Adaptation of OCR script (WebOCR) with Spellcheck (Aspell) to work with Japanese spellcheck. At the moment, my site works with European languages.
There isn't a Japanese aspell dictionary so this will be developed.
The basic process is to add a space between kanji (Unicode range U 4E00âÂÂU 9FBF) and other characters. This keeps kanji together in blocks and the other characters in blocks. This is to overcome the fact that Japanese does not use spaces.
When the user removes the spaces between the kanji and connected characters, the search can look for these strings of kanji and other character blocks.
When editing has finished, at the end, the spaces can be removed in a batch.
Used Code:
Complete zip package can be found at http://www.broken-notebook.com/spell_checker/.
Other source is Aspell, NHOCR and WeOCR.