feat(serverless): sync collections API pagination and text options

examples/serverless.rs

@@ -13,7 +13,8 @@
 
 use qdrant_client::qdrant::{PointStruct, QueryPointsBuilder, UpsertPointsBuilder};
 use qdrant_client::serverless::{
-    CollectionConfig, DenseVectorConfig, Distance, KeywordIndex, QdrantServerless,
+    CollectionConfig, DenseVectorConfig, Distance, KeywordIndex, ListCollectionsBuilder,
+    QdrantServerless,
 };
 use qdrant_client::{Payload, QdrantError};
 
@@ -38,12 +39,17 @@ async fn main() -> Result<(), QdrantError> {
             collection_name,
             CollectionConfig::new()
                 .dense_vector(DenseVectorConfig::new(4, Distance::Cosine))
-                .payload_index("color", KeywordIndex),
+                .payload_index("color", KeywordIndex::new()),
         )
         .await?;
     println!("create_collection: {result}");
 
-    println!("collections: {:?}", client.list_collections().await?);
+    println!(
+        "collections: {:?}",
+        client
+            .list_collections(ListCollectionsBuilder::new())
+            .await?
+    );
     println!("info: {:?}", client.get_collection(collection_name).await?);
 
     let points = vec![

proto/serverless_collections.proto

@@ -1,10 +1,13 @@
 // Source: https://github.com/qdrant/qdrant-cloud-public-api/blob/main/proto/qdrant/serverless/collections.proto
 // Renamed to serverless_collections.proto to sit alongside the regular collections.proto.
-// Regenerate with: cargo test --test serverless_protos -- --ignored
+// Client copy: buf.validate options are stripped (server-side only; wire format unchanged).
+// Regenerate with: cargo test --test serverless_protos -- --ignored --nocapture
+
 syntax = "proto3";
 
 package qdrant.serverless;
 
+
 // CollectionsService manages the collections of a qdrant serverless space.
 // Unlike the qdrant server API, it exposes only a simplified configuration:
 // the manager turns it into concrete qdrant settings, which are never
@@ -85,7 +88,15 @@ enum Tokenizer {
 }
 
 // Exact match on string values, e.g. `color: "red"`.
-message KeywordIndex {}
+message KeywordIndex {
+  // If set, enable prefix matching (`match: { "prefix": ... }`) on this field.
+  // Presence of this message enables prefix matching; it has no options yet.
+  optional KeywordPrefixParams prefix = 1;
+}
+
+// Prefix matching options for the keyword index. Has no options yet:
+// presence of this message enables prefix matching.
+message KeywordPrefixParams {}
 
 // Exact match and/or range filters on integers, e.g. `age: 25`. Both are on
 // by default; turning one off shrinks the index.
@@ -105,6 +116,35 @@ message UuidIndex {}
 // Range filters on RFC 3339 datetimes, e.g. `created_at: "2023-02-08T10:49:00Z"`.
 message DatetimeIndex {}
 
+// Tokens ignored by a full-text index. Language names match qdrant (e.g.
+// "english"); predefined lists and custom tokens are merged.
+message StopwordsSet {
+  // Languages whose predefined stopword lists to apply.
+  repeated string languages = 1;
+  // Extra stopwords to ignore, merged with the language lists.
+  repeated string custom = 2;
+}
+
+// Snowball stemming for a full-text index.
+message SnowballParams {
+  // Language for the snowball algorithm, e.g. "english".
+  string language = 1;
+}
+
+// Explicitly disable stemming (overrides any language default).
+message DisabledStemmer {}
+
+// Stemming algorithm for a full-text index. Unset: no stemming.
+message StemmingAlgorithm {
+  // Which stemming algorithm to use.
+  oneof stemming_params {
+    // Snowball stemmer for the given language.
+    SnowballParams snowball = 1;
+    // Explicitly disable stemming.
+    DisabledStemmer disabled = 2;
+  }
+}
+
 // Full-text filtering on string values.
 message TextIndex {
   // Tokenizer to split text with. Unset: WHITESPACE.
@@ -117,6 +157,12 @@ message TextIndex {
   optional uint64 min_token_len = 4;
   // Maximum token length to index.
   optional uint64 max_token_len = 5;
+  // Fold accented characters to ASCII. Default false.
+  optional bool ascii_folding = 6;
+  // Tokens to ignore at index and query time.
+  optional StopwordsSet stopwords = 7;
+  // Stemming algorithm. Unset: engine default (no stemming).
+  optional StemmingAlgorithm stemmer = 8;
 }
 
