Files
notarius/internal/framework/pipeline/runner_chunk_plan.go

349 lines
15 KiB
Go

package pipeline
import (
"context"
"fmt"
"path"
"time"
"gitea.maximumdirect.net/eric/notarius/internal/core/artifacts"
"gitea.maximumdirect.net/eric/notarius/internal/core/source"
"gitea.maximumdirect.net/eric/notarius/internal/framework/contracts"
)
type chunkPlanExecution struct {
accepted bool
chunks []source.Chunk
plan *source.ChunkPlan
warnings []contracts.Warning
rejection *contracts.RejectedOutput
lookup ChunkPlanDecision
record *ChunkPlanRecord
action string
summary artifacts.ChunkPlanSummary
validation *artifacts.ValidationSummary
}
type generatedChunkPlanCandidate struct {
plan source.ChunkPlan
chunks []source.Chunk
record ChunkPlanRecord
producerWarnings []contracts.Warning
terminal *attemptTerminalRecorder
}
func effectiveChunkCacheMode(mode ChunkCacheMode) ChunkCacheMode {
if mode == "" {
return ChunkCacheBypass
}
parsed, _ := ParseChunkCacheMode(string(mode))
return parsed
}
// Chunk-plan lookup and publication happen serially before lane workers start.
// The runner therefore does not add synchronization around the store.
func (r *Runner) runChunkPlan(ctx context.Context, input RunInput, doc *source.SourceDocument, sourceInput contracts.LLMInputMaterial, sessionID string) (chunkPlanExecution, error) {
mode := effectiveChunkCacheMode(input.ChunkCacheMode)
chunker := input.Prepared.chunker
result := chunkPlanExecution{
lookup: ChunkPlanDecision{Reason: "chunk plan lookup skipped"},
summary: artifacts.ChunkPlanSummary{
Mode: string(mode), SourceDigest: doc.Digest, RequestedModule: chunker.Key(), LookupStatus: "skipped",
LookupReason: "chunk plan lookup skipped", ValidationStatus: "not_run", PublicationStatus: "not_requested",
},
}
if mode == ChunkCacheAuto {
record, decision, err := input.ChunkPlans.Load(doc.Digest)
result.lookup = decision
result.summary.LookupStatus = chunkPlanLookupStatus(decision.Status)
result.summary.LookupReason = chunkPlanLookupReason(decision.Status)
if err != nil {
result.summary.LookupStatus = "skipped"
result.summary.LookupReason = chunkPlanLookupReason("")
return result, fmt.Errorf("load chunk plan: %w", err)
}
switch decision.Status {
case ChunkPlanHit:
plan, chunks, validationErr := validateAndMaterializeChunkPlan(doc, record.Plan)
if validationErr == nil {
report, err := r.validateChunkReport(ctx, doc, chunker.Key(), chunks, sourceInput, sessionID, input.pipeline.ChunkReferences.ReferenceSet, input.Metadata, input.Prepared.chunkValidators, 1, input.Debug)
if err != nil {
result.setValidation(report.Warnings(), nil, err)
return result, err
}
rejection := report.FirstRejection()
incomplete := firstIncompleteValidation(report)
if rejection == nil && incomplete == nil {
if err := result.setCandidate(record, "reused"); err != nil {
return result, fmt.Errorf("clone reused chunk plan record: %w", err)
}
result.plan = &plan
result.chunks = chunks
result.warnings = append(cloneWarnings(record.Warnings), report.Warnings()...)
result.accepted = true
result.setValidation(report.Warnings(), nil, nil)
cachedTerminal := producerAttemptTerminal{Action: producerTerminalAccepted, Validation: report}
summary := validationSummary(cachedTerminal, StageChunk, "", "", chunker.Key(), "", 0)
result.validation = &summary
return result, nil
}
if incomplete != nil && input.pipeline.ChunkValidationPolicy.ValidatorFailure == ValidatorFailureFailRun {
failure := validatorFailureError(*incomplete)
result.setValidation(report.Warnings(), nil, failure)
failedTerminal := producerAttemptTerminal{Action: producerTerminalFailed, Validation: report, ValidationIncomplete: true}
summary := validationSummary(failedTerminal, StageChunk, "", "", chunker.Key(), "", 0)
result.validation = &summary
return result, failure
}
// A cache hit is not model material. Its rejection is discarded and
// generation begins with the ordinary initial request below. Warnings
// from this discarded candidate are intentionally not promoted.
