Skip to content

HTTPS clone URL

Subversion checkout URL

You can clone with
or
.
Download ZIP
Interactive Twitter feed for Japanese word definitions
PHP Perl Erlang
branch: master

Fetching latest commit…

Cannot retrieve the latest commit at this time

Failed to load latest commit information.
conf
csv
img
logs
php
scripts
sql
.gitignore
INSTALL
LICENCE
README.md

README.md

Japxlate

What is it?

It's a Twitter bot that tweets out random Japanese words - with definitions - once an hour. It also attempts to define people's words (Japanese or English) that they Tweet in. Usage example sentences are occasionally Tweeted.

英語定義が欲しい場合は「@japxlate 日本語単語」をツイートしてください。リプライで教えます。Japxlate は日本語母語話者用にも便利!

What's new?

  • Finally comes with a script to process the EDICT dictionary file (EDICT2)
  • Updated INSTALL instructions
  • Full docblock comments for all PHP classes, methods and functions
  • Restructured class internals
  • Updated for Twitter API v1.1
  • Improved logging

How does it work?

The EDICT dictionary project (http://www.csse.monash.edu.au/~jwb/cgi-bin/wwwjdic.cgi) is used for the random words and to define people's words. The Tatoeba project (http://tatoeba.org) is used for usage example sentences. Abraham Williams' TwitterOAuth class (https://github.com/abraham/twitteroauth) is used to connect to Twitter.

What does it need to run?

  • Commandline PHP
  • mbstring extension for PHP
  • JSON extension for PHP
  • mysqli extension for PHP (but should be easy to alter everything for another MySQL extension or indeed another RDBMS)
  • MySQL (with Unicode ability)
  • The EDICT2 dictionary file
  • Some CSV files from Tatoeba.org
  • You'll need a Twitter account and then a "My application" (so you can get some OAuth access tokens etc). You'll want "Read, Write and Access direct messages" permission.

What's hot?

  • Source dictionary (EDICT) is of top quality
  • Interactive - Twitter users can use the bot as a word defining service
  • Bot owner can seed the next random word by using a to-self direct message
  • Easily modified for any other source dictionary

What's not?

  • Usage example sentences should match the previous random word, but sometimes don't
  • Direct message seeding can be iffy - I don't think Twitter likes it when users DM themselves!
  • Lacks any kind of "cashing in" mechanism for current trending hashtags (ie. defining the current trend words but in Japanese would be nifty)
  • As a timed robot it's pretty cheesy because all it does is call sleep() LOL!
  • Obviously you've got all your DB and Twitter credentials in plaintext in the configuration.php
  • Defining people's words needs to be less strict and "hit" the dictionary more
  • Definition formatting logic is duplicated in MANY places in the code - must consolidate
Something went wrong with that request. Please try again.