feat(query): Add support for single embedding retrieval with PGVector

Cargo.toml

@@ -53,10 +53,7 @@ arrow-array = { version = "52.2", default-features = false }
 arrow = { version = "52.2", default-features = false }
 parquet = { version = "52.2", default-features = false, features = ["async"] }
 redb = { version = "2.2" }
-sqlx = { version = "0.8.2", features = [
-  "postgres",
-  "uuid",
-], default-features = false }
+sqlx = { version = "0.8.2", features = ["postgres", "uuid"] }
 aws-config = "1.5"
 pgvector = { version = "0.4.0", features = ["sqlx"], default-features = false }
 aws-credential-types = "1.2"
@@ -87,6 +84,7 @@ tree-sitter-ruby = "0.23"
 tree-sitter-rust = "0.23"
 tree-sitter-typescript = "0.23"
 
+
 # Testing
 test-log = "0.2.16"
 testcontainers = { version = "0.23.0", features = ["http_wait"] }

examples/index_md_into_pgvector.rs

@@ -11,9 +11,40 @@ use swiftide::{
         },
         EmbeddedField,
     },
-    integrations::{self, pgvector::PgVector},
+    integrations::{self, fastembed::FastEmbed, pgvector::PgVector},
+    query::{self, answers, query_transformers, response_transformers},
+    traits::SimplePrompt,
 };
 
+async fn ask_query(
+    llm_client: impl SimplePrompt + Clone + 'static,
+    embed: FastEmbed,
+    vector_store: PgVector,
+    question: String,
+) -> Result<String, Box<dyn std::error::Error>> {
+    // By default the search strategy is SimilaritySingleEmbedding
+    // which takes the latest query, embeds it, and does a similarity search
+    //
+    // Pgvector will return an error if multiple embeddings are set
+    //
+    // The pipeline generates subquestions to increase semantic coverage, embeds these in a single
+    // embedding, retrieves the default top_k documents, summarizes them and uses that as context
+    // for the final answer.
+    let pipeline = query::Pipeline::default()
+        .then_transform_query(query_transformers::GenerateSubquestions::from_client(
+            llm_client.clone(),
+        ))
+        .then_transform_query(query_transformers::Embed::from_client(embed))
+        .then_retrieve(vector_store.clone())
+        .then_transform_response(response_transformers::Summary::from_client(
+            llm_client.clone(),
+        ))
+        .then_answer(answers::Simple::from_client(llm_client.clone()));
+
+    let result = pipeline.query(question).await?;
+    Ok(result.answer().into())
+}
+
 #[tokio::main]
 async fn main() -> Result<(), Box<dyn std::error::Error>> {
     tracing_subscriber::fmt::init();
@@ -62,6 +93,7 @@ async fn main() -> Result<(), Box<dyn std::error::Error>> {
     }
 
     tracing::info!("Starting indexing pipeline");
+
     indexing::Pipeline::from_loader(FileLoader::new(test_dataset_path).with_extensions(&["md"]))
         .then_chunk(ChunkMarkdown::from_chunk_range(10..2048))
         .then(MetadataQAText::new(llm_client.clone()))
@@ -70,6 +102,27 @@ async fn main() -> Result<(), Box<dyn std::error::Error>> {
         .run()
         .await?;
 
-    tracing::info!("PgVector Indexing test completed successfully");
+    tracing::info!("PgVector Indexing completed successfully");
+
+    for (i, question) in [
+            "What is SwiftIDE? Provide a clear, comprehensive summary in under 50 words.",
+            "How can I use SwiftIDE to connect with the Ethereum blockchain? Please provide a concise, comprehensive summary in less than 50 words.",
+        ]
+        .iter()
+        .enumerate()
+        {
+            let result = ask_query(
+                llm_client.clone(),
+                fastembed.clone(),
+                pgv_storage.clone(),
+                question.to_string(),
+            ).await?;
+            tracing::info!("*** Answer Q{} ***", i + 1);
+            tracing::info!("{}", result);
+            tracing::info!("===X===");
+        }
+
+    tracing::info!("PgVector Indexing & retrieval test completed successfully");
+
     Ok(())
 }

swiftide-integrations/src/pgvector/fixtures.rs

@@ -77,6 +77,8 @@ pub(crate) struct PgVectorTestData<'a> {
     pub metadata: Option<indexing::Metadata>,
     /// Vector embeddings with their corresponding fields
     pub vectors: Vec<(indexing::EmbeddedField, Vec<f32>)>,
+    pub expected_in_results: bool,
+    pub use_hybrid_search: bool,
 }
 
 impl<'a> PgVectorTestData<'a> {