Installation ============ Install using pip ----------------- The recommended method is to use a virtual environment. .. code-block:: console $ pip install xsdata[cli,lxml,soap] .. hint:: - Install the cli requirements for the code generator - Install the soap requirements for the builtin wsdl client - Install lxml if you want to use one of the lxml handlers/writers instead of the builtin python xml implementations. xsdata has a monthly release cycle, in order to use the latest updates you can also install directly from the git repo. .. code-block:: console $ pip install git+https://github.com/tefra/xsdata@master#egg=xsdata[cli,lxml] Install using conda ------------------- .. code-block:: console $ conda install -c conda-forge xsdata Verify installation ------------------- Verify installation using the cli entry point. .. cli:: xsdata --help Requirements ------------ .. admonition:: xsData relies on these awesome libraries and supports `python >= 3.6` :class: hint * `lxml `_ - XML parsing * `requests `_ - Webservice Default Transport * `click `_ - CLI entry point * `toposort `_ - Resolve class ordering * `jinja2 `_ - Code generation * `docformatter `_ - Code formatting .. warning:: In python 3.6 the typing module is flattening subclasses in unions, this may affect how values are converted. There is no official workaround because it's not very common, if you like monkey patching then take a look here :py:func:`typing._remove_dups_flatten` .. doctest:: >>> from dataclasses import dataclass >>> from typing import Union, get_type_hints ... >>> @dataclass ... class Example: ... value: Union[int, bool, str, float] ... >>> get_type_hints(Example) # doctest: +SKIP {'value': typing.Union[int, str, float]} >>> issubclass(bool, int) # doctest: +SKIP True