Consider adding the formality and glossary feature to DeepL

[BreakerZero]

At first sight, it seems that it is only a matter of adding some formated fields to the JSONRPC request:

For formality:

• Options available: "formal", "informal", null
• Even though it is a "premium" feature, DeepL returns the correct tone if entered (output obtained without the tone: "Vous êtes".):
  

For glossary (named dictionary in the json):

• The only pairs in the glossary that are sent are those where the target text is in the target language that the user has chosen.
• The glossary pairs are visibly sorted according to the order of the letters in the unicode table. (only the source text matters)
• The same source text cannot have several translations (the previous pair is overwritten if this is the case)
• A source text/target text pair is separated by '\t', two different pairs are separated by '\n'.
• Backslash is escaped if used.

Notes:

• browserType is optional.
• There must be another requirement to generate a correct glossary, because after some tests I get the error -32600 which corresponds to :Invalid Request 'Invalid commonJobParams'

[Animenosekai]

Thank you for reaching out and for this very detailed explanation !

I'm currently in vacation but I'll definitely look into it.

Also, I think that we'll add a "raw" attribute to the result models for the raw response and that we will make specific models for different translators.

[ZhymabekRoman]

I'm currently in vacation

Same thing, I won't be able to work on projects until September, if you need any help ping me. After rest we will start work on v3 😅

[Animenosekai]

Same thing, I won't be able to work on projects until September, if you need any help ping me. After rest we will start work on v3 😅

I might greatly progress into v3 in August, since I already made lots of changes in the website branch 😎

[ZhymabekRoman]

Done in #95. Just pass formality keyword argument to function:

from translatepy.translators.deepl import DeeplTranslateV2
dl = DeeplTranslateV2()
dl.translate("You are", "fr", formality="formal")

[Animenosekai]

This feature will come in the v3 (available now on the next branch)