Import a Yomiwa (iOS) word list into Anki.

toanki.py

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

# Explanation:
# Export a word list from Yomiwa iOS app to a file, convert to a tab-separated file, and import to Anki.
# Installation:
# 1. pip install jamdict
# 2. Follow jamdict's github page instructions to install the dictionary database.
# Usage:
# In Yomiwa, add words to a list.
# 1. Add example sentences by adding the word first, then by pressing analyze text and starring it.
# 2. Name the list 'anki' and choose 'export file' and move it to your computer.
# 3. Run this script in the same folder.
# 4. Import import_to_anki.txt into anki with tab-separated values setting.
#    The card type has to have the following fields:
#       kanji, reading, meaning, example sentence, pitch accent, tag
# 5. In the rare case that the word was not found in the dictionary but you added an example sentence, it will be added as a numeric ID. You can then manually check what the word is and fix it.
# 6. You can generate pitch accent graphs to the 'pitch accent' fields with some Anki plugin.

import re
import json
from jamdict import Jamdict

jam = Jamdict()
to_write = ''
word_count = 0
unfound_word_row = ''
tag = 'Yomiwa'

def finishRow(sentence = ''):
  # Add fields: Example sentence, pitch accent, tags
  # Assume there is already a tag before 'sentence' field to be added next.
  if len(sentence) == 0:
    return '\t' + '\t' + tag + '\n'
  else:
    return sentence + '\t' + '\t' + tag + '\n'

with open('anki_list.ymwlist', 'r') as f:
  data_str = f.read().replace('\n', '')
  data_str = re.sub('\n', ' ', data_str)
  # print(data_str)
  data = json.loads(data_str)['IDs']
  # print(data)

  # Newer versions of Yomiwa reverse the order.
  # Change order to oldest first.
  data.reverse()

  for index, entry in enumerate(data):
    if not isinstance(entry, int):
      # Skip over a non integer entry (= sentence) since it was handled together with the word.
      continue

    # Handle word.

    unfound_word_row = ''
    word_id = entry
    word = jam.lookup('id#' + str(word_id))
    word_count += 1

    if len(word.entries) == 0:
      # No definition found.
      unfound_word_row = str(word_id) + '\t\t\t'
    else:
      # Fill in word info.
      kana_form = word.entries[0].kana_forms[0]
      kanji_form = kana_form
      if len(word.entries[0].kanji_forms) > 0:
        kanji_form = word.entries[0].kanji_forms[0]
      senses = []
      # Strip out
      # - things like '((noun (common) (futsuumeishi)))'
      # - change '/' to ', '
      for s in word.entries[0].senses:
        s = str(s).replace('/', ', ')
        senses.append(re.sub(' [((](.*)[))]', '', s))
      meanings = ', '.join(senses)
      to_write += str(kanji_form) + '\t' + str(kana_form) + '\t' + meanings + '\t'

    # Handle example sentence.

    if index + 1 == len(data):
      # Last word had no example sentence.
      to_write += finishRow()
      continue

    sentence = data[index + 1]

    if not isinstance(sentence, int):
      # Example sentence
      if len(unfound_word_row) > 0:
        # Add the unfound word with a sentence, so the word can be manually added.
        # Spot by a number in the first field.
        to_write += unfound_word_row
      sentence = sentence.replace('\n', '')
      to_write += finishRow(sentence)
      continue
    else:
      # There was no sentence, just finish this row and read the entry as a word.
      to_write += finishRow()

f = open("import_to_anki.txt", "w")
f.write(to_write)
f.close()

print(str(word_count) + ' words exported to import_to_anki.txt')