Skip to content

HTTPS clone URL

Subversion checkout URL

You can clone with
or
.
Download ZIP
Commits on Jan 28, 2015
  1. @dams

    typo

    dams authored
  2. @dams
Commits on Jan 27, 2015
  1. @dams

    new generated documentation

    dams authored
  2. @dams

    documentation reorg

    dams authored
  3. @dams
  4. @dams
Commits on Jan 9, 2015
  1. @kappa

    Support topic "0" in psubscribe

    kappa authored
    A test and a basic '||' vs. 'defined' fix.
Commits on Oct 3, 2014
  1. @vsespb @dams

    better exception message for the fix of inconsistency with binary str…

    vsespb authored dams committed
    …ings (called "the unicode bug").
Commits on Aug 30, 2014
  1. @vsespb

    Fixing inconsistency with binary strings (called "the unicode bug").

    vsespb authored
    In perl, string which are identical (eq operator) should behave as same strings.
    Fixing inconsistency when binary strings with UTF-8 flag set and flag clear
    behave different way.
    
    Intention of the module to consider all input and output data as binary data,
    proper way to implement this does not involve call to _utf8_off, instead
    it just works in pure-perl.
    
    If data with wide characters is sent (i.e. characters with code > 255), it's can't
    be downgaded, thus cannot be considered as binary string, thus exception
    "Wide characters in syswrite" is thrown.
    
    Fixes #95
Commits on Jul 31, 2014
Commits on May 17, 2014
  1. @dams

    fix test

    dams authored
Commits on May 9, 2014
  1. @ikruglov

    use croak instead of die

    ikruglov authored
  2. @ikruglov
Commits on Apr 29, 2014
  1. @dams

    Update Redis.pm

    dams authored
    confess => croak
  2. @ikruglov

    optimize __try_read_sock()

    ikruglov authored
    As Pavel Shaydo noticed:
    The whole __try_read_sock subroutine looks a bit inefficient to me, you are
    reading data with MSG_PEEK, just to check if there's any, and then you
    throwing it away. Why not save it into {__buf} and avoid reading it again?
  3. @ikruglov

    Fix usage of MSG_DONTWAIT in Win32

    ikruglov authored
    WIN32 doesn't support MSG_DONTWAIT, need to swith fh to nonblockng mode manually.
    For Unix still use MSG_DONTWAIT because of fewer syscalls
Commits on Apr 26, 2014
  1. @dams

    Merge branch 'ikruglov/sysread_implementation'

    dams authored
    Conflicts:
    	lib/Redis.pm
  2. @dams

    fix doc for 'every' option, fix #81

    dams authored
Commits on Feb 17, 2014
  1. @dams

    more doc, more sentinels options

    dams authored
  2. @dams

    added sentinel timeouts

    dams authored
  3. @dams

    Merge branch 'sentinel2'

    dams authored
    Conflicts:
    	.gitignore
    	lib/Redis.pm
    	t/tlib/Test/SpawnRedisServer.pm
  4. @dams

    forgot to use exists in one place

    dams authored
  5. @dams

    don't use each, use foreach keys

    dams authored
  6. @dams

    Merge pull request #75 from xtab/master

    dams authored
    __try_read_sock: also check errno if read returns 0
  7. @dams
  8. @ikruglov

    Redis->new(server=>'') does not connect to localhost:6379 #77

    ikruglov authored
    Old logic causes subtle issues in some cases.
    Redis->new() remains behaves as used to be, connection to localhost:6379
Commits on Feb 1, 2014
  1. @xtab

    __try_read_sock: also check errno if read returns 0

    xtab authored
    In some cases (and notably FreeBSD), a read() on a socket which doesn't
    have data ready sets errno but returns 0.  This confuses the library
    into thinking that we hit an end-of-file because the server closed the
    connection.
    
    Therefore, we need to check errno even when read() returns 0; in
    particular, EAGAIN means read() returned no data, but the socket is
    still alive and should not be closed.
  2. @dams

    documentation

    dams authored
Commits on Jan 30, 2014
  1. @dams

    emphasis even more

    dams authored
Commits on Jan 29, 2014
  1. @ikruglov
  2. @vsespb

    Clarification for (p)unsubscribe commands.

    vsespb authored
    it seems that unsubscribe/punsubscribe needs reference to previous callbacks
    Fixes #69
Commits on Jan 21, 2014
  1. @dams
Commits on Jan 14, 2014
  1. @avar
Commits on Jan 12, 2014
  1. @rgs
  2. @rgs
Something went wrong with that request. Please try again.