feat: validate svglide stage artifacts

2026-07-08 01:51:27 +08:00 · 2026-07-03 02:09:17 +08:00
parent 15e7ab8b66
commit d4e074a494
3 changed files with 401 additions and 0 deletions
--- a/internal/svglide/schema.go
+++ b/internal/svglide/schema.go
@@ -0,0 +1,277 @@
+package svglide
+
+import (
+	"bytes"
+	"encoding/json"
+	"fmt"
+	"io"
+	"math/big"
+	"path/filepath"
+	"regexp"
+	"strings"
+)
+
+type liteJSONSchema struct {
+	Type                 string                    `json:"type"`
+	Required             []string                  `json:"required"`
+	AdditionalProperties *bool                     `json:"additionalProperties"`
+	Properties           map[string]liteJSONSchema `json:"properties"`
+	Items                *liteJSONSchema           `json:"items"`
+	MinItems             *int                      `json:"minItems"`
+	Enum                 []string                  `json:"enum"`
+	Pattern              string                    `json:"pattern"`
+}
+
+var stageOutputSchemaPaths = map[string]string{
+	"request/request.json":         "schemas/request.schema.json",
+	"request/source_manifest.json": "schemas/source_manifest.schema.json",
+	"research/sources.json":        "schemas/sources.schema.json",
+	"brief/design_brief.json":      "schemas/design_brief.schema.json",
+	"brief/visual_system.json":     "schemas/visual_system.schema.json",
+	"outline/deck.json":            "schemas/deck.schema.json",
+	"content/slide_content.json":   "schemas/slide_content.schema.json",
+	"assets/assets_plan.json":      "schemas/assets_plan.schema.json",
+	"receipts/lint.json":           "schemas/lint.schema.json",
+	"receipts/preview.json":        "schemas/preview.schema.json",
+}
+
+func ValidateStageOutputs(root string) error {
+	safeRoot, run, err := readRun(root)
+	if err != nil {
+		return err
+	}
+	stage, err := currentStage(run)
+	if err != nil {
+		return err
+	}
+	for _, output := range stage.Outputs {
+		if hasGlobMeta(output) || strings.ToLower(filepath.Ext(output)) != ".json" {
+			continue
+		}
+		schemaPath, ok := stageOutputSchemaPaths[output]
+		if !ok {
+			continue
+		}
+		if err := validateStageOutputSchema(safeRoot, output, schemaPath); err != nil {
+			return err
+		}
+	}
+	return nil
+}
+
+func validateStageOutputSchema(safeRoot, artifactPath, schemaPath string) error {
+	artifactRaw, err := readRunRegularArtifact(safeRoot, artifactPath)
+	if err != nil {
+		return fmt.Errorf("%s: read artifact: %w", artifactPath, err)
+	}
+	schemaRaw, err := readRunRegularArtifact(safeRoot, schemaPath)
+	if err != nil {
+		return fmt.Errorf("%s: read schema %s: %w", artifactPath, schemaPath, err)
+	}
+	schema, err := decodeLiteJSONSchema(schemaRaw)
+	if err != nil {
+		return fmt.Errorf("%s: schema %s: %w", artifactPath, schemaPath, err)
+	}
+	value, err := decodeJSONValue(artifactRaw)
+	if err != nil {
+		return fmt.Errorf("%s: invalid JSON: %w", artifactPath, err)
+	}
+	if err := validateJSONValue(schema, value, ""); err != nil {
+		return fmt.Errorf("%s: %w", artifactPath, err)
+	}
+	return nil
+}
+
+func decodeLiteJSONSchema(raw []byte) (liteJSONSchema, error) {
+	var schema liteJSONSchema
+	decoder := json.NewDecoder(bytes.NewReader(raw))
+	if err := decoder.Decode(&schema); err != nil {
+		return liteJSONSchema{}, fmt.Errorf("invalid JSON: %w", err)
+	}
+	if err := rejectTrailingJSON(decoder); err != nil {
+		return liteJSONSchema{}, err
+	}
+	return schema, nil
+}
+
+func decodeJSONValue(raw []byte) (any, error) {
+	decoder := json.NewDecoder(bytes.NewReader(raw))
+	decoder.UseNumber()
+	var value any
+	if err := decoder.Decode(&value); err != nil {
+		return nil, err
+	}
+	if err := rejectTrailingJSON(decoder); err != nil {
+		return nil, err
+	}
+	return value, nil
+}
+
+func rejectTrailingJSON(decoder *json.Decoder) error {
+	var extra any
+	if err := decoder.Decode(&extra); err != io.EOF {
+		if err == nil {
+			return fmt.Errorf("contains trailing JSON value")
+		}
+		return err
+	}
+	return nil
+}
+
+func validateJSONValue(schema liteJSONSchema, value any, fieldPath string) error {
+	switch schema.Type {
+	case "":
+		return nil
+	case "object":
+		return validateJSONObject(schema, value, fieldPath)
+	case "array":
+		return validateJSONArray(schema, value, fieldPath)
