Add Japanese data - #209
Conversation
tiffanyannedo
left a comment
tiffanyannedo
left a comment
There was a problem hiding this comment.
Adding suggestions to the JA set. Translations have been verified and committed directly to the branch
| "북측 주차장" | ||
| ], | ||
| "ja": [ | ||
| "ガレージエー", |
There was a problem hiding this comment.
It looks like 'garage a' and 'main garage' is used in a number of garage selections. Not sure if this is wanted?
There was a problem hiding this comment.
I think this is worth addressing, but separately. The aliases are a little bit overlapping and I think that could be a real problem even in English
There was a problem hiding this comment.
That's fine, just wanted to bring it up. The aliases provided for JA are fine if we keep the same 'base'. Can take a look again when we address it separately
| "first_name": "美穂", | ||
| "last_name": "渡辺", | ||
| "companion": { | ||
| "first_name": "佑樹", |
There was a problem hiding this comment.
@shama7gupta In Japan it is not customary for spouses to share a last name or for the wife to take on the husband's last name. It would be more realistic if there is also a last_name field for the companion
There was a problem hiding this comment.
not sure this could be easily handled, because it informs the user goal and all that, and the specific language on this record is a woman calling about her husband. This should be minor enough to allow?
There was a problem hiding this comment.
That's a good point, that could require some major restructuring of the data. It's not a super realistic scenario but it should be fine for this sort of evaluation
Resolves conflicts from the parallel locale additions in #208 (Dutch) and #207 (Italian). All three PRs appended a new locale key at the same position in the same files, so every conflict was a false conflict resolved by keeping all languages. The 27 conflicted JSON data files were resolved with a three-way structural merge, which reported zero real disagreements. Verified that stripping the "ja" keys from the result reproduces origin/main exactly, that all 663 Japanese keys survive, and that the nl/it/ko/en/de/hi/es/fr key sets are byte-identical to main.
3 participants
Adds japanese data to medical_hr, itsm, and airline domains.