Summary

All of the Option sorters in Jira ignore the fact that a custom field can have multiple contexts.
Because of that, they sort only using a sequence number, sometimes grouping items that shouldn't be grouped, and sometimes incorrectly removing results.

Environment

• Using Two Dimensional Filter Statistics Gadget
• Select List with Multiple contexts

Steps to reproduce

One of the scenarios in which an invalid Option sorter is used:

1. Set up custom field named "New Select", that is a Select List (any type)
   • Create options A, B in this sequence
2. Create some issues for some project to assign the value
   • 1 with New select = A
   • 1 with New select = B
3. Create Two Dimensional Filter Statistics Gadget and point to a filter containing these issues (e.g. New select is not empty)
4. Create new context on "New Select" field, name it context "Second context"
   • Select issue type = any
   • Select all projects in the project context
   • Create new options B, C, A in this sequence
5. Create new issues with different values of New Select (B, C, or A)
6. Refresh the gadget

 Expected results

All of the values are displayed properly, being grouped by the option's name, even if the options are assigned to a different context

Actual results

The gadget shows weird results, e.g. two rows for the value "B", while there are also issues with values C and A.