Files
notarius/internal/framework/semanticreconcile/preparation_test.go

387 lines
16 KiB
Go

package semanticreconcile
import (
"crypto/sha256"
"encoding/hex"
"encoding/json"
"math"
"reflect"
"strings"
"testing"
"gitea.maximumdirect.net/eric/notarius/internal/core/source"
)
func TestPrepareBuildsContiguousCandidatesAndOwnedSourceContext(t *testing.T) {
document := &source.SourceDocument{ID: "private-source-id", Units: []source.SourceUnit{
{ID: 40, Kind: "narration", Text: "zero"},
{ID: 10, Kind: "speech", Text: "one", Metadata: map[string]any{"speaker": map[string]any{"name": "Mira"}}},
{ID: 70, Kind: "speech", Text: "two"},
{ID: 20, Kind: "narration", Text: "three"},
{ID: 90, Kind: "speech", Text: "four"},
}}
references := []source.SourceRef{
{SourceID: document.ID, StartUnitID: 90, EndUnitID: 90},
{SourceID: document.ID, StartUnitID: 10, EndUnitID: 20},
{SourceID: document.ID, StartUnitID: 10, EndUnitID: 20},
}
candidates := []Candidate{
{Label: "The Tavern", SourceRefs: append([]source.SourceRef(nil), references...)},
{Label: "The Tavern", SourceRefs: append([]source.SourceRef(nil), references...)},
{Label: "Broken", SourceRefs: []source.SourceRef{{SourceID: document.ID, StartUnitID: 20, EndUnitID: 10}}},
}
before := cloneCandidates(candidates)
preparation, err := Prepare(document, candidates, Limits{
ContextRadius: 1,
MaximumCandidates: len(candidates),
MaximumMaterialBytes: 10000,
})
if err != nil {
t.Fatal(err)
}
if preparation.Disposition() != Ready {
t.Fatalf("Disposition() = %v, want Ready", preparation.Disposition())
}
if !reflect.DeepEqual(candidates, before) {
t.Fatalf("Prepare() mutated candidates: %#v", candidates)
}
if got, want := preparation.CandidateMappings(), []CandidateMapping{
{CandidateID: 1, CandidatePosition: 0},
{CandidateID: 2, CandidatePosition: 1},
}; !reflect.DeepEqual(got, want) {
t.Fatalf("CandidateMappings() = %#v, want %#v", got, want)
}
materials := preparation.Materials()
if len(materials) != 2 {
t.Fatalf("materials = %#v, want candidates and transcript", materials)
}
if _, ok := materials[candidateInputName]; !ok {
t.Fatal("candidate material is missing")
}
if _, ok := materials[transcriptInputName]; !ok {
t.Fatal("transcript material is missing")
}
for name, material := range materials {
if material.Name != name || material.MediaType != "application/json" || material.OriginURI != "" || material.SizeBytes != int64(len(material.Content)) {
t.Fatalf("material %q metadata = %#v", name, material)
}
digest := sha256.Sum256(material.Content)
if want := "sha256:" + hex.EncodeToString(digest[:]); material.Digest != want {
t.Fatalf("material %q digest = %q, want %q", name, material.Digest, want)
}
}
candidateContent := append([]byte(nil), materials[candidateInputName].Content...)
var candidatePayload candidateInput
if err := json.Unmarshal(candidateContent, &candidatePayload); err != nil {
t.Fatal(err)
}
wantCandidates := []visibleCandidate{
{CandidateID: 1, Label: "The Tavern", SourceRefs: []sourceRange{{StartUnitID: 10, EndUnitID: 20}, {StartUnitID: 90, EndUnitID: 90}}},
{CandidateID: 2, Label: "The Tavern", SourceRefs: []sourceRange{{StartUnitID: 10, EndUnitID: 20}, {StartUnitID: 90, EndUnitID: 90}}},
}
if !reflect.DeepEqual(candidatePayload.Candidates, wantCandidates) {
t.Fatalf("candidate payload = %#v, want %#v", candidatePayload.Candidates, wantCandidates)
}
var candidateObjects struct {
Candidates []map[string]json.RawMessage `json:"candidates"`
}
if err := json.Unmarshal(candidateContent, &candidateObjects); err != nil {
t.Fatal(err)
}
for _, candidate := range candidateObjects.Candidates {
if len(candidate) != 3 || candidate["candidate_id"] == nil || candidate["label"] == nil || candidate["source_refs"] == nil {
t.Fatalf("model-facing candidate fields = %#v", candidate)
}
}
combined := string(materials[candidateInputName].Content) + string(materials[transcriptInputName].Content)
for _, forbidden := range []string{document.ID, "application_entity_id", "private-entity-id"} {
if strings.Contains(combined, forbidden) {
