Currently if the code that calls .find_in_batches modifies the yielded array in place then .find_in_batches can enter an infinite loop searching with ruby object ids in the database instead of the primary key of records in the database. This happens because it naively assumes the yielded array hasn't been modified before calling #id on the last object in the array. And ruby (1.8 at least) alias' #id to #object_id so an integer is still returned no matter what the last object is.
By moving finding the #id of the last object before yielding the array it means the calling code can do whatever it wants to the array in terms of modifying it in place, and .find_in_batches doesn't care.
- If doing a count on a relation that has an :include and a :join, it does a distinct even though it should not.
- When doing a count on a relation that has an :include, it always falls back to a old style left join when performing the count. Looks like it was broken here:
b9599502c9
if a query contains a limit or an offset, ActiveRecord::FinderMethods#find_last had inconsistent behavior.
If the records were loaded, it returned the last record in the cached list.
If they were not, it reversed the order of the query and changed the limit to one.
If the earlier limit was less than the total matching the query in the db,
it would return a different record than if the records had been cached.
This commit changes find_last so that it loads the records when getting the
last record on a query containing a limit or an offset, which makes the behavior consistent.
* master: (51 commits)
order is not guaranteed by this select, so add an order and call first!
oracle stores this with microseconds, so convert to seconds before comparing
make sure that active connections are not cleared during test when an exception happens
clearing active connections in the ConnectionManagement middleware if an exception happens
proxy body responses so we close database connections after body is flushed
Pass the proper method_name instead of hardcoding to action_name.
Quote find_in_batches ORDER BY clause [#6620 state:resolved]
Delegate first!, last!, any? and many? to scoped
Dont call authenticate_or_request_with_http_basic twice
Remove 'warning: ambiguous first argument' when running ActionPack tests
Change exists? so that it doesn't instantiate records [#6127 state:resolved]
Move mapper_test to the appropriate location
Update the wildcard route to be non-greedy by default, therefore be able to match the (.:format) segment [#6605 state:resolved]
Fix examples
Added Base.http_basic_authenticate_with to do simple http basic authentication with a single class method call [DHH]
make sure we have an active database connection before running each connection management test
adding active_connections? to the connection pool for finding open connections
adding active_connection? to the connection pool
testing app delegation from the ConnectionManagement middleware
namespacing connection management tests. ❤️
...
* master:
Do not show optional (.:format) block for wildcard route [#6605 state:resolved]
pushing id insertion and prefetch primary keys down to Relation#insert
use prepared statements to fetch the last insert id
escaping binary data encoding when inserting to sqlite3. Thanks Naruse! [#6559 state:resolved]
schemas set by set_table_name are respected by the mysql adapter. [#5322 state:resolved]
Reapply extensions when using except and only
SJIS is an alias to Windows-31J in ruby trunk. Use SHIFT_JIS for this test
Improved resolver docs a bit
[action_view] docs for FileSystemResolver
[action_view] added custom patterns to template resolver
