<IFinlyticLogger<NewsMqttClient>>();
await finlyticLogger.LogInfoAsync(SettingKeys.NewsChannel, "Received RPC news_Get request. Correlation: {CorrelationId}", correlationId);
int limit = 20; int limit = 20;
int offset = 0; int offset = 0;
@@ -155,9 +176,7 @@ public class NewsMqttClient : ManagedMqttClient, IHostedService
{ {
try try
{ {
// Direktes Deserialisieren über das DailyNewsRequest-DTO
var req = JsonSerializer.Deserialize(payload, FinlyticJsonSerializerContext.Default.DailyNewsRequest); var req = JsonSerializer.Deserialize(payload, FinlyticJsonSerializerContext.Default.DailyNewsRequest);
if (req != null) if (req != null)
{ {
limit = req.Limit > 0 ? req.Limit : 20; limit = req.Limit > 0 ? req.Limit : 20;
@@ -167,7 +186,6 @@ public class NewsMqttClient : ManagedMqttClient, IHostedService
searchQuery = req.Query; searchQuery = req.Query;
date = req.Date; date = req.Date;
// Status anpassen, falls HasSentiment gesetzt ist
if (req.HasSentiment == true && string.IsNullOrEmpty(status)) if (req.HasSentiment == true && string.IsNullOrEmpty(status))
{ {
status = "Analyzed"; status = "Analyzed";
@@ -176,44 +194,35 @@ public class NewsMqttClient : ManagedMqttClient, IHostedService
} }
catch (Exception ex) catch (Exception ex)
{ {
_logger.LogWarning(ex, "Failed to parse DailyNewsRequest payload on news_Get"); await finlyticLogger.LogWarningAsync(SettingKeys.NewsChannel, ex, "Failed to parse DailyNewsRequest payload on news_Get");
} }
} }
try try
{ {
using var scope = _scopeFactory.CreateScope();
var dbService = scope.ServiceProvider.GetRequiredService<INewsDbService>(); var dbService = scope.ServiceProvider.GetRequiredService<INewsDbService>();
var articles = await dbService.GetFilteredNewsAsync(limit, offset, isin, date, status, searchQuery); var articles = await dbService.GetFilteredNewsAsync(limit, offset, isin, date, status, searchQuery);
var dtos = (await Task.WhenAll(articles.Select(a => MapToDtoAsync(a)))).ToList(); var dtos = (await Task.WhenAll(articles.Select(a => MapToDtoAsync(a)))).ToList();
string responseTopic = $"services/response/news_Get/{correlationId}"; string responseTopic = $"services/response/news_Get/{correlationId}";
_logger.LogInformation("Publishing RPC response to {ResponseTopic} with {Count} articles.", responseTopic, await finlyticLogger.LogInfoAsync(SettingKeys.NewsChannel, "Publishing RPC response to {ResponseTopic} with {Count} articles.", responseTopic, dtos.Count);
dtos.Count);
await PublishAsync(responseTopic, dtos); await PublishAsync(responseTopic, dtos);
} }
catch (Exception ex) catch (Exception ex)
{ {
_logger.LogError(ex, "Failed to compile RPC response for news_Get"); await finlyticLogger.LogErrorAsync(SettingKeys.NewsChannel, ex, "Failed to compile RPC response for news_Get");
// Antworte mit leerer Liste, um RPC-Timeouts im Gateway zu vermeiden
try try
{ {
string responseTopic = $"services/response/news_Get/{correlationId}"; string responseTopic = $"services/response/news_Get/{correlationId}";
await PublishAsync(responseTopic, new List<NewsArticleDto>()); await PublishAsync(responseTopic, new List<NewsArticleDto>());
} }
catch catch { }
{
}
} }
} }
private async Task OnGetNewsByIdAsync(string payload, string correlationId) private async Task OnGetNewsByIdAsync(string payload, string correlationId)
{ {
_logger.LogInformation("Received RPC news_GetById request. Correlation: {CorrelationId}", correlationId);