t.Fatalf("model material leaked %q: %s", forbidden, combined)
}
}
var transcript transcriptInput
if err := json.Unmarshal(materials[transcriptInputName].Content, &transcript); err != nil {
t.Fatal(err)
}
if len(transcript.Windows) != 1 || len(transcript.Windows[0].Units) != len(document.Units) {
t.Fatalf("windows = %#v, want one coalesced source window", transcript.Windows)
}
for index, wantID := range []int{40, 10, 70, 20, 90} {
if transcript.Windows[0].Units[index].ID != wantID {
t.Fatalf("unit %d id = %d, want %d", index, transcript.Windows[0].Units[index].ID, wantID)
}
}
if transcript.Windows[0].Units[0].Cited {
t.Fatal("radius-only unit marked cited")
}
for index := 1; index < len(transcript.Windows[0].Units); index++ {
if !transcript.Windows[0].Units[index].Cited {
t.Fatalf("evidence unit %d was not marked cited", index)
}
}
candidates[0].SourceRefs[0].StartUnitID = 40
document.Units[1].Metadata["speaker"].(map[string]any)["name"] = "changed"
if got := preparation.Materials()[candidateInputName].Content; !reflect.DeepEqual(got, candidateContent) {
t.Fatalf("candidate material changed through caller input: %s", got)
}
var retained transcriptInput
if err := json.Unmarshal(preparation.Materials()[transcriptInputName].Content, &retained); err != nil {
t.Fatal(err)
}
if got := retained.Windows[0].Units[1].Metadata["speaker"].(map[string]any)["name"]; got != "Mira" {
t.Fatalf("retained metadata = %v, want Mira", got)
}
returnedMappings := preparation.CandidateMappings()
returnedMappings[0].CandidatePosition = 99
returnedMaterials := preparation.Materials()
candidateMaterial := returnedMaterials[candidateInputName]
candidateMaterial.Content[0] = '['
returnedMaterials[candidateInputName] = candidateMaterial
delete(returnedMaterials, transcriptInputName)
if preparation.CandidateMappings()[0].CandidatePosition != 0 || !json.Valid(preparation.Materials()[candidateInputName].Content) || len(preparation.Materials()) != 2 {
t.Fatal("preparation accessors exposed retained data")
}
}
func TestPrepareRedactsInvalidSourceMetadata(t *testing.T) {
const sensitiveKey = "sensitive-metadata-key"
document := &source.SourceDocument{ID: "private-source", Units: []source.SourceUnit{
{ID: 1, Text: "private transcript", Metadata: map[string]any{sensitiveKey: math.NaN()}},
{ID: 2, Text: "other private transcript"},
}}
candidates := []Candidate{
{Label: "Private One", SourceRefs: []source.SourceRef{{SourceID: document.ID, StartUnitID: 1, EndUnitID: 1}}},
{Label: "Private Two", SourceRefs: []source.SourceRef{{SourceID: document.ID, StartUnitID: 2, EndUnitID: 2}}},
}
_, err := Prepare(document, candidates, DefaultLimits())
if err == nil || !strings.Contains(err.Error(), "invalid source metadata") {
t.Fatalf("Prepare() error = %v, want redacted metadata failure", err)
}
for _, forbidden := range []string{sensitiveKey, document.ID, document.Units[0].Text, candidates[0].Label, "non-finite", "float64"} {
if strings.Contains(err.Error(), forbidden) {
t.Fatalf("Prepare() error leaked %q: %v", forbidden, err)
}
}
}
func TestPrepareFiltersUnsafeCandidatesAndCoalescesAdjacentWindows(t *testing.T) {
document := &source.SourceDocument{ID: "session", Units: []source.SourceUnit{
{ID: 9}, {ID: 3}, {ID: 8}, {ID: 1}, {ID: 7},
}}
candidates := []Candidate{
{Label: "One", SourceRefs: []source.SourceRef{{SourceID: document.ID, StartUnitID: 3, EndUnitID: 3}}},
{Label: "Two", SourceRefs: []source.SourceRef{{SourceID: document.ID, StartUnitID: 8, EndUnitID: 8}}},
{Label: "Missing", SourceRefs: []source.SourceRef{{SourceID: document.ID, StartUnitID: 99, EndUnitID: 99}}},
{Label: "Foreign", SourceRefs: []source.SourceRef{{SourceID: "other", StartUnitID: 1, EndUnitID: 1}}},
{Label: "No references"},
{Label: "Partly invalid", SourceRefs: []source.SourceRef{
{SourceID: document.ID, StartUnitID: 1, EndUnitID: 1},
{SourceID: document.ID, StartUnitID: 100, EndUnitID: 100},
}},
}
limits := Limits{ContextRadius: 0, MaximumCandidates: len(candidates), MaximumMaterialBytes: 10000}
preparation, err := Prepare(document, candidates, limits)
