Add support for more languages for mnemonics

[wfjvdham]

This pr will add support for french, japanese, czech and portuguese. It does this by using a newer version (v0.21) of the python-mnemonic package. The tests are updated to check if the new languages are created and detected correctly.

Removed this part:

specter-desktop/src/cryptoadvance/specter/util/mnemonic.py

Lines 53 to 60 in a052373

	# If we get French as language we have to double check whether it is not really an English mnemonic because of overlaps in the wordlists
	if language == "french":
	count = 0
	for word in words_as_list:
	if word in WORDLIST:
	count += 1
	if count == 12:
	language = "english"

The detect_language() function takes into account ambiguity and returns an error if a language cannot successfully be determined in the new version of python-mnemonic.

[netlify]

✅ Deploy Preview for specter-desktop-docs canceled.

Name	Link
🔨 Latest commit	a198f3f
🔍 Latest deploy log	https://app.netlify.com/sites/specter-desktop-docs/deploys/66462d87441a250008c84cbc

[k9ert]

I guess this needs deletion.

[k9ert]

Thanks for your contribution and sorry for the late feedback.

[wfjvdham]

np, thanks!

[k9ert]

This is now released: https://github.com/cryptoadvance/specter-desktop/releases/tag/v2.0.3

[tadeubas]

I think we shouldn't encourage users to store mnemonics in other languages to avoid dead ends because of the lack of support from other wallets and possible mistakes by the users. A quantitative study with 990 Bitcoin users by Krombholz et al. reported that 22.5% of respondents had lost Bitcoins at least once. The majority of incidents was caused by user mistakes (43.2%).