Fix issue with parsing non-string values in the process_docs function.

VERSION

-1.10.1
+1.10.2

tests/test_mlp.py

 import json
-
 import pytest
 import regex as re
 
@@ -218,4 +217,13 @@ def test_sentences_not_separated_with_newline(mlp: MLP):
     assert mlp_result["lemmas"] == 'siin olema üks lause . see olema teine lause .'
 
 
+def test_parsing_non_text_value_in_dictionary(mlp: MLP):
+    result = mlp.process_docs([{"non_text_field": 1324331241}], doc_paths=["non_text_field"])
+    for key in result:
+        assert "mlp" not in key
+
 
+def test_parsing_empty_list_in_dictionary(mlp: MLP):
+    result = mlp.process_docs([{"empty_list_field": []}], doc_paths=["empty_list_field"])
+    for key in result:
+        assert "mlp" not in key

texta_mlp/mlp.py

 import logging
 import os
 import pathlib
-import shutil
-from typing import List
-from urllib.parse import urlparse
-from urllib.request import urlopen
 import regex as re
+import shutil
 import stanza
 from bs4 import BeautifulSoup
 from langdetect import detect
 from pelecanus import PelicanJson
+from typing import List
+from urllib.parse import urlparse
+from urllib.request import urlopen
 
 from texta_mlp.document import Document
 from texta_mlp.entity_mapper import EntityMapper
@@ -323,10 +323,14 @@ class MLP:
         """
         wrapper = PelicanJson(document)
         doc_path_as_list = doc_path.split(".")
-        doc_texts = wrapper.safe_get_nested_value(doc_path_as_list, default=[])
-        doc_texts = [] if doc_texts is None else doc_texts
-        doc_texts = [doc_texts] if isinstance(doc_texts, str) else doc_texts
-        return doc_texts
+        content = wrapper.safe_get_nested_value(doc_path_as_list, default=[])
+        if content and isinstance(content, str):
+            return [content]
+        # Check that content is non-empty list and there are only stings in the list.
+        elif content and isinstance(content, list) and all([isinstance(list_content, str) for list_content in content]):
+            return content
+        else:
+            return []
 
 
     def process_docs(self, docs: List[dict], doc_paths: List[str], analyzers=["all"]):