remove_texts added to Elasticsearch (#526)

* Added remove_texts on base vectorsearch

* Added remove_texts on elasticsearch vectorsearch

---------

Co-authored-by: Gregorio Galante <[email protected]>

lib/langchain/vectorsearch/base.rb

@@ -124,6 +124,11 @@ def update_texts(...)
       raise NotImplementedError, "#{self.class.name} does not support updating texts"
     end
 
+    # Method supported by Vectorsearch DB to delete a list of texts from the index
+    def remove_texts(...)
+      raise NotImplementedError, "#{self.class.name} does not support deleting texts"
+    end
+
     # Method supported by Vectorsearch DB to search for similar texts in the index
     def similarity_search(...)
       raise NotImplementedError, "#{self.class.name} does not support similarity search"

lib/langchain/vectorsearch/elasticsearch.rb

@@ -75,6 +75,17 @@ def update_texts(texts: [], ids: [])
       es_client.bulk(body: body)
     end
 
+    # Remove a list of texts from the index
+    # @param ids [Array<Integer>] The list of ids to delete
+    # @return [Elasticsearch::Response] from the Elasticsearch server
+    def remove_texts(ids: [])
+      body = ids.map do |id|
+        {delete: {_index: index_name, _id: id}}
+      end
+
+      es_client.bulk(body: body)
+    end
+
     # Create the index with the default schema
     # @return [Elasticsearch::Response] Index creation
     def create_default_schema

spec/langchain/vectorsearch/base_spec.rb

@@ -41,6 +41,12 @@
     end
   end
 
+  describe "#remove_texts" do
+    it "raises an error" do
+      expect { subject.remove_texts }.to raise_error(NotImplementedError)
+    end
+  end
+
   describe "#similarity_search" do
     it "raises an error" do
       expect { subject.similarity_search }.to raise_error(NotImplementedError)

spec/langchain/vectorsearch/elasticsearch_spec.rb

@@ -51,6 +51,19 @@
     end
   end
 
+  describe "#remove_texts" do
+    it "removes respective document" do
+      es_body = [
+        {delete: {_index: "langchain", _id: 1}}
+      ]
+
+      allow_any_instance_of(::Elasticsearch::Client).to receive(:bulk).with(body: es_body)
+      expect_any_instance_of(::Elasticsearch::Client).to receive(:bulk).with(body: es_body).once
+
+      subject.remove_texts(ids: [1])
+    end
+  end
+
   describe "#default_vector_settings" do
     it "returns default vector settings" do
       expect(subject.default_vector_settings).to eq({type: "dense_vector", dims: 384})