exact search endpoint

This commit is contained in:
Priec
2026-04-29 00:40:36 +02:00
parent 5de1cd7623
commit 1ceab57f3b
5 changed files with 124 additions and 5 deletions

View File

@@ -4,6 +4,7 @@ package komp_ac.search;
service Searcher { service Searcher {
rpc SearchTable(SearchRequest) returns (SearchResponse); rpc SearchTable(SearchRequest) returns (SearchResponse);
rpc ExactSearchTable(SearchRequest) returns (SearchResponse);
} }
message SearchRequest { message SearchRequest {

Binary file not shown.

View File

@@ -140,6 +140,27 @@ pub mod searcher_client {
.insert(GrpcMethod::new("komp_ac.search.Searcher", "SearchTable")); .insert(GrpcMethod::new("komp_ac.search.Searcher", "SearchTable"));
self.inner.unary(req, path, codec).await self.inner.unary(req, path, codec).await
} }
pub async fn exact_search_table(
&mut self,
request: impl tonic::IntoRequest<super::SearchRequest>,
) -> std::result::Result<tonic::Response<super::SearchResponse>, tonic::Status> {
self.inner
.ready()
.await
.map_err(|e| {
tonic::Status::unknown(
format!("Service was not ready: {}", e.into()),
)
})?;
let codec = tonic::codec::ProstCodec::default();
let path = http::uri::PathAndQuery::from_static(
"/komp_ac.search.Searcher/ExactSearchTable",
);
let mut req = request.into_request();
req.extensions_mut()
.insert(GrpcMethod::new("komp_ac.search.Searcher", "ExactSearchTable"));
self.inner.unary(req, path, codec).await
}
} }
} }
/// Generated server implementations. /// Generated server implementations.
@@ -159,6 +180,10 @@ pub mod searcher_server {
&self, &self,
request: tonic::Request<super::SearchRequest>, request: tonic::Request<super::SearchRequest>,
) -> std::result::Result<tonic::Response<super::SearchResponse>, tonic::Status>; ) -> std::result::Result<tonic::Response<super::SearchResponse>, tonic::Status>;
async fn exact_search_table(
&self,
request: tonic::Request<super::SearchRequest>,
) -> std::result::Result<tonic::Response<super::SearchResponse>, tonic::Status>;
} }
#[derive(Debug)] #[derive(Debug)]
pub struct SearcherServer<T> { pub struct SearcherServer<T> {
@@ -279,6 +304,49 @@ pub mod searcher_server {
}; };
Box::pin(fut) Box::pin(fut)
} }
"/komp_ac.search.Searcher/ExactSearchTable" => {
#[allow(non_camel_case_types)]
struct ExactSearchTableSvc<T: Searcher>(pub Arc<T>);
impl<T: Searcher> tonic::server::UnaryService<super::SearchRequest>
for ExactSearchTableSvc<T> {
type Response = super::SearchResponse;
type Future = BoxFuture<
tonic::Response<Self::Response>,
tonic::Status,
>;
fn call(
&mut self,
request: tonic::Request<super::SearchRequest>,
) -> Self::Future {
let inner = Arc::clone(&self.0);
let fut = async move {
<T as Searcher>::exact_search_table(&inner, request).await
};
Box::pin(fut)
}
}
let accept_compression_encodings = self.accept_compression_encodings;
let send_compression_encodings = self.send_compression_encodings;
let max_decoding_message_size = self.max_decoding_message_size;
let max_encoding_message_size = self.max_encoding_message_size;
let inner = self.inner.clone();
let fut = async move {
let method = ExactSearchTableSvc(inner);
let codec = tonic::codec::ProstCodec::default();
let mut grpc = tonic::server::Grpc::new(codec)
.apply_compression_config(
accept_compression_encodings,
send_compression_encodings,
)
.apply_max_message_size_config(
max_decoding_message_size,
max_encoding_message_size,
);
let res = grpc.unary(method, req).await;
Ok(res)
};
Box::pin(fut)
}
_ => { _ => {
Box::pin(async move { Box::pin(async move {
let mut response = http::Response::new( let mut response = http::Response::new(

