input check

haystack/preview/components/embedders/sentence_transformers_text_embedder.py

@@ -62,6 +62,11 @@ def warm_up(self):
     @component.output_types(embeddings=List[np.ndarray])
     def run(self, texts: List[str]):
         """Embed a list of strings."""
+        if not isinstance(texts, list) or not isinstance(texts[0], str):
+            raise ValueError(
+                "SentenceTransformersTextEmbedder expects a list of strings as input."
+                "In case you want to embed Documents, please use the SentenceTransformersDocumentEmbedder."
+            )
         self.warm_up()
         texts_to_embed = [self.prefix + text + self.suffix for text in texts]
         embeddings = self.embedding_backend.embed(

test/preview/components/embedders/test_sentence_transformers_text_embedder.py

@@ -82,3 +82,18 @@ def test_run(self):
         assert len(embeddings) == len(texts)
         for embedding in embeddings:
             assert isinstance(embedding, np.ndarray)
+
+    @pytest.mark.unit
+    def test_run_wrong_input_format(self):
+        embedder = SentenceTransformersTextEmbedder(model_name_or_path="model")
+        embedder.embedding_backend = MagicMock()
+        # embedder.embedding_backend.embed = lambda x, **kwargs: list(np.random.rand(len(x), 16))
+
+        string_input = "text"
+        list_integers_input = [1, 2, 3]
+
+        with pytest.raises(ValueError, match="SentenceTransformersTextEmbedder expects a list of strings as input"):
+            embedder.run(texts=string_input)
+
+        with pytest.raises(ValueError, match="SentenceTransformersTextEmbedder expects a list of strings as input"):
+            embedder.run(texts=list_integers_input)