+	case "string":
+		return validateJSONString(schema, value, fieldPath)
+	case "integer":
+		if !isJSONInteger(value) {
+			return fmt.Errorf("field %s expected integer, got %s", displayFieldPath(fieldPath), jsonValueType(value))
+		}
+		return nil
+	case "boolean":
+		if _, ok := value.(bool); !ok {
+			return fmt.Errorf("field %s expected boolean, got %s", displayFieldPath(fieldPath), jsonValueType(value))
+		}
+		return nil
+	default:
+		return fmt.Errorf("field %s uses unsupported schema type %q", displayFieldPath(fieldPath), schema.Type)
+	}
+}
+
+func validateJSONObject(schema liteJSONSchema, value any, fieldPath string) error {
+	object, ok := value.(map[string]any)
+	if !ok {
+		return fmt.Errorf("field %s expected object, got %s", displayFieldPath(fieldPath), jsonValueType(value))
+	}
+	for _, required := range schema.Required {
+		if _, ok := object[required]; !ok {
+			return fmt.Errorf("field %s is required", joinFieldPath(fieldPath, required))
+		}
+	}
+	if schema.AdditionalProperties != nil && !*schema.AdditionalProperties {
+		for name := range object {
+			if _, ok := schema.Properties[name]; !ok {
+				return fmt.Errorf("field %s is not allowed by additionalProperties:false", joinFieldPath(fieldPath, name))
+			}
+		}
+	}
+	for name, propertySchema := range schema.Properties {
+		child, ok := object[name]
+		if !ok {
+			continue
+		}
+		if err := validateJSONValue(propertySchema, child, joinFieldPath(fieldPath, name)); err != nil {
+			return err
+		}
+	}
+	return nil
+}
+
+func validateJSONArray(schema liteJSONSchema, value any, fieldPath string) error {
+	array, ok := value.([]any)
+	if !ok {
+		return fmt.Errorf("field %s expected array, got %s", displayFieldPath(fieldPath), jsonValueType(value))
+	}
+	if schema.MinItems != nil && len(array) < *schema.MinItems {
+		return fmt.Errorf("field %s has %d items, want minItems %d", displayFieldPath(fieldPath), len(array), *schema.MinItems)
+	}
+	if schema.Items == nil {
+		return nil
+	}
+	for i, item := range array {
+		if err := validateJSONValue(*schema.Items, item, joinArrayFieldPath(fieldPath, i)); err != nil {
+			return err
+		}
+	}
+	return nil
+}
+
+func validateJSONString(schema liteJSONSchema, value any, fieldPath string) error {
+	text, ok := value.(string)
+	if !ok {
+		return fmt.Errorf("field %s expected string, got %s", displayFieldPath(fieldPath), jsonValueType(value))
+	}
+	if len(schema.Enum) > 0 {
+		for _, allowed := range schema.Enum {
+			if text == allowed {
+				return nil
+			}
+		}
+		return fmt.Errorf("field %s value %q is not in enum %v", displayFieldPath(fieldPath), text, schema.Enum)
+	}
+	if schema.Pattern != "" {
+		matched, err := regexp.MatchString(schema.Pattern, text)
+		if err != nil {
+			return fmt.Errorf("field %s has invalid pattern %q: %w", displayFieldPath(fieldPath), schema.Pattern, err)
+		}
+		if !matched {
+			return fmt.Errorf("field %s value %q does not match pattern %q", displayFieldPath(fieldPath), text, schema.Pattern)
+		}
+	}
+	return nil
+}
+
+func isJSONInteger(value any) bool {
+	switch typed := value.(type) {
+	case json.Number:
+		return isCanonicalJSONInteger(typed.String())
+	case int, int8, int16, int32, int64, uint, uint8, uint16, uint32, uint64:
+		return true
+	default:
+		return false
+	}
+}
+
+func isCanonicalJSONInteger(value string) bool {
+	if value == "" || strings.ContainsAny(value, ".eE") {
+		return false
+	}
+	var parsed big.Int
+	_, ok := parsed.SetString(value, 10)
+	return ok
+}
+
+func jsonValueType(value any) string {
+	switch value.(type) {
+	case nil:
+		return "null"
+	case map[string]any:
+		return "object"
+	case []any:
+		return "array"
+	case string:
+		return "string"
+	case json.Number, float64:
+		return "number"
+	case bool:
+		return "boolean"
+	default:
+		return fmt.Sprintf("%T", value)
+	}
+}
+
+func joinFieldPath(parent, name string) string {
+	if parent == "" {
+		return name
+	}
+	return parent + "." + name
+}
+
+func joinArrayFieldPath(parent string, index int) string {
+	if parent == "" {
+		return fmt.Sprintf("[%d]", index)
+	}
+	return fmt.Sprintf("%s[%d]", parent, index)
+}
+
+func displayFieldPath(path string) string {
+	if path == "" {
+		return "$"
+	}
+	return path
+}
--- a/internal/svglide/schema_test.go
