* Updated scheduler to use v2 API by default
* Updated scheduling for post/page resource types
* Extended base method to take options param with token and jwt options
* Updated token expiration to 6 hours after publish/blog start time to allow retries
- Currently, we create a request ID for internal use if one isn't set & this is used in logs
- If a custom request ID is set via X-Request-ID header, this gets logged, however, we don't return this with the response
- Means that a custom ID gets lost on the way back out, and makes tracing requests through a system trickier
- This change ensures that if X-Request-ID is set on the request, it is also set on the response so that requests can be properly traced
- It's easy to set this in e.g. nginx so that the feature becomes available - Ghost doens't need to do this
- Note: also split request id handling out into new middleware
no-issue
This also adds a basic check before handing of to the members-ssr
module, this should make logs a little less noisy and only log warnings
if a token was passed and that token was invalid/incorrect.
no issue
- v0.1 is ☠️ so there's no longer any use of client auth
- removes all code related to `clients` and `client_trusted_domains`
- noops the "add backup client" migration in 1.7 because the referenced fixture no longer exists causing migrations and consequently all regression tests to fail
no issue
- adds `config:redirects` config option that defaults to `true`
- when set to `false`
- `/ghost/` will 404 on the front-end when a separate admin url is configured
- all `{resource}/edit/` URLs on the front-end will 404
no issue
- Removed v0.1 controllers
- Removed 0.1 API unit tests
- Removed 0.1 API app and mount point
- Removed leftover use of v0.1 in entry-lookup test suite
- Removed frontend client API enpoints and related code (middleware)
- Fixed prev/next test suites to use v2 API
- Set default API version to explicit v2 in UrlUtils
- Removed v0.1 API regex from public files middleware
no issue
- we used to redirect paths such as `/logout/` and `/signin/` to the admin but they are no longer desired
- with the introduction of members these URLs can be confused with front-end member related actions
- we want to be able to optionally "turn off" redirects to the admin to help mask the admin url when it's configured to be separate to the front-end
no issue
- `vhost` as used in b46f9b1dc2 does not pass down the `trust proxy` setting to child apps so it's required to be called explicitly in each child app
- fixed URLs being output as http:// instead of https:// when the front-end is accessed with `x-forwarded-proto: 'https'`
no issue
- added our theme error handling middleware to {admin}/content/ so that cache headers are properly set for 404s
- only registered {admin}/content when a separate admin url is configured so that we're not overriding {site}/content
no issue
- Session controllers were using API v1 http method which bypassed "frame" introduced with API v2.
- Changes here are just a long-awaited cleanup to allow completely remove v0.1 code
no issue
- As v0.1 API is dropped there is no need to keep an API client around
- Removed references to ghost-sdk in regression test suite
- Removed routes to /public/ghost-sdk.js
- Removed reference to ghost-sdk in grunt build process
refs https://github.com/TryGhost/Ghost/issues/11083
- the `/api/v2/admin/site/` endpoint is "public" and as such was not using the `authAdminApi` middleware stack so it did not act like other API endpoints with protocol or trailing-slash redirects
- adds `publicAdminApi` middleware array and uses it for the `/site/` endpoint in both v2 and canary API versions
closes https://github.com/TryGhost/Ghost/issues/11078
Problem:
- the admin client makes an XHR request to the `/private/` endpoint when a private site is configured
- when a separate admin URL is configured this was causing 500 errors in the admin client because missing CORS headers on the endpoint was causing browsers to abort the request
- browsers will also look at the CORS headers on any resources that are the result of a redirect and abort the request if they do not allow cross-origin requests, this means allowing all requests on `/private/` is not enough
Solution:
- uses the `cors` middleware with a dynamic options function for the whole of the front-end site app
- dynamic options function allows the following requests through:
- same-origin (browsers and non-browser agents will not send an `Origin` header)
- origin is `localhost` or `127.0.0.1` with any protocol and port
- origin matches the configured `url` hostname+port on any protocol
- origin matches the configured `admin:url` hostname+port on any protocol
no issue
- when too many login attempts were detected for the `/private/` form we were throwing 500 errors instead of the more appropriate 429 error that we use everywhere else for "too many request" type errors
no-issue
* Installed members-api@0.5.0 members-ssr@0.3.1
* Supported multiple members-forms
* Used members canary api
* Added GET handler to /members/ssr for id token
The identity token will be used to ensure that a payment is linked to the correct member
* Added stripe.js to ghost_head when members enabled
* Added basic support for linking to stripe checkout
* Removed listener to title and icon settings changes
* Added stripe subscription config
closes#10060
- Implemented scheduling for posts and pages
- Added cache invalidation when scheduling
- Refactored admin token eneration function to accept existing key as parameter in tests
- Added Ghost Scheduler Integration fixture
- Added fixture for permissions for post publish action
- Migrated getScheduled method to v2
- Did not add support for 'from' and 'to' parameters as they were not used by DefaultScheduler
- This method needs rethinking in a long run as it's an ugly hack and should rather become proper endpoint that returns JSON data instead of models
- Removed unused auth middleware from v2 routes
- Added internal scheduler role
- Implemetnted transactions in v2 frame
- This takes into account scenario mentioned in c93f03b87e
- Specifically:
>if two queries happening in a transaction we have to signalise
knex/mysql that we select for an update
otherwise the following case happens:
you fetch posts for an update
a user requests comes in and updates the post (e.g. sets title to "X")
you update the fetched posts, title would get overriden to the old one
refs #10060
- Migrated authentication.resetPassword method to v2
- Migrated authentication.acceptInvitation method to v2
- Migrated authentication.setup method to v2
- Added missing test coverage for "setupUpdate" method
- Migrated authentication.updateSetup method to v2
- Migrated authentication.isInvitation method to v2
- Migrated authentication.isSetup method to v2
- Removed unused 'setup.completed' event as it wasn's used anywhere in the system and has been complicating the logic unnecessarily
- Without the event, it's possible to simplify sendNotification method to just use email address of the user
- Added email sending check to v0.1 test suite
- Refactored sendNotification method to just use email address as parameter
- Renamed sendNotification to sendWelcomeMail
- The only thing the method does now is sending welcome mail, so new naming seems natural :)
no-issue
Previously we were using the error logger, which is probably a bit
extreme for these errors. This also removes the stacktrace from the logs
so we don't enter fresh hell whilst developing/looking through logs.
refs #10932
Previously we were only applying the cors middleware to the options
preflight request, which meant that if the request errored, the cors
headers would not be applied, resulting in the client being unable to
read response data. This applies the cors middleware to _all_ requests
to the Admin API.
