Cyrene_v2/internal/runtime/tool/coerce.go

package tool

import (
	"encoding/json"
	"math"
	"regexp"
	"strconv"
	"strings"
)

// maxJSONCoerceLen caps how long a string we will attempt to JSON-parse when
// coercing a model-emitted string into an array or object. It mirrors the cap
// OpenClaw uses so a huge string argument fails fast instead of being parsed.
const maxJSONCoerceLen = 64 * 1024

// jsonNumberRe matches a JSON number token; used to decide whether a model
// string such as "3.5" may be safely coerced to a number.
var jsonNumberRe = regexp.MustCompile(`^[+-]?(?:\d+\.?\d*|\.\d+)(?:[eE][+-]?\d+)?$`)

// coerceWithSchema mirrors OpenClaw's coerceWithJsonSchema: it walks the
// schema's allOf/anyOf/oneOf, coerces a value whose type does not yet match,
// then recurses into object properties and array items. It returns the value it
// was given (mutating maps/slices in place) so validation sees coerced types.
func coerceWithSchema(value any, schema map[string]any) any {
	next := value

	if arr, ok := schema["allOf"].([]any); ok {
		for _, raw := range arr {
			if s, ok := raw.(map[string]any); ok {
				next = coerceWithSchema(next, s)
			}
		}
	}
	if arr, ok := schema["anyOf"].([]any); ok {
		next = coerceWithUnion(next, arr)
	}
	if arr, ok := schema["oneOf"].([]any); ok {
		next = coerceWithUnion(next, arr)
	}

	types := schemaTypes(schema)
	if len(types) > 0 && !matchesAnyType(next, types) {
		for _, t := range types {
			if coerced, ok := coercePrimitiveByType(next, t); ok {
				next = coerced
				break
			}
		}
	}

	if isObject(next) {
		applyObjectCoercion(next.(map[string]any), schema)
	}
	if isArray(next) {
		applyArrayCoercion(next.([]any), schema)
	}
	return next
}

// coerceWithUnion tries each union member in turn, cloning the value and
// coercing against that member, returning the first candidate that validates.
// A null value is returned as-is when some member accepts null so it is not
// mangled into the empty string by a string branch.
func coerceWithUnion(value any, schemas []any) any {
	if value == nil {
		for _, raw := range schemas {
			s, ok := raw.(map[string]any)
			if !ok || !contains(schemaTypes(s), "null") {
				continue
			}
			if compiled, err := compileSchema(s); err == nil && compiled.Validate(value) == nil {
				return value
			}
		}
	}
	for _, raw := range schemas {
		s, ok := raw.(map[string]any)
		if !ok {
			continue
		}
		candidate := coerceWithSchema(deepCopy(value), s)
		if compiled, err := compileSchema(s); err == nil && compiled.Validate(candidate) == nil {
			return candidate
		}
	}
	return value
}

// applyObjectCoercion recurses into declared properties and, when
// additionalProperties is itself a schema, into undeclared keys.
func applyObjectCoercion(value map[string]any, schema map[string]any) {
	props, _ := schema["properties"].(map[string]any)
	for name, raw := range props {
		ps, ok := raw.(map[string]any)
		if !ok {
			continue
		}
		if val, present := value[name]; present {
			value[name] = coerceWithSchema(val, ps)
		}
	}
	if ap, ok := schema["additionalProperties"].(map[string]any); ok {
		for name, val := range value {
			if _, defined := props[name]; !defined {
				value[name] = coerceWithSchema(val, ap)
			}
		}
	}
}

// applyArrayCoercion recurses into array items, handling the schematic,
// tuple-in-items and 2020-12 prefixItems forms.
func applyArrayCoercion(value []any, schema map[string]any) {
	switch items := schema["items"].(type) {
	case map[string]any:
		for i := range value {
			value[i] = coerceWithSchema(value[i], items)
		}
	case []any:
		for i, item := range items {
			if i >= len(value) {
				break
			}
			if is, ok := item.(map[string]any); ok {
				value[i] = coerceWithSchema(value[i], is)
			}
		}
	}
	if prefix, ok := schema["prefixItems"].([]any); ok {
		for i, item := range prefix {
			if i >= len(value) {
				break
			}
			if is, ok := item.(map[string]any); ok {
				value[i] = coerceWithSchema(value[i], is)
			}
		}
	}
}

// coercePrimitiveByType maps a requested JSON type to its coercion helper; when
// no helper can apply it returns the value unchanged.
func coercePrimitiveByType(value any, typ string) (any, bool) {
	switch typ {
	case "number":
		return coerceNumber(value)
	case "integer":
		return coerceInteger(value)
	case "boolean":
		return coerceBoolean(value)
	case "string":
		return coerceString(value)
	case "array":
		return coerceArray(value)
	case "object":
		return coerceObject(value)
	case "null":
		return coerceNull(value)
	default:
		return value, false
	}
}

func coerceNumber(value any) (any, bool) {
	switch v := value.(type) {
	case nil:
		return 0, true
	case string:
		if trimmed := strings.TrimSpace(v); trimmed != "" {
			if n, ok := parseNumber(trimmed); ok {
				return n, true
			}
		}
	case bool:
		if v {
			return 1, true
		}
		return 0, true
	}
	return value, false
}

