Don't add third party settings when geocoder method is none

Nice! Do you think it would be worth writing a test for that?

Hello,

What is the status of this patch ? Is it still valid ?

Thanks.

Definitely still valid.

However, patch #2 doesn't apply cleanly to the end of the 8.x-2.x branch, so here's a trivial re-roll.

Before the patch, here's an example config from a test site for a field that has nothing to do with addresses or geocoding:

langcode: en
status: true
dependencies:
  config:
    - field.storage.node.field_summary
    - node.type.layout_page
  module:
    - geocoder_field
    - text
third_party_settings:
  geocoder_field:
    method: none
    field: title
    plugins: {  }
    dumper: wkt
    delta_handling: default
    failure:
      handling: preserve
      status_message: true
      log: true
id: node.layout_page.field_summary
field_name: field_summary
entity_type: node
bundle: layout_page
label: Summary
description: 'One or two sentence summary of this page, for example to use when sharing to social media.'
required: false
translatable: false
default_value: {  }
default_value_callback: ''
settings: {  }
field_type: text_long

Notice all the cruft about geocoder_field? Both in third_party_settings, and also the fact the field config now *depends* on geocoder_field (even though it's not related to geocoder at all!).

After applying the patch and re-saving the field settings, the config export now looks like this:

langcode: en
status: true
dependencies:
  config:
    - field.storage.node.field_summary
    - node.type.layout_page
  module:
    - text
id: node.layout_page.field_summary
field_name: field_summary
entity_type: node
bundle: layout_page
label: Summary
description: 'One or two sentence summary of this page, for example to use when sharing to social media.'
required: false
translatable: false
default_value: {  }
default_value_callback: ''
settings: {  }
field_type: text_long

MUCH better! No bogus dependency, and no ignored third_party_settings related to functionality this field has nothing to do with.

p.s. By default, d.o's 'Git command' is now using me as the author. Please don't. This is @AndyF's code and they should get credited as the patch author, not me.

I just tested ... and indeed still make sense. Tnx. Just committed in the last dev.