if err != nil || preparation.Disposition() != Ready {
t.Fatalf("Prepare() disposition = %v, error = %v", preparation.Disposition(), err)
}
if got, want := preparation.CandidateMappings(), []CandidateMapping{
{CandidateID: 1, CandidatePosition: 0},
{CandidateID: 2, CandidatePosition: 1},
}; !reflect.DeepEqual(got, want) {
t.Fatalf("CandidateMappings() = %#v, want %#v", got, want)
}
var transcript transcriptInput
if err := json.Unmarshal(preparation.Materials()[transcriptInputName].Content, &transcript); err != nil {
t.Fatal(err)
}
if len(transcript.Windows) != 1 || len(transcript.Windows[0].Units) != 2 || transcript.Windows[0].Units[0].ID != 3 || transcript.Windows[0].Units[1].ID != 8 {
t.Fatalf("windows = %#v, want adjacent source-order units coalesced", transcript.Windows)
}
oneCandidate, err := Prepare(document, candidates[:1], limits)
if err != nil {
t.Fatal(err)
}
if got, want := oneCandidate.CandidateMappings(), []CandidateMapping{{CandidateID: 1, CandidatePosition: 0}}; oneCandidate.Disposition() != InsufficientCandidates || !reflect.DeepEqual(got, want) || len(oneCandidate.Materials()) != 0 {
t.Fatalf("Prepare(one candidate) = disposition %v, mappings %#v, materials %#v", oneCandidate.Disposition(), got, oneCandidate.Materials())
}
nilPreparation, err := Prepare(nil, candidates, limits)
if err != nil {
t.Fatal(err)
}
if nilPreparation.Disposition() != InsufficientCandidates || len(nilPreparation.CandidateMappings()) != 0 || len(nilPreparation.Materials()) != 0 {
t.Fatalf("Prepare(nil) = disposition %v, mappings %#v, materials %#v", nilPreparation.Disposition(), nilPreparation.CandidateMappings(), nilPreparation.Materials())
}
}
func TestPrepareValidatesLimitsBeforeBuildingMaterials(t *testing.T) {
if err := DefaultLimits().Validate(); err != nil {
t.Fatalf("DefaultLimits().Validate() error = %v", err)
}
cycle := map[string]any{}
cycle["self"] = cycle
document := &source.SourceDocument{ID: "session", Units: []source.SourceUnit{
{ID: 1, Metadata: cycle}, {ID: 2},
}}
candidates := candidatesForEveryUnit(document)
tests := []struct {
name string
limits Limits
want string
}{
{name: "negative radius", limits: Limits{ContextRadius: -1, MaximumCandidates: 2, MaximumMaterialBytes: 100}, want: "radius"},
{name: "zero candidates", limits: Limits{ContextRadius: 0, MaximumCandidates: 0, MaximumMaterialBytes: 100}, want: "candidates"},
{name: "negative candidates", limits: Limits{ContextRadius: 0, MaximumCandidates: -1, MaximumMaterialBytes: 100}, want: "candidates"},
{name: "zero bytes", limits: Limits{ContextRadius: 0, MaximumCandidates: 2, MaximumMaterialBytes: 0}, want: "bytes"},
{name: "negative bytes", limits: Limits{ContextRadius: 0, MaximumCandidates: 2, MaximumMaterialBytes: -1}, want: "bytes"},
}
for _, test := range tests {
t.Run(test.name, func(t *testing.T) {
if _, err := Prepare(document, candidates, test.limits); err == nil || !strings.Contains(err.Error(), test.want) {
t.Fatalf("Prepare() error = %v, want %q validation", err, test.want)
}
})
}
}
func TestPrepareEnforcesCandidateLimitBeforeRenderingContext(t *testing.T) {
cycle := map[string]any{}
cycle["self"] = cycle
document := &source.SourceDocument{ID: "session", Units: []source.SourceUnit{
{ID: 1, Text: "one", Metadata: cycle},
{ID: 2, Text: "two"},
{ID: 3, Text: "three"},
}}
candidates := candidatesForEveryUnit(document)
limits := Limits{ContextRadius: 0, MaximumCandidates: 2, MaximumMaterialBytes: 10000}
exceeded, err := Prepare(document, candidates, limits)
if err != nil {
t.Fatalf("Prepare(over limit) error = %v; context should not be rendered", err)
}
if exceeded.Disposition() != LimitExceeded || len(exceeded.CandidateMappings()) != 3 || len(exceeded.Materials()) != 0 {
t.Fatalf("Prepare(over limit) = disposition %v, mappings %#v, materials %#v", exceeded.Disposition(), exceeded.CandidateMappings(), exceeded.Materials())
}
document.Units[0].Metadata = nil
exact, err := Prepare(document, candidates[:2], limits)
if err != nil || exact.Disposition() != Ready {
t.Fatalf("Prepare(at limit) = disposition %v, error %v", exact.Disposition(), err)
