cloud_llm_client.rs

  1pub mod predict_edits_v3;
  2
  3use std::str::FromStr;
  4use std::sync::Arc;
  5
  6use anyhow::Context as _;
  7use serde::{Deserialize, Serialize};
  8use strum::{Display, EnumIter, EnumString};
  9use uuid::Uuid;
 10
 11/// The name of the header used to indicate which version of Zed the client is running.
 12pub const ZED_VERSION_HEADER_NAME: &str = "x-zed-version";
 13
 14/// The name of the header used to indicate when a request failed due to an
 15/// expired LLM token.
 16///
 17/// The client may use this as a signal to refresh the token.
 18pub const EXPIRED_LLM_TOKEN_HEADER_NAME: &str = "x-zed-expired-token";
 19
 20/// The name of the header used to indicate when a request failed due to an outdated LLM token.
 21///
 22/// A token is considered "outdated" when we can't parse the claims (e.g., after adding a new required claim).
 23///
 24/// This is distinct from [`EXPIRED_LLM_TOKEN_HEADER_NAME`] which indicates the token's time-based validity has passed.
 25/// An outdated token means the token's structure is incompatible with the current server expectations.
 26pub const OUTDATED_LLM_TOKEN_HEADER_NAME: &str = "x-zed-outdated-token";
 27
 28/// The name of the header used to indicate the usage limit for edit predictions.
 29pub const EDIT_PREDICTIONS_USAGE_LIMIT_HEADER_NAME: &str = "x-zed-edit-predictions-usage-limit";
 30
 31/// The name of the header used to indicate the usage amount for edit predictions.
 32pub const EDIT_PREDICTIONS_USAGE_AMOUNT_HEADER_NAME: &str = "x-zed-edit-predictions-usage-amount";
 33
 34pub const EDIT_PREDICTIONS_RESOURCE_HEADER_VALUE: &str = "edit_predictions";
 35
 36/// The name of the header used to indicate the minimum required Zed version.
 37///
 38/// This can be used to force a Zed upgrade in order to continue communicating
 39/// with the LLM service.
 40pub const MINIMUM_REQUIRED_VERSION_HEADER_NAME: &str = "x-zed-minimum-required-version";
 41
 42/// The name of the header used by the client to indicate to the server that it supports receiving status messages.
 43pub const CLIENT_SUPPORTS_STATUS_MESSAGES_HEADER_NAME: &str =
 44    "x-zed-client-supports-status-messages";
 45
 46/// The name of the header used by the server to indicate to the client that it supports sending status messages.
 47pub const SERVER_SUPPORTS_STATUS_MESSAGES_HEADER_NAME: &str =
 48    "x-zed-server-supports-status-messages";
 49
 50/// The name of the header used by the client to indicate that it supports receiving xAI models.
 51pub const CLIENT_SUPPORTS_X_AI_HEADER_NAME: &str = "x-zed-client-supports-x-ai";
 52
 53/// The maximum number of edit predictions that can be rejected per request.
 54pub const MAX_EDIT_PREDICTION_REJECTIONS_PER_REQUEST: usize = 100;
 55
 56#[derive(Debug, PartialEq, Clone, Copy, Serialize, Deserialize)]
 57#[serde(rename_all = "snake_case")]
 58pub enum UsageLimit {
 59    Limited(i32),
 60    Unlimited,
 61}
 62
 63impl FromStr for UsageLimit {
 64    type Err = anyhow::Error;
 65
 66    fn from_str(value: &str) -> Result<Self, Self::Err> {
 67        match value {
 68            "unlimited" => Ok(Self::Unlimited),
 69            limit => limit
 70                .parse::<i32>()
 71                .map(Self::Limited)
 72                .context("failed to parse limit"),
 73        }
 74    }
 75}
 76
 77#[derive(Debug, Clone, Copy, PartialEq)]
 78pub enum Plan {
 79    V2(PlanV2),
 80}
 81
 82#[derive(Debug, Clone, Copy, Default, PartialEq, Serialize, Deserialize)]
 83#[serde(rename_all = "snake_case")]
 84pub enum PlanV2 {
 85    #[default]
 86    ZedFree,
 87    ZedPro,
 88    ZedProTrial,
 89}
 90
 91#[derive(
 92    Debug, PartialEq, Eq, Hash, Clone, Copy, Serialize, Deserialize, EnumString, EnumIter, Display,
 93)]
 94#[serde(rename_all = "snake_case")]
 95#[strum(serialize_all = "snake_case")]
 96pub enum LanguageModelProvider {
 97    Anthropic,
 98    OpenAi,
 99    Google,
100    XAi,
101}
102
103#[derive(Debug, Clone, Serialize, Deserialize)]
104pub struct PredictEditsBody {
105    #[serde(skip_serializing_if = "Option::is_none", default)]
106    pub outline: Option<String>,
107    pub input_events: String,
108    pub input_excerpt: String,
109    #[serde(skip_serializing_if = "Option::is_none", default)]
110    pub speculated_output: Option<String>,
111    /// Whether the user provided consent for sampling this interaction.
