[Bug]: Table widget always shows a vertical scrollbar despite fitting exactly one page of rows

[RobWilsonTas]

Is there an existing issue for this?

• I have searched the existing issues

Description

The table widget is always showing a vertical scrollbar even when pagination is enabled

Changing the widget height does not resolve the issue. No matter how the widget is resized, the table appears to calculate one extra row (variable pageSize), resulting in a permanent scrollbar.

I.e if the table widget is 12.5 rows in height, it'll try to show 13 rows instead of 12, so you have to scroll the extra 0.5.

The top comment here mentioned this issue 2 years ago.
https://www.reddit.com/r/nocode/comments/1b6777a/a_review_of_appsmith/

Steps To Reproduce

Create a Table widget with pagination enabled and connected it to some data.
Set the page size whatever amount of rows.
Resize the table widget to various heights.
Observe that a vertical scrollbar is always present, especially with a row height of short.

Public Sample App

No response

Environment

Production

Severity

Low (Cosmetic UI issues)

Issue video log

No response

Version

v2.1

[linear-code]

APP-15693

[jiuyige]

I’d like to work on this issue.

[RobWilsonTas]

I’d like to work on this issue.

That would be great

Fyi I think there are two levels to this, first is that the variable pageSize seems to be calculated slightly too high, something like (pseudo code):

pageSize = (widgetHeight - topBarHeight - columnHeaderHeight) / rowHeight

when it should be something like

pageSize = Math.floor((widgetHeight - topBarHeight - columnHeaderHeight) / rowHeight)

The second level is how there are more complicated situations where a column might have "Cell wrapping" enabled, which, in an ideal world, would cause pageSize to be calculated on the fly. This would be a lower priority to fix up though, so just the first part would be great to have fixed.

[jiuyige]

I’d like to work on this issue.

That would be great

Fyi I think there are two levels to this, first is that the variable pageSize seems to be calculated slightly too high, something like (pseudo code):

pageSize = (widgetHeight - topBarHeight - columnHeaderHeight) / rowHeight

when it should be something like

pageSize = Math.floor((widgetHeight - topBarHeight - columnHeaderHeight) / rowHeight)

The second level is how there are more complicated situations where a column might have "Cell wrapping" enabled, which, in an ideal world, would cause pageSize to be calculated on the fly. This would be a lower priority to fix up though, so just the first part would be great to have fixed.

Thanks for clarifying. I’ll keep this change scoped to the fixed-row-height
case by flooring the calculated pageSize and adding regression coverage for
all three row-height modes.
I’ll leave dynamic page-size calculation for wrapped cells out of this PR as
suggested.