Limit the entries shown by default

[birtles]

As per this review: https://addons.mozilla.org/en-US/firefox/addon/10ten-ja-reader/reviews/1905937/

Copying the full text here for posterity:

I installed this extension as an alternative to RIkaikun as someone trying to transition from using Chromium-based browsers, back to firefox. 10ten basically acts the same, which I appreciate, but my one issue makes it nearly impossible to use: it shows way, way, way too many definitions.

Whereas Rikaikun shows you only the most relevant definition for the kanji compound or kana-word on hover, 10ten seems to show you every single definition for every kanji in the compound AND THEN the kana that comprising the reading of that kanji. It ALSO will take katana or hiragana words, and break them down into their individual kana and then start spitting out unrelated kanji readings so the list explodes to the point of being unusable.

For example, here is a screenshot of what I get when I hover over the word チェンソー. Whereas Rikaikun would simply give me the result for チェンソー, 10ten gives me the results for チェ、血、地、針、千…endlessly going down each kanji with "chi" as a reading.

https://i.imgur.com/kjbyAoI.png

What about お披露目？Same problem, where Rikaikun simply brings up the definition for that word alone, 10ten floods the screen with the definitions for 尾、御、緒…

https://i.imgur.com/YrwXwHP.png

Other entries that were especially bothersome: に関して brings up results for 二、荷，煮 and goes on... Kanji compounds like 実際 Are hammered by every possible definition for the usage of 実、Same with any word beginning with 人、No matter how relevant all of those usages Could possibly be to the compound in use.

Turning off dictionaries until I only have a single one turned on seems to do nothing to change these results.

I really do like this extension and how easy to use and read is. I'm sorry for such a long review. I just really wanted to lay this issue out as clearly as possible, because I do think it makes the extension much harder for people to use.

Thank you for taking the time to read this.

[birtles]

As I mentioned in my reply there, sometimes those single kanji entries are useful, e.g. when looking up 非の打ち所がない

That said, now that the popup should be scrollable on all devices, perhaps we could define an "above the fold" cut-off so that you need to scroll to see the remaining entries based on some heuristic like only showing the longest matches (or any substring matches that are still significantly long).

Doing this might also allow us to actually return more results as requested in #555.

We might also filter out some results like single-kanji results that only match on reading (not on kanji) so that, for example, we don't show the entry for 血 just because you looked up a word that started with a kana ち/チ.

Given that I've never heard anyone complain about this before, it might be best to make any change in behavior here able to be disabled with a setting.