string responseTopic = $"services/response/news_GetById/{correlationId}"; string responseTopic = $"services/response/news_GetById/{correlationId}";
if (string.IsNullOrWhiteSpace(payload)) if (string.IsNullOrWhiteSpace(payload))
{ {
await PublishAsync(responseTopic, (object?)null); await PublishAsync(responseTopic, (object?)null);
@@ -222,9 +231,7 @@ public class NewsMqttClient : ManagedMqttClient, IHostedService
try try
{ {
var request = var request = JsonSerializer.Deserialize<ArticleRequest>(payload, FinlyticJsonSerializerContext.Default.ArticleRequest);
JsonSerializer.Deserialize<ArticleRequest>(payload,
FinlyticJsonSerializerContext.Default.ArticleRequest);
var targetIdStr = request?.ArticleId ?? request?.Id; var targetIdStr = request?.ArticleId ?? request?.Id;
if (Guid.TryParse(targetIdStr, out var articleId)) if (Guid.TryParse(targetIdStr, out var articleId))
@@ -241,33 +248,25 @@ public class NewsMqttClient : ManagedMqttClient, IHostedService
} }
} }
} }
catch (Exception ex) catch { }
{
_logger.LogError(ex, "Failed to execute RPC news_GetById");
}
await PublishAsync(responseTopic, (object?)null); await PublishAsync(responseTopic, (object?)null);
} }
private async Task OnGetPendingNewsAsync(string payload, string correlationId) private async Task OnGetPendingNewsAsync(string payload, string correlationId)
{ {
_logger.LogInformation("Received RPC news_GetPending request. Correlation: {CorrelationId}", correlationId);
int limit = 10; int limit = 10;
if (!string.IsNullOrWhiteSpace(payload)) if (!string.IsNullOrWhiteSpace(payload))
{ {
try try
{ {
using var doc = JsonDocument.Parse(payload); using var doc = JsonDocument.Parse(payload);
if (doc.RootElement.TryGetProperty("limit", out var limitProp) && if (doc.RootElement.TryGetProperty("limit", out var limitProp) && limitProp.TryGetInt32(out var parsedLimit))
limitProp.TryGetInt32(out var parsedLimit))
{ {
limit = Math.Min(parsedLimit, 10); limit = Math.Min(parsedLimit, 10);
} }
} }
catch catch { }
{
}
} }
try try
@@ -279,25 +278,17 @@ public class NewsMqttClient : ManagedMqttClient, IHostedService
var dtos = (await Task.WhenAll(pendingArticles.Take(limit).Select(a => MapToDtoAsync(a)))).ToList(); var dtos = (await Task.WhenAll(pendingArticles.Take(limit).Select(a => MapToDtoAsync(a)))).ToList();
string responseTopic = $"services/response/news_GetPending/{correlationId}"; string responseTopic = $"services/response/news_GetPending/{correlationId}";
_logger.LogInformation("Publishing RPC news_GetPending response to {ResponseTopic} with {Count} articles.",
responseTopic, dtos.Count);
await PublishAsync(responseTopic, dtos); await PublishAsync(responseTopic, dtos);
} }
catch (Exception ex) catch { }
{
_logger.LogError(ex, "Failed to publish RPC pending news response");
}
} }
private async Task OnUpdateNewsStatusAsync(string payload, string correlationId) private async Task OnUpdateNewsStatusAsync(string payload, string correlationId)
{ {
_logger.LogInformation("Received RPC news_UpdateStatus request. Correlation: {CorrelationId}", correlationId);
UpdateNewsStatusResponse response; UpdateNewsStatusResponse response;
try try
{ {
var request = JsonSerializer.Deserialize<UpdateNewsStatusRequest>(payload, var request = JsonSerializer.Deserialize<UpdateNewsStatusRequest>(payload, FinlyticJsonSerializerContext.Default.UpdateNewsStatusRequest);