}
}
func TestPrepareStopsRenderingContextWhenMaterialLimitIsExceeded(t *testing.T) {
cycle := map[string]any{}
cycle["self"] = cycle
document := &source.SourceDocument{ID: "session", Units: []source.SourceUnit{
{ID: 1, Text: strings.Repeat("oversized", 100)},
{ID: 2, Text: "must not be inspected"},
}}
candidates := candidatesForEveryUnit(document)
base, err := Prepare(document, candidates, Limits{
ContextRadius: 0,
MaximumCandidates: len(candidates),
MaximumMaterialBytes: 10000,
})
if err != nil || base.Disposition() != Ready {
t.Fatalf("Prepare(base) = disposition %v, error %v", base.Disposition(), err)
}
candidateBytes := len(base.Materials()[candidateInputName].Content)
document.Units[1].Metadata = cycle
limited, err := Prepare(document, candidates, Limits{
ContextRadius: 0,
MaximumCandidates: len(candidates),
MaximumMaterialBytes: candidateBytes + 64,
})
if err != nil {
t.Fatalf("Prepare(limited) error = %v; rendering should stop at the material bound", err)
}
if limited.Disposition() != LimitExceeded || len(limited.Materials()) != 0 {
t.Fatalf("Prepare(limited) = disposition %v, materials %#v, want bounded skip", limited.Disposition(), limited.Materials())
}
}
func TestPrepareAcceptsExactCombinedByteLimitAndSkipsOneOver(t *testing.T) {
document := &source.SourceDocument{ID: "session", Units: []source.SourceUnit{
{ID: 1, Text: "one"},
{ID: 2, Text: "two"},
}}
candidates := candidatesForEveryUnit(document)
baseLimits := Limits{ContextRadius: 0, MaximumCandidates: len(candidates), MaximumMaterialBytes: 10000}
base, err := Prepare(document, candidates, baseLimits)
if err != nil || base.Disposition() != Ready {
t.Fatalf("Prepare(base) = disposition %v, error %v", base.Disposition(), err)
}
materials := base.Materials()
totalBytes := len(materials[candidateInputName].Content) + len(materials[transcriptInputName].Content)
exactLimits := baseLimits
exactLimits.MaximumMaterialBytes = totalBytes
exact, err := Prepare(document, candidates, exactLimits)
if err != nil || exact.Disposition() != Ready {
t.Fatalf("Prepare(exact bytes) = disposition %v, error %v", exact.Disposition(), err)
}
oneOverLimits := exactLimits
oneOverLimits.MaximumMaterialBytes--
oneOver, err := Prepare(document, candidates, oneOverLimits)
if err != nil || oneOver.Disposition() != LimitExceeded || len(oneOver.Materials()) != 0 {
t.Fatalf("Prepare(one over) = disposition %v, materials %#v, error %v", oneOver.Disposition(), oneOver.Materials(), err)
}
}
func TestPrepareSerializationIsDeterministic(t *testing.T) {
document := &source.SourceDocument{ID: "session", Units: []source.SourceUnit{
{ID: 5, Text: "five", Metadata: map[string]any{"z": 1, "a": []any{"first", "second"}}},
{ID: 2, Text: "two", Metadata: map[string]any{"nested": map[string]any{"b": true, "a": false}}},
}}
candidates := candidatesForEveryUnit(document)
limits := Limits{ContextRadius: 0, MaximumCandidates: len(candidates), MaximumMaterialBytes: 10000}
first, err := Prepare(document, candidates, limits)
if err != nil {
t.Fatal(err)
}
second, err := Prepare(document, cloneCandidates(candidates), limits)
if err != nil {
t.Fatal(err)
}
for _, name := range []string{candidateInputName, transcriptInputName} {
firstMaterial := first.Materials()[name]
secondMaterial := second.Materials()[name]
if !reflect.DeepEqual(firstMaterial, secondMaterial) {
t.Fatalf("material %q is not deterministic:\n%#v\n%#v", name, firstMaterial, secondMaterial)
}
}
}
func candidatesForEveryUnit(document *source.SourceDocument) []Candidate {
candidates := make([]Candidate, len(document.Units))
for index, unit := range document.Units {
candidates[index] = Candidate{
Label: "candidate",
SourceRefs: []source.SourceRef{{
SourceID: document.ID,
StartUnitID: unit.ID,
EndUnitID: unit.ID,
}},
}
}
return candidates
}
func cloneCandidates(candidates []Candidate) []Candidate {
cloned := append([]Candidate(nil), candidates...)
for index := range cloned {
cloned[index].SourceRefs = append([]source.SourceRef(nil), candidates[index].SourceRefs...)
}
return cloned
}