ensure the LEFT JOIN behaves as expected #623
Conversation
The WHERE clause will filter out these rows because NULL values do not satisfy either the IS NULL or the greater than conditions, effectively excluding them from the result set. Turning the LEFT JOIN into an INNER JOIN for rows where NULL values would have been expected. * move includeDeleted checks from query to join, where the filter will be applied * copy filters over and add filter as necessary, on case deleted in not included * keep includeDeleted checks within query aswell, as null checks now can exists in where clause aswell. * adjust tests to reflect changes in code.
remove conformity to equatable for query action
…L optimization incorporate changes in test cases
bwdmr
force-pushed
the
leftjointimestamp
branch
from
7c4f2c7 to
88fbec7
Compare
bwdmr
force-pushed
the
leftjointimestamp
branch
from
33236f1 to
983cd12
Compare
Codecov Report❌ Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## main #623 +/- ##
==========================================
+ Coverage 24.22% 24.42% +0.19%
==========================================
Files 149 149
Lines 9348 9373 +25
==========================================
+ Hits 2265 2289 +24
- Misses 7083 7084 +1
🚀 New features to boost your workflow:
|
|
It appears that the Mongo driver (or Mongo itself) can't handle this. @Joannis, can you provide any insight? |
| } | ||
| } | ||
|
|
||
There was a problem hiding this comment.
Excess whitespace change, please remove
|
support for complex joins with filters is not implemented yet. joins with filter would require a pipeline as in the following example db.orders.aggregate([
{ $lookup: {
from: "customers",
let: {
orderId: "$customerId",
orderDate: "$orderDate"
},
pipeline: [
{ $match: {
$expr: {
$and: [
{ $eq: ["$_id", "$$orderId"] },
{ $gt: ["$memberSince", "$$orderDate"] },
{ $eq: ["$isVerified", true] }
]
}
}}
],
as: "customerInfo"
}}
])
|
* Add missing targets to api-docs.yml * Add missing targets to pathsToInvalidate
…or#628) The CompositeIDProperty was force-unwrapping its value during database input, which would crash if the value was nil. This could happen during initial inserts where the composite ID is meant to be set by a database trigger. Now checks if value exists before attempting to input it.
|
A solution for the Mongo driver is being worked on (though not by me, as I neither have nor want any knowledge of Mongo 😅) |
can this be merged. i believe the mongo driver to be deprecated |
|
Although the Mongo driver is deprecated for use in new Fluent 4 projects, it is unfortunately still in use by existing projects, and merging this would break normal usage for those users. |
|
can we proceed. support for the mongo community could be included in a separate pr |
4 participants
The WHERE clause will filter out these rows because NULL values do not satisfy either the IS NULL or the greater than conditions, effectively excluding them from the result set.
Turning the LEFT JOIN into an INNER JOIN for rows where NULL values would have been expected.