aboutsummaryrefslogtreecommitdiffstats
path: root/activerecord
Commit message (Collapse)AuthorAgeFilesLines
* speed up match?Aaron Patterson2013-10-091-0/+1
|
* expose the root node and call itAaron Patterson2013-10-091-21/+11
|
* just skip the join if it's already thereAaron Patterson2013-10-091-14/+10
| | | | Remove the annoying `find_or_build` stuff.
* speed up finding existing nodesAaron Patterson2013-10-091-3/+1
| | | | | The node must be a child of the parent passed in, so only search children of the parent node
* make node search more efficientAaron Patterson2013-10-094-3/+43
| | | | | | Rather than search every node in the tree, comparing that node and all of its parents every time, start at the root from both sides and work our way down the tree
* remove == so we can see where walking up parents occursAaron Patterson2013-10-094-25/+15
|
* push parent up to the superclassAaron Patterson2013-10-093-7/+11
| | | | We always want a linked list back to the root node.
* convert JoinBase to a tree and remove the Node classAaron Patterson2013-10-092-42/+30
|
* names are guaranteed to be symbolsAaron Patterson2013-10-081-1/+1
|
* eliminate function that is only used in one placeAaron Patterson2013-10-081-7/+2
|
* we will always detect the same node, so just pass the node inAaron Patterson2013-10-081-7/+2
|
* we do not need to to_s the name all the timeAaron Patterson2013-10-081-5/+3
|
* parent is guaranteed to be the sameAaron Patterson2013-10-081-4/+1
| | | | | since we are searching a tree, the current level should be consistent, so we don't have to inspect the parent class / table
* reduce number of comparisons and array allocationsAaron Patterson2013-10-081-11/+10
| | | | | Just inspecting the children nodes will reduce the number of comparisons we have to do (since it is limited to the child size)
* rename the variable to make more sense with a treeAaron Patterson2013-10-081-7/+7
|
* walk the tree rather than generating a hash and walking itAaron Patterson2013-10-081-21/+10
|
* do not convert the tree to a list just for the first nodeAaron Patterson2013-10-081-1/+1
|
* generate the association hash from the top downAaron Patterson2013-10-081-17/+15
| | | | | This prevents us from walking back up the parent tree once for each leaf.
* store join parts in a treeAaron Patterson2013-10-081-16/+45
|
* join parts in `build` are *always* new, so don't bother checkingAaron Patterson2013-10-081-1/+2
|
* stop looking up the reflection multiple timesAaron Patterson2013-10-081-6/+5
|
* cache join associations and join base on the stackAaron Patterson2013-10-081-4/+11
| | | | No reason to decompose the join objects on every iteration
* use the API rather than accessing underlying data structuresAaron Patterson2013-10-081-1/+1
|
* associations is always a hashAaron Patterson2013-10-081-19/+13
|
* keys are always symbolsAaron Patterson2013-10-081-2/+4
|
* we can get a list of reflections from the join parts, so let's removeAaron Patterson2013-10-081-3/+5
| | | | the @reflections ivar.
* transform the association input so we can avoid type checks later.Aaron Patterson2013-10-081-17/+31
| | | | | We should consider moving the input munging outside the class instantiation
* we know the keys are scalars, so avoid some typecheckingAaron Patterson2013-10-081-8/+12
|
* this method is only ever called with a reflectionAaron Patterson2013-10-081-6/+2
| | | | Stop typechecking the parameter. We know the type.
* don't access the reflections hashAaron Patterson2013-10-082-3/+7
| | | | The reflections hash is supposed to be private, so let's use the API.
* build association tree thing on demand.Aaron Patterson2013-10-081-6/+11
| | | | | `instantiate` is only called once, so there is no reason to cache the data structure on the instance and mutate itself.
* hide join_parts internals from other classesAaron Patterson2013-10-082-17/+17
|
* decouple JoinAssociation from JoinDependencyAaron Patterson2013-10-082-10/+5
|
* @associations can never have an array, so remove the dead codeAaron Patterson2013-10-081-4/+0
|
* typecast records returned from the db rather than to_sing everythingAaron Patterson2013-10-082-12/+21
|
* extract the string handling part to a method.Aaron Patterson2013-10-071-15/+15
| | | | | | We know the structure passed in to the `construct` method will be a hash, so we don't need to test it all the time. The key value will be a symbol or string, so handle it with the special method
* remove more dead codeAaron Patterson2013-10-071-2/+0
| | | | | We control the inputs to `construct`, so the wrong type shouldn't be passed in ever.
* this code is never used, remove itAaron Patterson2013-10-071-4/+0
|
* log the statement name along with the SQLAaron Patterson2013-10-044-6/+20
|
* log every sql statement, even when they errorAaron Patterson2013-10-041-25/+25
|
* prepare the statement inside the begin / rescue blockAaron Patterson2013-10-041-4/+4
|
* rm LogIntercepterAaron Patterson2013-10-041-15/+0
|
* stop adding singleton methods to the SQLite3 connectionAaron Patterson2013-10-041-11/+11
|
* stop adding singleton methods to the mysql2 adapterAaron Patterson2013-10-043-22/+22
|
* stop adding singleton methods to the PG connectionAaron Patterson2013-10-041-11/+26
|
* wrap logging around the actual query call itself.Aaron Patterson2013-10-042-50/+50
| | | | This is to be consistent with the way the mysql2 adapter times queries
* add regression test for set_inverse_instance on add_to_targetArthur Neves2013-10-041-0/+13
|
* inject down the hash cacheAaron Patterson2013-10-031-3/+2
|
* hashes are ordered, so we do not need to sortAaron Patterson2013-10-031-3/+3
|
* Always pass parameters to `build`Aaron Patterson2013-10-031-2/+2
| | | | | We always know the parameters in advance, so we don't need to build it in to the method signature
generated by cgit v1.2.3 (git 2.25.1) at 2024-10-30 06:41:04 +0000