feat(cv-import): ground reviewed facts in revisioned career profiles

2026-10-26 14:30:14 +01:00
parent 3ea8ff9d62
commit e8f76a9dea
24 changed files with 1197 additions and 13 deletions
@@ -0,0 +1,71 @@
+using JobTrackerApi.Data;
+using Microsoft.EntityFrameworkCore;
+
+namespace JobTrackerApi.Services;
+
+public sealed class CvImportRetentionHostedService(IServiceScopeFactory scopes, AppPaths paths, ILogger<CvImportRetentionHostedService> logger) : BackgroundService
+{
+    protected override async Task ExecuteAsync(CancellationToken stoppingToken)
+    {
+        using var timer = new PeriodicTimer(TimeSpan.FromHours(1));
+        do
+        {
+            try
+            {
+                await using var scope = scopes.CreateAsyncScope();
+                await SweepAsync(scope.ServiceProvider.GetRequiredService<JobTrackerContext>(), paths, DateTimeOffset.UtcNow, stoppingToken);
+            }
+            catch (OperationCanceledException) when (stoppingToken.IsCancellationRequested) { break; }
+            catch (Exception) { logger.LogWarning("CV import retention could not complete; it will retry on the next sweep."); }
+        } while (await timer.WaitForNextTickAsync(stoppingToken));
+    }
+
+    public static async Task<int> SweepAsync(JobTrackerContext db, AppPaths paths, DateTimeOffset now, CancellationToken token)
+    {
+        var runs = await db.CvExtractionRuns.IgnoreQueryFilters().Include(r => r.Artifact)
+            .Where(r => r.ParserVersion == "docling-v1" && r.Artifact != null && r.Artifact.StoragePath != "")
+            .OrderBy(r => r.Id).Take(1000).ToListAsync(token);
+        var root = Path.GetFullPath(paths.CvArtifactsRoot) + Path.DirectorySeparatorChar;
+        var removed = 0;
+        foreach (var run in runs)
+        {
+            token.ThrowIfCancellationRequested();
+            if (run.StartedAtUtc > now.AddDays(-7) && (!(run.Status is "cancelled" or "failed" or "discarded") || run.StartedAtUtc > now.AddDays(-1))) continue;
+            var artifact = run.Artifact!;
+            if (artifact.OwnerUserId != run.OwnerUserId) continue;
+            var path = Path.GetFullPath(artifact.StoragePath);
+            if (!path.StartsWith(root, OperatingSystem.IsWindows() ? StringComparison.OrdinalIgnoreCase : StringComparison.Ordinal)) continue;
+            // Revoke the operation before deleting its source; a late worker cannot publish a result.
+            var subject = run.Id.ToString(System.Globalization.CultureInfo.InvariantCulture);
+            var operations = await db.UserOperations.IgnoreQueryFilters().Where(o => o.OwnerUserId == run.OwnerUserId && o.SubjectType == CvProcessingQueue.SubjectType && o.SubjectId == subject).ToListAsync(token);
+            foreach (var operation in operations.Where(o => !Models.OperationStatuses.IsTerminal(o.Status)))
+            {
+                operation.Status = Models.OperationStatuses.Cancelled;
+                operation.CancellationRequestedAtUtc = now.UtcDateTime;
+                operation.CompletedAtUtc = now.UtcDateTime;
+            }
+            await db.SaveChangesAsync(token);
+            try
+            {
+                if (File.Exists(path) && (File.GetAttributes(path) & FileAttributes.ReparsePoint) != 0) continue;
+                File.Delete(path);
+            }
+            catch (IOException) { continue; }
+            catch (UnauthorizedAccessException) { continue; }
+            artifact.StoragePath = "";
+            run.DocumentJson = null;
+            run.RawExtractedText = null;
+            run.NormalizedText = null;
+            if (run.Status != "applied")
+            {
+                run.StructuredProfileJson = null;
+                if (run.Status is "pending_review" or "queued" or "running") run.Status = "expired";
+                run.CompletedAtUtc ??= now;
+                run.ErrorMessage = "The source document expired. Upload it again to start a new review.";
+            }
+            removed++;
+        }
+        await db.SaveChangesAsync(token);
+        return removed;
+    }
+}