Permission labels are indexed by type text in SOLR

[Zharktas]

CKAN Version if known (or site URL)

2.7, master

Please describe the expected behaviour

Organization Admin should see organizations private datasets even when SOLR indexing has been changed to accommodate different languages. The permission label should not be index by language rules.

Please describe the actual behaviour

If indexing rules are changed for example to use EdgeNGram, the permission label id's are indexed by these rules and organization admin, editor or member can not see organizations private datasets.

What steps can be taken to reproduce the issue?

Apply EdgeNGram to solr indexing.

Create organization.
Create user.
Add user to organization.
Create private Dataset to organization.

Possible fix

Use type textgen instead of text in the default schema.xml.

[wardi]

@Zharktas sounds right to me. Mind submitting a PR?

[TkTech]

Use string not textgen, which in CKAN's schema is not the same as the old Solr 1.* textgen FieldType (confusing I know).

[Zharktas]

@wardi submitted minor PR for this.