[ja] Refactor Japanese transforms names and add English descriptions

[Lyroxide]

In tandem with #1000

Closes #1053

This PR changes the following:

• Add robust English descriptions (except for ksb). Made an attempt to show all meanings of each inflection and how it can be conjugated.
• Change names for:

adv -> -ku: adv is simply unclear.

polite -> -masu: to standardise suffix naming convention

past -> -ta: it has meanings other than past tense (a big beginner trap)

masu stem -> continuative: again, only to be found in English translations. ます語幹 is not a thing, but the 連用形.

noun -> -sa: to standardise naming

progressive or perfect -> -teiru: has meanings other than progressive or perfect

• Add Japanese i18n. This is not as robust, but then this feature is also not supported yet.

References:
imabi.org
新明解国語辞典 第八版
三省堂国語辞典 第八版
大辞林 第四版
JMdict

[StefanVukovic99]

japanese-transforms.test.js is failing because adv got renamed to adverb

[Lyroxide]

Yeah I know. Similar to adv, I'm thinking if other names are appropriate. It seems like there's no consistency.

we have polite which is the -masu conjugation but we also have 'masu stem', which is more accurate to say it is the continuative form or renyoukei, but i dont know which one is better to be the name (which is shown directly) or description

if you see some of the ja i18n, some of it are not exact translations of the english name.

let me know your opinions because im way past a beginner and i do not know what would a beginner think when they see stuff like polite or progressive or perfect as opposed to -ba

[Lyroxide]

Ready for review. The main description has been updated to show the changes.

[StefanVukovic99]

There's a ton of work here, thanks 🧎‍♂️
I see this as 4 separate concerns:

1. add descriptions
2. rename inflections
3. add causative inflections
4. remove negative imperative

it'd probably be a pain to decouple 1 and 2, but i think you can move 3 and 4 into their own PRs

[StefanVukovic99]

I'd like to hear more opinions on this one, since choosing whether we'll refer to this as masu stem or continuative impacts a lot of the other descriptions. I think it's worth keeping it as masu stem since that seems more typical for beginner content

[Lyroxide]

This is a hill I will die on to have this change. I don't know what beginner content you're referring to, but not everyone learnt grammar from the same resources that teach 'masu stem.' In fact, I never knew what was that until I used Yomichan. To clarify what is a stem,

書く
語幹＝書
未然形＝か、こ
連用形＝き
終止形＝く
連体形＝く
仮定形＝け
命令形＝け

stem is the unchanging part of the word after inflection, i.e. the common part in its paradigm. Here, clearly the stem is 書, while some other sources would say 'kak' is the stem. How one would find the stem in English or other languages is irrelevant; in Japanese, nobody calls the 'part where you conjugate masu then drop the masu' the 'masu stem'.

Well, I don't want the deinflections to be littered with only linguistically accurate words, but this is a misnomer that I can't close an eye to.

[StefanVukovic99]

I don't think the nomenclature per se is very important, but this is a change to existing user-facing stuff so want to think it through.

For beginner content, the tmw guide/grammar recommends cure dolly, tae kim, DoJG.
Dolly calls it い-stem, tae calls it the stem. Both mention masu stem but don't like it:

(dolly) In English it's sometimes called the "ます stem", which is not inaccurate because of course
it does, among many, many other things, connect the helper-verb "ます".
I don't like this expression because it stems from the whole ecology where
the so-called ます-form of verbs, which means verbs shifted into the い-stem
and having the "ます" helper verb attached to them, were the base-form of verbs.

(tae) This is often called the ます-stem in Japanese textbooks but we will call it just the stem because it is used in many more conjugations than just its ます -form.

Dolly mentions 連用形 a bit, neither mentions continuative. DoJG seems to call it Vmasu.

All of the options are bad in their own way...

[codspeed-hq]

CodSpeed Performance Report

Merging #1007 will not alter performance

_{Comparing Lyroxide:add-descriptions (d9696a7) with master (1d2a915)}

Summary

✅ 5 untouched benchmarks

[github-actions]

⚠️ Visual differences introduced by this PR; please validate if they are desirable.

View Playwright Report (note: open the "playwright-report" artifact)