Sentencegator - small tool that give you sentences based on your WK level

tepmex · October 30, 2013, 9:52pm

今日は皆さん！

I wrote small tool on Google Go programming language. Its tool looking in Tanuka Corpus list of sentences(over 150k items) for those which contains only your WK vocab items and kana symbols.

https://github.com/TepMex/Sentencegator/releases

HOW TO USE
0. Download and unzip archive.
1. Run executable file(Sentencegator.exe), and Sentencegator binds on 80 port of your machine.
2. Open in your browser http://localhost 3. Use web interface NOTE: If firewall or any security program asks your permission, accept it
B-line is some additional information about sentence(A-line).

Example sentence with B-line:

A: またいつか風のように走るんだ。 Someday I’ll run like the wind.

B: 又{また} 何時か{いつか} 風(かぜ) の様に[01]{のように} 走る[01] のだ{んだ}

Also, since version 2.0 Sentencegator contains KanjiStats module.
KanjiStats useful if you want to know how many kanji you need to learn for reading your favorite ranobe or something else.
You choose japanese text file on your local machine(only utf8 encoding supports), and see full statistics about this.

See your API Key: http://www.wanikani.com/account

Have fun

Smasher816 · October 30, 2013, 10:01pm

Awesome tool. I got 971 sentences at level 3. It looks like I need to start studying my grammar some more.

Two comments.
1) Any chance of a linux version? (it works in Wine, but where is the fun in that :P)
2) Is there any way to (easily) turn this into a custom anki deck?

Janitha · October 30, 2013, 10:05pm

Amazing!!! I can’t wait to try out this later this evening. Thank you so much, tepmex!!!

Gabriel · October 30, 2013, 10:08pm

When I open Sentencegator.exe, it seems to load the sentences and all, but then it just closes, and the result.txt is blank.

CarnivalQueue · October 30, 2013, 10:09pm

It always makes me a little sad to see something awesome like this and not be technically savvy enough to use it.

tepmex · October 30, 2013, 10:14pm

Smasher816 said... Awesome tool. I got 971 sentences at level 3. It looks like I need to start studying my grammar some more.

Two comments.
1) Any chance of a linux version? (it works in Wine, but where is the fun in that :P)
2) Is there any way to (easily) turn this into a custom anki deck?

Thanks for feedback!

I will try to compile it for linux soon. Hope that no have problems with it.

About export to anki - all sentences is TAB-separated(japanese-TAB-english), i think export it to anki deck is possible. And in my plans is write output result in CSV.

Juichiro · October 30, 2013, 10:16pm

The Tanaka Corpus is useless. This has been repeatedly said. The sentences are no better than Machine translations (probably not even better than Google translations). If you were to find an actual corpus of reliable data I would take my hat off to you.

tepmex · October 30, 2013, 10:18pm

Gabriel said... When I open Sentencegator.exe, it seems to load the sentences and all, but then it just closes, and the result.txt is blank.

Kingcarnie said...It always makes me a little sad to see something awesome like this and not be technically savvy enough to use it. :(

Press Win+R, type %PATH%/Sentencegator --apik=%API_KEY%, press Enter
%PATH% - path to foldier where you unzip Sentencegator
%API_KEY% - your WaniKani Api key, you can see it here

tepmex · October 30, 2013, 10:25pm

Juichiro said... The Tanaka Corpus is useless. This has been repeatedly said. The sentences are no better than Machine translations (probably not even better than Google translations). If you were to find an actual corpus of reliable data I would take my hat off to you.

I not use the original Tanuka Corpus files, which contain very big amount of mistakes. I use edited version of this, and its looks normal, at least better than machine translation =).

Juichiro · October 30, 2013, 10:28pm

tepmex said...
Juichiro said... The Tanaka Corpus is useless. This has been repeatedly said. The sentences are no better than Machine translations (probably not even better than Google translations). If you were to find an actual corpus of reliable data I would take my hat off to you.

I not use the original Tanuka Corpus files, which contain very big amount of mistakes. I use edited version of this, and its looks normal, at least better than machine translation =).

Ah, great. Link kudasai?

Gabriel · October 30, 2013, 10:32pm

tepmex said...
Gabriel said... When I open Sentencegator.exe, it seems to load the sentences and all, but then it just closes, and the result.txt is blank.

Kingcarnie said...It always makes me a little sad to see something awesome like this and not be technically savvy enough to use it. :(
Press Win+R, type %PATH%/Sentencegator --apik=%API_KEY%, press Enter
%PATH% - path to foldier where you unzip Sentencegator
%API_KEY% - your WaniKani Api key, you can see it here

Thanks, it worked. :)

5,387 sentences. Nice!

tepmex · October 30, 2013, 10:35pm

Ah, great. Link kudasai?

http://www.csse.monash.edu.au/~jwb/examples.utf.gz

I guess this DB used in WWWJDIC and jisho.org

Segmelsian · October 30, 2013, 10:59pm

Its worth noting that if you have the GO tools installed on your system you can run it using go run main.go --apik=$APIKEY
This at least works in Max OSX 10.9, and thus I’m assuming would work similarly for whatever distributions are supported by Go.
That being said this looks great thanks

tosirius · October 30, 2013, 11:25pm

Haven’t tried it as of yet but this looks great! I was hoping to find something like this. It’s great learning the grammar and all but you lose it quickly if you cant find any examples.

Breathless · October 31, 2013, 12:35am

"The version isn’t compatible with your Windows version.[…]"
Windows 7 Ultimate SP1 32bits

petitcrabe · October 31, 2013, 12:41am

This looks very promising. I opened my .txt file and it’s a HUGE wall of text. Anyway to get the sentences in an organized way?

galymzhan · October 31, 2013, 1:04am

Compile under linux with “go build main.go -o sentencegator”, then execute “./sentencegator --apik=APIKEY”.
Thanks for this

dlemay · October 31, 2013, 1:04am

I opened it using Notepad++. I would also recommend opening it with Excel. Import the data as tab delimited values so it will split Japanese and English in separate cells. Putting it in a spreadsheet also allows you to create an Anki deck with those sentences.

Besamel · October 31, 2013, 1:14am

Thanks for this.

Smasher816 · October 31, 2013, 1:17am

Segmelsian said... Its worth noting that if you have the GO tools installed on your system you can run it using go run main.go --apik=$APIKEY
This at least works in Max OSX 10.9, and thus I'm assuming would work similarly for whatever distributions are supported by Go.
That being said this looks great thanks :)

Thanks. I just figured out there is a go package for my linux distro.

Using "$ go run main.go --apik=***" works great :D