Handle <math> tags with attributes

[BoboTiG]

Required by #2319.

• Word: https://en.wiktionary.org/wiki/Euclidean_metric
• Wikicode: <math style="vertical-align:+0%;">...</math>

As of now, our regexp won't match this code:

ebook-reader-dict/wikidict/utils.py

Line 505 in 6be5847

text = sub(r"<math>([^<]+)</math>", partial(convert_math, word=word), text)

I think we could do a first pass in clean() to remove all attributes: <math style="vertical-align:+0%;">...</math> -> <math>...</math>