refs https://github.com/TryGhost/Team/issues/527
refs https://github.com/TryGhost/Ghost/issues/10790
- Frontent has to have as few as possible coupling points with the Ghost Server API. By design that point has been a "proxy.api" property that will become more and more constraint in the future based to limit the surface of frontend interaction with servers's API
- Removing `.../server/api` requires in favor of using a proxy decreases direct coupling
- This isn't really a "service" - it's a set of utilities for working with labs flags
- It's also required all over the place, and doesn't require anything that isn't shared
- Therefore, it should live in shared
- This isn't really a "service" - it's a set of utilities for working with labs flags
- It's also required all over the place, and doesn't require anything that isn't shared
- Therefore, it should live in shared
- This is a precursor to trying to split apart into:
- model events + webhooks system which makes sense
- frontend events which should be independent or removed
- maybe some concept of a settings manager that we can use in various places to bind logic 🤔
- other usages of events that should be refactored to not use events
refs: 076ad99593
- as of 076ad99593 we no longer use the error property of the active theme anywhere
- cleaning up and removing this usage reduces the code pathways and makes the init fn a bit clearer
- The router bootstrap is no longer allowed to fetch it's own settings, but rather is passed them
- This moves the call to the site routes.js file, which isn't much better but it's a start
- The goal is to always pass these in from the boot process, or from the bridge reloader
- Reduced the number of levels in our debug naming in the frontend
- Unified components like "themes" and "routing" under one name
- Should help to make debug slightly more useful again
refs https://github.com/TryGhost/Team/issues/781
refs 813d288eb2#
- The 500 error was introduced through a refed commit long time ago when (probably?) there were no other safeguards preventing from serving content through a theme with errors. Since than we have multiple safeguards when aploading/activating the theme with errors and the default handling when such error occurs is more graceful - a 400 with specific error details is shown
- We need this change to land before bumping gscan that introduces more suphisticated error detaction in theme templates. Otherwise, people upgrading to new version and having an error on an unused template or somewhere undetected previously woul end up with a bricked site showing a generic 500 - not a great experience!
refs https://github.com/TryGhost/Team/issues/845
refs 517d2abc5c
- updated router response formatting functions and `{{#get}}` helper response handling to make any `feature_image_caption` properties in the response a `SafeString` instance so triple-curlies are not needed when using the property in themes
- The main aim here is to end up with a simple and clear public API for the meta module
- Secondarily, we want to make it a bit clearer which bits don't really belong here so we can see what to do with them
- To achieve this, the main logic has been moved into get-meta (although there's still some logic here which needs moving further)
- The index.js now has a small clear public API, and the proxy, which is the only way this is consumed, is able to use the public API directly
- This function is quite different to the others, as it generates an excerpt from HTML (truncating)
- Most functions in the meta data folder just contain content negotiation logic, like if post then feature_image else cover_image type things
- This function is more like a library and shouldn't live in Ghost, it should probably be in @tryghost/helpers
- It's definitely something we'd love to rewrite to work better tooooo
- getting rid of instances of new Error as we should always use @tryghost/errors
- Whilst here, got rid of i18n but discovered the messages were missing!
- This is my fault, they disappeared when I removed external apps and clearly removed too much: 8c1a0b8d0c (diff-0f5cc40aa8906a1be1bad2002a35361bbf9e766e46b3b29be10f4f479265426a)
- Therefore, I have restored these messages in the places where they were used, except amp_content, where I have written a new message, as the message that was there was not relevant
- This is part of the quest to separate the frontend and server & get rid of all the places where there are cross-requires
- At the moment the settings cache is one big shared cache used by the frontend and server liberally
- This change doesn't really solve the fundamental problems, as we still depend on events, and requires from inside frontend
- However it allows us to control the misuse slightly better by getting rid of restricted requires and turning on that eslint ruleset
- At the moment the bootstrap.start method asks the settings service for its settings
- This couples the routing and settings services together - when maybe we want to use a different method to generate settings
- By passing the settings to the routing service at the right time, we open up possibilities for refactoring
- The main goal here is getting this settings related code out of the routing service as it really doesn't belong there
- This settings file is used purely by the API to get and set files - its not really anything to do with actual routing
- This file calls out to the bridge to do a reload, which helps decouple slightly
- More refactoring is needed to get rid of the urlService dependency
- Note this file is really similar to the redirects one, it would be good to merge them
- At the moment the bootstrap.start method asks the settings service for its settings
- This couples the routing and settings services together - when maybe we want to use a different method to generate settings
- By passing the settings to the routing service at the right time, we open up possibilities for refactoring
- broken down large function into smaller functions to reduce repeated code
- try to make this and the redirects equivalent look similar
- this code is the getter and setter for the API
- TODO: I think this can be further refactored into a settings file class
- Allows for slight decoupling of API and frontend with route settings being updated
- Activate theme now calls the same codepath to reload the frontend
- Yet another step on the path to make it possible to init/reload/run the frontend independently from the server
- Part of the effort to split Ghost down into smaller, decoupled pieces
- Moved out our internal validator tooling to a separate library
- Replaced all usage of our own tooling and validatorjs directly with @tryghost/validator
- Removed the validatorjs dependency and removed the renovate pin
- This gives us a consistant, smaller, clearer public API for validations
- It will eventually be used on Ghost Admin too
- This way we can start getting up to date with validator whilst not increasing build size
no issue
The only pieces of Ghost-Ignition used in Ghost were debug and
logging. Both of these modules have been superceded by the Framework
monorepo, and all usages of Ignition have now been removed, replaced
with @tryghost/debug and @tryghost/logging.
