Ember OSF
This repo contains code for interacting with the OSF APIv2 inside of an Ember app. See addon API Docs for a list of what this addon provides.
Contributing
Please read the CONTRIBUTING.md
Installing tools to develop and modify this addon
See the yarn docs for installing and using yarn. These instructions will prepare your environment if you plan to modify and test this addon in isolation.
-
git clone
this repository yarn install --frozen-lockfile
bower install
Using this addon in another Ember app
For production use
Other applications that wish to consume this addon should add the following line manually to the consuming Ember app's
package.json
file, then run yarn install
and bower install
inside that app.
"ember-osf": "git+https://github.com/CenterForOpenScience/ember-osf.git#NewestCommitHashGoesHere",
In the future, we will provide an installable npm
package to simplify this process.
For local development
- Clone the repository:
git clone https://github.com/CenterForOpenScience/ember-osf.git
- From the consuming Ember app:
- install the addon and it's dependencies:
ember install ../ember-osf
- this generates a config/local.yml file (see 'Configuration' below)
- link the app for local development:
npm link ../ember-osf
- Import code from ember-osf like:
import Ember from 'ember';
import OsfTokenLoginRouteMixin from 'ember-osf/mixins/osf-token-login-route';
export default Ember.Route.extend(OsfLoginRouteMixin);
Note: Running ember install will automatically install many bower and npm dependencies for ember-osf.
Configuration
Specifying configuration information
local.yml settings
If for some reason you don't have a config/local.yml you can generate one. To do this:
ember generate ember-osf
Ember-osf needs certain configuration variables to run. This is usually done via a config file structured as follows:
OAUTH_SCOPES: osf.full_write
REDIRECT_URI: http://localhost:4200/login
CLIENT_ID: null
PERSONAL_ACCESS_TOKEN: null
You will need to fill out options for each backend you want to use (see 'Running' below). We recommend using the 'test' backend for development and testing as it is the most stable of our environments. When configuring your application, make sure that your login redirect uri is correct. If it needs a trailing slash, be sure to include a trailing slash!
Edit the new file (installed in the config directory) and set:
-
CLIENT_ID
to the client id of your developer application -
PERSONAL_ACCESS_TOKEN
to the newly generated token (Only required or recognized for the LOCAL backend; do not set this value for staging, production, or test backends) -
REDIRECT_URI
: Must exactly match the redirect URI used to register the OAuth developer application. Default value is appropriate for local development usingember server
, with a login page at/login
Because of the potential for this file to include sensitive information, we strongly recommend adding this file to
.gitignore
for your project.
Alternate option: Environment variables
If you do not wish to use file-based configuration, any of the settings above can be overridden individually as environment variables with the same name as appears in the config file. Eg
BACKEND=test CLIENT_ID=gibberish ember server
If you provide a setting in both the config file and an environment variable, the environment variables take precedence.
You can always override auth-related settings, but attempts to override server URLs will be ignored unless you
explicitly specify BACKEND=env
(see "Running" for example).
Using the API
Most apps that use this addon will authorize requests via OAuth2. As may be apparent from the CLIENT_ID
and
REDIRECT_URI
settings above, you will need to create a developer application
on the relevant version of the OSF, and provide the appropriate settings for your app.
Advanced usage: Selecting an authorization type
We expect that most projects based on ember-osf
will authenticate via OAuth 2.0 ("Token Login"); the addon is
configured to use this out of the box, so long as you provide your own login page based on the appropriate mixins.
This is the most effective way for third-party applications to work with our services.
If you are developing an application that will be hosted under the osf.io
domain, you may wish to use cookie-based
authentication instead. In that rare case, add the following lines to your config/environment.js
file:
var authorizationType = 'cookie';
ENV.authorizationType = authorizationType;
ENV['ember-simple-auth'] = {
authorizer: `authorizer:osf-${authorizationType}`,
authenticator: `authenticator:osf-${authorizationType}`
};
Running the OSF Locally (optional)
For local development, you will need to be running the OSF APIv2.
To connect to the APIv2 while using fakecas, you
will need to generate a personal access token on your local OSF instance here-
go ahead and grant full privilege access to all scopes (the osf.full_write
option).
Using this addon
Ember Data: Using the OSF models
The models, serializers, adapters bundled in this addon with be available to you automatically. For example, simply do:
this.store.findAll('node')
to fetch all nodes (or at least the first page of results). If you need to fetch many results, see the API docs for information about how to handle pagination. Ember-osf also provides support for paginated relationship requests via a third-party addon.
MathJax
We use MathJax to make math look pretty in all browsers. If you want this in your application, copy this section into the <head>
element of your index.html
file:
<script type="text/javascript"
src="//cdnjs.cloudflare.com/ajax/libs/mathjax/2.6.1/MathJax.js?config=TeX-AMS-MML_HTMLorMML">
</script>
<script type="text/x-mathjax-config">
MathJax.Hub.Config({
tex2jax: {inlineMath: [['$','$'], ['\\(','\\)']], processEscapes: true},
skipStartupTypeset: true
});
</script>
Advanced: using components and styles
Some of the ember-osf components require additional configuration to take advantage of premade widgets or styles.
Some of these settings, as well as recommended best practices, are gathered together in a demonstration app that consumes this addon.
Running
Using the addon with a specific server
We recommend developers target our test server:
- test (
test
): matches production features, very stable
Ember-osf also ships with builtin support for several other servers:
- local (
local
): for developers running the OSF stack locally - staging (
stage
): contains bleeding edge features, but less stable - staging2 (
stage2
) orstaging3
(stage3
): Staging servers that run specific feature branches, usually for functionality that is being tested for longer or not targeted for immediate release - production (
prod
): The main osf.io site. This is a good choice for apps you deploy, but please be a good citizen and avoid using the production server for your test data.
Then (using test as an example) run:
BACKEND=test ember server
and visit your app at http://localhost:4200.
Note: When run from within the ember-osf
repository, this command runs the dummy app contained in /tests/dummy
Using a custom OSF backend
In certain circumstances, you may wish to use a custom set of servers not known to the ember-osf
addon.
You can elect to specify your server URLs individually, by specifying BACKEND=env
and passing additional environment
variables/config file entries. For example:
BACKEND=env OSF_URL=https://e.io/ OSF_API_URL=https://api.e.io OSF_RENDER_URL=https://mfr.e.io/render OSF_FILE_URL=https://files.e.io OSF_HELP_URL=https://help.e.io OSF_COOKIE_LOGIN_URL=https://accounts.e.io/login OSF_OAUTH_URL=https://accounts.e.io/oauth2/authorize ember server
Running Tests
-
yarn test
will run all tests used by Travis -
ember test --server
will run just ember tests, and reload any time that code is changed
For more information on using ember-cli, visit http://www.ember-cli.com/.
Distribution / releases
The information below applies only to package maintainers, and most users will not need this.
The following commands can be used to publish a new release of ember-osf
:
-
yarn run bump-version [<newversion> | major | minor | patch]
: Verify that the code is in a releasable state, increment the version number without generating a new git commit, and update documentation. See docs for recognized options. -
yarn run make-release
: Uses git flow to prepare a new release. You must be on the develop branch, commit all changes, and have rungit init
in this folder at least one time (ever). -
yarn publish
: Publish a new version of the package to the NPM registry. It is highly recommended that you do this from a fresh checkout of the repo, and validate the package contents before uploading. You must be a recognized NPM collaborator or this command will fail.