narratio/internal/artifacts/analyze_evidence.go

package artifacts

import (
	"crypto/sha256"
	"encoding/hex"
	"fmt"
	"io"
	"path/filepath"
	"strings"

	"gitea.maximumdirect.net/eric/narratio/internal/artifactmodel"
	"gitea.maximumdirect.net/eric/narratio/internal/fileops"
	"gitea.maximumdirect.net/eric/narratio/internal/manifest"
	"gitea.maximumdirect.net/eric/narratio/internal/pathsafe"
)

// AnalyzeEvidenceState distinguishes a verified current configured artifact
// from every form of unavailable evidence.
type AnalyzeEvidenceState string

const (
	AnalyzeEvidenceCurrent    AnalyzeEvidenceState = "current"
	AnalyzeEvidenceNonCurrent AnalyzeEvidenceState = "non_current"
)

// AnalyzeEvidence is the read-only result of inspecting one configured
// artifact's manifest record and durable output.
type AnalyzeEvidence struct {
	State         AnalyzeEvidenceState
	Reason        string
	SourceID      string
	Path          string
	ProducerRunID string
	Contract      *artifactmodel.ContractMetadata
	Checksum      string
	Size          int64
}

// InspectAnalyzeEvidence verifies that one configured artifact has supported,
// current manifest evidence for the exact canonical bytes on disk.
func InspectAnalyzeEvidence(
	paths SessionPaths,
	m *manifest.Manifest,
	key string,
	configured ConfiguredArtifactDefinition,
) AnalyzeEvidence {
	normalizedKey := strings.TrimSpace(key)
	sourceID := ConfiguredArtifactSourceID(normalizedKey)
	nonCurrent := func(reason string) AnalyzeEvidence {
		return AnalyzeEvidence{State: AnalyzeEvidenceNonCurrent, Reason: reason, SourceID: sourceID}
	}

	if m == nil {
		return nonCurrent("analyze manifest evidence is absent; regenerate the artifact")
	}
	stageRecord := m.Stages["analyze"]
	if stageRecord == nil || stageRecord.Name != "analyze" {
		return nonCurrent("analyze manifest evidence is absent; regenerate the artifact")
	}
	if !stageRecord.HasVersionedAnalyzeState() {
		return nonCurrent("analyze manifest evidence is legacy or unsupported; regenerate the artifact")
	}
	record, ok := stageRecord.AnalyzeArtifacts[normalizedKey]
	if !ok {
		return nonCurrent("configured artifact has no analyze manifest record; regenerate the artifact")
	}
	if record.Status != manifest.AnalyzeArtifactCurrent {
		return nonCurrent(fmt.Sprintf("configured artifact manifest status is %q; regenerate the artifact", record.Status))
	}
	if err := manifest.ValidateAnalyzeArtifactCollection(
		stageRecord.AnalyzeStateVersion,
		map[string]manifest.AnalyzeArtifactRecord{normalizedKey: record},
	); err != nil {
		return nonCurrent("configured artifact manifest evidence is malformed; regenerate the artifact")
	}

	configuredPath, err := pathsafe.NormalizeRelativeDestination(strings.TrimSpace(configured.OutputPath))
	if err != nil {
		return nonCurrent("configured artifact output path is unsafe; correct the configuration")
	}
	if record.Output.LocalPath != configuredPath {
		return nonCurrent("configured artifact manifest path differs from current configuration; regenerate the artifact")
	}
	if record.Output.SourceID != sourceID || record.Output.Kind != "scriptorium_artifact" {
		return nonCurrent("configured artifact manifest identity is incompatible; regenerate the artifact")
	}

	file, err := fileops.OpenConfinedRegularFile(paths.Root, configuredPath)
	if err != nil {
		return nonCurrent("configured artifact output is missing or unsafe; regenerate the artifact")
	}
	defer file.Close()
	info, err := file.Stat()
	if err != nil || !info.Mode().IsRegular() {
		return nonCurrent("configured artifact output is not a safe regular file; regenerate the artifact")
	}
	hash := sha256.New()
	size, err := io.Copy(hash, file)
	if err != nil {
		return nonCurrent("configured artifact output could not be verified; regenerate the artifact")
	}
	if size != info.Size() || size != record.OutputSize {
		return nonCurrent("configured artifact output size differs from manifest evidence; regenerate the artifact")
	}
	if hex.EncodeToString(hash.Sum(nil)) != record.Output.Checksum {
		return nonCurrent("configured artifact output checksum differs from manifest evidence; regenerate the artifact")
	}

	return AnalyzeEvidence{
		State:         AnalyzeEvidenceCurrent,
		SourceID:      sourceID,
		Path:          filepath.Join(paths.Root, filepath.FromSlash(configuredPath)),
		ProducerRunID: record.ProducerRunID,
		Contract:      cloneArtifactContract(record.Output.Contract),
		Checksum:      record.Output.Checksum,
		Size:          record.OutputSize,
	}
}

// HydrateAnalyzeArtifacts makes configured sources available only from
// validated current manifest evidence. It never mutates manifest state.
func (c *ArtifactCatalog) HydrateAnalyzeArtifacts(
	paths SessionPaths,
	m *manifest.Manifest,
	configured map[string]ConfiguredArtifactDefinition,
) {
	if c == nil || len(configured) == 0 {
		return
	}
	for _, entry := range c.ListConfigured() {
		definition, ok := configured[entry.ConfiguredKey]
		if !ok {
			continue
		}
		evidence := InspectAnalyzeEvidence(paths, m, entry.ConfiguredKey, definition)
		if evidence.State != AnalyzeEvidenceCurrent {
			continue
		}
		_ = c.markAvailableFromAnalyzeManifest(
			evidence.SourceID, evidence.Path, evidence.ProducerRunID,
			evidence.Checksum, evidence.Size, evidence.Contract,
		)
	}
}