summaryrefslogtreecommitdiffstats
path: root/src/client/vertexai.rs
diff options
context:
space:
mode:
authordev <dev@localhost>2026-10-21 22:14:10 +0000
committerdev <dev@localhost>2026-10-21 22:14:10 +0000
commit49a527fc22447b85a0fe6fb945a6ece4e18de260 (patch)
treefac6a9f64f2de6fbdeac0e9df6fc73b67fe8a951 /src/client/vertexai.rs
parent710f0daff34e4ae02cf906841117449bcf114198 (diff)
downloadaichat-49a527fc22447b85a0fe6fb945a6ece4e18de260.tar.gz
Thread idempotency key through the request path
Propagate the per-request Idempotency-Key (generated once in the retry loop) through request construction so every retry attempt reuses the same key rather than minting a new one. - Add field to ChatCompletionsData. - Generate a single key per logical request in chat_completions and chat_completions_streaming. - Thread the key through request_builder (macro-based clients), Bedrock, and VertexAI so the header is attached once and reused on retries.
Diffstat (limited to 'src/client/vertexai.rs')
-rw-r--r--src/client/vertexai.rs9
1 files changed, 6 insertions, 3 deletions
diff --git a/src/client/vertexai.rs b/src/client/vertexai.rs
index 16b55d1..5d88280 100644
--- a/src/client/vertexai.rs
+++ b/src/client/vertexai.rs
@@ -44,8 +44,9 @@ impl Client for VertexAIClient {
prepare_gcloud_access_token(client, self.name(), &self.config.adc_file).await?;
let model = self.model();
let model_category = ModelCategory::from_str(model.real_name())?;
+ let idempotency_key = data.idempotency_key.clone();
let request_data = prepare_chat_completions(self, data, &model_category)?;
- let builder = self.request_builder(client, request_data);
+ let builder = self.request_builder(client, request_data, idempotency_key.as_deref());
match model_category {
ModelCategory::Gemini => gemini_chat_completions(builder, model).await,
ModelCategory::Claude => claude_chat_completions(builder, model).await,
@@ -62,8 +63,9 @@ impl Client for VertexAIClient {
prepare_gcloud_access_token(client, self.name(), &self.config.adc_file).await?;
let model = self.model();
let model_category = ModelCategory::from_str(model.real_name())?;
+ let idempotency_key = data.idempotency_key.clone();
let request_data = prepare_chat_completions(self, data, &model_category)?;
- let builder = self.request_builder(client, request_data);
+ let builder = self.request_builder(client, request_data, idempotency_key.as_deref());
match model_category {
ModelCategory::Gemini => {
gemini_chat_completions_streaming(builder, handler, model).await
@@ -84,7 +86,7 @@ impl Client for VertexAIClient {
) -> Result<Vec<Vec<f32>>> {
prepare_gcloud_access_token(client, self.name(), &self.config.adc_file).await?;
let request_data = prepare_embeddings(self, data)?;
- let builder = self.request_builder(client, request_data);
+ let builder = self.request_builder(client, request_data, None);
embeddings(builder, self.model()).await
}
}
@@ -316,6 +318,7 @@ pub fn gemini_build_chat_completions_body(
top_p,
functions,
stream: _,
+ idempotency_key: _,
} = data;
let system_message = extract_system_message(&mut messages);