refs: https://github.com/TryGhost/Team/issues/1599
refs: f3d5d9cf6b
- this commit adds the concept of a frontend data service, intended for passing data to the frontend from the server in a clean way. This is the start of a new & improved pattern, to hopefully reduce coupling
- the newly added internal frontend key is then exposed through this pattern so that the frontend can make use of it
- the first use case is so that portal can use it to talk to the content API instead of having weird endpoints for portal
- this key will also be used by other internal scripts in future, it's public and therefore safe to expose, but it's meant for internal use only and therefore is not exposed in a generic way e.g. as a helper
refs: https://github.com/TryGhost/Toolbox/issues/229
- Removed the final usage (logging only) of config.get('api:versions') and removed the config
- This is one of the last references to API versioning in the codebase & we don't need it anymore
- Cleaned up a random comment whilst searching for refs
- updated usage of url-utils.urlFor to work with v3
Co-authored-by: Renovate Bot <bot@renovateapp.com>
Co-authored-by: Hannah Wolfe <github.erisds@gmail.com>
- In the interest of simplifying and cleaning up this code which affects Ghost's boot process
- Moved config to DI, injecting only known properties. It's expected that you must restart the server to pick up new config
- Even in tests!
- Got rid of unused bluebird code, but also reworked some promise-based code to be simpler using util.promisify
- following on from removing api versioning logic from the frontend, it's possible to make more sense of what's happening
- this commit first introduces a proper jsdoc'd object that gets passed through the frontent load & reload flow
- that object contains the urlService and optionally our routeSettings processed from routes.yaml
- additionally, we were passing around a start boolean, which told the routerManager whether to just init, or init+start
- with this refactor, we always pass in the routeSettings when we want to do init+start, so we no longer need a boolean
- The refactor itself moves logic from the reload function in site.js and urlservice + routesettings fetching logic from routes.js
into the reloadFrontend function in bridge.js.
- This makes it clearer to see what happens when we call reloadFrontend.
- This commit also makes it clearer to see what is happening with the route settings, where they are needed and why
- Ideally we'd also clean up the weird dupliated logic and somewhat unnecessary routes.js file
refs: https://github.com/TryGhost/Toolbox/issues/228
- we are getting rid of the concept of api versions from Ghost
- this means getting rid of them from the frontend as well, and from themes
refs https://github.com/TryGhost/Toolbox/issues/280
- This is continuation of putting pieces together to allow Ghost notifying owner and admin users about version mismatch errors
- This changeset adds the api-version-compatibility intialization during the boot (needed to be able to pospone the send email initialization so that it's testeable from the e2e tests)
- There's also a data service which handles fetching/saving notifications and fetching emails of users that should be notified
refs https://github.com/TryGhost/Toolbox/issues/254
- it's useful for us to know how long it takes the theme service to
init, especially as it is dealing with a lot of user-provided files
outside of the Ghost codebase
refs https://ghost.slack.com/archives/C02G9E68C/p1647599592576139
refs https://ghost.slack.com/archives/C02G9E68C/p1647620250625909
Issue: `Cannot destructure property 'fromRegex' of 'this.redirectsredirectId]' as it is undefined.` is being thrown, only when running all tests.
Cause: duplicate redirects are added to a redirectManager, and not cleared correctly in the redirectManager, which throws an error when removing one of the duplicate redirects.
Because the same redirectManager is used, multiple event listeners are connected to the same redirectManager. So when the offer service has been initialised multiple times, multiple listeners are added, which create a redirect for every newly created offer... to the same redirectManager.
So there are three possible fixes for the same problem (would be best to fix them all):
- Create a new redirectManager every time the offer service is initialised
- Figure out a way to remove DomainEvents subscribers between tests
- Don't add the same redirect id multiple times to redirectIds in addRedirect from the express-dynamic-redirects package
This commit contains a fix for the first solution.
