[MGNLUI-1805] Leaving the focus on a cell being edited expands the row it belongs to Created: 29/Jun/13  Updated: 04/Sep/13  Resolved: 04/Jul/13

Status: Closed
Project: Magnolia UI
Component/s: tree/list
Affects Version/s: 5.0
Fix Version/s: 5.0.1

Type: Bug Priority: Critical
Reporter: Federico Grilli Assignee: Mikaël Geljić
Resolution: Fixed Votes: 0
Labels: None
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified

Issue Links:
Relates
relates to MGNLUI-2018 TreeTable should not expand the node ... Closed
Template:
Acceptance criteria:
Empty
Task DoD:
[ ]* Doc/release notes changes? Comment present?
[ ]* Downstream builds green?
[ ]* Solution information and context easily available?
[ ]* Tests
[ ]* FixVersion filled and not yet released
[ ]  Architecture Decision Record (ADR)
Bug DoR:
[ ]* Steps to reproduce, expected, and actual results filled
[ ]* Affected version filled
Date of First Response:

 Description   

Sorry for the clumsy summary but I could not word it better. I think this the very issue which makes the inplace editing tree still rather awkward to use.

To reproduce

  • go to config app
  • expand modules and, without expanding it before, dbl-click on 'core' just to make the cell editable
  • now try to select the next row, i.e. 'ui-contentapp'

Expected outcome

  • 'ui-contentapp' row is selected and core row isn't expanded

Actual outcome

  • the core node is unnecessarily expanded, whilst the 'ui-contentapp' row isn't selected. It looks like for a moment it gets selected but eventually the 'core' row is selected again, then expanded and finally unselected.


 Comments   
Comment by Jozef Chocholacek [ 02/Jul/13 ]

The

Node root = session.getRootNode();
Node visibleRoot = root.addNode(NODE_ROOT_ITEM_ID);
Node parent = visibleRoot.addNode(NODE_PARENT);
Node node = parent.addNode(NODE);
node.addNode(NODE_CHILD);
node.setProperty(NODE_PROPERTY, "112");

part is the same in each test - so it should go to the setUp() method.

Comment by Mikaël Geljić [ 02/Jul/13 ]

I do not agree.
1. Assertions are run against these local variables, might not be as much legible (nor as safe) if I turn these into fields.
2. It conforms to the Behavior Driven Development style that applies to our unit tests; doesn't make sense to trade //GIVEN blocks away for a few LoC.
3. Further tests are free to use a different test repo structure.

Comment by Mikaël Geljić [ 04/Jul/13 ]

1. One UI test had to be ignored with first solution
2. Side effect was also that when adding a property or subnode in config app, node would no longer expand automatically.

Comment by Mikaël Geljić [ 04/Jul/13 ]

Fixed by no longer triggering workbench select from the browser presenter on editing nodes. Their itemId (identifier - used to be path) doesn't change anymore when editing.

We keep that workbench reselect for properties - itemId still changes when renaming, but it has no bad effect on editing, properties cannot be expanded anyway.

Generated at Mon Feb 12 08:50:19 CET 2024 using Jira 9.4.2#940002-sha1:46d1a51de284217efdcb32434eab47a99af2938b.