[Rust Frontend] Align Rust allowed_token_ids validation with Python (#46348)

Co-authored-by: Bugen Zhao <i@bugenzhao.com>
Signed-off-by: reidliu41 <reid201711@gmail.com>
Signed-off-by: Bugen Zhao <i@bugenzhao.com>
This commit is contained in:
Reid
2026-06-23 08:32:33 +00:00
committed by GitHub
co-authored by Bugen Zhao
parent accaa434f3
commit 2d721ab5d8
6 changed files with 138 additions and 27 deletions
+12 -2
View File
@@ -102,7 +102,7 @@ fn is_request_validation_error(error: &vllm_text::Error) -> bool {
vllm_text::Error::PromptTooLong { .. }
| vllm_text::Error::EmptyPromptTokenIds { .. }
| vllm_text::Error::Logprobs(_)
| vllm_text::Error::OutOfVocab(_)
| vllm_text::Error::TokenIds(_)
| vllm_text::Error::InvalidThinkingTokenBudget
// An empty tokenized prompt detected later, at request prepare
// time, surfaces through the transparent Llm wrapper.
@@ -188,7 +188,7 @@ mod tests {
#[test]
fn out_of_vocab_validation_maps_to_invalid_request() {
let error = vllm_text::Error::OutOfVocab(vllm_text::OutOfVocabError {
let error = vllm_text::Error::TokenIds(vllm_text::TokenIdsError::OutOfVocab {
parameter: "logprob_token_ids",
token_ids: vec![1000],
vocab_size: 1000,
@@ -197,6 +197,16 @@ mod tests {
assert_eq!(api_error.status_code(), StatusCode::BAD_REQUEST);
}
#[test]
fn empty_allowed_token_ids_maps_to_invalid_request() {
let error = vllm_text::Error::TokenIds(vllm_text::TokenIdsError::EmptyAllowedTokenIds);
let api_error = text_submit_error("failed to submit completion request", error);
assert_eq!(api_error.status_code(), StatusCode::BAD_REQUEST);
let response = api_error.to_error_response();
assert_eq!(response.error.error_type, "invalid_request_error");
assert!(response.error.message.contains("allowed_token_ids"));
}
#[test]
fn other_submit_errors_stay_internal() {
let error = vllm_text::Error::Tokenizer("backend exploded".to_string());
+72
View File
@@ -2237,6 +2237,42 @@ async fn invalid_request_returns_openai_error() {
assert_eq!(json["error"]["type"], "invalid_request_error");
}
#[tokio::test(flavor = "multi_thread", worker_threads = 2)]
#[serial]
async fn chat_completions_empty_allowed_token_ids_returns_openai_error() {
let mut app = test_app().await;
let response = app
.call(
Request::builder()
.method("POST")
.uri("/v1/chat/completions")
.header("content-type", "application/json")
.body(Body::from(
json!({
"model": "Qwen/Qwen1.5-0.5B-Chat",
"stream": false,
"messages": [{"role": "user", "content": "hello"}],
"allowed_token_ids": []
})
.to_string(),
))
.expect("build request"),
)
.await
.expect("call app");
assert_eq!(response.status(), StatusCode::BAD_REQUEST);
let body = to_bytes(response.into_body(), usize::MAX).await.expect("read body");
let json: serde_json::Value = serde_json::from_slice(&body).expect("decode json");
assert_eq!(json["error"]["type"], "invalid_request_error");
assert!(
json["error"]["message"]
.as_str()
.expect("message string")
.contains("allowed_token_ids should not be empty")
);
}
#[tokio::test(flavor = "multi_thread", worker_threads = 2)]
#[serial]
async fn non_stream_chat_returns_json_response() {
@@ -2972,6 +3008,42 @@ async fn completions_invalid_request_returns_openai_error() {
assert_eq!(json["error"]["type"], "invalid_request_error");
}
#[tokio::test(flavor = "multi_thread", worker_threads = 2)]
#[serial]
async fn completions_empty_allowed_token_ids_returns_openai_error() {
let mut app = test_app().await;
let response = app
.call(
Request::builder()
.method("POST")
.uri("/v1/completions")
.header("content-type", "application/json")
.body(Body::from(
json!({
"model": "Qwen/Qwen1.5-0.5B-Chat",
"prompt": "hello",
"stream": false,
"allowed_token_ids": []
})
.to_string(),
))
.expect("build request"),
)
.await
.expect("call app");
assert_eq!(response.status(), StatusCode::BAD_REQUEST);
let body = to_bytes(response.into_body(), usize::MAX).await.expect("read body");
let json: serde_json::Value = serde_json::from_slice(&body).expect("decode json");
assert_eq!(json["error"]["type"], "invalid_request_error");
assert!(
json["error"]["message"]
.as_str()
.expect("message string")
.contains("allowed_token_ids should not be empty")
);
}
#[tokio::test(flavor = "multi_thread", worker_threads = 2)]
#[serial]
async fn non_stream_completions_return_json_response() {