112    #[serde(default, alias = "data_collection_permission")]
113    pub can_collect_data: bool,
114    #[serde(skip_serializing_if = "Option::is_none", default)]
115    pub diagnostic_groups: Option<Vec<(String, serde_json::Value)>>,
116    /// Info about the git repository state, only present when can_collect_data is true.
117    #[serde(skip_serializing_if = "Option::is_none", default)]
118    pub git_info: Option<PredictEditsGitInfo>,
119    /// The trigger for this request.
120    #[serde(default)]
121    pub trigger: PredictEditsRequestTrigger,
122}
123
124#[derive(Default, Debug, Clone, Copy, Serialize, Deserialize)]
125pub enum PredictEditsRequestTrigger {
126    Testing,
127    Diagnostics,
128    Cli,
129    #[default]
130    Other,
131}
132
133#[derive(Debug, Clone, Serialize, Deserialize)]
134pub struct PredictEditsGitInfo {
135    /// SHA of git HEAD commit at time of prediction.
136    #[serde(skip_serializing_if = "Option::is_none", default)]
137    pub head_sha: Option<String>,
138    /// URL of the remote called `origin`.
139    #[serde(skip_serializing_if = "Option::is_none", default)]
140    pub remote_origin_url: Option<String>,
141    /// URL of the remote called `upstream`.
142    #[serde(skip_serializing_if = "Option::is_none", default)]
143    pub remote_upstream_url: Option<String>,
144}
145
146#[derive(Debug, Clone, Serialize, Deserialize)]
147pub struct PredictEditsResponse {
148    pub request_id: String,
149    pub output_excerpt: String,
150}
151
152#[derive(Debug, Clone, Serialize, Deserialize)]
153pub struct AcceptEditPredictionBody {
154    pub request_id: String,
155}
156
157#[derive(Debug, Clone, Deserialize)]
158pub struct RejectEditPredictionsBody {
159    pub rejections: Vec<EditPredictionRejection>,
160}
161
162#[derive(Debug, Clone, Serialize)]
163pub struct RejectEditPredictionsBodyRef<'a> {
164    pub rejections: &'a [EditPredictionRejection],
165}
166
167#[derive(Debug, Clone, Serialize, Deserialize, PartialEq)]
168pub struct EditPredictionRejection {
169    pub request_id: String,
170    #[serde(default)]
171    pub reason: EditPredictionRejectReason,
172    pub was_shown: bool,
173}
174
175#[derive(Default, Debug, Clone, Copy, Serialize, Deserialize, PartialEq)]
176pub enum EditPredictionRejectReason {
177    /// New requests were triggered before this one completed
178    Canceled,
179    /// No edits returned
180    Empty,
181    /// Edits returned, but none remained after interpolation
182    InterpolatedEmpty,
183    /// The new prediction was preferred over the current one
184    Replaced,
185    /// The current prediction was preferred over the new one
186    CurrentPreferred,
187    /// The current prediction was discarded
188    #[default]
189    Discarded,
190}
191
192#[derive(Debug, PartialEq, Eq, PartialOrd, Ord, Hash, Clone, Copy, Serialize, Deserialize)]
193#[serde(rename_all = "snake_case")]
194pub enum CompletionIntent {
195    UserPrompt,
196    ToolResults,
197    ThreadSummarization,
198    ThreadContextSummarization,
199    CreateFile,
200    EditFile,
201    InlineAssist,
202    TerminalInlineAssist,
203    GenerateGitCommitMessage,
204}
205
206#[derive(Debug, Serialize, Deserialize)]
207pub struct CompletionBody {
208    #[serde(skip_serializing_if = "Option::is_none", default)]
209    pub thread_id: Option<String>,
210    #[serde(skip_serializing_if = "Option::is_none", default)]
211    pub prompt_id: Option<String>,
212    #[serde(skip_serializing_if = "Option::is_none", default)]
213    pub intent: Option<CompletionIntent>,
214    pub provider: LanguageModelProvider,
215    pub model: String,
216    pub provider_request: serde_json::Value,
217}
218
219#[derive(Debug, PartialEq, Clone, Serialize, Deserialize)]
220#[serde(rename_all = "snake_case")]
221pub enum CompletionRequestStatus {
222    Queued {
223        position: usize,
224    },
225    Started,
226    Failed {
227        code: String,
228        message: String,
229        request_id: Uuid,
230        /// Retry duration in seconds.