FinlyticJsonSerializerContext.Default.UpdateNewsStatusRequest);
if (request != null && request.Id != Guid.Empty) if (request != null && request.Id != Guid.Empty)
{ {
using var scope = _scopeFactory.CreateScope(); using var scope = _scopeFactory.CreateScope();
@@ -313,7 +304,6 @@ public class NewsMqttClient : ManagedMqttClient, IHostedService
} }
catch (Exception ex) catch (Exception ex)
{ {
_logger.LogError(ex, "Failed to execute RPC news_UpdateStatus");
response = new UpdateNewsStatusResponse(false, ex.Message); response = new UpdateNewsStatusResponse(false, ex.Message);
} }
@@ -321,42 +311,99 @@ public class NewsMqttClient : ManagedMqttClient, IHostedService
await PublishAsync(responseTopic, response); await PublishAsync(responseTopic, response);
} }
private async Task OnSettingsGetAllAsync(string correlationId)
{
using var scope = _scopeFactory.CreateScope();
var finlyticLogger = scope.ServiceProvider.GetRequiredService<IFinlyticLogger<NewsMqttClient>>();
var settingsService = scope.ServiceProvider.GetRequiredService<ISettingsService>();
await finlyticLogger.LogInfoAsync(SettingKeys.MqttChannel, "[FinlyticNews] [Settings_GetAll] Retrieving all dynamic settings via reflection [CorrelationId: {CorrelationId}]", correlationId);
try
{
var settings = await settingsService.GetAllRegisteredSettingsAsync(new[] { typeof(SettingKeys) });
var responseTopic = $"services/response/news_settings_GetAll/{correlationId}";
await PublishAsync(responseTopic, settings);
await finlyticLogger.LogInfoAsync(SettingKeys.MqttChannel, "[FinlyticNews] [Settings_GetAll] Published {Count} settings to '{ResponseTopic}'", settings.Count, responseTopic);
}
catch (Exception ex)
{
await finlyticLogger.LogErrorAsync(SettingKeys.MqttChannel, ex, "[FinlyticNews] [Settings_GetAll] Failed to retrieve settings.");
}
}
private async Task OnSettingsUpdateAsync(string payload, string correlationId)
{
if (string.IsNullOrWhiteSpace(payload)) return;
using var scope = _scopeFactory.CreateScope();
var finlyticLogger = scope.ServiceProvider.GetRequiredService<IFinlyticLogger<NewsMqttClient>>();
var settingsService = scope.ServiceProvider.GetRequiredService<ISettingsService>();
await finlyticLogger.LogInfoAsync(SettingKeys.MqttChannel, "[FinlyticNews] [Settings_Update] Processing settings update RPC [CorrelationId: {CorrelationId}]", correlationId);
try
{
Dictionary<string, object?>? updates = null;
try
{
updates = JsonSerializer.Deserialize<Dictionary<string, object?>>(payload);
}
catch
{
var list = JsonSerializer.Deserialize<List<DynamicSettingDto>>(payload);
if (list != null)
{
updates = new Dictionary<string, object?>();
foreach (var item in list) updates[item.Key] = item.Value;
}
}
if (updates != null && updates.Count > 0)
{
await settingsService.UpdateSettingsAsync(updates);
await finlyticLogger.LogInfoAsync(SettingKeys.MqttChannel, "[FinlyticNews] [Settings_Update] Successfully updated {Count} settings in database and cache.", updates.Count);
}
var currentSettings = await settingsService.GetAllRegisteredSettingsAsync(new[] { typeof(SettingKeys) });
var responseTopic = $"services/response/news_settings_Update/{correlationId}";
await PublishAsync(responseTopic, currentSettings);
}
catch (Exception ex)
{
await finlyticLogger.LogErrorAsync(SettingKeys.MqttChannel, ex, "[FinlyticNews] [Settings_Update] Failed to update settings.");
