That's a good idea, and worth keeping in mind. Pulling just the ids off of a schedule would be an extremely light process, even for one with 10,000 or more terms in it, so it's definitely possible.
Does the "schedule" endpoint (the one that gives you a list of your schedules) really not give you schedule type (kanji/grammar/word/etc.)? Am I missing something? I'd like to extract the IDs and names for kanji schedules only. You can do it for lists, so I expected the same for schedules.
I had this problem, for my app, how i solved it was caching all schedules I had in a database, then for each schedule, I run 1 page of its contents and check if it had a "japanese" field or a "kanji_full", to keep only sentences and words. I guess you could check if it has an "Onyonmi" or "kunyomi" field. I understand this is not as practical as having a selector built in the call though.
I thought about doing that, but decided to ask first, since it felt a bit extreme... I guess I'll either do that or display all schedules. Trying to load the wrong schedule type doesn't really break anything, it's just not the best in terms of UX. Probably more of a me problem, since I have so many.
Wouldn’t it be more useful to be able to query the schedule endpoint by schedule type? That would not only save you from doing the filter on the client, but insulate clients from future changes in schedule classification.
I've built a small app that will sync your Renshuu vocabulary and grammar and then use an OpenAI provider to generate challenge sentences for you based on your skill level. It will either generate an English sentence for you to translate to Japanese, or a Japanese sentence for you to translate back to English. It can show you furigana and supports clicking words to see their definitions. It prefers definitions that came directly from Renshuu when available.
I'm looking for people to help test it on Android or Linux right now. Maybe Windows, but I care more about mobile. No iOS yet - I don't have a Mac, and I'd need to get one to use their dev tooling.
Caveats and requirements:
* You need your own OpenAI provider API key and endpoint. If you don't know what this means, this probably isn't for you. * AI is not perfect, and some OpenAI-compatible providers are better than others, so there should be no expectation that your translations are accurately scored by the LLM. * You need to understand how to install Android apps from untrusted sources or how to sideload apps with ADB
There's a button to copy the app logs to email them if you see any funny behaviors. Right now, it doesn't always do a great job picking the correct reading for kanji, and sometimes the dictionary lookup choices are bad.
If you are interested in testing, please DM me so that I can send you a link to download the APK. I will not be posting the download information here.
Note that this is not meant to replace Renshuu or steal anything from Renshuu. This is a companion app to give you "randomly" generated translation practice.
Note: since others have already commented in the other thread, please view that one for my comments on this posting (they were kind enough to reach out to me for permission to post this).