It also moved the offers service initialising before the frontend (to `initServicesForFrontend`)
refs https://github.com/TryGhost/Toolbox/issues/163
- as of Node 15, unhandled rejections will exit the process so if
Ghost is running on Node 15+ and encounters one, it will kill Ghost
- if Sentry is enabled, it will add a handler for the event that will
send it to Sentry but the logging is sent to stdout/stderr, which means
we lose it in Ghost logs
- this commit adds a process handler for the `unhandledRejection` event
which will log the reason to Ghost logs and prevent Ghost from
exiting
refs https://github.com/TryGhost/Toolbox/issues/175
- we're going to be making some changes in v5 wrt supported databases
- we needed a way of detecting the difference between MySQL 5 + 8,
MariaDB etc
- I've created `@tryghost/database-info`, which is a small wrapper
around `knex`, which returns this information
- this commit:
- adds the library to Ghost
- initializes the DB info library upon boot
- updates the Admin API /config/ endpoint and UpdateCheck to return
the new string - `mysql5`, `mysql8` etc
refs: https://github.com/TryGhost/Toolbox/issues/146
Switched to @tryghost/logging instead of passing around the library. The main sticking points of this change are jobs. When jobs are launched we don't want them to use a separate @tryghost/logging instance because they would start parallel rotation jobs. @tryghost/logging v2.x passes all logs to the parent process if run in a child process, so that we can use the same patterns in jobs and the rest of the codebase.
refs https://github.com/TryGhost/Toolbox/issues/152
- Because the root app module was initialized only once per runtime it caused all the express apps to stack on each other causing all sorts of strange behavior when trying to test redirects/vhost mounts etc. Lesson here: be very cautious of how the module is initialized, an explicit function is almost always a better way!
refs 3c7a8dead4
- The idea is a rip-off from the referenced commit. It allows to isolate initialization logic to "frontend" mounted on top of a "parent"
- Gives path to simplify the express-mock tests
Co-authored-by: Hannah Wolfe <erisds@gmail.com>
refs https://github.com/TryGhost/Toolbox/issues/152
- This stops the mounting of the admin and frontend from being buried deep in express initialisation
- Instead it's explicit, which makes two things almost possible:
1. we can potentially boot the frontend or backend independently
2. we can pass services and settings loaded during boot into the frontend
- This needs more work, but we can start to group all the frontend code together
- Meanwhile we also need to rip apart the routing and url services to decouple the frontend from the backend fully
- BABY STEPS!
Co-authored-by: Hannah Wolfe <erisds@gmail.com>
refs: TryGhost/Toolbox#147
* Replaces all references to isIgnitionError with isGhostError
* Switches use of GhostError to InternalServerError - as GhostError is no longer public
There are places where InternalServerError is not the valid error, and new errors should be added to the @tryghost/errors package to ensure that we can use semantically correct errors in those cases.
refs https://github.com/TryGhost/Toolbox/issues/139
- The custom redirects services belong in the initServicesForFrontend because frontend depends on these to function properly. When placed in general init section the middleware would not get initialized properly before it's used by the "frontend express app"
- Reduced our maintenance middleware code down to the bare minimum!
- We have an old maintenance middleware in place to handle when a site is forcibly put into maintenance mode, or the urlService hasn't finished booting
- This maintenance middleware was mounted on every sub app, instead of globally for reasons I no longer remember
- Recently, we introduced a new, static version of maintenence middleware to show during the boot process so we can get the server started earlier & not drop requests
- This version has its own HTML template and doesn't depend on any of Ghost's error rendering code
- To simplify and help with decoupling, this commit merges the two middleware, so that the new independent & static middleware renders its template for any one of the 3 possible maintenance modes
- It only needs to exist in the top level app 🙌
TODO: move the maintenance middleware to its own file/package so it's not part of the app.js as that is weird
- Card asset reloading was incorrectly only happening if the API version changed 🙈
- In addition, having an init function was redundant, as theme activation happens on boot
- This meant that the card assets were being generated twice on boot
- Instead, we now only generate them on theme activation, which covers the boot case and simplifies all the logic
refs https://github.com/TryGhost/Toolbox/issues/135
- Looking closer into the reason why the test was failing without "forceStart" revealed that the server only start was overoptimized - "initServicesForFrontend" should be a part of a backend as those are generic theme services.
refs https://github.com/TryGhost/Toolbox/issues/135
- The reason the test **settings** test was failing when the force start flag was removed in the **custom themes** was the bridge! The bridge was trying to execute function on the frontend when the boot was done without initializing the frontend. The setting test was changing locale and the timezone which triggered events calling up on frontend components - we clearly don't want to do this when the instance is booted without the frontend
- To make event initialization conditional moved it to the "init". This way the event listeners are only set up when we boot with the "frontend" flag set to true
no issue
- `startsWith` method is way easier to read and understand. also, **probably** has better performance comparing to building up a regexp and then matching
refs https://github.com/TryGhost/Toolbox/issues/135
- These flags are meant to control initialization of sections of the boot sequence depending on the needs - with or without bakend (API)/frontend (public handlebars site)
- Ideally these flags should not be passed deep into the components, and if the are (like in the web/parent/app case) it's a smell that we need to move things up into the boot process!
