𢦏, a phonetic component which is not a radical and is also not used as a character, is not present in the kanji dictionary. That’s as one would expect; however, if you search for 𢦏, you get this:
Note that the above is 𠮟, not 叱. 𠮟 was not previously in the kanji dictionary; now it is, but it seems to have been added in a half-broken way.
There are also dozens of irregular vocabulary items, none of which belong there. A few of them could arguably be regular readings, but most of them are simply garbage.
This is very interesting. I do not have time to look into it, but I have to wonder if the data encoding of the database column in mysql is causing these to come up as the same code point.
Let me ask again, just to make sure there is no misunderstanding. @マイコー, did you administratively remove a post for security reasons? Would you like me to send you the information again via DM?
Good, I didn’t think so, but I couldn’t rule it out.
In response to your comment that the problem might be in MySQL, I speculated that the issue most likely was tied to error handling. Specifically, I advised you to check the code that runs after a null query result for type conversion errors that might strip away bits.
Yea, not sure where that post went. Anyway, It's not..any of that, although I appreciate the suggestion!
So, while the collision is most likely due to the encoding, most of those kanji are not in renshuu's database to begin with, so when I fix it (not now), it'll simply return a "kanji not found", and not the actual kanji. They can be added, but due to the rarity of these kanji, it's simply not something I can work on at the moment.
I'm pretty sure that none of those were part of the original kanji imports that I made into renshuu's database. While there is also the code point issue that is tied into the collation used by the database, even if that wasn't happening, most if not all of these simply are not in renshuu's database to begin with, and I do not expect to have the time to add them in the near future.
Okay, I think I’ve figured out the vocabulary artifacts at least. These all seem to be dictionary entries that reference a missing character, usually as an alternate spelling. Adding these ten or so missing characters would be an easy workaround that would hide the most glaring symptoms of the bug and make it much less likely that the average user will encounter it.