}
}
private async Task OnConfigUpdatedAsync(string payload) private async Task OnConfigUpdatedAsync(string payload)
{ {
_logger.LogInformation("Received config update event for FinlyticNews.");
try try
{ {
using var doc = JsonDocument.Parse(payload); using var doc = JsonDocument.Parse(payload);
if (doc.RootElement.TryGetProperty("settings", out var settingsProp)) if (doc.RootElement.TryGetProperty("settings", out var settingsProp))
{ {
var dict = JsonSerializer.Deserialize<Dictionary<string, string>>(settingsProp.GetRawText()); var dict = JsonSerializer.Deserialize<Dictionary<string, object?>>(settingsProp.GetRawText());
if (dict != null && dict.Count > 0) if (dict != null && dict.Count > 0)
{ {
using var scope = _scopeFactory.CreateScope(); using var scope = _scopeFactory.CreateScope();
var settingsDb = scope.ServiceProvider.GetRequiredService<ISettingsDbService>(); var settings = scope.ServiceProvider.GetRequiredService<ISettingsService>();
await settingsDb.UpdateSettingsFromDictionaryAsync(dict); await settings.UpdateSettingsAsync(dict);
_logger.LogInformation("Successfully persisted {Count} updated settings.", dict.Count);
} }
} }
} }
catch (Exception ex) catch { }
{
_logger.LogError(ex, "Error processing MQTT config update event");
}
} }
private async Task OnHealthPingAsync(string[] segments, string correlationId) private async Task OnHealthPingAsync(string[] segments, string correlationId)
{ {
// Zerlegt den Topic-Pfad z. B. services/request/health_Ping/FinlyticNews/{correlationId}
bool isForMe = segments.Length >= 5 && segments[3].Equals("FinlyticNews", StringComparison.OrdinalIgnoreCase); bool isForMe = segments.Length >= 5 && segments[3].Equals("FinlyticNews", StringComparison.OrdinalIgnoreCase);
if (isForMe) if (isForMe)
{ {
string respTopic = $"services/response/health_Ping/{correlationId}"; string respTopic = $"services/response/health_Ping/{correlationId}";
await PublishAsync(respTopic, await PublishAsync(respTopic, new ServiceHealthResponse("FinlyticNews", "Online", DateTime.UtcNow, "Connected"));
new ServiceHealthResponse("FinlyticNews", "Online", DateTime.UtcNow, "Connected"));
_logger.LogInformation("Responded to live health_Ping RPC request [CorrelationId: {CorrelationId}].", using var scope = _scopeFactory.CreateScope();
correlationId); var finlyticLogger = scope.ServiceProvider.GetRequiredService<IFinlyticLogger<NewsMqttClient>>();
await finlyticLogger.LogInfoAsync(SettingKeys.HealthPingChannel, "Responded to live health_Ping RPC request [CorrelationId: {CorrelationId}].", correlationId);
} }
} }
@@ -375,41 +422,31 @@ public class NewsMqttClient : ManagedMqttClient, IHostedService
var targetId = a.Id.ToString(); var targetId = a.Id.ToString();
IsinAnalysisEntry? sentimentEntry = null; IsinAnalysisEntry? sentimentEntry = null;
// 1. Snappy Local Disk Check for Article File var articlePath = Path.Combine(Directory.GetCurrentDirectory(), "data", "summaries", "articles", $"{targetId}.json");
var articlePath = Path.Combine(Directory.GetCurrentDirectory(), "data", "summaries", "articles",
$"{targetId}.json");
if (File.Exists(articlePath)) if (File.Exists(articlePath))
{ {
try try
{ {
var json = await File.ReadAllTextAsync(articlePath); var json = await File.ReadAllTextAsync(articlePath);
