Skip to content

HTTPS clone URL

Subversion checkout URL

You can clone with
or
.
Download ZIP
Tag: Sereal-Decoder…
Commits on Sep 13, 2012
  1. @tsee

    Changelog, version bump

    tsee authored
  2. @tsee

    Implement snappy_threshold option

    tsee authored
    Defaults to 1k - less data and we won't even try to compress.
Commits on Sep 12, 2012
  1. @tsee
Commits on Sep 11, 2012
  1. @demerphq

    fix comment

    demerphq authored
  2. @tsee

    Fix typo in example

    tsee authored
  3. @tsee
  4. @tsee

    Version bumps and change logs

    tsee authored
  5. @demerphq
  6. @demerphq

    Merge pull request #1 from audreyt/patch-1

    demerphq authored
    two small typos
  7. @audreyt

    two small typos

    audreyt authored
  8. @demerphq

    move use Data::Dumper down so we load it after we load the tests

    demerphq authored
    Loading DD on some versions of Perl seems to change the result of
    !0, which means we fail tests. This is probably a perl bug.
    
    Anyway, this changes the tests to make sure we load DD after we load
    the test set, so we get the right data structures in memory.
  9. @tsee

    Add test for warn_unknown

    tsee authored
  10. @demerphq
  11. @tsee

    Promote TODO test

    tsee authored
  12. @demerphq

    define the macro inline where it is used, and then undef it afterwards

    demerphq authored
    this macro is intimate to one routine only and exists only for refactoring, so
    move it local to where it is used, and undef it after the last use so it is clear
    it is not for general use (and next to where it is used).
  13. @demerphq
  14. @tsee

    Fix Solaris build failure

    tsee authored
    Probably, including the right header would do the same thing...
  15. @tsee
  16. @tsee
  17. @tsee

    Promote TODO test

    tsee authored
Commits on Sep 10, 2012
  1. @demerphq

    Do not die on undef PVMG's

    demerphq authored
    We end up dieing because the SV we are dumping is a magic SV, which
    we have called mg_get() on, which has been legitimately set to undef.
    
    I am not sure if there is a better way to detect whether PVMGs have been
    handled ok. We could maybe rework this logic further, but it seems ok
    now.
  2. @tsee

    WIP: Attempt to fix/handle unknown SV types better

    tsee authored
    This is still broken. Specifically "breaks" support for tied hashes (and
    possibly arrays and scalars). Not clear that supporting them was
    strictly correct before.
  3. @tsee

    Implement (stringify|undef|warn)_unknown options to the Encoder

    tsee authored
    Basically, if we find unknown data structures (references, really), we
    have several options: die (default), emit undef, or stringify the data
    structure. If either of the latter two, we optionally emit a warning,
    too.
    
    Not being too strict with your input data makes a lot of sense if you're
    working on the system that is used to track problems. You never want
    that flat-out dying rather than just making do.
  4. @tsee

    Remove outdated comment

    tsee authored
  5. @tsee
  6. @tsee
  7. @tsee
  8. @tsee
  9. @tsee

    Minor spec fixes (cosmetics)

    tsee authored
  10. @tsee

    README updates and fixes

    tsee authored
Commits on Sep 9, 2012
  1. @tsee
Commits on Sep 8, 2012
  1. @demerphq
  2. @demerphq
  3. @demerphq
  4. @demerphq
Something went wrong with that request. Please try again.