View File

@@ -20,6 +20,12 @@ const INDEX_ROOT: &str = "./tantivy_indexes";
const DEFAULT_RESULT_LIMIT: usize = 5; const DEFAULT_RESULT_LIMIT: usize = 5;
const SEARCH_RESULT_LIMIT: usize = 100; const SEARCH_RESULT_LIMIT: usize = 100;
#[derive(Clone, Copy)]
enum SearchMode {
Fuzzy,
Exact,
}
pub struct SearcherService { pub struct SearcherService {
pub pool: PgPool, pub pool: PgPool,
} }
@@ -165,7 +171,11 @@ async fn resolve_search_targets(
} }
// Query building // Query building
fn build_query(index: &Index, normalized_query: &str) -> Result<Option<BooleanQuery>, Status> { fn build_query(
index: &Index,
normalized_query: &str,
mode: SearchMode,
) -> Result<Option<BooleanQuery>, Status> {
let schema = index.schema(); let schema = index.schema();
let prefix_edge_field = schema let prefix_edge_field = schema
.get_field("prefix_edge") .get_field("prefix_edge")
@@ -182,6 +192,24 @@ fn build_query(index: &Index, normalized_query: &str) -> Result<Option<BooleanQu
return Ok(None); return Ok(None);
} }
if matches!(mode, SearchMode::Exact) {
let exact_parser = QueryParser::for_index(index, vec![prefix_full_field]);
let exact_query_str = if words.len() == 1 {
normalized_query.to_string()
} else {
format!("\"{}\"", normalized_query)
};
let exact_query = exact_parser
.parse_query(&exact_query_str)
.map_err(|e| Status::internal(format!("Failed to build exact query: {}", e)))?;
return Ok(Some(BooleanQuery::new(vec![(
Occur::Must,
Box::new(exact_query),
)])));
}
let mut query_layers: Vec<(Occur, Box<dyn Query>)> = Vec::new(); let mut query_layers: Vec<(Occur, Box<dyn Query>)> = Vec::new();
// Layer 1: prefix // Layer 1: prefix
@@ -276,6 +304,7 @@ async fn search_target(
pool: &PgPool, pool: &PgPool,
target: &SearchTarget, target: &SearchTarget,
query_str: &str, query_str: &str,
mode: SearchMode,
) -> Result<Vec<Hit>, Status> { ) -> Result<Vec<Hit>, Status> {
if !target.index_path.exists() { if !target.index_path.exists() {
return Ok(vec![]); return Ok(vec![]);
@@ -286,7 +315,7 @@ async fn search_target(
register_slovak_tokenizers(&index) register_slovak_tokenizers(&index)
.map_err(|e| Status::internal(format!("Failed to register Slovak tokenizers: {}", e)))?; .map_err(|e| Status::internal(format!("Failed to register Slovak tokenizers: {}", e)))?;
let Some(master_query) = build_query(&index, &normalize_slovak_text(query_str))? else { let Some(master_query) = build_query(&index, &normalize_slovak_text(query_str), mode)? else {
return Ok(vec![]); return Ok(vec![]);
}; };
@@ -360,6 +389,23 @@ impl Searcher for SearcherService {
async fn search_table( async fn search_table(
&self, &self,
request: Request<SearchRequest>, request: Request<SearchRequest>,
) -> Result<Response<SearchResponse>, Status> {
self.run_search(request, SearchMode::Fuzzy).await
}
async fn exact_search_table(
&self,
request: Request<SearchRequest>,
) -> Result<Response<SearchResponse>, Status> {
self.run_search(request, SearchMode::Exact).await
}
}
impl SearcherService {
async fn run_search(
&self,
request: Request<SearchRequest>,
mode: SearchMode,
) -> Result<Response<SearchResponse>, Status> { ) -> Result<Response<SearchResponse>, Status> {
let req = request.into_inner(); let req = request.into_inner();
let profile_name = req.profile_name.trim(); let profile_name = req.profile_name.trim();
@@ -404,7 +450,7 @@ impl Searcher for SearcherService {
// Merge per-table hits // Merge per-table hits
let mut hits = Vec::new(); let mut hits = Vec::new();
for target in &targets { for target in &targets {
hits.extend(search_target(&self.pool, target, query).await?); hits.extend(search_target(&self.pool, target, query, mode).await?);
} }
hits.sort_by(|left, right| right.score.total_cmp(&left.score)); hits.sort_by(|left, right| right.score.total_cmp(&left.score));
@@ -413,7 +459,11 @@ impl Searcher for SearcherService {
} }
info!( info!(
"Processed search for profile '{}' (table scope: {}). Returning {} hits.", "Processed {} search for profile '{}' (table scope: {}). Returning {} hits.",
match mode {
SearchMode::Fuzzy => "fuzzy",
SearchMode::Exact => "exact",
},
profile_name, profile_name,
requested_table.unwrap_or("*"), requested_table.unwrap_or("*"),
hits.len() hits.len()

2
server

Submodule server updated: df65bbf8f3...b26adc0cb0