This is the repository for the data that is used to generate the book "An introduction to Japanese", previously found at http://www.amazon.com/dp/9081507117 and other stores.
I've taken this book off the market (because I cancelled my account with my publisher and will be looking at a new way to do publishing) but the book is still available in PDF form.
I've been running it as a dokuwiki for a few years on http://grammar.nihongoresources.com and that has proven to be useful, but progressively harder to maintain as dokuwiki changes its APIs, and my content relies on specific dokuwiki plugins for which I don't have time to actually update them to work with the new versions.
So, I'll be trying to rewrite it to a self-contained website that you can file github issues for, with a php compiler to turn the raw data into .pdf data.
Interesting fact: just because someone's front door is open doesn't mean you have the right to take their stuff, and in the same vein: just because I'm hosting this on a public repository does not give you the right to use the code and raw data for your own purposes.
This book is free for the general public in PDF form, and available as the more traditional (affordable, gasp!) paper textbook version at book retailers, but this is a product, not a project: all code and data is owned by me, Mike "Pomax" Kamermans, and I reserve all rights. You expressly do not have permission to start compiling your own version (except to test the compilation from plain text to .pdf), nor do you have permission to distribute this code or data yourself.
You are, however, welcome to help improve the text, or suggesting compile improvements, by filing issues or submitting pull requests. Contributions deserve acknowledgements in the book's acknowledgement section.
Live site: https://Pomax.github.io/nrGrammar
Compilation of book relies on some older technologies in part because it was initially created back in 2008 when PHP was still the king of "get shit done fast" scripting langauges for people who needed one scripting language to do everything from CLI to web content (was has since been replaced by JavaScript), and LuaTex was still four years away from a version 1, making the only sane TeX choice XeLaTeX (because XeTeX is natively utf8, rather than needing special instructions just to understand utf8 documents).
Localised content is housed in ./data/pages, where each
locale uses its own locale code as directory name. Inside
of those, all files should follow the filename convention
as used in the en-GB directory, in order for the compilation
scripts to find them.
Translations should follow the spirit of the text, not the letter of text: if an idiom or illustratrive piece of text does not work in the language you're targeting, please do not translate it literally. The worst thing you can do when translating is to take text that works in one language, and force-translating it to a weird, clunky text in a different language. Sometimes that might mean rewriting an entire paragraph, or even more than one, in order for the discourse to flow naturally in your target language: that is fine. It is in fact infinitely preferable over a literal translation that isn't up to whatever is considered university-level langauge in the country/region you're localizing for.