+++ b/internal/svglide/schema_test.go
@@ -0,0 +1,120 @@
+package svglide
+
+import (
+	"os"
+	"path/filepath"
+	"strings"
+	"testing"
+)
+
+func TestValidateStageOutputsRejectsMissingRequiredField(t *testing.T) {
+	initStatusTestRun(t)
+	if err := os.WriteFile(filepath.Join("demo", "request", "request.json"), []byte(`{"title":"Demo"}`), 0o644); err != nil {
+		t.Fatal(err)
+	}
+
+	err := ValidateStageOutputs("demo")
+	if err == nil {
+		t.Fatal("expected schema validation error")
+	}
+	if !strings.Contains(err.Error(), "request/request.json") || !strings.Contains(err.Error(), "input") {
+		t.Fatalf("error = %v, want path and missing field", err)
+	}
+}
+
+func TestValidateStageOutputsAcceptsCurrentRequestArtifacts(t *testing.T) {
+	initStatusTestRun(t)
+
+	if err := ValidateStageOutputs("demo"); err != nil {
+		t.Fatal(err)
+	}
+}
+
+func TestValidateStageOutputsRejectsInvalidValidatePreviewRepairReceipts(t *testing.T) {
+	tests := []struct {
+		name  string
+		setup func(t *testing.T)
+		path  string
+	}{
+		{
+			name: "lint",
+			setup: func(t *testing.T) {
+				t.Helper()
+				if err := os.WriteFile(filepath.Join("demo", "receipts", "lint.json"), []byte(`{"status":"failed"}`), 0o644); err != nil {
+					t.Fatal(err)
+				}
+			},
+			path: "receipts/lint.json",
+		},
+		{
+			name: "preview",
+			setup: func(t *testing.T) {
+				t.Helper()
+				if err := os.WriteFile(filepath.Join("demo", "receipts", "lint.json"), []byte(`{"status":"passed","issues":[]}`), 0o644); err != nil {
+					t.Fatal(err)
+				}
+				if err := os.WriteFile(filepath.Join("demo", "receipts", "preview.json"), []byte(`{"status":"passed","slides":[{"path":"slides/01.svg","rendered":"yes"}]}`), 0o644); err != nil {
+					t.Fatal(err)
+				}
+			},
+			path: "receipts/preview.json",
+		},
+	}
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			initStatusTestRun(t)
+			setCurrentStageForStatusTest(t, StageValidatePreviewRepair)
+			tt.setup(t)
+
+			err := ValidateStageOutputs("demo")
+			if err == nil {
+				t.Fatal("expected schema validation error")
+			}
+			if !strings.Contains(err.Error(), tt.path) {
+				t.Fatalf("error = %v, want path %s", err, tt.path)
+			}
+		})
+	}
+}
+
+func TestValidateStageOutputsRejectsNonCanonicalIntegers(t *testing.T) {
+	for _, pages := range []string{"8.0", "8e0", "0.99999999999999999"} {
+		t.Run(pages, func(t *testing.T) {
+			initStatusTestRun(t)
+			raw := `{"title":"Demo","input":"source.md","pages":` + pages + `}`
+			if err := os.WriteFile(filepath.Join("demo", "request", "request.json"), []byte(raw), 0o644); err != nil {
+				t.Fatal(err)
+			}
+
+			err := ValidateStageOutputs("demo")
+			if err == nil {
+				t.Fatal("expected schema validation error")
+			}
+			if !strings.Contains(err.Error(), "request/request.json") || !strings.Contains(err.Error(), "pages") {
+				t.Fatalf("error = %v, want path and pages field", err)
+			}
+		})
+	}
+}
+
+func TestCompleteCurrentStageRejectsInvalidCurrentStageOutputSchema(t *testing.T) {
+	initStatusTestRun(t)
+	if err := os.WriteFile(filepath.Join("demo", "request", "source_manifest.json"), []byte(`{"sources":[{"path":"source.md","type":"remote"}]}`), 0o644); err != nil {
+		t.Fatal(err)
+	}
+
+	_, err := CompleteCurrentStage("demo")
+	if err == nil {
+		t.Fatal("expected schema validation error")
+	}
+	run := readStatusTestRunFile(t)
+	if run.CurrentStage != StageRequest {
+		t.Fatalf("run.CurrentStage = %q, want %q", run.CurrentStage, StageRequest)
+	}
+	if got := stageStatus(t, run, StageRequest); got == StatusDone {
+		t.Fatalf("request stage status = %q, want not %q", got, StatusDone)
+	}
+	if _, statErr := os.Stat(filepath.Join("demo", "receipts", "request.json")); !os.IsNotExist(statErr) {
+		t.Fatalf("receipt should not be written, stat err = %v", statErr)
+	}
+}
--- a/internal/svglide/stage.go
+++ b/internal/svglide/stage.go
@@ -31,6 +31,10 @@ func CompleteCurrentStage(root string) (StatusReport, error) {
 		return StatusReport{}, fmt.Errorf("current stage %q missing outputs: %s", stage.Name, strings.Join(missingOutputs, ", "))
 	}

+	if err := ValidateStageOutputs(root); err != nil {
+		return StatusReport{}, err
+	}
+
 	if err := writeStageReceipt(safeRoot, StageReceipt{
 		Stage:     stage.Name,
 		Status:    StatusDone,