chore: fix no lock build warnings and CI timeouts (#2650)

Example CI failures:
- publish build timeout:
https://github.com/lancedb/lancedb/actions/runs/17626482881/job/50084552906
- doc test build timeout:
https://github.com/lancedb/lancedb/actions/runs/17627058590/job/50086456818
This commit is contained in:
Jack Ye
2025-09-11 15:30:35 -07:00
committed by GitHub
parent f7d78c3420
commit 0ebc8d45a8
12 changed files with 28 additions and 26 deletions

View File

@@ -86,11 +86,11 @@ rand = { version = "0.9", features = ["small_rng"] }
random_word = { version = "0.4.3", features = ["en"] }
uuid = { version = "1.7.0", features = ["v4"] }
walkdir = "2"
aws-sdk-dynamodb = { version = "1.38.0" }
aws-sdk-s3 = { version = "1.38.0" }
aws-sdk-kms = { version = "1.37" }
aws-config = { version = "1.0" }
aws-smithy-runtime = { version = "1.3" }
aws-sdk-dynamodb = { version = "1.55.0" }
aws-sdk-s3 = { version = "1.55.0" }
aws-sdk-kms = { version = "1.48.0" }
aws-config = { version = "1.5.10" }
aws-smithy-runtime = { version = "1.9.1" }
datafusion.workspace = true
http-body = "1" # Matching reqwest
rstest = "0.23.0"

View File

@@ -45,10 +45,10 @@ use crate::{
pub trait EmbeddingFunction: std::fmt::Debug + Send + Sync {
fn name(&self) -> &str;
/// The type of the input data
fn source_type(&self) -> Result<Cow<DataType>>;
fn source_type(&self) -> Result<Cow<'_, DataType>>;
/// The type of the output data
/// This should **always** match the output of the `embed` function
fn dest_type(&self) -> Result<Cow<DataType>>;
fn dest_type(&self) -> Result<Cow<'_, DataType>>;
/// Compute the embeddings for the source column in the database
fn compute_source_embeddings(&self, source: Arc<dyn Array>) -> Result<Arc<dyn Array>>;
/// Compute the embeddings for a given user query

View File

@@ -75,11 +75,11 @@ impl EmbeddingFunction for BedrockEmbeddingFunction {
"bedrock"
}
fn source_type(&self) -> Result<Cow<DataType>> {
fn source_type(&self) -> Result<Cow<'_, DataType>> {
Ok(Cow::Owned(DataType::Utf8))
}
fn dest_type(&self) -> Result<Cow<DataType>> {
fn dest_type(&self) -> Result<Cow<'_, DataType>> {
let n_dims = self.model.ndims();
Ok(Cow::Owned(DataType::new_fixed_size_list(
DataType::Float32,

View File

@@ -144,11 +144,11 @@ impl EmbeddingFunction for OpenAIEmbeddingFunction {
"openai"
}
fn source_type(&self) -> Result<Cow<DataType>> {
fn source_type(&self) -> Result<Cow<'_, DataType>> {
Ok(Cow::Owned(DataType::Utf8))
}
fn dest_type(&self) -> Result<Cow<DataType>> {
fn dest_type(&self) -> Result<Cow<'_, DataType>> {
let n_dims = self.model.ndims();
Ok(Cow::Owned(DataType::new_fixed_size_list(
DataType::Float32,

View File

@@ -407,11 +407,11 @@ impl EmbeddingFunction for SentenceTransformersEmbeddings {
"sentence-transformers"
}
fn source_type(&self) -> crate::Result<std::borrow::Cow<arrow_schema::DataType>> {
fn source_type(&self) -> crate::Result<std::borrow::Cow<'_, arrow_schema::DataType>> {
Ok(Cow::Owned(DataType::Utf8))
}
fn dest_type(&self) -> crate::Result<std::borrow::Cow<arrow_schema::DataType>> {
fn dest_type(&self) -> crate::Result<std::borrow::Cow<'_, arrow_schema::DataType>> {
let (n_dims, dtype) = self.compute_ndims_and_dtype()?;
Ok(Cow::Owned(DataType::new_fixed_size_list(
dtype,

View File

@@ -242,17 +242,15 @@ pub struct OptimizeStats {
/// Describes what happens when a vector either contains NaN or
/// does not have enough values
#[derive(Clone, Debug, Default)]
#[allow(dead_code)] // https://github.com/lancedb/lancedb/issues/992
enum BadVectorHandling {
/// An error is returned
#[default]
Error,
#[allow(dead_code)] // https://github.com/lancedb/lancedb/issues/992
/// The offending row is droppped
Drop,
#[allow(dead_code)] // https://github.com/lancedb/lancedb/issues/992
/// The invalid/missing items are replaced by fill_value
Fill(f32),
#[allow(dead_code)] // https://github.com/lancedb/lancedb/issues/992
/// The invalid items are replaced by NULL
None,
}

View File

@@ -341,10 +341,10 @@ impl EmbeddingFunction for MockEmbed {
fn name(&self) -> &str {
&self.name
}
fn source_type(&self) -> Result<Cow<DataType>> {
fn source_type(&self) -> Result<Cow<'_, DataType>> {
Ok(Cow::Borrowed(&self.source_type))
}
fn dest_type(&self) -> Result<Cow<DataType>> {
fn dest_type(&self) -> Result<Cow<'_, DataType>> {
Ok(Cow::Borrowed(&self.dest_type))
}
fn compute_source_embeddings(&self, source: Arc<dyn Array>) -> Result<Arc<dyn Array>> {