-
Notifications
You must be signed in to change notification settings - Fork 66
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
refactor!: run data generators only for visible columns #6798
base: main
Are you sure you want to change the base?
refactor!: run data generators only for visible columns #6798
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Regarding the utility of this change: This sort of moves the performance issue to a different use-case, which is showing and hiding columns. In that case it now needs to reload all data whenever a single column is made visible. Previously it would load everything up front, but then doesn't have to reload everything if you show a single column.
What is everyone's thought on this?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
In theory, hiding and unhiding columns shouldn't need to result in a data provider call, but only the data generators should need to be rerun for the already loaded items.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I'm pretty sure that whoever created this issue wouldn't expect additional data provider calls as part of this change.
Since there's a trade-off, maybe it could be turned into an option. Though it's a bit challenging to communicate what this even does.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This nasty prototype method in Grid does the job without additional data provider calls, but it requires reflection. I hope there's a better way (maybe storing the value in Grid.setRequestedRange):
private void regenerateData() {
try {
var communicator = getDataCommunicator();
var requestedRangeField = communicator.getClass()
.getDeclaredField("requestedRange");
requestedRangeField.setAccessible(true);
var requestedRange = (Range) requestedRangeField
.get(communicator);
for (int i = requestedRange.getStart(); i < requestedRange.getEnd(); i++) {
var item = communicator.getItem(i);
if (item != null) {
getDataCommunicator().refresh(item);
}
}
} catch (Exception e) {
// ignore
}
}
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Might be related: #6713
.map(dataGenerator -> grid | ||
.addDataGenerator((DataGenerator) dataGenerator)) | ||
.orElse(null); | ||
grid.getDataProvider().refreshAll(); |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
If the app calls refreshAll
before making a column visible, that results in two calls to the data provider now.
In general I'm uncertain about calling refreshAll
in beforeClientResponse
, which is the same phase that data communicator uses to prepare data to send to the client. Is there a reason that this has to be called in beforeClientResponse
, or could this already be called in setVisible
?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I agree that it would be better to do it that way. However, directly calling in setVisible
does not work and I'm not sure what would be a good way to refactor the refresh.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Added a fix and the test for it to handle this case.
Would making the data generator run conditionally based on the column's visibility fix the original issue? |
9c98369
to
982bb82
Compare
Adapted the suggestion. It works nicely and is cleaner. |
Quality Gate passedIssues Measures |
Description
Warning
Possibly behaviour altering change.
Unlikely to break an existing app.
Currently, the value provider in any type of created grid column runs regardless of its visibility. This can cause unnecessary load if the tasks are heavy.
This PR updates the rendering logic in the columns to run data generators only if the column is visible before the client response. This also avoids sending unnecessary data to the client.
The downside is that there is an extra refresh when a column is set visible.
Fixes #6737
Type of change
Checklist