Fix for #16 and a few bonus changes

[Mithgol]

• Moved compatibility checks out of methods (fixes for the sake of speed, move compatibility checks out of your methods #16).
• Used window[type + 'Array'] instead of all[type + 'Array'], otherwise several tests fail in Firefox 13 with all is not defined.
• Restructured the first three tests to avoid the Expected at least one assertion, but none were run - call expect(0) to accept zero assertions message.
• As MDN docs say, arguments.callee is expensive, and forbidden in ECMAScript 5 strict mode, and should be avoided in favour of the function's name unless it's anonymous. Having considered that, I used this instanceof jDataView instead of this instanceof arguments.callee in the constructor.

[vjeux]

I'm sorry for the delay but I've been really busy the last few weeks, I'm going to take a look at it as soon as I find some time :)

Thanks for working on it, that's awesome btw!

[Mithgol]

Thank you for your attention. I understand that this pull request is larger than the other two that you merged today and thus it will take more time to evaluate.

By the way, the 28a3ab9 commit brings here several lines of code from the already merged #19 to avoid a merge conflict in msysgit that I encountered today.

[Mithgol]

The e1c16be fix for #22 is flawed (does not pass the tests).

Do not merge yet.

[Mithgol]

I came to an understanding that the tests were written to expect little-endian defaults. That defaults, however, are removed in e1c16be.

You may merge this pull request immediately, just remember to rewrite the multi-byte tests and the jParser after that.

[Mithgol]

Of course, this should be done carefully.

I guess the jDataView's version has to be pulled — and then the new jParser made to require that new version.

[Mithgol]

Tests are done, and now jParser is the only thing that remains to be rewritten after #22 is fixed.

[vjeux]

I'm going to sleep, I'll do that tomorrow :) Thanks a lot for the help on it

[Mithgol]

:)

[Mithgol]

jParser would be corrected by vjeux/jParser#1 and that fix seems backwards-compatible (contrary to what I feared).

Thus the version bump to 1.1.0 here does not seem necessary for jParser, but may still be handy for other modules out there.

[vjeux]

if we are in node and use an ArrayBuffer, it's going to crash as window does not exist.

[Mithgol]

Fixed, thanks.

[Mithgol]

After #16 and #23 is closed, the only real value of this commit is that it contains a fix for #22, but a pull request for that should be separate.

Closing.

[Mithgol]

That new (separate) pull request is #24.