In d6_field_formatter_settings, there is the following check.
process:
# We skip field types that don't exist because they weren't migrated by the
# field migration.
field_type_exists:
-
plugin: migration_lookup
migration: d6_field
source:
- field_name
-
plugin: skip_on_empty
method: row
-
plugin: extract
index:
- 1
This fails to properly skip the processing the row. If the migration_lookup does not finds a record for field_name, it returns an array containing two null values. Such array fails the skip_on_empty validation. A similar check is part of d6_field_instance and d6_field_instance_widget_settings. In both cases, the order of the skip_on_empty and extract plugins is inverted. That takes care of properly skipping the row when the migration_lookup does not return a migrated record.
d6_field_formatter_settings inverted the process plugins in
#2560671: The Migration process plugin should not skip rows →
.
On a D6 migration that skipped field storage, run the field formatter settings migration.
Invert the order of the skip_on_empty and extract plugins to be aligned with d6_field_instance and d6_field_instance_widget_settings.
Review patch.
Write tests.
None.
Not sure if the migration plugin is a public API.
None.
None.
Needs work
11.0 🔥
The change is currently missing an automated test that fails when run with the original code, and succeeds when the bug has been fixed.
We might need the to review the skip logic in the view_mode and bundle process pipelines. Let's see what the test bot reports about the patch first.
Thank you for reporting.
Can we get a test case showing the issue?