func coerceInteger(value any) (any, bool) {
	switch v := value.(type) {
	case nil:
		return 0, true
	case string:
		if trimmed := strings.TrimSpace(v); trimmed != "" {
			if n, ok := parseNumber(trimmed); ok && math.Trunc(n) == n {
				return n, true
			}
		}
	case bool:
		if v {
			return 1, true
		}
		return 0, true
	}
	return value, false
}

func coerceBoolean(value any) (any, bool) {
	switch v := value.(type) {
	case nil:
		return false, true
	case string:
		if v == "true" {
			return true, true
		}
		if v == "false" {
			return false, true
		}
	case float64:
		if v == 1 {
			return true, true
		}
		if v == 0 {
			return false, true
		}
	}
	return value, false
}

func coerceString(value any) (any, bool) {
	switch v := value.(type) {
	case nil:
		return "", true
	case bool:
		return strconv.FormatBool(v), true
	case float64:
		return strconv.FormatFloat(v, 'f', -1, 64), true
	case float32:
		return strconv.FormatFloat(float64(v), 'f', -1, 32), true
	case int:
		return strconv.Itoa(v), true
	case int64:
		return strconv.FormatInt(v, 10), true
	case json.Number:
		return v.String(), true
	}
	return value, false
}

func coerceArray(value any) (any, bool) {
	s, ok := value.(string)
	if !ok || len(s) > maxJSONCoerceLen || strings.TrimSpace(s) == "" {
		return value, false
	}
	var parsed any
	if err := json.Unmarshal([]byte(s), &parsed); err != nil {
		return value, false
	}
	if arr, ok := parsed.([]any); ok {
		return arr, true
	}
	return value, false
}

func coerceObject(value any) (any, bool) {
	s, ok := value.(string)
	if !ok || len(s) > maxJSONCoerceLen || strings.TrimSpace(s) == "" {
		return value, false
	}
	var parsed any
	if err := json.Unmarshal([]byte(s), &parsed); err != nil {
		return value, false
	}
	if obj, ok := parsed.(map[string]any); ok {
		return obj, true
	}
	return value, false
}

func coerceNull(value any) (any, bool) {
	switch v := value.(type) {
	case string:
		if strings.TrimSpace(v) == "" {
			return nil, true
		}
	case float64:
		if v == 0 {
			return nil, true
		}
	case bool:
		if !v {
			return nil, true
		}
	}
	return value, false
}

// parseNumber parses a strictly JSON-shaped numeric token (no hex, Inf, NaN).
func parseNumber(s string) (float64, bool) {
	if !jsonNumberRe.MatchString(s) {
		return 0, false
	}
	f, err := strconv.ParseFloat(s, 64)
	if err != nil || math.IsInf(f, 0) || math.IsNaN(f) {
		return 0, false
	}
	return f, true
}

// schemaTypes returns the type names declared by a schema (string or union).
func schemaTypes(schema map[string]any) []string {
	switch t := schema["type"].(type) {
	case string:
		return []string{t}
	case []any:
		out := make([]string, 0, len(t))
		for _, item := range t {
			if s, ok := item.(string); ok {
				out = append(out, s)
			}
		}
		return out
	default:
		return nil
	}
}

// matchesAnyType reports whether value already satisfies one of the type names.
func matchesAnyType(value any, types []string) bool {
	for _, t := range types {
		if matchesJSONType(value, t) {
			return true
		}
	}
	return false
}

// matchesJSONType reports whether value satisfies a single JSON type name.
func matchesJSONType(value any, typ string) bool {
	switch typ {
	case "number":
		return isNumber(value)
	case "integer":
		return isInteger(value)
	case "boolean":
		_, ok := value.(bool)
		return ok
	case "string":
		_, ok := value.(string)
		return ok
	case "null":
		return value == nil
	case "array":
		return isArray(value)
	case "object":
		return isObject(value)
	default:
		return false
	}
}

// isObject reports whether v is a JSON object (map[string]any).
func isObject(v any) bool {
	_, ok := v.(map[string]any)
	return ok
}

// isArray reports whether v is a JSON array ([]any).
func isArray(v any) bool {
	_, ok := v.([]any)
	return ok
}

// contains reports whether slice has s.
func contains(slice []string, s string) bool {
	for _, item := range slice {
		if item == s {
			return true
		}
	}
	return false
}

// isNumber reports whether v is any integer or floating-point numeric type.
func isNumber(v any) bool {
	switch v.(type) {
	case json.Number, float32, float64,
		int, int8, int16, int32, int64,
		uint, uint8, uint16, uint32, uint64:
		return true
	default:
		return false
	}
}

// isInteger reports whether v is an integer type or a floating-point / json
// Number that holds an integral value.
func isInteger(v any) bool {
	switch x := v.(type) {
	case int, int8, int16, int32, int64,
		uint, uint8, uint16, uint32, uint64:
		return true
	case float32:
		return math.Trunc(float64(x)) == float64(x)
	case float64:
		return math.Trunc(x) == x
	case json.Number:
		_, err := strconv.ParseInt(x.String(), 10, 64)
		return err == nil
	default:
		return false
	}
}