Added the preprocessor directives for changing the default number base of literals

I've added the preprocessor directives base_decimal and base_hexadecimal, that can be used to change how the assembler assembler numerical literals for which the base is not specified. The default is hexadecimal, which I don't like (but changing it would break almost all programs), so I added the preprocessor directive to make it possible to change that default.

Improved the behaviour in case of a server error during upload

@abdrd made it so that, in the event of a server error, the error is logged into the console using console.error. Now I've improved that by making the errors be shown to the user with alert.

Made the Download Hexadecimal button check whether the assembler has been run

I've made the "Download Hexadecimal" button check whether the assembler has been run since the last modification of the assembly code, and warn the user if it hasn't been.

Added support for octal numerals

So, I've added the support for octal numerals, which can be used by appending 'o. I've also added an example program that uses them.

Fixed a broken link to the Abidin Durdu GitHub profile and speeded up fetching the JSON

So, I've speeded up fetching the JSON by using response.json() instead of response.text() and then JSON.parse(), and I've also fixed a broken link to the Abidin Durdu's GitHub profile.

Minor refactoring of the core of the assembler

I've done some minor refactoring of the assembler.js file. It's still pretty untidy, though.

Appearance improvements and syntax highlighting improvements

I've made the background change its position when the user scrolls down. As well, I did some improvements on the syntax highlighter.

Added the support for a few functions in the preprocessor

I've added the support for a few functions in the preprocessor: mod(), bitand(), bitor(), and invertBits().

Fixed a bug causing internal assembler errors in the tokenizer

With the previous release, I introduced a bug that the tokenizer crashes with an out-of-bounds array access on the following program:

start: load s0, s0

This release fixes that. Let's hope there aren't any more such critical bugs left...

Fixed a bug in the tokenizer

I've fixed a bug in the tokenizer which prevented Fibonacci Sequence and Assembler Test from being assembled.