chore: centralize embedding errors, retrieval strategy, and test DB helpers.

Replace anyhow in embedding production code with EmbeddingError, move
RetrievalStrategy into common config, and deduplicate Surreal test setup
via common::test_utils.
This commit is contained in:
Per Stark
2026-05-29 14:35:07 +02:00
parent e3bb2935d0
commit d3443d4153
17 changed files with 366 additions and 304 deletions
+13 -13
View File
@@ -2,8 +2,11 @@ use common::storage::types::conversation::SidebarConversation;
use common::storage::{db::SurrealDbClient, store::StorageManager};
use common::utils::embedding::EmbeddingProvider;
use common::utils::template_engine::{ProvidesTemplateEngine, TemplateEngine};
use common::{create_template_engine, storage::db::ProvidesDb, utils::config::AppConfig};
use retrieval_pipeline::{reranking::RerankerPool, RetrievalStrategy};
use common::{
create_template_engine, storage::db::ProvidesDb,
utils::config::{AppConfig, RetrievalStrategy},
};
use retrieval_pipeline::reranking::RerankerPool;
use std::collections::HashMap;
use std::sync::{
atomic::{AtomicUsize, Ordering},
@@ -16,6 +19,7 @@ use tracing::debug;
use crate::{OpenAIClientType, SessionStoreType};
#[derive(Clone)]
/// Shared application state for HTML handlers and middleware.
pub struct HtmlState {
pub db: Arc<SurrealDbClient>,
pub openai_client: Arc<OpenAIClientType>,
@@ -31,7 +35,7 @@ pub struct HtmlState {
#[derive(Clone)]
struct ConversationArchiveCacheEntry {
conversations: Vec<SidebarConversation>,
conversations: Arc<[SidebarConversation]>,
expires_at: Instant,
}
@@ -72,23 +76,19 @@ impl HtmlState {
}
pub fn retrieval_strategy(&self) -> RetrievalStrategy {
self.config
.retrieval_strategy
.as_deref()
.and_then(|value| value.parse().ok())
.unwrap_or(RetrievalStrategy::Default)
self.config.resolved_retrieval_strategy()
}
pub async fn get_cached_conversation_archive(
&self,
user_id: &str,
) -> Option<Vec<SidebarConversation>> {
) -> Option<Arc<[SidebarConversation]>> {
let now = Instant::now();
let should_evict_expired = {
let cache = self.conversation_archive_cache.read().await;
if let Some(entry) = cache.get(user_id) {
if entry.expires_at > now {
return Some(entry.conversations.clone());
return Some(Arc::clone(&entry.conversations));
}
true
} else {
@@ -107,7 +107,7 @@ impl HtmlState {
pub async fn set_cached_conversation_archive(
&self,
user_id: &str,
conversations: Vec<SidebarConversation>,
conversations: Arc<[SidebarConversation]>,
) {
let now = Instant::now();
let mut cache = self.conversation_archive_cache.write().await;
@@ -235,10 +235,10 @@ mod tests {
cache.insert(
user_id.to_string(),
ConversationArchiveCacheEntry {
conversations: vec![SidebarConversation {
conversations: Arc::from([SidebarConversation {
id: "conv-1".to_string(),
title: "A stale chat".to_string(),
}],
}]),
expires_at: Instant::now() - Duration::from_secs(1),
},
);
+3 -27
View File
@@ -20,6 +20,7 @@ use crate::{
auth_middleware::RequireUser,
response_middleware::{HtmlError, TemplateResponse},
},
utils::truncate::{first_non_empty_line, truncate_with_ellipsis},
};
/// Serde deserialization decorator to map empty Strings to None,
@@ -41,31 +42,6 @@ fn source_id_suffix(source_id: &str) -> String {
source_id[start..].to_string()
}
fn truncate_label(value: &str, max_chars: usize) -> String {
let mut end = None;
for (count, (idx, _)) in value.char_indices().enumerate() {
if count == max_chars {
end = Some(idx);
break;
}
}
match end {
Some(idx) => format!("{}...", &value[..idx]),
None => value.to_string(),
}
}
fn first_non_empty_line(text: &str, max_chars: usize) -> Option<String> {
for line in text.lines() {
let trimmed = line.trim();
if !trimmed.is_empty() {
return Some(truncate_label(trimmed, max_chars));
}
}
None
}
#[derive(Deserialize)]
struct UrlInfoLabel {
#[serde(default)]
@@ -121,7 +97,7 @@ fn build_source_label(row: &SourceLabelRow) -> String {
if let Some(context) = row.context.as_ref() {
let trimmed = context.trim();
if !trimmed.is_empty() {
return truncate_label(trimmed, MAX_LABEL_CHARS);
return truncate_with_ellipsis(trimmed, MAX_LABEL_CHARS);
}
}
@@ -131,7 +107,7 @@ fn build_source_label(row: &SourceLabelRow) -> String {
let category = row.category.trim();
if !category.is_empty() {
return truncate_label(category, MAX_LABEL_CHARS);
return truncate_with_ellipsis(category, MAX_LABEL_CHARS);
}
format!("Text snippet: {}", source_id_suffix(&row.id))