sentimentEntry = JsonSerializer.Deserialize<IsinAnalysisEntry>(json, sentimentEntry = JsonSerializer.Deserialize<IsinAnalysisEntry>(json, FinlyticJsonSerializerContext.Default.IsinAnalysisEntry);
FinlyticJsonSerializerContext.Default.IsinAnalysisEntry);
}
catch
{
} }
catch { }
} }
// 2. ISIN Summary File Fallback
if (sentimentEntry == null && a.MatchedAssets != null && a.MatchedAssets.Count > 0) if (sentimentEntry == null && a.MatchedAssets != null && a.MatchedAssets.Count > 0)
{ {
foreach (var asset in a.MatchedAssets) foreach (var asset in a.MatchedAssets)
{ {
if (string.IsNullOrWhiteSpace(asset.Isin)) continue; if (string.IsNullOrWhiteSpace(asset.Isin)) continue;
var isinPath = Path.Combine(Directory.GetCurrentDirectory(), "data", "summaries", "isin", var isinPath = Path.Combine(Directory.GetCurrentDirectory(), "data", "summaries", "isin", $"{asset.Isin.Trim()}.json");
$"{asset.Isin.Trim()}.json");
if (File.Exists(isinPath)) if (File.Exists(isinPath))
{ {
try try
{ {
var json = await File.ReadAllTextAsync(isinPath); var json = await File.ReadAllTextAsync(isinPath);
var isinDoc = JsonSerializer.Deserialize<IsinSentimentSummaryDto>(json, var isinDoc = JsonSerializer.Deserialize<IsinSentimentSummaryDto>(json, FinlyticJsonSerializerContext.Default.IsinSentimentSummaryDto);
FinlyticJsonSerializerContext.Default.IsinSentimentSummaryDto); var match = isinDoc?.Analyses?.FirstOrDefault(entry => string.Equals(entry.Article?.ArticleId?.Trim(), targetId, StringComparison.OrdinalIgnoreCase));
var match = isinDoc?.Analyses?.FirstOrDefault(entry =>
string.Equals(entry.Article?.ArticleId?.Trim(), targetId,
StringComparison.OrdinalIgnoreCase));
if (match != null) if (match != null)
{ {
@@ -417,9 +454,7 @@ public class NewsMqttClient : ManagedMqttClient, IHostedService
break; break;
} }
} }
catch catch { }
{
}
} }
} }
} }
@@ -432,10 +467,7 @@ public class NewsMqttClient : ManagedMqttClient, IHostedService
confidence = finbertResult.Confidence; confidence = finbertResult.Confidence;
} }
} }
catch (Exception ex) catch { }
{
_logger.LogTrace(ex, "[MapToDtoAsync] Sentiment fetch skipped for article {Id}", a.Id);
}
return new NewsArticleDto return new NewsArticleDto
{ {
+24
View File
@@ -0,0 +1,24 @@
using FinlyticCore.Models.Settings;
namespace FinlyticNews.Util;
public static class SettingKeys
{
// --- Logging-Kanäle ---
public static readonly SettingKey<bool> NewsChannel = new("Logging.Channel.News", true);
public static readonly SettingKey<bool> MqttChannel = new("Logging.Channel.MQTT", true);
public static readonly SettingKey<bool> HealthPingChannel = new("Logging.Channel.Health", true);
// --- Scraping & Feed-Konfiguration ---
public static readonly SettingKey<int> ScrapeIntervalMinutes = new("Scraping.IntervalMinutes", 15);
public static readonly SettingKey<int> MaxArticlesPerFeed = new("Scraping.MaxArticlesPerFeed", 20);
public static readonly SettingKey<bool> EnableAutoScraping = new("Feature.EnableAutoScraping", true);
public static readonly SettingKey<int> HttpTimeoutSeconds = new("Scraping.HttpTimeoutSeconds", 20);
// --- KI & Sentiment-Konfiguration ---
public static readonly SettingKey<int> FinBertBatchSize = new("AI.FinBertBatchSize", 8);
public static readonly SettingKey<double> MinSentimentConfidence = new("AI.MinSentimentConfidence", 0.65);
// --- Daten-Retention & Cleanup ---
public static readonly SettingKey<int> ArticleRetentionDays = new("Data.ArticleRetentionDays", 90);
}