[Graph] Fix graph saved object references

[flash1293]

Fixes #34989
Fixes #34997

This PR fixes saved object references for Graph and makes them importable/exportable.

The problem

Right now, graph does use the reference system, but it's not storing the proper id, but the title of the index pattern instead:

[
  { "type": "index-pattern", "name": "indexPattern_0", "id": "titleOfThePattern*" }
]

In the serialized state, an indexPatternRefName is stored:

{
  "vertices": [...],
  "indexPatternRefName": "indexPattern_0"
}

In x-pack/plugins/graph/public/services/persistence/saved_workspace_references.ts the reference is placed inside the state.

This part of the code wasn't touched in this PR, as it's already functioning as it should, the only issue being it contains the title instead of the id of the index pattern.

If a user would try to export/import a workspace saved object, they would get errors because the referenced index pattern can't be found.

The solution

This PR resolves this by doing the following changes:

• Migrate all existing saved objects to remove the broken reference and add a legacyIndexPatternRef property (x-pack/plugins/graph/server/saved_objects/migrations.ts)
• If a saved graph workspace is loaded, migrateLegacyIndexPatternRef (x-pack/plugins/graph/public/services/persistence/deserialize.ts) is called, which tries to map the index pattern title to the id by searching for it in a list of all index pattern which is anyway fetched for the Graph UI. If it is successful, legacyIndexPatternRef is removed, and the index pattern id is stored in the deserialized workspace state
• If the workspace is saved again, the id is placed in the references array instead of the title via the existing logic in (see x-pack/plugins/graph/public/services/persistence/serialize.ts and x-pack/plugins/graph/public/services/persistence/saved_workspace_references.ts)
• New workspaces and legacy workspaces which are opened and saved at least once have consistent references, while unopened workspaces don't have any references at all. If they are exported and imported without a matching index pattern by name, an error message is shown

Other changes

• xpack.graph.loadWorkspace.missingIndexPatternErrorMessage got improved by including the name of the index pattern causing the problem (either if loading it fails for some reason or if a legacy workspace is opened and the legacyIndexPatternRef can't be mapped)
• In x-pack/plugins/graph/server/saved_objects/graph_workspace.ts workspace saved objects are whitelisted for importing/exporting
• In x-pack/plugins/graph/server/plugin.ts a new UI capability "show" is added to make it possible to enable the edit link in the saved object management page for graph workspaces if the user has at least "read" permissions for the Graph app

[elasticmachine]

Pinging @elastic/kibana-app (Team:KibanaApp)

[dej611]

Code reviewed 👍 . Only minor comments

[flash1293]

@elasticmachine merge upstream

[kibanamachine]

💚 Build Succeeded

• continuous-integration/kibana-ci/pull-request
• Commit: 4bc46d9

Metrics [docs]

Async chunks

Total size of all lazy-loaded chunks that will be downloaded as the user navigates the app

id	before	after	diff
graph	1.3MB	1.3MB	+856.0B

Distributable file count

id	before	after	diff
default	46981	47741	+760

Saved Objects .kibana field count

Every field in each saved object type adds overhead to Elasticsearch. Kibana needs to keep the total field count below Elasticsearch's default limit of 1000 fields. Only specify field mappings for the fields you wish to search on or query. See https://www.elastic.co/guide/en/kibana/master/development-plugin-saved-objects.html#_mappings

id	before	after	diff
graph-workspace	9	10	+1

History

• 💔 Build #93086 failed 74989ee
• 💚 Build #92779 succeeded 55667df

To update your PR or re-run it, just comment with:
@elasticmachine merge upstream