231        retry_after: Option<f64>,
232    },
233    UsageUpdated {
234        amount: usize,
235        limit: UsageLimit,
236    },
237    ToolUseLimitReached,
238}
239
240#[derive(Serialize, Deserialize)]
241#[serde(rename_all = "snake_case")]
242pub enum CompletionEvent<T> {
243    Status(CompletionRequestStatus),
244    Event(T),
245}
246
247impl<T> CompletionEvent<T> {
248    pub fn into_status(self) -> Option<CompletionRequestStatus> {
249        match self {
250            Self::Status(status) => Some(status),
251            Self::Event(_) => None,
252        }
253    }
254
255    pub fn into_event(self) -> Option<T> {
256        match self {
257            Self::Event(event) => Some(event),
258            Self::Status(_) => None,
259        }
260    }
261}
262
263#[derive(Serialize, Deserialize)]
264pub struct WebSearchBody {
265    pub query: String,
266}
267
268#[derive(Debug, Serialize, Deserialize, Clone)]
269pub struct WebSearchResponse {
270    pub results: Vec<WebSearchResult>,
271}
272
273#[derive(Debug, Serialize, Deserialize, Clone)]
274pub struct WebSearchResult {
275    pub title: String,
276    pub url: String,
277    pub text: String,
278}
279
280#[derive(Serialize, Deserialize)]
281pub struct CountTokensBody {
282    pub provider: LanguageModelProvider,
283    pub model: String,
284    pub provider_request: serde_json::Value,
285}
286
287#[derive(Serialize, Deserialize)]
288pub struct CountTokensResponse {
289    pub tokens: usize,
290}
291
292#[derive(Debug, PartialEq, Eq, Hash, Clone, Serialize, Deserialize)]
293pub struct LanguageModelId(pub Arc<str>);
294
295impl std::fmt::Display for LanguageModelId {
296    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
297        write!(f, "{}", self.0)
298    }
299}
300
301#[derive(Debug, Clone, PartialEq, Serialize, Deserialize)]
302pub struct LanguageModel {
303    pub provider: LanguageModelProvider,
304    pub id: LanguageModelId,
305    pub display_name: String,
306    pub max_token_count: usize,
307    pub max_token_count_in_max_mode: Option<usize>,
308    pub max_output_tokens: usize,
309    pub supports_tools: bool,
310    pub supports_images: bool,
311    pub supports_thinking: bool,
312    #[serde(default)]
313    pub supports_streaming_tools: bool,
314    /// Only used by OpenAI and xAI.
315    #[serde(default)]
316    pub supports_parallel_tool_calls: bool,
317}
318
319#[derive(Debug, Serialize, Deserialize)]
320pub struct ListModelsResponse {
321    pub models: Vec<LanguageModel>,
322    pub default_model: Option<LanguageModelId>,
323    pub default_fast_model: Option<LanguageModelId>,
324    pub recommended_models: Vec<LanguageModelId>,
325}
326
327#[derive(Debug, PartialEq, Serialize, Deserialize)]
328pub struct CurrentUsage {
329    pub edit_predictions: UsageData,
330}
331
332#[derive(Debug, PartialEq, Serialize, Deserialize)]
333pub struct UsageData {
334    pub used: u32,
335    pub limit: UsageLimit,
336}
337
338#[cfg(test)]
339mod tests {
340    use pretty_assertions::assert_eq;
341    use serde_json::json;
342
343    use super::*;
344
345    #[test]
346    fn test_plan_v2_deserialize_snake_case() {
347        let plan = serde_json::from_value::<PlanV2>(json!("zed_free")).unwrap();
348        assert_eq!(plan, PlanV2::ZedFree);
349
350        let plan = serde_json::from_value::<PlanV2>(json!("zed_pro")).unwrap();
351        assert_eq!(plan, PlanV2::ZedPro);
352
353        let plan = serde_json::from_value::<PlanV2>(json!("zed_pro_trial")).unwrap();
354        assert_eq!(plan, PlanV2::ZedProTrial);
355    }
356
357    #[test]
358    fn test_usage_limit_from_str() {
359        let limit = UsageLimit::from_str("unlimited").unwrap();
360        assert!(matches!(limit, UsageLimit::Unlimited));
361
362        let limit = UsageLimit::from_str(&0.to_string()).unwrap();
363        assert!(matches!(limit, UsageLimit::Limited(0)));
364
365        let limit = UsageLimit::from_str(&50.to_string()).unwrap();
366        assert!(matches!(limit, UsageLimit::Limited(50)));
367
368        for value in ["not_a_number", "50xyz"] {
369            let limit = UsageLimit::from_str(value);
370            assert!(limit.is_err());
371        }
372    }
373}