}
result.lookup = ChunkPlanDecision{Status: ChunkPlanInvalid, Reason: chunkPlanLookupReason(ChunkPlanInvalid)}
result.summary.LookupStatus = "invalid"
result.summary.LookupReason = chunkPlanLookupReason(ChunkPlanInvalid)
case ChunkPlanMissing, ChunkPlanInvalid:
// Generate below.
default:
return result, fmt.Errorf("load chunk plan returned unsupported decision status")
}
}
if mode == ChunkCacheAuto || mode == ChunkCacheRefresh {
result.summary.PublicationStatus = "not_published"
}
terminal, err := runProducerAttempts(ctx, producerAttemptConfig{
Retries: input.pipeline.Chunk.Retries,
Policy: input.pipeline.ChunkValidationPolicy,
AllowStructuralRetry: input.pipeline.ChunkExecutionClass == contracts.ExecutionClassLLMBacked,
}, func(attemptCtx context.Context, request producerAttemptRequest) (producerAttemptOutput, error) {
attempt := request.Number
attemptStarted := time.Now().UTC()
attemptPath := path.Join("chunk", fmt.Sprintf("attempt-%02d", attempt))
attemptCtx, llmScope := withDebugLLMScope(attemptCtx, attemptPath)
attemptTerminal := newAttemptTerminalRecorder(input.Debug, attemptPath, "chunk", llmScope, debugTimedEnvelope{Stage: string(StageChunk), ModuleKey: chunker.Key(), Attempt: attempt, AttemptKind: string(request.Kind), StartedAt: attemptStarted})
requestMetadata, metadataErr := cloneMetadata(input.Metadata)
if metadataErr != nil {
return producerAttemptOutput{}, attemptTerminal.record(nil, fmt.Errorf("clone chunk request metadata: %w", metadataErr))
}
chunkResult, callErr := chunker.Plan(attemptCtx, contracts.ChunkRequest{
Source: doc, SourceInput: sourceInput.Clone(), SessionID: sessionID,
References: CloneReferenceSet(input.pipeline.ChunkReferences.ReferenceSet),
LLMProfile: input.pipeline.Chunk.LLMProfile, StructuredOutputRepairAttempts: cloneStructuredOutputRepairAttempts(input.pipeline.Chunk.StructuredOutputRepairAttempts), Correction: request.Correction, Metadata: requestMetadata,
})
if callErr != nil {
return producerAttemptOutput{}, attemptTerminal.record(nil, fmt.Errorf("chunk source with chunker %q: %w", chunker.Key(), callErr))
}
plan, chunks, validationErr := validateAndMaterializeChunkPlan(doc, chunkResult.Plan)
if validationErr != nil {
attemptErr := fmt.Errorf("validate chunk plan from chunker %q: %w", chunker.Key(), validationErr)
payload := map[string]any{"plan": debugChunkPlanEnvelope(chunkResult.Plan), "warnings": debugWarningEnvelopes(chunkResult.Warnings)}
return producerAttemptOutput{}, attemptTerminal.record(payload, fmt.Errorf("%w: %v", contracts.ErrInvalidStructuredOutput, attemptErr))
}
planDigest, digestErr := source.DigestChunkPlan(plan)
if digestErr != nil {
return producerAttemptOutput{}, attemptTerminal.record(nil, fmt.Errorf("digest generated chunk plan: %w", digestErr))
}
producerMetadata, _, metadataErr := moduleManifestMetadata(chunker)
if metadataErr != nil {
return producerAttemptOutput{}, attemptTerminal.record(nil, fmt.Errorf("clone chunker manifest metadata: %w", metadataErr))
}
profile := ""
if input.pipeline.ChunkExecutionClass == contracts.ExecutionClassLLMBacked {
profile = input.pipeline.Chunk.LLMProfile
}
candidate := ChunkPlanRecord{
SchemaVersion: ChunkPlanSchemaVersion, SourceDigest: doc.Digest, PlanDigest: planDigest,
Plan: source.CloneChunkPlan(plan),
Producer: ChunkPlanProducer{
InputModule: input.Prepared.input.Key(), ChunkModule: chunker.Key(), LLMProfile: profile,
References: append([]artifacts.ReferenceProvenance(nil), referenceTargetProvenance(input.pipeline.ChunkReferences)...),
