response.rs 7.34 KB
Newer Older
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
// Responses API response types

use crate::protocols::openai::responses::request::ResponsesRequest;
use crate::protocols::openai::responses::types::*;
use serde::{Deserialize, Serialize};
use std::collections::HashMap;

fn generate_response_id() -> String {
    format!("resp_{}", uuid::Uuid::new_v4().simple())
}

fn current_timestamp() -> i64 {
    std::time::SystemTime::now()
        .duration_since(std::time::UNIX_EPOCH)
        .unwrap_or_else(|_| std::time::Duration::from_secs(0))
        .as_secs() as i64
}

#[derive(Debug, Clone, Deserialize, Serialize)]
pub struct ResponsesResponse {
    /// Response ID
    #[serde(default = "generate_response_id")]
    pub id: String,

    /// Object type
    #[serde(default = "default_object_type")]
    pub object: String,

    /// Creation timestamp
    #[serde(default = "current_timestamp")]
    pub created_at: i64,

    /// Model name
    pub model: String,

    /// Output items
    #[serde(default)]
    pub output: Vec<ResponseOutputItem>,

    /// Response status
    pub status: ResponseStatus,

    /// Usage information
    #[serde(skip_serializing_if = "Option::is_none")]
    pub usage: Option<UsageInfo>,

    /// Whether parallel tool calls are enabled
    #[serde(default = "default_true")]
    pub parallel_tool_calls: bool,

    /// Tool choice setting
    #[serde(default = "default_tool_choice")]
    pub tool_choice: String,

    /// Available tools
    #[serde(default)]
    pub tools: Vec<ResponseTool>,
}

fn default_object_type() -> String {
    "response".to_string()
}

fn default_true() -> bool {
    true
}

fn default_tool_choice() -> String {
    "auto".to_string()
}

impl ResponsesResponse {
    /// Create a response from a request
    #[allow(clippy::too_many_arguments)]
    pub fn from_request(
        request: &ResponsesRequest,
        _sampling_params: &HashMap<String, serde_json::Value>,
        model_name: String,
        created_time: i64,
        output: Vec<ResponseOutputItem>,
        status: ResponseStatus,
        usage: Option<UsageInfo>,
    ) -> Self {
        Self {
            id: request.request_id.clone(),
            object: "response".to_string(),
            created_at: created_time,
            model: model_name,
            output,
            status,
            usage,
            parallel_tool_calls: request.parallel_tool_calls,
            tool_choice: match request.tool_choice {
                ToolChoice::Auto => "auto".to_string(),
                ToolChoice::Required => "required".to_string(),
                ToolChoice::None => "none".to_string(),
            },
            tools: request.tools.clone(),
        }
    }

    /// Create a new response with default values
    pub fn new(request_id: String, model: String, status: ResponseStatus) -> Self {
        Self {
            id: request_id,
            object: "response".to_string(),
            created_at: current_timestamp(),
            model,
            output: Vec::new(),
            status,
            usage: None,
            parallel_tool_calls: true,
            tool_choice: "auto".to_string(),
            tools: Vec::new(),
        }
    }

    /// Add an output item to the response
    pub fn add_output(&mut self, item: ResponseOutputItem) {
        self.output.push(item);
    }

    /// Set the usage information
    pub fn set_usage(&mut self, usage: UsageInfo) {
        self.usage = Some(usage);
    }

    /// Update the status
    pub fn set_status(&mut self, status: ResponseStatus) {
        self.status = status;
    }

    /// Check if the response is complete
    pub fn is_complete(&self) -> bool {
        matches!(self.status, ResponseStatus::Completed)
    }

    /// Check if the response is in progress
    pub fn is_in_progress(&self) -> bool {
        matches!(self.status, ResponseStatus::InProgress)
    }

    /// Check if the response failed
    pub fn is_failed(&self) -> bool {
        matches!(self.status, ResponseStatus::Failed)
    }

    /// Check if the response was cancelled
    pub fn is_cancelled(&self) -> bool {
        matches!(self.status, ResponseStatus::Cancelled)
    }

    /// Check if the response is queued
    pub fn is_queued(&self) -> bool {
        matches!(self.status, ResponseStatus::Queued)
    }

    /// Convert usage to OpenAI Responses API format
    pub fn usage_in_response_format(
        &self,
    ) -> Option<crate::protocols::openai::responses::types::ResponseUsage> {
        self.usage.as_ref().map(|usage| usage.to_response_usage())
    }

    /// Get the response as a JSON value with usage in response format
    pub fn to_response_format(&self) -> serde_json::Value {
        let mut response = serde_json::to_value(self).unwrap_or(serde_json::Value::Null);

        // Convert usage to response format if present
        if let Some(usage) = &self.usage {
            if let Ok(usage_value) = serde_json::to_value(usage.to_response_usage()) {
                response["usage"] = usage_value;
            }
        }

        response
    }
}

// ============= Helper Functions =============

impl ResponseOutputItem {
    /// Create a new message output item
    pub fn new_message(
        id: String,
        role: String,
        content: Vec<ResponseContentPart>,
        status: String,
    ) -> Self {
        Self::Message {
            id,
            role,
            content,
            status,
        }
    }

    /// Create a new reasoning output item
    pub fn new_reasoning(
        id: String,
        summary: Vec<String>,
        content: Vec<ResponseReasoningContent>,
        status: Option<String>,
    ) -> Self {
        Self::Reasoning {
            id,
            summary,
            content,
            status,
        }
    }

    /// Create a new function tool call output item
    pub fn new_function_tool_call(
        id: String,
        name: String,
        arguments: String,
        output: Option<String>,
        status: String,
    ) -> Self {
        Self::FunctionToolCall {
            id,
            name,
            arguments,
            output,
            status,
        }
    }
}

impl ResponseContentPart {
    /// Create a new text content part
    pub fn new_text(
        text: String,
        annotations: Vec<String>,
        logprobs: Option<crate::protocols::openai::common::ChatLogProbs>,
    ) -> Self {
        Self::OutputText {
            text,
            annotations,
            logprobs,
        }
    }
}

impl ResponseReasoningContent {
    /// Create a new reasoning text content
    pub fn new_reasoning_text(text: String) -> Self {
        Self::ReasoningText { text }
    }
}

impl UsageInfo {
    /// Create a new usage info with token counts
    pub fn new(prompt_tokens: u32, completion_tokens: u32, reasoning_tokens: Option<u32>) -> Self {
        Self {
            prompt_tokens,
            completion_tokens,
            total_tokens: prompt_tokens + completion_tokens,
            reasoning_tokens,
            prompt_tokens_details: None,
        }
    }

    /// Create usage info with cached token details
    pub fn new_with_cached(
        prompt_tokens: u32,
        completion_tokens: u32,
        reasoning_tokens: Option<u32>,
        cached_tokens: u32,
    ) -> Self {
        Self {
            prompt_tokens,
            completion_tokens,
            total_tokens: prompt_tokens + completion_tokens,
            reasoning_tokens,
            prompt_tokens_details: Some(PromptTokenUsageInfo { cached_tokens }),
        }
    }
}