Hello! I was interested in constructing a kanji schedule for myself with an order based around the "parts" of each kanji. For example, the character 待 has the parts [彳,寺,土,寸] in Renshuu's dictionary. I would need a master list of all the top-level parts of every kanji (at least joyo kanji), but I can't seem to find one online. There are several websites which list them for each Kanji, but no master list. Likewise, Jim Breen's "kradfile" only lists the radicals within each kanji, not kanji within kanji (e.g. 寺 within 待). I was wondering if anyone knew of a good way to acquire such a list. I suppose that I could use Renshuu's api, but the api doesn't seem to be built for bulk requests, and I feel icky about the idea of sending thousands of requests over the course of several days to avoid the daily request limit.
I’m pretty sure the parts file is part of the kanjidict download, but I can’t give you the details. Michael probably knows.
However, I can’t say I recommend this approach. Lists built off of radicals and phonetic parts make sense to me. I have invested many hours in constructing such lists for my own use, which you are welcome to take advantage of. However, I don’t see any point in making a list of all the characters that have a horizontal line, for example.
KanjiVG would be a step up from the parts file. 頑張って if you want to look into it.
Kanji schedules do have a radicals vector, so all you would need is to make a kanji schedule on what you're learning and just practice with the radicals from there. I don't see a vector that exists that covers parts/components in addition to that. In my view, that's nothing to worry about.
I would say the reason for that is it's because radicals are most favored out of convenience, due to the fact they're usually mainly based on the Kangxi Dictionary as a resource that's been around for hundreds of years. That's where there's various lists on radicals, but not everything that would make up every part/component. So this is not that difficult to find anywhere, depending on your tastes.
The thing is, there are 214 radicals, and not even everyone has memorized them in their entirety. IMO, putting parts/components on top of that would just be too much to be humanly possible to deal with, and you already have so much kanji on your plate, which does carry more meaning, and this is not quite easy to memorize given that there are still some kanji that are uncommon enough to remember. Besides, IMO, parts/components themselves are not all that significant to learn independently. They just help to learn kanji instead, especially with the evolving nature of kanji which forms their own etymologies which does not have bearing on any part/component itself.
In addition to the 214 radicals, there are approximately 500 phonetic components (in Zhengzhang’s system: I haven’t made an exact count, but have almost 500 lists already with no end in sight. The Baxtar-Sagart system apparently has even more.)
When you account for the fact that most people consider a goal of learning 1000 characters to be ambitious, the argument for studying 700+ components seems less than compelling.
Thank you for the input, everyone! The KanjiVG repository seems to contain what I was looking for. I have considered your comments about the efficacy of such a strategy. I still want to try something like this for the fun of it, but perhaps take a more measured approach. My inspiration came from trying to learn unknown kanji for vocab that I already knew, and thinking that there should be a more modular way of memorizing the more complex kanji. I think what I will try is taking all of the unstudied kanji from studied vocab, and specifically targeting components of those kanji that are 4+ stroke joyo kanji themselves. I think that would be a good balance between modularization and practicality.
In the off chance that you may find it useful, here's the list that the "filter by parts" filter for the dictionary on Renshuu uses: (organized by number of strokes).