Metadata: producerMetadata,
},
Warnings: cloneWarnings(chunkResult.Warnings), CreatedAt: time.Now().UTC(),
}
return producerAttemptOutput{Value: generatedChunkPlanCandidate{plan: plan, chunks: chunks, record: candidate, producerWarnings: cloneWarnings(chunkResult.Warnings), terminal: &attemptTerminal}, Candidate: chunkResult.ModelCandidate, Warnings: cloneWarnings(chunkResult.Warnings)}, nil
}, func(validationCtx context.Context, output producerAttemptOutput) (validationReport, error) {
candidate, ok := output.Value.(generatedChunkPlanCandidate)
if !ok {
return validationReport{}, fmt.Errorf("chunk attempt candidate has incompatible type")
}
report, validationErr := r.validateChunkReport(validationCtx, doc, chunker.Key(), candidate.chunks, sourceInput, sessionID, input.pipeline.ChunkReferences.ReferenceSet, input.Metadata, input.Prepared.chunkValidators, candidate.terminal.envelope.Attempt, input.Debug)
attemptWarnings := append(cloneWarnings(output.Warnings), report.Warnings()...)
payload := map[string]any{
"plan": debugChunkPlanEnvelope(candidate.plan), "materialized_chunks": debugSourceChunkEnvelopes(candidate.chunks),
"warnings": debugWarningEnvelopes(attemptWarnings), "rejection": debugRejectedOutputPtr(chunkRejection(report, candidate.terminal.envelope.Attempt, chunker.Key())),
}
if validationErr != nil {
return report, candidate.terminal.record(payload, validationErr)
}
return report, candidate.terminal.record(payload, nil)
})
terminalSummary := validationSummary(terminal, StageChunk, "", "", chunker.Key(), "", 0)
if debugErr := writeProducerTerminalDebug(input.Debug, "chunk/terminal.json", terminal, input.pipeline.ChunkValidationPolicy, terminalSummary); debugErr != nil {
return result, debugErr
}
if err != nil {
if result.summary.ValidationStatus == "not_run" {
result.summary.ValidationStatus = "error"
}
return result, err
}
if terminal.Action == producerTerminalRejected {
result.rejection = terminal.Rejection
if result.rejection != nil {
result.rejection.Stage = string(StageChunk)
result.rejection.ModuleKey = chunker.Key()
}
result.warnings = cloneWarnings(terminal.Warnings)
result.validation = &terminalSummary
if result.rejection != nil {
result.rejection.Validation = cloneValidationSummaryPtr(result.validation)
}
result.setValidation(terminal.Validation.Warnings(), result.rejection, nil)
return result, nil
}
candidate, ok := terminal.Value.(generatedChunkPlanCandidate)
if !ok {
return result, fmt.Errorf("chunk attempt terminal has incompatible value")
}
if err := result.setCandidate(candidate.record, "generated"); err != nil {
return result, fmt.Errorf("clone generated chunk plan record: %w", err)
}
if mode == ChunkCacheRefresh {
result.action = "refreshed"
result.summary.Action = "refreshed"
}
if mode == ChunkCacheBypass {
result.action = "bypassed"
result.summary.Action = "bypassed"
}
result.accepted = true
result.plan = &candidate.plan
result.chunks = candidate.chunks
result.warnings = cloneWarnings(terminal.Warnings)
result.validation = &terminalSummary
result.setValidation(terminal.Validation.Warnings(), nil, nil)
if terminal.ValidationIncomplete {
result.summary.ValidationStatus = "incomplete"
}
if (mode == ChunkCacheAuto || mode == ChunkCacheRefresh) && !terminal.ValidationIncomplete {
record, cloneErr := cloneChunkPlanRecord(*result.record)
if cloneErr != nil {
return result, fmt.Errorf("clone chunk plan record for publication: %w", cloneErr)
}
