feat: update localrecall and use postgresql as default (#394)

Signed-off-by: Ettore Di Giacinto <[email protected]>

Author: mudler

docker-compose.amd.yaml

@@ -16,6 +16,11 @@ services:
       file: docker-compose.yaml
       service: dind
 
+  localrecall-postgres:
+    extends:
+      file: docker-compose.yaml
+      service: localrecall-postgres
+
   localrecall:
     extends:
       file: docker-compose.yaml

docker-compose.intel.yaml

@@ -17,6 +17,11 @@ services:
       file: docker-compose.yaml
       service: dind
 
+  localrecall-postgres:
+    extends:
+      file: docker-compose.yaml
+      service: localrecall-postgres
+
   localrecall:
     extends:
       file: docker-compose.yaml

docker-compose.nvidia.yaml

@@ -22,6 +22,11 @@ services:
       file: docker-compose.yaml
       service: dind
 
+  localrecall-postgres:
+    extends:
+      file: docker-compose.yaml
+      service: localrecall-postgres
+
   localrecall:
     extends:
       file: docker-compose.yaml

docker-compose.yaml

@@ -26,18 +26,41 @@ services:
       - ./volumes/backends:/backends
       - ./volumes/images:/tmp/generated/images
 
+  localrecall-postgres:
+    image: quay.io/mudler/localrecall:${LOCALRECALL_VERSION:-v0.5.2}-postgresql
+    environment:
+      - POSTGRES_DB=localrecall
+      - POSTGRES_USER=localrecall
+      - POSTGRES_PASSWORD=localrecall
+    ports:
+      - 5432:5432
+    volumes:
+      - postgres_data:/var/lib/postgresql/data
+    healthcheck:
+      test: ["CMD-SHELL", "pg_isready -U localrecall"]
+      interval: 10s
+      timeout: 5s
+      retries: 5
+
   localrecall:
-    image: quay.io/mudler/localrecall:main
+    image: quay.io/mudler/localrecall:${LOCALRECALL_VERSION:-v0.5.2}
+    depends_on:
+      localrecall-postgres:
+        condition: service_healthy
+      localai:
+        condition: service_started
     ports:
       - 8080
     environment:
-      - COLLECTION_DB_PATH=/db
+      - DATABASE_URL=postgresql://localrecall:localrecall@localrecall-postgres:5432/localrecall?sslmode=disable
+      - VECTOR_ENGINE=postgres
       - EMBEDDING_MODEL=granite-embedding-107m-multilingual
       - FILE_ASSETS=/assets
       - OPENAI_API_KEY=sk-1234567890
       - OPENAI_BASE_URL=http://localai:8080
+      - HYBRID_SEARCH_BM25_WEIGHT=0.5
+      - HYBRID_SEARCH_VECTOR_WEIGHT=0.5
     volumes:
-      - ./volumes/localrag/db:/db
       - ./volumes/localrag/assets/:/assets
 
   localrecall-healthcheck:
@@ -102,3 +125,6 @@ services:
       - "host.docker.internal:host-gateway"
     volumes:
       - ./volumes/localagi/:/pool
+
+volumes:
+  postgres_data:

pkg/localrag/client.go

@@ -90,6 +90,32 @@ func (c *WrappedClient) Store(s string) error {
 	return c.Client.Store(c.collection, f)
 }
 
+// apiResponse is the standardized LocalRecall API response wrapper (since 3f73ff3a).
+type apiResponse struct {
+	Success bool            `json:"success"`
+	Message string         `json:"message,omitempty"`
+	Data    json.RawMessage `json:"data,omitempty"`
+	Error   *apiError       `json:"error,omitempty"`
+}
+
+type apiError struct {
+	Code    string `json:"code"`
+	Message string `json:"message"`
+	Details string `json:"details,omitempty"`
+}
+
+// parseAPIError reads the response body and returns an error from the API response or a generic message.
+func parseAPIError(resp *http.Response, body []byte, fallback string) error {
+	var wrap apiResponse
+	if err := json.Unmarshal(body, &wrap); err == nil && wrap.Error != nil {
+		if wrap.Error.Details != "" {
+			return fmt.Errorf("%s: %s", wrap.Error.Message, wrap.Error.Details)
+		}
+		return errors.New(wrap.Error.Message)
+	}
+	return fmt.Errorf("%s: %s", fallback, string(body))
+}
+
 // Result represents a single result from a query.
 type Result struct {
 	ID        string
@@ -153,7 +179,8 @@ func (c *Client) CreateCollection(name string) error {
 	defer resp.Body.Close()
 
 	if resp.StatusCode != http.StatusCreated {
-		return errors.New("failed to create collection")
+		body, _ := io.ReadAll(resp.Body)
+		return parseAPIError(resp, body, "failed to create collection")
 	}
 
 	return nil
@@ -176,17 +203,30 @@ func (c *Client) ListCollections() ([]string, error) {
 	}
 	defer resp.Body.Close()
 
+	body, err := io.ReadAll(resp.Body)
+	if err != nil {
+		return nil, err
+	}
+
 	if resp.StatusCode != http.StatusOK {
-		return nil, errors.New("failed to list collections")
+		return nil, parseAPIError(resp, body, "failed to list collections")
 	}
 
