kanji.py

import json
import sys
import xml.etree.ElementTree as ET

def main():
    # preprocess
    kanji_info = {}

    # from kanji dict xml file
    tree = ET.parse('./res/kanji.xml')
    root = tree.getroot()
    for ch in root.iter('character'):
        info = {}
        info['codepoint'] = ch.find('codepoint/cp_value[@cp_type="ucs"]').text
        stroke_count = ch.find('misc/stroke_count')
        info['stroke_count'] = int(stroke_count.text) if stroke_count is not None else 0
        cp = int(info['codepoint'], 16)
        kanji_info[cp] = info

    print >> sys.stderr, 'Finish processing Kanjidict file'

    # from jlpt files
    jlpt = []
    for level in range(5):
        with open('./res/jlpt-n{:d}.txt'.format(level+1)) as f:
            kanji = [kanji_info[ord(unicode(ch.strip(), 'utf8'))] for ch in f if len(ch) > 0]
            jlpt.append({'title': 'N%d' % (level+1), 'characters': kanji})

    # with open('./res/radical.txt') as f:
    #     kanji = [kanji_info[k] for k in [ord(unicode(ch.strip(), 'utf8')) for ch in f if len(ch) > 0] if k in kanji_info]
    #     jlpt["rad"] = kanji

    json.dump(jlpt, sys.stdout, separators=(',', ':'), sort_keys=True)

    print >> sys.stderr, 'Finish processing JLPT files'


if __name__ == '__main__':
    sys.exit(int(main() or 0))