package validators import ( "encoding/json" "strings" "testing" "gitea.maximumdirect.net/eric/audita/v3/internal/core/chunking" "gitea.maximumdirect.net/eric/audita/v3/internal/core/glossary" "gitea.maximumdirect.net/eric/audita/v3/internal/core/schema" "gitea.maximumdirect.net/eric/audita/v3/internal/framework/proposals" ) func buildContextItems(req Request) ([]LLMValidationItem, []Decision) { req.Section = req.sectionLookup() items, immediate := buildValidationItems(req) context := validationContextFor(req) for i := range items { items[i].NeighboringSegments, items[i].GlossaryEntries = context(items[i]) } return items, immediate } func TestValidationContextCarriesAdjacentEvidenceAndRelevantDefinitions(t *testing.T) { req := makeReq([]proposals.EnrichedCorrectionProposal{mk(7, "gestures", "Jesters")}) req.WorkingTranscript = &schema.Transcript{Segments: []schema.Segment{ {ID: 1, Speaker: "Host", Text: "The performers arrived."}, {ID: 2, Speaker: "Guest", Text: "Those gestures are back."}, {ID: 3, Speaker: "Host", Text: "Lyra recognized the troupe."}, {ID: 4, Text: "Elsewhere."}, }} req.CandidateProposal[0].TargetSegmentID = 2 req.Glossary = []glossary.Entry{ {ID: "jesters", Name: "Jesters", Description: "A traveling troupe"}, {ID: "lyra", Name: "Lyra", Description: "The host"}, {ID: "unrelated", Name: "Unrelated", Description: "Not mentioned"}, } payload, immediate := buildContextItems(req) if len(immediate) != 0 || len(payload) != 1 { t.Fatalf("payload: %+v immediate: %+v", payload, immediate) } item := payload[0] if len(item.NeighboringSegments) != 2 || item.NeighboringSegments[0].Position != "preceding" || item.NeighboringSegments[1].Position != "following" || len(item.GlossaryEntries) != 2 || item.GlossaryEntries[0].Name != "Jesters" || item.GlossaryEntries[1].Name != "Lyra" { t.Fatalf("context: %+v", item) } variables, err := BuildValidationVariables(payload, "") if err != nil { t.Fatal(err) } var rendered []LLMValidationItem if err := json.Unmarshal([]byte(variables["PayloadJSON"]), &rendered); err != nil { t.Fatal(err) } if strings.Contains(variables["PayloadJSON"], `"segment_id":`) || rendered[0].CorrectionIndex != 1 || payload[0].CorrectionIndex != 7 { t.Fatalf("leaked internal identity or mutated request: %s", variables["PayloadJSON"]) } if rendered[0].Speaker != "Guest" || len(rendered[0].NeighboringSegments) != 2 || rendered[0].NeighboringSegments[0].Speaker != "Host" || rendered[0].GlossaryEntries[0].Description != "A traveling troupe" { t.Fatalf("rendered context: %+v", rendered) } // The context-free endpoint must not reach beyond the current chunk. req.WorkingTranscript.Segments = req.WorkingTranscript.Segments[1:2] req.WorkingTranscript.Segments[0].Speaker = "" req.Glossary = nil isolated, _ := buildContextItems(req) if isolated[0].Speaker != "" || len(isolated[0].NeighboringSegments) != 0 || len(isolated[0].GlossaryEntries) != 0 { t.Fatalf("unexpected external context: %+v", isolated) } } func TestValidationContextFitsBudgetWithoutDiscardingCandidate(t *testing.T) { estimator := chunking.NewSimpleTokenEstimator() base := LLMValidationItem{CorrectionIndex: 7, OriginalText: "teh", CorrectedText: "the", OriginalSegmentText: "teh cat", CorrectedSegmentText: "the cat"} required, err := estimateBatchTokens(estimator, []LLMValidationItem{base}) if err != nil { t.Fatal(err) } base.Speaker = "Guest speaker with a long identifying label" withSpeaker, err := estimateBatchTokens(estimator, []LLMValidationItem{base}) if err != nil || withSpeaker <= required { t.Fatalf("speaker missing from token accounting: before=%d after=%d err=%v", required, withSpeaker, err) } required = withSpeaker enriched := base enriched.NeighboringSegments = []NeighboringSegment{{Position: "following", Text: strings.Repeat("long context ", 300)}} enriched.GlossaryEntries = []glossary.Entry{{Name: "cat", Description: strings.Repeat("long definition ", 300)}} fit, err := fitValidationContext(enriched, required, estimator, nil) if err != nil { t.Fatal(err) } fitted := fit.item tokens, err := estimateBatchTokens(estimator, []LLMValidationItem{fitted}) if err != nil || tokens > required || fitted.Speaker != base.Speaker || fitted.OriginalSegmentText != base.OriginalSegmentText || fitted.CorrectedSegmentText != base.CorrectedSegmentText { t.Fatalf("fitted: %+v tokens=%d err=%v", fitted, tokens, err) } full, _ := estimateBatchTokens(estimator, []LLMValidationItem{enriched}) batches, err := ChunkLLMValidationItems([]LLMValidationItem{enriched, enriched}, full, estimator) if err != nil || len(batches) != 2 { t.Fatalf("context must count toward batch budget: %+v %v", batches, err) } } func TestValidationContextSelectsGlossaryForEachReplacement(t *testing.T) { req := makeReq([]proposals.EnrichedCorrectionProposal{mk(7, "gestures", "Jesters"), mk(11, "gestures", "Actors")}) req.WorkingTranscript = &schema.Transcript{Segments: []schema.Segment{ {ID: 1, Text: "Those gestures are back."}, {ID: 2, Text: "They met Lyra."}, }} req.Glossary = []glossary.Entry{{ID: "lyra", Name: "Lyra"}, {ID: "actors", Name: "Actors"}, {ID: "jesters", Name: "Jesters"}} payload, immediate := buildContextItems(req) if len(immediate) != 0 || len(payload) != 2 { t.Fatalf("payload: %+v immediate: %+v", payload, immediate) } for i, name := range []string{"Jesters", "Actors"} { entries := payload[i].GlossaryEntries if len(entries) != 2 || entries[0].Name != "Lyra" || entries[1].Name != name { t.Fatalf("candidate %d context: %+v", i, entries) } } }