Some time ago I wrote about an alternative to migrations to set up contentful spaces. This was based upon desired state configuration. I no longer work at the company that had that utility so can’t ask to open source it. These are the notes that I am going to use to recreate the utility from scratch. These notes are based from my memory of how Contentful works. Desired State Configuration is far easier to manage than migrations. You only need to edit one file that is kept in line with source control. If there are multiple spaces using the same configuration it will bring them into alignment.
The basic idea is that you have a declaration file that shows how the contentful space is meant to be defined. This covers the content types in the content model. When the utility is run it attempts to upgrade the space to match the declaration.
It will add types that don’t exist first. These will be a single field version of the type. This is sufficient to use them as a reference for other types.
Next it will ensure that the fields on each of the types match the definition given in the configuration. Fields that don’t exists will be added. Fields that already exist will be updated if it is possible to upgrade. Contentful has rules about how a type can change: you can’t turn a string into a number or a float into an integer. Fields that have been removed will be marked as disabled. It can’t work around the hard limit of 50 fields per content type so you may need to manually delete existing disabled fields.
The third part is to apply the “Editor Interface” to any field that has it specified. This will allow the specific UI rules to be applied.
Contentful is eventually consistent with the type updates so you need to wait after making an update to ensure that the read-side will have applied the update. This also means that the migration itself will need to carefully rate-limit it’s actions.
To get started you need a contentful space and a cma token. These need to be generated manually from: https://app.contentful.com/spaces/{insert space id here}/api/cma_tokens
For this example I’ll start using the sample blog space that you can generate with a free account.
This is the curl command that I use to extract the details:
curl https://api.contentful.com/spaces/{insert space id here}/environments/master/content_types -H "Authorization: Bearer {insert token here}"
This returns some json (I may have manually adapted this space from the defaults):
{
"sys": {
"type": "Array"
},
"total": 3,
"skip": 0,
"limit": 100,
"items": [
{
"sys": {
"space": {
"sys": {
"type": "Link",
"linkType": "Space",
"id": "e6p9dxi1frd5"
}
},
"id": "blogPost",
"type": "ContentType",
"createdAt": "2019-05-22T13:33:53.326Z",
"updatedAt": "2020-05-14T17:28:43.402Z",
"environment": {
"sys": {
"id": "master",
"type": "Link",
"linkType": "Environment"
}
},
"publishedVersion": 3,
"publishedAt": "2020-05-14T17:28:43.402Z",
"firstPublishedAt": "2019-05-22T13:33:54.668Z",
"createdBy": {
"sys": {
"type": "Link",
"linkType": "User",
"id": "4akz83hNhoZtJH86B6riKE"
}
},
"updatedBy": {
"sys": {
"type": "Link",
"linkType": "User",
"id": "4akz83hNhoZtJH86B6riKE"
}
},
"publishedCounter": 2,
"version": 4,
"publishedBy": {
"sys": {
"type": "Link",
"linkType": "User",
"id": "4akz83hNhoZtJH86B6riKE"
}
}
},
"displayField": "title",
"name": "Blog Post",
"description": null,
"fields": [
{
"id": "title",
"name": "Title",
"type": "Symbol",
"localized": false,
"required": true,
"validations": [],
"disabled": false,
"omitted": false
},
{
"id": "slug",
"name": "Slug",
"type": "Symbol",
"localized": false,
"required": true,
"validations": [],
"disabled": false,
"omitted": false
},
{
"id": "heroImage",
"name": "Hero Image",
"type": "Link",
"localized": false,
"required": true,
"validations": [],
"disabled": false,
"omitted": false,
"linkType": "Asset"
},
{
"id": "description",
"name": "Description",
"type": "Text",
"localized": false,
"required": true,
"validations": [],
"disabled": false,
"omitted": false
},
{
"id": "body",
"name": "Body",
"type": "Text",
"localized": false,
"required": true,
"validations": [],
"disabled": false,
"omitted": false
},
{
"id": "author",
"name": "Author",
"type": "Link",
"localized": false,
"required": false,
"validations": [],
"disabled": false,
"omitted": false,
"linkType": "Entry"
},
{
"id": "publishDate",
"name": "Publish Date",
"type": "Date",
"localized": false,
"required": true,
"validations": [],
"disabled": false,
"omitted": false
},
{
"id": "tags",
"name": "Tags",
"type": "Array",
