chore: ingestion-pipeline refactor, sort technical debt, rustfmt

evaluations/src/pipeline/context.rs

@@ -14,7 +14,7 @@ use common::{
     utils::embedding::EmbeddingProvider,
 };
 use retrieval_pipeline::{
-    pipeline::{StageTimings, RetrievalConfig},
+    pipeline::{RetrievalConfig, StageTimings},
     reranking::RerankerPool,
 };
 
@@ -122,11 +122,15 @@ impl<'a> EvaluationContext<'a> {
     }
 
     pub fn slice(&self) -> Result<&slice::ResolvedSlice<'a>> {
-        self.slice.as_ref().ok_or_else(|| anyhow!("slice has not been prepared"))
+        self.slice
+            .as_ref()
+            .ok_or_else(|| anyhow!("slice has not been prepared"))
     }
 
     pub fn db(&self) -> Result<&SurrealDbClient> {
-        self.db.as_ref().ok_or_else(|| anyhow!("database connection missing"))
+        self.db
+            .as_ref()
+            .ok_or_else(|| anyhow!("database connection missing"))
     }
 
     pub fn descriptor(&self) -> Result<&snapshot::Descriptor> {
@@ -142,15 +146,23 @@ impl<'a> EvaluationContext<'a> {
     }
 
     pub fn openai_client(&self) -> Result<Arc<Client<async_openai::config::OpenAIConfig>>> {
-        Ok(Arc::clone(self.openai_client.as_ref().ok_or_else(|| anyhow!("openai client missing"))?))
+        Ok(Arc::clone(
+            self.openai_client
+                .as_ref()
+                .ok_or_else(|| anyhow!("openai client missing"))?,
+        ))
     }
 
     pub fn corpus_handle(&self) -> Result<&corpus::CorpusHandle> {
-        self.corpus_handle.as_ref().ok_or_else(|| anyhow!("corpus handle missing"))
+        self.corpus_handle
+            .as_ref()
+            .ok_or_else(|| anyhow!("corpus handle missing"))
     }
 
     pub fn evaluation_user(&self) -> Result<&User> {
-        self.eval_user.as_ref().ok_or_else(|| anyhow!("evaluation user missing"))
+        self.eval_user
+            .as_ref()
+            .ok_or_else(|| anyhow!("evaluation user missing"))
     }
 
     #[allow(clippy::arithmetic_side_effects)]
@@ -168,7 +180,8 @@ impl<'a> EvaluationContext<'a> {
     }
 
     pub fn into_summary(self) -> Result<EvaluationSummary> {
-        self.summary.ok_or_else(|| anyhow!("evaluation summary missing"))
+        self.summary
+            .ok_or_else(|| anyhow!("evaluation summary missing"))
     }
 }
 

evaluations/src/pipeline/stages/run_queries.rs

@@ -10,7 +10,7 @@ use crate::eval::{
     CaseSummary, RetrievedSummary,
 };
 use retrieval_pipeline::{
-    pipeline::{self, StageTimings, RetrievalConfig},
+    pipeline::{self, RetrievalConfig, StageTimings},
     reranking::RerankerPool,
 };
 use tokio::sync::Semaphore;
@@ -169,10 +169,10 @@ pub(crate) async fn run_queries(
                 let query_start = Instant::now();
 
                 debug!(question_id = %question_id, "Evaluating query");
-                let query_embedding =
-                    embedding_provider.embed(&question).await.with_context(|| {
-                        format!("generating embedding for question {question_id}")
-                    })?;
+                let query_embedding = embedding_provider
+                    .embed(&question)
+                    .await
+                    .with_context(|| format!("generating embedding for question {question_id}"))?;
                 let reranker = match rerank_pool.as_ref() {
                     Some(pool) => pool.checkout().await,
                     None => None,
@@ -204,8 +204,10 @@ pub(crate) async fn run_queries(
                 let mut match_rank = None;
                 let answers_lower: Vec<String> =
                     answers.iter().map(|ans| ans.to_ascii_lowercase()).collect();
-                let expected_chunk_ids_set: HashSet<&str> =
-                    expected_chunk_ids.iter().map(std::string::String::as_str).collect();
+                let expected_chunk_ids_set: HashSet<&str> = expected_chunk_ids
+                    .iter()
+                    .map(std::string::String::as_str)
+                    .collect();
                 let chunk_id_required = has_verified_chunks;
                 let mut entity_hit = false;
                 let mut chunk_text_hit = false;
@@ -304,15 +306,12 @@ pub(crate) async fn run_queries(
                     None
                 };
 
-                Ok::<
-                    (
-                        usize,
-                        CaseSummary,
-                        Option<CaseDiagnostics>,
-                        StageTimings,
-                    ),
-                    anyhow::Error,
-                >((idx, summary, diagnostics, stage_timings))
+                Ok::<(usize, CaseSummary, Option<CaseDiagnostics>, StageTimings), anyhow::Error>((
+                    idx,
+                    summary,
+                    diagnostics,
+                    stage_timings,
+                ))
             }
         })
         .buffer_unordered(concurrency)

evaluations/src/pipeline/stages/summarize.rs

@@ -13,7 +13,11 @@ use super::super::{
 };
 use super::{map_guard_error, StageResult};
 
-#[allow(clippy::too_many_lines, clippy::arithmetic_side_effects, clippy::cast_precision_loss)]
+#[allow(
+    clippy::too_many_lines,
+    clippy::arithmetic_side_effects,
+    clippy::cast_precision_loss
+)]
 pub(crate) async fn summarize(
     machine: EvaluationMachine<(), QueriesFinished>,
     ctx: &mut EvaluationContext<'_>,