Files
narratio/internal/artifacts/extraction_catalog.go

100 lines
3.5 KiB
Go

package artifacts
import (
"os"
"path/filepath"
"strings"
"gitea.maximumdirect.net/eric/narratio/internal/artifactmodel"
"gitea.maximumdirect.net/eric/narratio/internal/manifest"
)
// MaxExtractionPayloadBytes bounds one validated Notarius lane or index payload.
const MaxExtractionPayloadBytes int64 = 64 * 1024 * 1024
const (
extractStageName = "extract"
extractionLaneKind = "notarius_lane"
extractionIndexKind = "notarius_index"
extractionMetadataRun = "narratio_run_id"
extractionMetadataRoot = "bundle_root"
)
// HydrateExtractionArtifacts marks extraction sources available only when the current
// manifest contains one complete, internally consistent, succeeded extraction bundle.
// Invalid, stale, incomplete, or unsafe records leave every extraction source unavailable.
func (c *ArtifactCatalog) HydrateExtractionArtifacts(
paths SessionPaths,
m *manifest.Manifest,
configured map[string]ExtractionArtifactDefinition,
) {
if c == nil || m == nil || len(configured) == 0 {
return
}
proof := InspectExtractionEvidence(paths, m, configured)
if proof.State != ExtractionEvidenceValid {
return
}
for sourceID, path := range proof.Outputs {
_ = c.markAvailableFromExtractManifest(sourceID, path, proof.ProducerRunID)
}
}
func compatibleCatalogExtractionContract(got *artifactmodel.ContractMetadata, want ExtractionArtifactDefinition) bool {
return got != nil && got.MediaType == want.MediaType && got.SchemaID == want.SchemaID &&
got.SchemaVersion == want.SchemaVersion && (want.ModuleKey == "" || got.ModuleKey == want.ModuleKey)
}
func compatibleCatalogExtractionProvenance(
got *artifactmodel.ExternalProvenance,
runID, pipelineID string,
want ExtractionArtifactDefinition,
) bool {
return got != nil && got.System == "notarius" && got.RunID == runID && got.PipelineID == pipelineID &&
pipelineID == strings.TrimSpace(want.PipelineID) && got.ArtifactID == want.LaneID
}
func safeExistingExtractionDirectory(sessionRoot, bundleRoot string) bool {
if !pathWithinExtractionRoot(sessionRoot, bundleRoot) || !safeExtractionComponents(sessionRoot, bundleRoot) {
return false
}
info, err := os.Lstat(bundleRoot)
return err == nil && info.IsDir() && info.Mode()&os.ModeSymlink == 0
}
func safeExtractionComponents(root, target string) bool {
relative, err := filepath.Rel(filepath.Clean(root), filepath.Clean(target))
if err != nil || relative == "." || relative == ".." || strings.HasPrefix(relative, ".."+string(filepath.Separator)) {
return false
}
current := filepath.Clean(root)
for _, part := range strings.Split(relative, string(filepath.Separator)) {
current = filepath.Join(current, part)
info, err := os.Lstat(current)
if err != nil || info.Mode()&os.ModeSymlink != 0 {
return false
}
}
return true
}
func pathWithinExtractionRoot(root, target string) bool {
relative, err := filepath.Rel(filepath.Clean(root), filepath.Clean(target))
return err == nil && relative != "." && relative != ".." && !strings.HasPrefix(relative, ".."+string(filepath.Separator))
}
func safeExtractionPathSegment(value string) bool {
return value != "" && value != "." && value != ".." && filepath.Base(value) == value &&
!strings.ContainsAny(value, `/\\`)
}
func extractionMetadataString(metadata map[string]any, key string) string {
value, _ := metadata[key].(string)
return strings.TrimSpace(value)
}
func extractionReceiptIdentity(metadata map[string]any) (string, string) {
receipt, _ := metadata["receipt"].(map[string]any)
return extractionMetadataString(receipt, "run_id"), extractionMetadataString(receipt, "pipeline_id")
}