Drop unecessary hypothesis usage

This adds A LOT of time to test execution, for little benefit. While
having random data is good, in reality, most of it was just numbers, and
we rarely got very diverse data.

These hand-picked samples are as good to catch any regressions or issues
with servers.

tests/storage/__init__.py

@@ -5,12 +5,10 @@ from urllib.parse import quote as urlquote
 from urllib.parse import unquote as urlunquote
 
 import pytest
-from hypothesis import settings
 
 from .. import assert_item_equals
 from .. import EVENT_TEMPLATE
 from .. import normalize_item
-from .. import printable_characters_strategy
 from .. import TASK_TEMPLATE
 from .. import VCARD_TEMPLATE
 from vdirsyncer import exceptions
@@ -297,10 +295,15 @@ class StorageTests:
 
     @pytest.mark.parametrize(
         "value",
-        [None]
-        + [
-            printable_characters_strategy.example()
-            for _ in range(settings.get_profile(settings._current_profile).max_examples)
+        [
+            None,
+            "",
+            "Hello there!",
+            "Österreich",
+            "中国",
+            "한글",
+            "42a4ec99-b1c2-4859-b142-759112f2ca50",
+            "فلسطين",
         ],
     )
     def test_metadata_normalization(self, requires_metadata, s, value):