Google only has 17 results for "権利をかける", it seems, if you click through to the second page of results

Yahoo! returns 16 results

Hehe, yeah :-)

I'm not sure which engine Yahoo! Japan uses at the moment

Ah, Yahoo! Japan uses Google now

That explains it

search.yahoo.co.jp/…

@Earthliŋ What makes you guess that?

I'm not sure how they get the results from Google.

But there may be another explanation

There isn't enough information to guess that it's outdated or cached

It's a possible explanation, though

Hehe.

Well, I haven't worked there for a few years now

So I don't know what they do anymore

Yeah

You can tell, because I'm one of the only people on the planet who bothers typing the exclamation point

Hah

Yahoo! Japan is actually a separate company from Yahoo!

Unlike Google Japan, which is just Google

It was the first intl that Yahoo! set up. They made sure to own a controlling interest in all of the others where possible (I think China is a special case due to laws about Chinese companies being owned by foreign companies)

But Yahoo! Japan is special! :-)

Also, happily, Yahoo! Japan is one of the few places where Yahoo! Auctions took off (it bombed almost everywhere else, though it did okay in a couple other intls like Singapore)

So I was part of the team that worked on that :-)

Because Yahoo! Auctions got there first

Same reason there's no Yahoo! Auctions in the US

Because eBay got there first

It's easy to corner that sort of market with a network effect

No, but I used to live with someone who did

Haha.

See you later!

source-error is gone!

@TokyoNagoya We were just making guesses because we didn't know. I don't think anyone is trying to say they know more Japanese than you

I'm just learning, and part of the process of learning is discussing the language :-)

Only the top three kanji make the cut in the ACA's list (which is mostly limited to the most frequent 3500 kanji)

And of those three, the second and third are almost always read with readings other than シュツ...

出 is the only one on the 常用漢字表 with シュツ listed

None of the remaining five have any words listed in EDICT with the シュツ reading. I don't see your 刃恤 word anywhere... What is it?

So it seems like practically speaking, 出 is the only kanji read シュツ, with very rare exceptions

Hmm...

初 is also read chū in Mandarin, but that's ショ in Japanese

Baxter-Sagart's Middle Chinese transcriptions and Old Chinese reconstructions indicate that 初 had a different vowel back then

That's all I've got :-)

I don't really know why there aren't more kanji read シュツ

@Earthliŋ There are lots of frequency lists. My first resort is to check this list‌​, but it only covers 3500 kanji

Some of the other lists have maybe less reliable data, but they cover more kanji

Or they just collected it in different ways

So I made a composite list to try to improve the coverage

It's not put together in a rigorous fashion or anything :-)

But if you want to use it for some reason, it's here: quarplet.com/composite.txt

@Earthliŋ I don't know of a list with those characteristics, but with certain caveats on the sources used for the data, you could assemble it yourself

@Earthliŋ The Wikipedia list has raw frequency: shang.kapsi.fi/kanji

Well, Wikipedia isn't exactly a balanced corpus

Oh, nice! Did you set up a script to do it?

Did you notice you can use some RE syntax on BCCWJ?

You can search for 権利を, and then in the 後文脈 field type ^.け

Then you get all the results where the following context begins with any character and then け

And then you can click the 後文脈 header to sort by it

@Earthliŋ To make the extra fields appear, click the こちら link

kotonoha.gr.jp/shonagon/…

It should work if you have a recent session cookie

Otherwise you'll probably get booted back to the intro page

1. Install LiveHttpHeaders

2. Perform a query

3. Look in the LiveHttpHeaders header output for the POSTed content. It'll be under a content-length line.

Because the form data isn't being put directly into the URL (via GET), it's natural to conclude that it's being submitted via POST

Although you could verify this rather than inferring it by looking at the page source, or by looking through the DOM etc.

It looks like this for me: query_string=%E5%B0%86&lcontext_regex=&rcontext_regex=&media=%E6%9B%B8%E7%B1%8D&‌​media=%E9%9B%91%E8%AA%8C&media=%E6%96%B0%E8%81%9E&media=%E7%99%BD%E6%9B%B8&media=‌​%E6%95%99%E7%A7%91%E6%9B%B8&media=%E5%BA%83%E5%A0%B1%E7%B4%99&media=Yahoo%21%E7%9‌​F%A5%E6%81%B5%E8%A2%8B&media=Yahoo%21%E3%83%96%E3%83%AD%E3%82%B0&media=%E9%9F%BB%‌​E6%96%87&media=%E6%B3%95%E5%BE%8B&media=%E5%9B%BD%E4%BC%9A%E4%BC%9A%E8%AD%B0%E9%8‌​C%B2&entire_period=1

That's URI-escaped, of course. That's fine. I removed %E5%B0%86 (which was 将, the character I searched for) and replaced it with a URI-escaped version of 未

Then I changed that to GET by appending it to the URL after ?

It turns out, the server accepts the query via both GET and POST. (This is not a given. In this case it worked, but in another case this might not.)

This gave me the basic prototype URL. The first three fields are what you'd expect, query string and left and right REs for matching context. The remainder of the arguments specify which subcorpora to query--you can't leave these out. I didn't test whether entire_period was required or not, but it presumably does what it says

(I haven't tried restricting queries in BCCWJ to specific years)

So, you could do this same thing yourself. But you'll notice that it's stateful--it only works with the session cookie it creates when you agree to the terms.

Well, then you can't give someone a link directly to it, because you can't put the data in the URL. But you can still grab the data yourself with whatever tool you want.

Like, curl for example.

If you were going to write a script to query for each kanji in a list, you could still do it if it required POST.

There are any number of tools. You could put a script together in perl or python, and you could use either a library (use libcurl from perl via WWW:Curl or python via pycurl) or just call out to an appropriate command-line tool (curl, wget, etc.)

Sure. You can access MySQL (is that what you're using?) via perl's DBD::mysql or via python's MySQLdb

I've been using perl for most of my life so it's second-nature to me, but I think I'd recommend you start with python if you're starting from zero

I tend to write small one-off scripts in perl still

Remember that if you automate access to someone's webserver that you're a guest. Don't hammer it--put in a generous delay between requests, like what you'd do as a human

Yay

Hehe!

You can do it!