tests for tokeniser and types

Cover WordTokeniser (Unicode, empty input, punctuation, multiple scripts)
and validation types (immutability, edge cases, failure summary).

tests/conftest.py

@@ -0,0 +1,20 @@
+"""Shared pytest fixtures for Veritext tests."""
+
+import pytest
+
+from veritext.core.tokenisation import WordTokeniser
+
+
+@pytest.fixture
+def word_tokeniser() -> WordTokeniser:
+    return WordTokeniser()
+
+
+@pytest.fixture
+def word_tokeniser_no_lowercase() -> WordTokeniser:
+    return WordTokeniser(lowercase=False)
+
+
+@pytest.fixture
+def word_tokeniser_keep_punctuation() -> WordTokeniser:
+    return WordTokeniser(remove_punctuation=False)