add readme

Some raw and crappy documentation.

fix formatted text parsing

earlier, situations when there were strings like 'bold\ntext\n' that
needed to be treated as bold text were converted in a wrong way (e.g.
like '**bold\ntext\n**', which is breaking markup because of newline
characters before closing asterisks), but now it's kind of fixed.

also there's now support for emails and <pre> blocks.

add tests

i've found some issues with formatted text, this could be seen in output
(which is not included now). there are some complications with parsing
newline characters, but i'll fix it later.

also script is now usable like command-line application. you can specify
the directory in which source files are located.

to make some tests you need to run
$ python parse.py tests

this will create directory tests/formatted_posts, where markdown posts
will be populated.

if directory is not specified, it assumes that script is in the source
    folder. if needed json file could not be found, it exits with code
    1.

add option to reassign directories for media

now it's done through a variables photo_dir and media_dir, these should
be filled in before running and they are intended to fit your specific
setup. e.g. mine uses /assets folder for such file storage.

add raw parser

handles:
  - photo
  - text with bold, italic, underlined, stroke style and with links.
  - audio files

not handles:
  - tags
  - replies
  - forwarded messages