fix missing encoding indication in pretty mode

[fmigneault]

When using pretty=True, the encoding="UTF-8" indication is missing:

Json2xml({"data": [123]}, pretty=True).to_xml()
'<?xml version="1.0" ?>\n<all>\n\t<data type="list">\n\t\t<item type="int">123</item>\n\t</data>\n</all>\n'

However, it is present when pretty=False:

Json2xml({"data": [123]}, pretty=False).to_xml()
b'<?xml version="1.0" encoding="UTF-8" ?><all><data type="list"><item type="int">123</item></data></all>'

Since in pretty=True mode, the encoding remains UTF-8, the contents should still indicate it.
The proposed fix produces the following:

Json2xml({"data": [123]}, pretty=True).to_xml() 
'<?xml version="1.0" encoding="UTF-8"?>\n<all>\n\t<data type="list">\n\t\t<item type="int">123</item>\n\t</data>\n</all>\n'

Diff emphasis before/after with rendered whitespaces:

- <?xml version="1.0" ?>
+ <?xml version="1.0" encoding="UTF-8"?>
<all>
	<data type="list">
		<item type="int">123</item>
	</data>
</all>

[vinitkumar]

Thanks for your contribution @fmigneault. Could you also add tests accompanying this change, so that there are no future regressions with this?

[vinitkumar]

Actually, I will just add the test myself. Thank you so much for catching this. 🙇🏼

[vinitkumar]

@fmigneault Just made a new release with your fix. Feel free to upgrade and let me know if all works fine?

[fmigneault]

@vinitkumar
Thanks for the quick integration.
I will test it out in the following days.