The Kontent.ai CLI helps you when you need to change content models within your Kontent.ai projects and migrate existing content to match the changes. The CLI provides you with guidance on how to write and run migration scripts.
NOTE: The Kontent.ai CLI tool supports only Javascript files, so if you write your migrations in Typescript or any other language you have to transpile your code before running.
The Kontent.ai CLI requires Node 10+ and npm 6+, and uses the Kontent.ai Management SDK to manipulate content in your projects.
npm install -g @kontent-ai/cli
The current version of the CLI is useful for creating and running migration templates. Let's go through creating your first migration for a Kontent.ai project.
When you need to add new features to your project and app, it's better to verify the changes in a separate non-production environment. In Kontent.ai, clone your project from the list of your projects.
To improve the learning curve of our new CLI, we've prepared a Kontent.ai CLI boilerplate with examples on how to use the CLI. Clone the boilerplate GitHub repository on your drive. In the next step, you'll run a migration script from the boilerplate's Migrations
directory.
Open a command line and navigate to the root of the boilerplate folder (should be migrations-boilerplate
) and execute the following commands:
# Navigates to the root of the Kontent.ai CLI boilerplate folder.
cd ./migrations-boilerplate
npm install
# Registers an environment (a pair of two keys, a project ID and API key used to manage the project) for migrations.
kontent environment add --name DEV --api-key <Api key> --project-id <Project ID> (Use the copy of your production project from the first step)
# Runs a specific migration.
npm run migrate 01_sample_init_createBlogType
Kontent.ai CLI supports only running JavaScript migration files so in case you want to write in TypesSript, CoffeScript or in any other language you have to transpile your code before running. In the case of TypeScript, you may use this example from Kontent.ai CLI boilerplate
That's it! You've run your first Kontent.ai migration. This migration created a content type called Blog that contains three text elements named Title, Author and Text. The sample migration is written in TypeScript.
The boilerplate is configured to transpile TypeScript migrations into plain JavaScript so that the Kontent.ai CLI can execute the migrations. Note that if you don't want to use TypeScript for your migrations, it's fine to write the migrations directly in JavaScript.
You should now be able to go through the other boilerplate sample migrations. The migration scripts in the Migrations directory all focus on one scenario – replacing a piece of text from the Author text element with Author content items, which contain more information about the author. This way you can replace the texts within your items by more complex objects containing, for example, images and rich text.
You can use similar approach for your own specific scenarios. For example, imagine you need to add display information to the images inserted in your articles. You may want to specify relative size or caption for each image. To do this, you would need to open each article and replace every image with a component that would contain the image and a few elements for image metadata. You'd create small migration scripts for separate parts of the scenario (such as creating a new type, updating the articles, and so on) and the migrations will take care of the process for all articles within your project.
Use the --help
parameter to display the help section for CLI tool.
kontent --help
Combine the --help
parameter with a specific command to learn more about that command.
kontent migration add --help
The supported commands are divided into groups according to their target, at this first version there are just to spaces "migration" and "environment" containing following commands:
-
environment add
– Store information about the environment locally.- The environment is defined as a named pair of values. For example, "DEV" environment can be defined as a pair of a specific project ID and Management API key. This named pair of values is stored within your local repository in a configuration file named
.environments.json
. - You can specify a named pair of project ID and Management API key using these options:
--project-id <your project ID> --api-key <management api key> --name <name of the environment>
.
- The environment is defined as a named pair of values. For example, "DEV" environment can be defined as a pair of a specific project ID and Management API key. This named pair of values is stored within your local repository in a configuration file named
-
migration add --name <migration name>
– Generates a script file (in JavaScript or TypeScript) for running a migration on a Kontent.ai project.-
The file is stored in the
Migrations
directory within the root of your repository. -
Add your migration script in the body of the
run
function using the Kontent.ai Management SDK that was injected via theapiClient
parameter. -
Add your rollback script in the body of the
rollback
function using the Kontent.ai Management SDK that was injected via theapiClient
parameter. -
To choose between JavaScript and TypeScript when generating the script file, use the
--template-type
option, such as--template-type "javascript"
. -
The migration template contains an
order
property that is used to run a batch of migrations (range or all) in the specified order. Order can be one of the two types -number
ordate
.- Ordering by
number
has a higher priority. Theorder
must be a unique positive integer or zero. There may be gaps between migrations, for example, the following sequence is perfectly fine 0,3,4,5,10 - Ordering by
date
has a lower priority. To add date ordering use the switch option-d
. The CLI will generate a new file which name consists of the date in UTC and the name you have specified. Moreover, the propertyorder
inside the file will be set to the Date accordingly. - Executing all migrations will firstly migrate migrations with orders specified by number and only then migrations with order specified by date.