- This is a minor bugbare, but it will affect some configuration I'm about to do for c8
- I've been wanting to do it for ages, middleware is plural all on it's own so it's an odd affectation in our codebase
- This also only exists in 2 places, everywhere else we use "middleware"
- Sadly it did result in a lot of churn as I did a full find and replace, but consistency is king!
- The version util is now required in logging as well as sentry and migrations
- I can see us needing it in config too, so put it straight to the first item
- we do this so that the debug statements can tell us how long each step took as we look to optimise everything
- At the moment there's no way to see in the logs when the URL Service finally finishes
- This is the moment when Ghost stops serving 503s
- Adding this log line so it's clear to see, which migh be useful whilst were refactorising
refs https://github.com/TryGhost/Toolbox/issues/125
- The in-memory url objects would be persisted in the contents folder allowing to take advantage of the cache during the next instance start
- Requires the new @tryghost/minifier package
- Adds a new service that will handle taking config from the theme and optionally including assets for Koenig editor cards
- It supports both css and js as cards may need one or both
- For any given config, the tool can find the matching files to include and concat and minify them into one file per type
- Currently has an override in place so that this is not yet customisable in the theme - will remove this override when we're ready for the feature
refs https://github.com/TryGhost/Members/commit/9e59f5a9
Since we have a DynamicRedirectManager for handling adding/removing
redirects at runtime, we no longer need the custom-redirects middleware.
The redirects service does however need an init method now to add the
custom redirects at Ghost boot, so it's been refactored into our Class &
DI pattern.
refs https://github.com/TryGhost/Ghost/issues/13380
- Now that i18n.t has been removed everywhere, we can cleanup the final usages
- Still TODO: merge the i18n logic into themeI18n, and get rid of shared/i18n entirely
no issue
Boot-time metrics are really valuable for understanding how releases affect the performance of Ghost. The new @tryghost/metrics package lets us ship these metrics to different sources using the same configuration framework as @tryghost/logging introduced.
refs c33b596e9c
- initializing the themes service before the custom theme settings service meant that `customThemeSettings.api` was not available when it's called as part of theme initialization
- swapped the boot order around so that the supporting service initialization is performed before the full themes service is initialized
refs https://github.com/TryGhost/Team/issues/1104
- bumped `@tryghost/custom-theme-settings-service` so it throws a more appropriate `ValidationError` when setting keys don't exist or a select value is not known
- changed the custom theme settings service to have a `.init()` method which creates an instance of the service under `.api` so that we're able to create the instance at a particular point in the boot process when we know the models have been initialised
- there were problems in tests because the service was being initialised through the require chain before models were initialised through the boot process
- fixed incorrect `camelCase` of resource name in API responses
refs https://github.com/TryGhost/Team/issues/1090
This updates the Members & Offers modules to provide support for using
Offers in Stripe Checkout.
Members module now needs a handle to the Offers module repository, and
as such we have had to reorder the services boot order.
refs https://github.com/TryGhost/Team/issues/1083
This adds the Offers service which will allow us to build out an Admin
API to communicate with it. As well as wiring up the redirects to the
site app.
- The helper registration code is "framework" code and very specific
- At the moment the "theme engine" is full of lots of disparate theme related stuff
- I'm trying to make the frontend framework code clearer and also expand it to make it more useful
- The helper system now also exposes 3 methods allowing you to register a directory, a helper or an alias
- I've updated the codebase to use these both for our core helpers and for "apps"
refs https://github.com/TryGhost/Team/issues/1083
The Offers service is going to need access to the StripeAPIService too,
so we need to move it out of the @tryghost/members-api module and make
it accessible to both.
refs https://linear.app/tryghost/issue/CORE-35/refactor-route-and-redirect-settings
- There's no reason for the boot to block the event by loading route settings sychronously
- The only leftover use of a sync loader might also be refactored in some way to avoid blocking the event loo - for example by caching the value on the service layer.
refs https://linear.app/tryghost/issue/CORE-35/refactor-route-and-redirect-settings
- Frontend is not meant to know about the underlying source of the "routes" configuration, so any reads/edits/validations are being moved into a backend service. This should also simplify the coupling of the backend with the frontend where the latter will get a JSON blob with all needed configuration during the boot
- Nother problem the "get" method had was hiding an underlying function it was doing - reading the file from the filesystem SYNCRONOUSLY. It might be a thing we need to do during the "web" app initialization, but there's no clear need to do this in a sync fassion during the bootup for example. Also having a more explicit name should help :)
