- mostly doxygen comments
- some alignment

closes #159

Refs #142

Refs #159

Refs #159

Refs #154

as null pointers, contained actual key (type of RRset, that is), which did not
comply with definitions, and therefore was producing warnings. Changed node test,
so that they reflect this fact. Skip list's function return_list() and list_length()
now obsolete. Refs #121

fixes #140

Added sorting test for node. Added some helper function in skip list, namely skip_length and skip_return_list. Refs #116

Wrong freeing of dname in test_dname_compare().
RDATA test count.
Comparing RDATA in test_rrset_rdata().
All tests count.

Was not initializing the position pointers.

Changes to both rrset and node APIs as well as test for those APIs. Node API and tests should be complete. Refs #108, #109, #110, #116

Added test for rdata_get_item().
Test for set_items() now checks all return values.

refs #117

Checking for bad input parameters in dnslib_rdata_set_items().
Comments adjusted.

refs #119

Somewhat complete rrset test implementation. Needs some polishing and possibly more sophisticated way of testing rrsigs.

- wire_size() was not counting the stored length octet in case of
  TEXT or BINARY.
- typo in to_wire()

refs #119

Modified also the helper functions fill_data() and check_data()
  - they are now storing all what is saved to the RDATA, so it
    can be used later for comparison.

refs #117

refs #119

refs #117

Tests now reflect the new RDATA API and use RR type descriptors
  to fill the structures with some data.

Added includes to dname tests.

refs #117

refs #119

Fixed enum dnslib_rr_type - must explicitly set numbers for
  types which do not follow the previous ones in numbering.
Fixed naming of mxrdtln enum items to contain prefix DNSLIB_.

The previous approach made converting RDATA  to wire format too
  complex. Now RDATA items distinguish only between domain name
  and other data. The format is given by a descriptor.

TODO: Consider saving the length of the data to the byte array
      always, so getting it would not require that huge switch.

refs #119