-	var collections []string
-	err = json.NewDecoder(resp.Body).Decode(&collections)
-	if err != nil {
-		return nil, err
+	var wrap apiResponse
+	if err := json.Unmarshal(body, &wrap); err != nil || !wrap.Success {
+		if wrap.Error != nil {
+			return nil, errors.New(wrap.Error.Message)
+		}
+		return nil, fmt.Errorf("invalid response: %w", err)
 	}
 
-	return collections, nil
+	var data struct {
+		Collections []string `json:"collections"`
+	}
+	if err := json.Unmarshal(wrap.Data, &data); err != nil {
+		return nil, err
+	}
+	return data.Collections, nil
 }
 
 // ListEntries lists all entries in a collection
@@ -206,17 +246,30 @@ func (c *Client) ListEntries(collection string) ([]string, error) {
 	}
 	defer resp.Body.Close()
 
+	body, err := io.ReadAll(resp.Body)
+	if err != nil {
+		return nil, err
+	}
+
 	if resp.StatusCode != http.StatusOK {
-		return nil, errors.New("failed to list entries")
+		return nil, parseAPIError(resp, body, "failed to list entries")
 	}
 
-	var entries []string
-	err = json.NewDecoder(resp.Body).Decode(&entries)
-	if err != nil {
-		return nil, err
+	var wrap apiResponse
+	if err := json.Unmarshal(body, &wrap); err != nil || !wrap.Success {
+		if wrap.Error != nil {
+			return nil, errors.New(wrap.Error.Message)
+		}
+		return nil, fmt.Errorf("invalid response: %w", err)
 	}
 
-	return entries, nil
+	var data struct {
+		Entries []string `json:"entries"`
+	}
+	if err := json.Unmarshal(wrap.Data, &data); err != nil {
+		return nil, err
+	}
+	return data.Entries, nil
 }
 
 // DeleteEntry deletes an entry in a collection
@@ -246,19 +299,30 @@ func (c *Client) DeleteEntry(collection, entry string) ([]string, error) {
 	}
 	defer resp.Body.Close()
 
+	body, err := io.ReadAll(resp.Body)
+	if err != nil {
+		return nil, err
+	}
+
 	if resp.StatusCode != http.StatusOK {
-		bodyResult := new(bytes.Buffer)
-		bodyResult.ReadFrom(resp.Body)
-		return nil, errors.New("failed to delete entry: " + bodyResult.String())
+		return nil, parseAPIError(resp, body, "failed to delete entry")
 	}
 
-	var results []string
-	err = json.NewDecoder(resp.Body).Decode(&results)
-	if err != nil {
-		return nil, err
+	var wrap apiResponse
+	if err := json.Unmarshal(body, &wrap); err != nil || !wrap.Success {
+		if wrap.Error != nil {
+			return nil, errors.New(wrap.Error.Message)
+		}
+		return nil, fmt.Errorf("invalid response: %w", err)
 	}
 
-	return results, nil
+	var data struct {
+		RemainingEntries []string `json:"remaining_entries"`
+	}
+	if err := json.Unmarshal(wrap.Data, &data); err != nil {
+		return nil, err
+	}
+	return data.RemainingEntries, nil
 }
 
 // Search searches a collection
@@ -289,17 +353,30 @@ func (c *Client) Search(collection, query string, maxResults int) ([]Result, err
 	}
 	defer resp.Body.Close()
 
+	body, err := io.ReadAll(resp.Body)
+	if err != nil {
+		return nil, err
+	}
+
 	if resp.StatusCode != http.StatusOK {
-		return nil, errors.New("failed to search collection")
+		return nil, parseAPIError(resp, body, "failed to search collection")
 	}
 
-	var results []Result
-	err = json.NewDecoder(resp.Body).Decode(&results)
-	if err != nil {
-		return nil, err
+	var wrap apiResponse
+	if err := json.Unmarshal(body, &wrap); err != nil || !wrap.Success {
+		if wrap.Error != nil {
+			return nil, errors.New(wrap.Error.Message)
+		}
+		return nil, fmt.Errorf("invalid response: %w", err)
 	}
 
-	return results, nil
+	var data struct {
+		Results []Result `json:"results"`
+	}
+	if err := json.Unmarshal(wrap.Data, &data); err != nil {
+		return nil, err
+	}
+	return data.Results, nil
 }
 
 // Reset resets a collection
@@ -320,9 +397,8 @@ func (c *Client) Reset(collection string) error {
 	defer resp.Body.Close()
 
 	if resp.StatusCode != http.StatusOK {
-		b := new(bytes.Buffer)
-		b.ReadFrom(resp.Body)
-		return errors.New("failed to reset collection: " + b.String())
+		body, _ := io.ReadAll(resp.Body)
+		return parseAPIError(resp, body, "failed to reset collection")
 	}
 
 	return nil
@@ -371,20 +447,8 @@ func (c *Client) Store(collection, filePath string) error {
 	defer resp.Body.Close()
 
 	if resp.StatusCode != http.StatusOK {
-		b := new(bytes.Buffer)
-		b.ReadFrom(resp.Body)
-
-		type response struct {
-			Error string `json:"error"`
-		}
-
-		var r response
-		err = json.Unmarshal(b.Bytes(), &r)
-		if err == nil {
-			return errors.New("failed to upload file: " + r.Error)
-		}
-
-		return errors.New("failed to upload file")
+		body, _ := io.ReadAll(resp.Body)
+		return parseAPIError(resp, body, "failed to upload file")
 	}
 
 	return nil