CrossRelay/internal/responses/types.go

// Package responses describes the OpenAI Responses API that Codex
// speaks and converts it to and from Chat Completions.
package responses

import (
	"bytes"
	"encoding/json"
	"strings"
)

// Request is the Responses API request body. Only the fields CrossRelay
// understands are modelled; everything else is ignored.
type Request struct {
	Model           string          `json:"model"`
	Instructions    string          `json:"instructions,omitempty"`
	Input           json.RawMessage `json:"input"`
	Tools           []Tool          `json:"tools,omitempty"`
	ToolChoice      json.RawMessage `json:"tool_choice,omitempty"`
	MaxOutputTokens int             `json:"max_output_tokens,omitempty"`
	Temperature     *float64        `json:"temperature,omitempty"`
	TopP            *float64        `json:"top_p,omitempty"`
	Stream          bool            `json:"stream,omitempty"`
}

// Tool is one tool definition. Type is "function" for regular tools;
// "custom" and "local_shell" are freeform forms Codex may send.
type Tool struct {
	Type        string          `json:"type"`
	Name        string          `json:"name"`
	Description string          `json:"description,omitempty"`
	Parameters  json.RawMessage `json:"parameters,omitempty"`
}

// Item is one entry of the input array. Only the fields belonging to
// Type are populated.
type Item struct {
	Type string `json:"type"` // message | function_call | function_call_output | reasoning
	Role string `json:"role,omitempty"`
	// message
	Content json.RawMessage `json:"content,omitempty"` // string or []ContentPart
	// function_call
	CallID    string `json:"call_id,omitempty"`
	Name      string `json:"name,omitempty"`
	Arguments string `json:"arguments,omitempty"`
	// function_call_output
	Output json.RawMessage `json:"output,omitempty"`
	// reasoning
	Summary json.RawMessage `json:"summary,omitempty"`
}

// ContentPart is one part of a message item's content.
type ContentPart struct {
	Type string `json:"type"` // input_text | output_text | input_image
	Text string `json:"text,omitempty"`
	// input_image
	ImageURL string `json:"image_url,omitempty"`
}

// Response is the Responses API response body.
type Response struct {
	ID                string      `json:"id"`
	Object            string      `json:"object"`
	CreatedAt         int64       `json:"created_at"`
	Status            string      `json:"status"`
	Model             string      `json:"model"`
	Output            []OutItem   `json:"output"`
	OutputText        string      `json:"output_text,omitempty"`
	Usage             *Usage      `json:"usage,omitempty"`
	IncompleteDetails *Incomplete `json:"incomplete_details,omitempty"`
	Error             *RespError  `json:"error,omitempty"`
}

// OutItem is one output entry: an assistant message or a function call.
type OutItem struct {
	ID        string        `json:"id,omitempty"`
	Type      string        `json:"type"` // message | function_call
	Role      string        `json:"role,omitempty"`
	Status    string        `json:"status,omitempty"`
	Content   []ContentPart `json:"content,omitempty"`
	CallID    string        `json:"call_id,omitempty"`
	Name      string        `json:"name,omitempty"`
	Arguments string        `json:"arguments,omitempty"`
}

// Usage reports token counts.
type Usage struct {
	InputTokens  int `json:"input_tokens"`
	OutputTokens int `json:"output_tokens"`
	TotalTokens  int `json:"total_tokens"`
}

// IncompleteDetails explains a non-completed response.
type Incomplete struct {
	Reason string `json:"reason"`
}

// RespError is the error payload of a failed response.
type RespError struct {
	Code    string `json:"code"`
	Message string `json:"message"`
}

// ErrorResponse is the Responses API error envelope.
type ErrorResponse struct {
	Error Detail `json:"error"`
}

// Detail is the error payload.
type Detail struct {
	Type    string `json:"type"`
	Message string `json:"message"`
}

// ParseInput decodes the input field, which is either a plain string
// or an array of items.
func ParseInput(raw json.RawMessage) ([]Item, error) {
	trimmed := bytes.TrimSpace(raw)
	if len(trimmed) == 0 || string(trimmed) == "null" {
		return nil, nil
	}
	if trimmed[0] == '"' {
		var s string
		if err := json.Unmarshal(trimmed, &s); err != nil {
			return nil, err
		}
		if s == "" {
			return nil, nil
		}
		return []Item{{Type: "message", Role: "user", Content: trimmed}}, nil
	}
	var items []Item
	if err := json.Unmarshal(trimmed, &items); err != nil {
		return nil, err
	}
	return items, nil
}

// ParseContent decodes a message item's content field into parts. A
// plain string becomes a single text part.
func ParseContent(raw json.RawMessage) ([]ContentPart, error) {
	trimmed := bytes.TrimSpace(raw)
	if len(trimmed) == 0 || string(trimmed) == "null" {
		return nil, nil
	}
	if trimmed[0] == '"' {
		var s string
		if err := json.Unmarshal(trimmed, &s); err != nil {
			return nil, err
		}
		if s == "" {
			return nil, nil
		}
		return []ContentPart{{Type: "input_text", Text: s}}, nil
	}
	var parts []ContentPart
	if err := json.Unmarshal(trimmed, &parts); err != nil {
		return nil, err
	}
	return parts, nil
}

// OutputText flattens a function_call_output output field, which is
// either a plain string or an array of text parts.
func OutputText(raw json.RawMessage) string {
	parts, err := ParseContent(raw)
	if err != nil {
		return ""
	}
	var sb strings.Builder
	for _, p := range parts {
		switch p.Type {
		case "output_text", "input_text", "text":
			sb.WriteString(p.Text)
		}
	}
	return sb.String()
}