Speach

Important changes
- Use ``chirptext`` >= 0.2a6.post1, support Python 3.10 and Python 3.11
- Infer timestamps and duratio of ref_annotations from linked annotations
- Fixed Symbolic Association linking bug
- Use explicit None checking for timestamps and duration info
- Add ``elan.Doc.clone()`` function
- Fixed missing lemmas bug in ``ttl_to_igrow()``
- Fixed ``_xml_tostring()`` method for Python < 3.8
- Other bug fixes

----------------

- 20022-03-09

- Cross-check with controlled vocabularies when creating annotations if possible

- 2022-03-07

- Allow to add new annotations to all 5 tier stereotypes

- None (default root tiers)
- Included In
- Time Subdivision
- Symbolic Subdivision
- Symbolic Association

- Support ``lxml`` when available

- Support ``pretty_print`` in ``elan.Doc.to_xml_str()`` and ``elan.Doc.to_xml_bin()``

- 2022-03-02

- Add ``speach.elan.create()`` function for creating a new ELAN file from scratch

- 2022-02-28

- Add ``LOCALE`` support
- Add Speach logo

PyPI permanent link: https://pypi.org/project/speach/0.1a14/

-------------

- 2022-01-14

- Use ``defusedxml`` automatically when available to parse XML for better security

-------------

- 2021-11-03

- Support controlled vocabularies editing
- Add new controlled vocabulary entries
- Remove controlled vocabulary entries
- Edit controlled vocabularies (values, desciptions, languages)
- Add crc32 helper functions to ``speach.media`` module

-------------

- 2021-08-26

- Add ``encoding`` option to ``eaf2csv`` command

-------------

- 2021-07-27

- Support editing ELAN media fields

- media_file
- time_units
- media_url
- mime_type
- relative_media_url