Superfluous insertion of whitespace in `text`

I expect the below to produce "foobarbaz" - but it produces "foo bar baz":

text(parsehtml("<em>foo</em>bar<em></em>baz").root)

The motivating case was:

text(parsehtml("<math><mrow><msub><mrow><mi>MoSe</mi></mrow><mrow><mn>2</mn></mrow></msub></mrow></math>").root)

which produces "MoSe 2" rather than "MoSe2".

The issue is with the apparently redundant ' ' in

Line 45 in afc2b2b

print(io, c.text, ' ')

Would it be acceptable to change this? If so, I can make a PR.