Commits on Mar 9, 2013
  1. Version 0.1.4

    committed Mar 9, 2013
  2. Updating README

    committed Mar 9, 2013
  3. Removing the perf folder.

    committed Mar 9, 2013
  4. Correcting tests.

    committed Mar 9, 2013
  5. Making splicing of SD happen backwards.

    Going "forwards" in the loop prevented multiple SD's (which are valid) from
    being parsed out.
    committed Mar 9, 2013
Commits on Mar 8, 2013
  1. Fixing RFC 5848 test.

    committed Mar 8, 2013
  2. Using valid SD in 5424 tests.

    committed Mar 8, 2013
  3. Replacing structured data parser.

    This is a faster than regex implementation that is using a semi-convoluted
    finite state machine which has room for optimisation, but strictly adhears to
    the specification and no longer does data chopping and hacking to get the values
    out.
    committed Mar 8, 2013
Commits on Mar 7, 2013
  1. Adding structured data test.

    This is also breaking thanks to the issue I found with our parser.
    committed Mar 7, 2013
  2. Adding tests to decode privals.

    In the course of this test I discovered that higher than pesmissable privals
    (>=191) were being "let through". They now return false.
    committed Mar 7, 2013
  3. Moving parser to use prototypes, like the producer.

    This will make unit tests possible, more consistent code and well, you get the
    idea. We're instantiating a new instance on export because this will mean that
    users won't have to change how they already use the library - it should just
    work as is.
    committed Mar 7, 2013
Commits on Mar 6, 2013
  1. Removing previous methods of tests.

    That's right folks, fixtures don't really make sense here. Instead I have opted
    to go through and validate almost* every value of each message that was in
    fixtures. This has caused me to find two problems, one with ISO8601 parsing but
    I blame other people for that, but also with the structured data parsing so as
    of this commit TESTS FAIL with FIXME left where they need fixing.
    
    * Times in 3164 tests are not immediately tested because they feature no year
    and no TZ, new Date('...') will use this TZ and it's known DST, and this year.
    Tests will break until I find a better way of doing things.
    committed Mar 6, 2013
Commits on Mar 3, 2013
  1. Removing original parse test.

    committed Mar 3, 2013
  2. Fixing invalid JSON in fixture

    committed Mar 3, 2013
  3. Fixing runner to work on any js file.

    No "test" prefix required. Will also not run itself, and not put any data out to
    STDOUT/STDERR unless something fails.
    committed Mar 3, 2013
  4. Adding a generic runner so all tests can be run.

    This has been modified from
    https://github.com/mikeal/request/blob/master/tests/run.js and modified in line
    with the rest of my code.
    committed Mar 3, 2013
  5. Removing snake case from vars

    committed Mar 3, 2013
Commits on Jan 9, 2013
  1. Version 0.1.3

    committed Jan 9, 2013
  2. Merge pull request #9 from stancarney/master

    Host/timestamp parsing fix for RFC3164 messages.
    committed Jan 8, 2013
  3. @stancarney

    Fix for #7.

    For RFC3164 messages the first 9 days of each month cause an extra space to appear between the month code and the day of the month. This fix checks to see if the extra space exists and removes it if it does.
    stancarney committed Jan 8, 2013