"localized": false,
"required": false,
"validations": [],
"disabled": false,
"omitted": false,
"items": {
"type": "Symbol",
"validations": [
{
"in": [
"general",
"javascript",
"static-sites"
]
}
]
}
},
{
"id": "backgroundColour",
"name": "BackgroundColour",
"type": "Symbol",
"localized": false,
"required": false,
"validations": [],
"disabled": false,
"omitted": false
}
]
},
{
"sys": {
"space": {
"sys": {
"type": "Link",
"linkType": "Space",
"id": "e6p9dxi1frd5"
}
},
"id": "person",
"type": "ContentType",
"createdAt": "2019-05-22T13:33:53.330Z",
"updatedAt": "2019-05-23T08:46:50.110Z",
"environment": {
"sys": {
"id": "master",
"type": "Link",
"linkType": "Environment"
}
},
"publishedVersion": 5,
"publishedAt": "2019-05-23T08:46:50.110Z",
"firstPublishedAt": "2019-05-22T13:33:54.691Z",
"createdBy": {
"sys": {
"type": "Link",
"linkType": "User",
"id": "4akz83hNhoZtJH86B6riKE"
}
},
"updatedBy": {
"sys": {
"type": "Link",
"linkType": "User",
"id": "4akz83hNhoZtJH86B6riKE"
}
},
"publishedCounter": 3,
"version": 6,
"publishedBy": {
"sys": {
"type": "Link",
"linkType": "User",
"id": "4akz83hNhoZtJH86B6riKE"
}
}
},
"displayField": "name",
"name": "Person",
"description": null,
"fields": [
{
"id": "name",
"name": "Name",
"type": "Symbol",
"localized": false,
"required": true,
"validations": [],
"disabled": false,
"omitted": false
},
{
"id": "title",
"name": "Title",
"type": "Symbol",
"localized": false,
"required": true,
"validations": [],
"disabled": false,
"omitted": false
},
{
"id": "company",
"name": "Company",
"type": "Symbol",
"localized": false,
"required": true,
"validations": [],
"disabled": false,
"omitted": false
},
{
"id": "shortBio",
"name": "Short Bio",
"type": "Text",
"localized": false,
"required": true,
"validations": [],
"disabled": false,
"omitted": false
},
{
"id": "email",
"name": "Email",
"type": "Symbol",
"localized": false,
"required": false,
"validations": [],
"disabled": false,
"omitted": false
},
{
"id": "phone",
"name": "Phone",
"type": "Symbol",
"localized": false,
"required": false,
"validations": [],
"disabled": false,
"omitted": false
},
{
"id": "facebook",
"name": "Facebook",
"type": "Symbol",
"localized": false,
"required": false,
"validations": [],
"disabled": false,
"omitted": false
},
{
"id": "twitter",
"name": "Twitter",
"type": "Symbol",
"localized": false,
"required": false,
"validations": [],
"disabled": false,
"omitted": false
},
{
"id": "github",
"name": "Github",
"type": "Symbol",
"localized": false,
"required": false,
"validations": [],
"disabled": false,
"omitted": false
},
{
"id": "image",
"name": "Image",
"type": "Link",
"localized": false,
"required": false,
"validations": [],
"disabled": false,
"omitted": false,
"linkType": "Asset"
},
{
"id": "favoriteIceCream",
"name": "favorite ice cream",
"type": "Symbol",
"localized": false,
"required": false,
"validations": [],
"disabled": false,
"omitted": false
},
{
"id": "petName",
"name": "pet name",
"type": "Symbol",
"localized": false,
"required": false,
"validations": [],
"disabled": false,
"omitted": false
}
]
},
{
"sys": {
"space": {
"sys": {
"type": "Link",
"linkType": "Space",
"id": "e6p9dxi1frd5"
}
},
"id": "update",
"type": "ContentType",
"createdAt": "2019-05-22T13:46:35.343Z",
"updatedAt": "2019-05-22T13:46:38.843Z",
"environment": {
"sys": {
"id": "master",
"type": "Link",
"linkType": "Environment"
}
},
"publishedVersion": 3,
"publishedAt": "2019-05-22T13:46:38.843Z",
"firstPublishedAt": "2019-05-22T13:46:35.942Z",
"createdBy": {
"sys": {
"type": "Link",
"linkType": "User",
"id": "4akz83hNhoZtJH86B6riKE"
}
},
"updatedBy": {
"sys": {
"type": "Link",
"linkType": "User",
"id": "4akz83hNhoZtJH86B6riKE"
}
},
"publishedCounter": 2,
"version": 4,
"publishedBy": {
"sys": {
"type": "Link",
"linkType": "User",
"id": "4akz83hNhoZtJH86B6riKE"
}
}
},
"displayField": null,
"name": "Update",
"description": "Latest update",
"fields": [
{
"id": "update",
"name": "Update",
"type": "RichText",
"localized": false,
"required": false,
"validations": [],
"disabled": false,
"omitted": false
}
]
}
]
}
It is unlikely that you will need to add paging to the content type analysis, but I will make the utility handle it.
The key information here is the list of items each with a list of fields and a specified displayField.
I’ll return to this topic as I develop this utility. At time of writing there is no code for this project.
Note that this content type uses readable identifiers for the fields. I would always recommend doing this as it greatly simplifies the use of this data.
Developer Rants 