record.Warnings = cloneWarnings(candidate.producerWarnings)
if err := input.ChunkPlans.Save(record); err != nil {
return result, fmt.Errorf("save chunk plan: %w", err)
}
result.summary.PublicationStatus = "published"
}
return result, nil
}
func chunkRejection(report validationReport, attempt int, moduleKey string) *contracts.RejectedOutput {
rejection := report.FirstRejection()
if rejection == nil {
return nil
}
return &contracts.RejectedOutput{Stage: string(StageChunk), ModuleKey: moduleKey, ValidatorName: rejection.validatorName, ReasonCode: rejection.reasonCode, Message: rejection.message, AttemptCount: attempt, DiagnosticArtifactPath: rejection.diagnosticPath}
}
func chunkPlanLookupStatus(status ChunkPlanStatus) string {
switch status {
case ChunkPlanHit:
return "hit"
case ChunkPlanMissing:
return "missing"
case ChunkPlanInvalid:
return "invalid"
default:
return "skipped"
}
}
func chunkPlanLookupReason(status ChunkPlanStatus) string {
switch status {
case ChunkPlanHit:
return "stored chunk plan is valid"
case ChunkPlanMissing:
return "chunk plan not found"
case ChunkPlanInvalid:
return "stored chunk plan is invalid"
default:
return "chunk plan lookup skipped"
}
}
func (result *chunkPlanExecution) setCandidate(record ChunkPlanRecord, action string) error {
cloned, err := cloneChunkPlanRecord(record)
if err != nil {
return err
}
result.record = &cloned
result.action = action
result.summary.Action = action
result.summary.SourceDigest = record.SourceDigest
result.summary.CandidateDigest = record.PlanDigest
return nil
}
func (result *chunkPlanExecution) setValidation(warnings []contracts.Warning, rejection *contracts.RejectedOutput, err error) {
switch {
case err != nil:
result.summary.ValidationStatus = "error"
case rejection != nil:
result.summary.ValidationStatus = "rejected"
case len(warnings) > 0:
result.summary.ValidationStatus = "approved_with_warnings"
default:
result.summary.ValidationStatus = "approved"
}
}
func cloneChunkPlanRecord(record ChunkPlanRecord) (ChunkPlanRecord, error) {
record.Plan = source.CloneChunkPlan(record.Plan)
record.Producer.References = append([]artifacts.ReferenceProvenance(nil), record.Producer.References...)
metadata, err := cloneMetadata(record.Producer.Metadata)
if err != nil {
return ChunkPlanRecord{}, fmt.Errorf("clone chunk plan producer metadata: %w", err)
}
record.Producer.Metadata = metadata
record.Warnings = cloneWarnings(record.Warnings)
return record, nil
}
func applyChunkPlanExecution(output *RunOutput, result chunkPlanExecution) error {
if output == nil {
return nil
}
summary := result.summary
output.ChunkPlan = &summary
if output.Manifest.ChunkPlan == nil {
return nil
}
manifest := output.Manifest.ChunkPlan
manifest.Action = result.action
if result.record == nil {
return nil
}
record := result.record
manifest.SourceDigest = record.SourceDigest
manifest.PlanDigest = record.PlanDigest
manifest.PlanSchemaVersion = record.SchemaVersion
manifest.ProducerInputModule = record.Producer.InputModule
manifest.ProducerModule = record.Producer.ChunkModule
manifest.ProducerLLMProfile = record.Producer.LLMProfile
manifest.ProducerReferences = append([]artifacts.ReferenceProvenance(nil), record.Producer.References...)
metadata, err := cloneMetadata(record.Producer.Metadata)
if err != nil {
return fmt.Errorf("clone chunk plan manifest producer metadata: %w", err)
}
manifest.ProducerMetadata = metadata
createdAt := record.CreatedAt
manifest.CreatedAt = &createdAt
return nil
}