audita/internal/framework/proposal_generation/module_proposal_test.go

package proposal_generation

import (
	"context"
	"encoding/json"
	"errors"
	"testing"

	"gitea.maximumdirect.net/eric/audita/v3/internal/core/config"
	"gitea.maximumdirect.net/eric/audita/v3/internal/core/schema"
	"gitea.maximumdirect.net/eric/audita/v3/internal/framework/contracts"
	"gitea.maximumdirect.net/eric/audita/v3/internal/framework/outcomes"
	"gitea.maximumdirect.net/eric/audita/v3/internal/framework/proposals"
	"gitea.maximumdirect.net/eric/audita/v3/internal/framework/structuredoutput"
	"gitea.maximumdirect.net/eric/audita/v3/internal/prompts"
)

func TestExecuteModuleProposalBuildsMessagesFromSectionAndDescription(t *testing.T) {
	client := &fakeStructuredClient{
		responses: []StructuredCorrectionSet{
			{Corrections: []StructuredCorrectionProposal{{SegmentReference: 1, OriginalText: "teh", CorrectedText: "the", Confidence: 0.9}}},
		},
	}
	section := contracts.SectionMetadata{Index: 7}
	cfg := config.Config{}
	cfg.Context.Description = "Hearing transcript with role titles."
	transcript := &schema.Transcript{Segments: []schema.Segment{{ID: 1, Speaker: "A", Start: 0, End: 1, Text: "teh"}}}
	glossary := &schema.Glossary{Entries: []schema.GlossaryEntry{{Name: "X"}}}

	var gotSectionIndex int
	var gotDescription string
	var gotTranscript *schema.Transcript
	var gotGlossary *schema.Glossary

	out, err := ExecuteModuleProposal(context.Background(), ModuleProposalRequest{
		ProposalRequest: contracts.ProposalRequest{
			ExecutionContext: contracts.ExecutionContext{
				Config:            &cfg,
				WorkingTranscript: transcript,
				Glossary:          glossary,
				Section:           &section,
			},
			RunSpec: contracts.ModuleRunSpec{
				ModuleKey:         "grammar",
				InstanceName:      "grammar",
				ReplacementPolicy: proposals.ReplacementPolicyRequireUnique,
			},
			LLMClient: client,
		},
		PromptID: prompts.PromptIDModuleGrammarProposal,
		BuildVariables: func(inTranscript *schema.Transcript, inGlossary *schema.Glossary, sectionIndex int, transcriptDescription string) (map[string]string, error) {
			gotSectionIndex = sectionIndex
			gotDescription = transcriptDescription
			gotTranscript = inTranscript
			gotGlossary = inGlossary
			return map[string]string{"SectionJSON": "[]"}, nil
		},
	})
	if err != nil {
		t.Fatalf("ExecuteModuleProposal error: %v", err)
	}
	if gotSectionIndex != 7 {
		t.Fatalf("section index: got=%d want=%d", gotSectionIndex, 7)
	}
	if gotDescription != cfg.Context.Description {
		t.Fatalf("transcript description: got=%q want=%q", gotDescription, cfg.Context.Description)
	}
	if gotTranscript != transcript {
		t.Fatalf("expected shared transcript pointer")
	}
	if gotGlossary != glossary {
		t.Fatalf("expected shared glossary pointer")
	}
	if len(client.calls) != 1 || client.calls[0].PromptID != prompts.PromptIDModuleGrammarProposal || client.calls[0].Context.SectionIndex == nil || *client.calls[0].Context.SectionIndex != 7 {
		t.Fatalf("unexpected stage name calls: %+v", client.calls)
	}
	if len(out.Proposals) != 1 || out.Proposals[0].CorrectedText != "the" {
		t.Fatalf("unexpected proposals: %+v", out)
	}
}

func TestExecuteModuleProposalValidatesInputs(t *testing.T) {
	if _, err := ExecuteModuleProposal(context.Background(), ModuleProposalRequest{}); err == nil {
		t.Fatalf("expected missing message builder error")
	}

	_, err := ExecuteModuleProposal(context.Background(), ModuleProposalRequest{
		BuildVariables: func(transcript *schema.Transcript, glossary *schema.Glossary, sectionIndex int, transcriptDescription string) (map[string]string, error) {
			return nil, nil
		},
	})
	if err == nil {
		t.Fatalf("expected empty prompt ID error")
	}