- By specifying range you can migrate either number-ordered migrations or date-numbered migrations. They can't be combined.
// Example migration template import {MigrationModule} from "@kontent-ai/cli"; const migration: MigrationModule = { order: 1, run: async (apiClient) => { // TODO: Your migration code. }, }; export default migration;
- Ordering by
-
-
migration run
- Runs a migration script specified by file name (option--name <file name>
), or runs multiple migration scripts in the order specified in the migration files (options--all
or--range
).- By adding
--range
you need to add value in form ofnumber:number
in case of number ordering or in the format ofTyyyy-mm-dd-hh-mm-ss:yyyy-mm-dd-hh-mm-ss
in case of date order.When using the range with dates, only the year value is mandatory and all other values are optional. It is fine to have a range specified like
T2023-01:2023-02
. It will take all migrations created in January of 2023. Notice the T at the beginning of the Date range. It helps to separate date ordering from number order. - You can execute a migration against a specific project (options
--project <YOUR_PROJECT_ID> --api-key <YOUR_MANAGEMENT_API_KEY>
) or environment stored in the local configuration file (option--environment <YOUR_ENVIRONMENT_NAME>
). - To execute your
rollback
scripts instead ofrun
scripts, use the switch option--rollback
or shortly-b
. - After each run of a migration script, the CLI logs the execution into a status file. This file holds data for the next run to prevent running the same migration script more than once. You can choose to override this behavior, for example for debugging purposes, by using the
--force
parameter.Note: For every migration there is only one record in the status file. Calling run/rollback continously overrides the that record with new data.
- You can choose whether you want to keep executing the migration scripts even if one migration script fails (option
--continue-on-error
) or whether you want to get additional information logged by HttpService into the console (option--log-http-service-errors-to-console
).
- By adding
-
backup --action [backup|restore|clean]
- This command enables you to use Kontent.ai backup manager- The purpose of this tool is to backup & restore Kontent.ai projects. This project uses CM API to both get & restore data.
You might want to implement your way to store information about migrations status. For instance, you would like to save it into DB such as MongoDB, Firebase, etc,... and not use the default JSON file. Therefore, we provide you with an option to implement functions readStatus
and saveStatus
. To do so, create a new file called plugins.js
at the root of your migrations project, and implement mentioned functions there. To fit into the required declarations, you can use the template below:
//plugins.js
exports.saveStatus = async (data) => {}
exports.readStatus = async () => {}
Note: Both functions must be implemented.
It is also possible to use Typescript. We have prepared types SaveStatusType
and ReadStatusType
to typesafe your functions. To create plugins in Typescript, create a file plugins.ts
and implement your functions there. We suggest using and implementing the template below:
//plugins.ts
import type { ReadStatusType, SaveStatusType } from "@kontent-ai/cli";
export const saveStatus: SaveStatusType = async (data: string) => {}
export const readStatus: ReadStatusType = async () => {}
Note: Don't forget to transpile
plugins.ts
intoplugins.js
otherwise your plugins will not work.
By default, we do not provide any additional logs from the HttpService. If you require these logs, you can change this behavior by using (option --log-http-service-errors-to-console
).
If you come across an error and you're not sure how to fix it, execute your migration script as follows and setup your debugger to the specified port.
node --inspect .\node_modules\@kontent-ai\cli\lib\index.js migration run -n 07_sample_migration_publish -e DEV
-
Writing migration scripts can involve a lot of repetitive work, especially when it requires getting different object types and iterating through them. That's why we've decided to continue improving the developer experience and focus on that in upcoming releases. We plan to reduce the code that you need to write to the bare minimum by providing you with a "command builder". This builder will allow you to write migrations using queries and callbacks that should be applied to every object selected by that query. For example, select content types, all items based on the types, and all variants of the items, and execute your callback function on them.
-
The tool isn't reserved only for migrations. A valid use case could also be Kontent.ai project data export and import, which could together with the possibility to clone/create/archive projects via the management API be a great way to e.g. run integration tests on the test environment that would be archived after the successful tests run.
Check out the contributing page to see the best places to file issues, start discussions, and begin contributing. We have lot of ideas on how to improve the CLI and developer experience with our product, but we'd love to hear from you so we can focus on your needs.