#!/usr/bin/env python
# -*- coding: utf-8 -*-

'''
This script is for online dictionaries.
The variation of dictionaries are
  English-Japanese, Japanese-English and Wikipedia
and they are based on the Web Service provided by EAST.

You can get original and more information at
  http://dejizo.jp/dev/index.html

But this script support only REST version,
and you can get the document about it at
  http://dejizo.jp/dev/rest.html

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Sample code

>>> import dejizo
>>> items = dejizo.get_items('EJdict', 'dict')
>>> items
[{'expr': u'dict.', 'id': u'011348'}, {'expr': u'dicta', 'id': u'011349'}, {'expr': u'Dictaphone', 'id': u'011350'}, {'expr': u'dictate', 'id': u'011351'}, {'expr': u'dictation', 'id': u'011352'}, {'expr': u'dictator', 'id': u'011353'}, {'expr': u'dictatorial', 'id': u'011354'}, {'expr': u'dictatorship', 'id': u'011355'}, {'expr': u'diction', 'id': u'011356'}, {'expr': u'dictionary', 'id': u'011357'}]
>>> detail = dejizo.get_detail('EJdict', '011348')
>>> detail
{'body': u'dictation\tdictator\tdictionary', 'head': u'dict.'}

'''

import urllib, xml.dom.minidom, unicodedata, re

ROOT = 'http://public.dejizo.jp/NetDicV09.asmx'

##
def _dom(url):
    '''
    Download XML and parse it to DOM
    '''
    f = urllib.urlopen(url)
    p = f.read()
    f.close()

    return xml.dom.minidom.parseString(p)

##
def _text(node):
    '''
    Extract text data from DOM element node
    '''
    o = u''

    for child in node.childNodes:
        if child.nodeType == child.TEXT_NODE:
            o += child.data
        elif child.hasChildNodes():
            o += _text(child)

    return o

##
def _norm(us):
    '''
    Normalize Unicode string
    '''
    us = unicodedata.normalize('NFKC', us).strip()
    us = re.compile('\s\s+').sub(u' ', us)

    return us

##
def get_items(dic, word, scope = 'HEADWORD', match = 'STARTWITH', size = 10):
    '''
    This function is used for getting list of words.
    It contains maps whose keys are "id" and "expr"

    You can set some parameters as below
     - dic: what dictionary you want to check: "EJdict", "EdictJE" or "wpedia"
     - scope: target scope to search: "HEADWORD" or "ANYWHERE"
     - match: how to match: "STARTWITH", "ENDWITH", "EXACT" or "CONTAIN"
     - size: how many results do you want: I don't know neither the min nor max

     In this function, the parameters are NOT validated, so be careful...
     See also "http://dejizo.jp/dev/rest.html" for detail.
    '''
    m = {'Dic': dic, 'Word': word.encode('utf-8'),
         'Scope': scope, 'Match': match,
         'Merge': 'AND', 'Prof': 'XHTML', 'PageSize': size, 'PageIndex': 0}

    url = '%s/SearchDicItemLite?%s' % (ROOT, urllib.urlencode(m))
    doc = _dom(url)

    ids = doc.getElementsByTagName('ItemID')
    spans = doc.getElementsByTagName('span')

    items = [{'id': _text(id), 'expr': _text(span)}
             for id, span in zip(ids, spans)]

    return items

##
def get_detail(dic, id):
    '''
    This function is used for getting detail of a word.

    You can and must set only "dic" and "id".
     - dic: what dictionary you want to check: "EJdict", "EdictJE" or "wpedia"
     - id: ID of the word. You can get it by the function "get_items()"
    '''
    m = {'Dic': dic, 'Item': id, 'Loc': '', 'Prof': 'XHTML'}

    url = '%s/GetDicItemLite?%s' % (ROOT, urllib.urlencode(m))
    doc = _dom(url)

    head = _norm(_text(doc.getElementsByTagName('Head')[0]))
    body = _norm(_text(doc.getElementsByTagName('Body')[0]))

    return {'head': head, 'body': body}

##
def _format_items(items):
    '''
    This function is used for showing items only in command line mode.
    '''
    o = [u'']
    for idx, item in enumerate(items):
        w = u' "%s":%d,' % (item['expr'], idx)
        if len(o[-1] + w) > 60: o.append(w)
        else:                   o[-1] += w

    print u'\n'.join(o)

##
def _main():
    '''
    Command line mode
    '''
    import sys

    available_dicts = {'ej': 'EJdict', 'je': 'EdictJE', 'wp': 'wpedia'}

    if len(sys.argv) < 3 or not available_dicts.has_key(sys.argv[1]):
        sys.exit('  Usage: %s ej|je|wp word' % sys.argv[0])
    else:
        dic = available_dicts[sys.argv[1]]
        word = unicode(sys.argv[2], 'utf-8', 'ignore')
        items = get_items(dic, word)

        if not items:
            sys.exit('Woops! You got no result!')

    if len(sys.argv) < 4:
        _format_items(items)
    else:
        id = items[int(sys.argv[3])]['id']
        item = get_detail(dic, id)
        print item['head'], ':'
        print item['body'].replace(u'\t', u'\n')

##
if __name__ == '__main__':
    _main()