New filter plugin - unicode_normalization (#3359)

* Initial commit

* Adding maintainer in BOTMETA

* Adding changelog fragment

* Updating filter_guide

* Applying initial review suggestions

docs/docsite/rst/filter_guide.rst

@@ -751,3 +751,34 @@ To extract ports from all clusters with name containing 'server1':
         server_name_query: "domain.server[?contains(name,'server1')].port"
 
 .. note:: while using ``starts_with`` and ``contains``, you have to use `` to_json | from_json `` filter for correct parsing of data structure.
+
+Working with Unicode
+---------------------
+
+`Unicode <https://unicode.org/main.html>`_ makes it possible to produce two strings which may be visually equivalent, but are comprised of distinctly different characters/character sequences. To address this ``Unicode`` defines `normalization forms <https://unicode.org/reports/tr15/>`_ which avoid these distinctions by choosing a unique character sequence for a given visual representation.
+
+You can use the ``community.general.unicode_normalize`` filter to normalize ``Unicode`` strings within your playbooks.
+
+.. code-block:: yaml+jinja
+
+    - name: Compare Unicode representations
+      debug:
+        msg: "{{ with_combining_character | community.general.unicode_normalize == without_combining_character }}"
+      vars:
+        with_combining_character: "{{ 'Mayagu\u0308ez' }}"
+        without_combining_character: Mayagüez
+
+This produces:
+
+.. code-block:: ansible-output
+
+    TASK [Compare Unicode representations] ********************************************************
+    ok: [localhost] => {
+        "msg": true
+    }
+
+The ``community.general.unicode_normalize`` filter accepts a keyword argument to select the ``Unicode`` form used to normalize the input string.
+
+:form: One of ``'NFC'`` (default), ``'NFD'``, ``'NFKC'``, or ``'NFKD'``. See the `Unicode reference <https://unicode.org/reports/tr15/>`_ for more information.
+
+.. versionadded:: 3.7.0