You can now download from https://sourceforge.net/projects/easyjasub/files/easyjasub-0.6.2/ a new build of easyjasub with up-to-date documentation and an extension of the summary HTML file produced from subtitles: now you obtain also a complete word list, useful for reviewing vocabulary
I could avoid forcing wkhtmltoimage usage if I managed to make use of any Java-only HTML rendering library; CSSBox (http://cssbox.sourceforge.net/) seems promising, unfortunately it does not seem to support <ruby> tag and I have troubles making it accept my external css file…
A bugfix release, an issue with file type detection was preventing successfull usage of the tool in Linux.
There are now Travis-CI continuous builds in https://travis-ci.org/riccardove/easyjasub
This release fixes various issues notified on Sourceforge ticket #1 and #2, it also requires by default the wkhtmltoimage program (http://sf.net/p/wkhtmltopdf).
I have had problems with rendering .png images with html2image or CSSBox (recently linked).
The BDSup2Sub is now embedded to directly produce IDX/SUB files, of course it may still be useful for you to install it if you want to fine-tune the conversion from BDN/XML subtitles.
This version is more verbose to help debugging issues.
in 0.4 version you get a raw translation of some words using the JMdict (http://www.edrdg.org/jmdict/j_jmdict.html) dictionary, lookup can be perfected but yet even a not-so-perfect translation may be useful…