Permalink
Switch branches/tags
Commits on Jan 25, 2011
  1. Release 0.08190

    ribasushi committed Jan 25, 2011
Commits on Jan 20, 2011
  1. Rewrite the check for nontrivial joinfree conditions to throw on both…

    ribasushi committed Jan 17, 2011
    … new_related and set_from_related
  2. Fix $object->search_related aliasing, change semantics of _resolve_co…

    ribasushi committed Jan 16, 2011
    …ndition
    
    Change the RV of _resolve_condition one last time. Now it checks the RV of the
    resolved coderef (if any), and chooses (preferrably) the join-free condition
    or the cross-table one. In list context returns a flag signifying if the
    resulting condition is available for standalone use (false) or requires the
    joins to remain (true)
  3. Pass relationship name to _resolve_condition explicitly

    ribasushi committed Jan 17, 2011
    The API is ambuguous at best, will need a cleanup some day, for now just do by
    tackling even more arguments to the method, yielding insanity like:
    _resolve_condition($cond, $obj, $f_rel, $f_rel) BLEH!
  4. use $self->{from_bind} for extra binds transport

    ruoso committed with ribasushi Dec 13, 2010
    The __overriden_select method was used to build extra binds
    that should be transported in the $self->{from_bind} array
    instead.
  5. Back out "support for prefetch from resultsource using extended_rels"

    ribasushi committed Dec 12, 2010
    Blindly taking the condition retuned by an incomplete (not
    all pieces of the hashref are supplied) call to the condition
    maker, and attempting to order by its keys assuming it is a
    hash is... naive
  6. do not use "me" on the related_resultset pessimization

    ruoso committed with ribasushi Nov 30, 2010
    This was requested by ribasushi, but it is still incomplete.
    Added a TODO test with the expected final query.
    
    Problem is: the alias is defined in $source->_resolve_join,
    which is completely disconnected from the search context
    so it cannot receive a different alias
  7. refuse to invoke create_related as class method

    ruoso committed with ribasushi Nov 24, 2010
    t/cdbi/09-has-many.t has a specific test for it.
  8. fix the alias in the identity column

    ruoso committed with ribasushi Nov 24, 2010
    When building the identity where clause, discover the current_source_alias
    to build the hash correctly
  9. eager check if the create_related is at all possible

    ruoso committed with ribasushi Nov 24, 2010
    When calling create_related, it should stop trying if the extended rel
    does not provide the complete relationship definition.
  10. When getting a related_resultset in a row with custom rels without th…

    ruoso committed with ribasushi Nov 24, 2010
    …e extended definition, we step back, get a rs equivalent to the current row and do a search_related there.
  11. makes search_related on extended rels without the optimized version w…

    ruoso committed with ribasushi Jun 2, 2010
    …ork. involves hacking the from attribute
  12. Most of the code reviewed... Missing ResultSet->populate, ResultSet->…

    ruoso committed with ribasushi Jun 2, 2010
    …find, Row->copy and Relationship::Accessor->add_relationship_accessor
  13. Reshape initial tests

    ribasushi committed Sep 12, 2009
  14. failing test

    monken committed with ribasushi Jun 30, 2009
  15. Subject to this branch

    monken committed with ribasushi Jun 30, 2009
Commits on Jan 19, 2011
  1. Release 0.08127

    abraxxa committed Jan 19, 2011
  2. Rewrap Changes before 0.08127

    ribasushi committed Jan 19, 2011