 // Geo radius / bounding box / polygon filters on `{lon, lat}` values.
@@ -211,9 +257,15 @@ message GetCollectionResponse {
   optional uint64 point_count = 3;
 }
 
-// Lists the caller's collections. The tenant travels in metadata, so there is
-// nothing to name here.
-message ListCollectionsRequest {}
+// Lists the caller's collections. The tenant travels in metadata.
+message ListCollectionsRequest {
+  // Maximum number of collections to return. Defaults to 20 and must not
+  // exceed 100.
+  optional uint32 limit = 1;
+  // Opaque token returned as `next_offset_token` by the previous page. Clients
+  // must not interpret this value.
+  optional string offset_token = 2;
+}
 
 // One collection in a listing.
 message CollectionSummary {
@@ -226,7 +278,9 @@ message CollectionSummary {
 
 // The caller's collections.
 message ListCollectionsResponse {
-  // Ordered by name. A collection whose creation never published a manifest is
-  // not listed: it is not servable.
+  // Collections in this page.
   repeated CollectionSummary collections = 1;
+  // Opaque token to pass as `offset_token` to retrieve the next page. Absent
+  // when there are no more results.
+  optional string next_offset_token = 2;
 }

src/serverless/client.rs

@@ -30,7 +30,9 @@ use crate::serverless::grpc::collections_service_client::CollectionsServiceClien
 use crate::serverless::grpc::{
     CreateCollectionRequest, DeleteCollectionRequest, GetCollectionRequest, ListCollectionsRequest,
 };
-use crate::serverless::models::{CollectionConfig, CollectionInfo, CollectionSummary};
+use crate::serverless::models::{
+    CollectionConfig, CollectionInfo, CollectionSummary, CollectionsList, ListCollections,
+};
 use crate::Qdrant;
 
 /// Default gRPC port for serverless when the URL omits an explicit port.
@@ -70,7 +72,7 @@ pub const DEFAULT_SERVERLESS_GRPC_PORT: u16 = 443;
 ///         "my-collection",
 ///         CollectionConfig::new()
 ///             .dense_vector(DenseVectorConfig::new(4, Distance::Cosine))
-///             .payload_index("color", KeywordIndex),
+///             .payload_index("color", KeywordIndex::new()),
 ///     )
 ///     .await?;
 ///
@@ -288,23 +290,51 @@ impl QdrantServerless {
         Ok(self.get_collection(collection_name).await?.exists)
     }
 
-    /// Lists the collections of the space.
+    /// Lists a page of collections in the space.
+    ///
+    /// Defaults to the server page size (20, max 100). Pass `offset_token` from a
+    /// previous response's `next_offset_token` to fetch the next page.
     ///
-    /// Returns summaries (name and eventually consistent point count), ordered by name.
-    pub async fn list_collections(&self) -> QdrantResult<Vec<CollectionSummary>> {
+    /// ```no_run
+    /// # use qdrant_client::serverless::{ListCollectionsBuilder, QdrantServerless};
+    /// # async fn run(client: QdrantServerless) -> Result<(), qdrant_client::QdrantError> {
+    /// let page = client
+    ///     .list_collections(ListCollectionsBuilder::new().limit(50))
+    ///     .await?;
+    /// if let Some(token) = page.next_offset_token {
+    ///     let next = client
+    ///         .list_collections(ListCollectionsBuilder::new().offset_token(token))
+    ///         .await?;
+    /// }
+    /// # Ok(())
+    /// # }
+    /// ```
+    pub async fn list_collections(
+        &self,
+        request: impl Into<ListCollections>,
+    ) -> QdrantResult<CollectionsList> {
+        let ListCollections {
+            limit,
+            offset_token,
+        } = request.into();
+        let request = ListCollectionsRequest {
+            limit,
+            offset_token,
+        };
+        let request = &request;
         self.with_collections_client(|mut api| async move {
-            let response = api
-                .list_collections(ListCollectionsRequest {})
-                .await?
-                .into_inner();
-            Ok(response
-                .collections
-                .into_iter()
-                .map(|c| CollectionSummary {
-                    collection_name: c.collection_name,
-                    point_count: c.point_count,
-                })
-                .collect())
+            let response = api.list_collections(request.clone()).await?.into_inner();
+            Ok(CollectionsList {
+                collections: response
+                    .collections
+                    .into_iter()
+                    .map(|c| CollectionSummary {
+                        collection_name: c.collection_name,
+                        point_count: c.point_count,
+                    })
+                    .collect(),
+                next_offset_token: response.next_offset_token,
+            })
         })
         .await
     }