git.stella-ops.org/src/Signals/StellaOps.Signals/Services/UnknownsIngestionService.cs

using System;
using System.Collections.Generic;
using System.Linq;
using System.Threading;
using System.Threading.Tasks;
using Microsoft.Extensions.Logging;
using StellaOps.Signals.Models;
using StellaOps.Signals.Persistence;

namespace StellaOps.Signals.Services;

internal sealed class UnknownsIngestionService : IUnknownsIngestionService
{
    private readonly IUnknownsRepository repository;
    private readonly TimeProvider timeProvider;
    private readonly ILogger<UnknownsIngestionService> logger;

    public UnknownsIngestionService(IUnknownsRepository repository, TimeProvider timeProvider, ILogger<UnknownsIngestionService> logger)
    {
        this.repository = repository ?? throw new ArgumentNullException(nameof(repository));
        this.timeProvider = timeProvider ?? throw new ArgumentNullException(nameof(timeProvider));
        this.logger = logger ?? throw new ArgumentNullException(nameof(logger));
    }

    public async Task<UnknownsIngestResponse> IngestAsync(UnknownsIngestRequest request, CancellationToken cancellationToken)
    {
        ArgumentNullException.ThrowIfNull(request);

        if (request.Subject is null)
        {
            throw new UnknownsValidationException("Subject is required.");
        }

        if (string.IsNullOrWhiteSpace(request.CallgraphId))
        {
            throw new UnknownsValidationException("callgraphId is required.");
        }

        if (request.Unknowns is null || request.Unknowns.Count == 0)
        {
            throw new UnknownsValidationException("Unknowns list must not be empty.");
        }

        var subjectKey = request.Subject.ToSubjectKey();
        if (string.IsNullOrWhiteSpace(subjectKey))
        {
            throw new UnknownsValidationException("Subject must include scanId, imageDigest, or component/version.");
        }

        var now = timeProvider.GetUtcNow();
        var normalized = new List<UnknownSymbolDocument>();

        foreach (var entry in request.Unknowns)
        {
            if (entry is null)
            {
                continue;
            }

            var hasContent = !(string.IsNullOrWhiteSpace(entry.SymbolId)
                && string.IsNullOrWhiteSpace(entry.CodeId)
                && string.IsNullOrWhiteSpace(entry.Purl)
                && string.IsNullOrWhiteSpace(entry.EdgeFrom)
                && string.IsNullOrWhiteSpace(entry.EdgeTo));

            if (!hasContent)
            {
                continue;
            }

            normalized.Add(new UnknownSymbolDocument
            {
                SubjectKey = subjectKey,
                CallgraphId = request.CallgraphId,
                SymbolId = entry.SymbolId?.Trim(),
                CodeId = entry.CodeId?.Trim(),
                Purl = entry.Purl?.Trim(),
                EdgeFrom = entry.EdgeFrom?.Trim(),
                EdgeTo = entry.EdgeTo?.Trim(),
                Reason = entry.Reason?.Trim(),
                CreatedAt = now,
                UpdatedAt = now,
                LastAnalyzedAt = now
            });
        }

        if (normalized.Count == 0)
        {
            throw new UnknownsValidationException("Unknown entries must include at least one symbolId, codeId, purl, or edge.");
        }

        await repository.UpsertAsync(subjectKey, normalized, cancellationToken).ConfigureAwait(false);
        logger.LogInformation("Stored {Count} unknown symbols for subject {SubjectKey}", normalized.Count, subjectKey);

        return new UnknownsIngestResponse
        {
            SubjectKey = subjectKey,
            UnknownsCount = normalized.Count
        };
    }
}