	_, err = ExecuteModuleProposal(context.Background(), ModuleProposalRequest{
		PromptID: "missing.prompt.id",
		BuildVariables: func(transcript *schema.Transcript, glossary *schema.Glossary, sectionIndex int, transcriptDescription string) (map[string]string, error) {
			return map[string]string{"SectionJSON": "[]"}, nil
		},
	})
	if err == nil {
		t.Fatalf("expected unknown prompt ID error")
	}
}

func TestExecuteModuleProposalPropagatesBuilderError(t *testing.T) {
	wantErr := errors.New("builder failed")
	_, err := ExecuteModuleProposal(context.Background(), ModuleProposalRequest{
		PromptID: prompts.PromptIDModuleGlossaryProposal,
		BuildVariables: func(transcript *schema.Transcript, glossary *schema.Glossary, sectionIndex int, transcriptDescription string) (map[string]string, error) {
			return nil, wantErr
		},
	})
	if !errors.Is(err, wantErr) {
		t.Fatalf("expected builder error, got %v", err)
	}
}

func TestExecuteModuleProposalPreservesFatalOperation(t *testing.T) {
	cause := errors.New("missing embedded asset")
	client := &fakeStructuredClient{err: &structuredoutput.Failure{Class: structuredoutput.FailureFatal, Reason: outcomes.ReasonSetupFailure, Message: "model setup failed", Cause: cause}}
	section := contracts.SectionMetadata{Index: 4}
	result, err := ExecuteModuleProposal(context.Background(), ModuleProposalRequest{
		ProposalRequest: contracts.ProposalRequest{
			ExecutionContext: contracts.ExecutionContext{Section: &section},
			RunSpec:          contracts.ModuleRunSpec{ModuleKey: "grammar", InstanceName: "grammar_2", ReplacementPolicy: proposals.ReplacementPolicyRequireUnique},
			LLMClient:        client,
		},
		PromptID: prompts.PromptIDModuleGrammarProposal,
		BuildVariables: func(*schema.Transcript, *schema.Glossary, int, string) (map[string]string, error) {
			return map[string]string{"SectionJSON": "[]"}, nil
		},
	})
	if !errors.Is(err, cause) || len(result.Operations) != 1 || result.Operations[0].Reference.ModuleInstance != "grammar_2" || result.Operations[0].Reference.SectionIndex != 4 {
		t.Fatalf("fatal operation lost: %+v, %v", result, err)
	}
}

func TestModuleProposalMapsLocalReferencesWithoutGuessing(t *testing.T) {
	transcript := &schema.Transcript{Segments: []schema.Segment{{ID: 71, Text: "teh cat"}, {ID: 903, Text: "uh hello"}}}
	client := &fakeStructuredClient{responses: []StructuredCorrectionSet{{Corrections: []StructuredCorrectionProposal{
		{SegmentReference: 2, OriginalText: "uh ", CorrectedText: "", Confidence: .9},
		{SegmentReference: 1, OriginalText: "teh", CorrectedText: "the", Confidence: .9},
		{SegmentReference: 2, OriginalText: "hello", CorrectedText: "Hello", Confidence: .9},
		{SegmentReference: 71, OriginalText: "teh", CorrectedText: "the", Confidence: .9},
	}}}}
	result, err := ExecuteModuleProposal(context.Background(), ModuleProposalRequest{
		ProposalRequest: contracts.ProposalRequest{
			ExecutionContext: contracts.ExecutionContext{WorkingTranscript: transcript},
			RunSpec:          contracts.ModuleRunSpec{ModuleKey: "grammar", InstanceName: "grammar", ReplacementPolicy: proposals.ReplacementPolicyRequireUnique},
			LLMClient:        client,
		},
		PromptID: prompts.PromptIDModuleGrammarProposal, BuildVariables: BuildProposalVariables,
	})
	if err != nil {
		t.Fatal(err)
	}
	var payload struct {
		Segments []struct {
			ID   int    `json:"id"`
			Text string `json:"text"`
		} `json:"segments"`
	}
	if err := json.Unmarshal([]byte(client.calls[0].Variables["SectionJSON"]), &payload); err != nil {
		t.Fatal(err)
	}
	if len(payload.Segments) != 2 || payload.Segments[0].ID != 1 || payload.Segments[1].ID != 2 || payload.Segments[1].Text != "uh hello" {
		t.Fatalf("request references: %+v", payload)
	}
	if len(result.Proposals) != 4 {
		t.Fatalf("proposals: %+v", result)
	}
	for i, id := range []int{903, 71, 903, 0} {
		if result.Proposals[i].TargetSegmentID != id {
			t.Fatalf("mapped proposal %d: %+v", i, result.Proposals[i])
		}
	}
	if result.Proposals[0].CorrectedText != "" || transcript.Segments[0].ID != 71 || transcript.Segments[1].ID != 903 {
		t.Fatal("mapping changed deletion or source identities")
	}
}