closes https://github.com/TryGhost/Team/issues/761
With multiple products, each product can have an active monthly/yearly price, so we no longer store the monthly/yearly price ids in global settings but instead store them in product table directly. This means we need to update our global `@price` helper to also use the updated schema and use the monthly/yearly prices from product table instead of settings data.
refs: https://github.com/TryGhost/Team/issues/759
- wired up a matchHelper feature flag & used the labsEnabledHelper tool to gate the helper
- added a first version of the match helper, which is intended to replace the has helper
- this is an experimental helper and may or may not make it to GA
- match is a simple comparison helper, right now it does a very basic equals or not equals comparison
- much more functionality is needed to reach parity with has
refs https://github.com/TryGhost/Team/issues/708
- Defaults to an empty array on `@products` so we have valid data
(product should be null if products isn't)
- This is the first step toward supporting multiple products at the
theme level
closes https://github.com/TryGhost/Ghost/issues/12986
refs 1345268089
As part of changes in 4.6, the default price ids for monthly/yearly prices are stored in new settings - `members_monthly_price_id`, `members_yearly_price_id` - which are used to determine current active prices for the site from list of all existing prices. While the last commit updated the prices to use the settings, the data for currency was still used from non-zero prices instead of the new settings value.
- Updated tests to check price currency
closes https://github.com/TryGhost/Ghost/issues/12980
closes https://github.com/TryGhost/Team/issues/730
As part of changes in 4.6, the default price ids for monthly/yearly prices are stored in new settings - `members_monthly_price_id`, `members_yearly_price_id` - which are used to determine current active prices for the site from list of all existing prices. The `@price` helper was incorrectly still relying on the old logic for active monthly/yearly price using the first active price with matching nickname, and resulted in showing incorrect price data on the theme.
- Updated tests to check price data using settings value
no-issue
Since we now allow archiving prices, we should filter them out from
being considered the monthly or yearly plan, as they are unable to be
subscribed to.
no-issue
Themes which use the `@price` data will have a 400 error if they are not
setup prices. This adds default price data so that the theme will not
error.
refs https://github.com/TryGhost/Team/issues/668
Since we no longer store price data in the settings we must use the api
to read the stripe prices for the default price, so that we can maintain
backwards compatibility for the `@price` data in themes.
- we need the basePath concept for the main i18n class so we can pull it out into a module
- we already had this in the themeI18n class, so I just had to move it up
- also I added a default of __dirname, so we don't have to declare this constantly in the tests
- Reworking the location of i18n in boot has fixed the main error
- However, many of our tests depend on i18n being loaded but don't explicitly call init
- There are many ways we could fix this in our tests, but I don't want to spend more time on this now
- final preparation for moving i18n out of Ghost core
- logging is passed in via DI
- theme i18n needs a config value, but no need to pass all of config for one parameter, a better pattern is to pass the one value needed
- preparation for moving the base class out of Ghost
- refactored so that all the logic for file loading and fallbacks live in the base class
- theme i18n now only overrides init with the properties it needs, filepath generation and error handling
- this makes it much easier to move the i18n file out, and eventually have theme i18n live elsewhere too
- also prepares for using DI for logging
- when activating a theme, we need to load the current locale
- this request used to be buried deep in the themeI18n init call
- now we surface it in the bridge and pass it down, which is closer to what we want to do with eventually initialising the frontend
with everything it needs up front (or not initialising it, if it isn't needed)
- in the related helpers we depend on the site.locale value instead of proxy -> themeI18n -> settingsCache drastically simplifying the code and removing deep requires
- site.locale is updated via middleware and can be relied upon
- the core i18n library and theme i18n library have slightly different methods of getting a candidate string
- both of them use forms of jsonpath, meaning they both require jsonpath as a dependency
- to try to get to a point of being able to rip more things out of ghost, we want to have less dependencies
- so instead of overloading the method, we pass in a stringMode as an argument
- eventually we might not need an overloaded class for themeI18n at all, which would simplify the codebase
- preparation for using DI instead of requires, so we can move this out of Ghost
- have done this for both the main i18n and theme i18n file
- refactored the constructor
refs 829e8ed010
- i18n is used everywhere but only requires shared or external packages, therefore it's a good candidate for living in shared
- this reduces invalid requires across frontend and server, and lets us use it everywhere until we come up with a better option
- Having these as destructured from the same package is hindering refactoring now
- Events should really only ever be used server-side
- i18n should be a shared module for now so it can be used everywhere until we figure out something better
- Having them seperate also allows us to lint them properly
refs: bf0823c9a2
refs: ae86254972
- continuing the work of splitting up the theme service into logical components
Themes Service
- The serverside theme service now serves just the API and boot
- It loads the theme and passes it to the theme-engine via the bridge
This achieves the bare minimum goal of removing all the cross requires between server and frontend around themes
There is still a lot more to do to achieve an ideal architecture here as laid out in ae86254972
refs: bf0823c9a2
- continuing the work of splitting up the theme service into logical components
- am about to move the theme service to core/server so it should require i18n directly
refs: bf0823c9a2
- continuing the work of splitting up the theme service into logical components
- this file is not part of the theme engine so it should use the bridge not the engine
- am about to move the theme service to core/server so this will make even more sense then
refs: bf0823c9a2
- continuing the work of splitting up the theme service into logical components
Theme activations are a trickier piece of the theme split puzzle because they are called from the API and theme service on boot in different ways.
Activations require a theme to have been validated at different levels. Validations are also tricky - do they belong to the theme engine, or the theme service?
There are then several different flows for activations:
- On Boot
- API "activate" call
- API override on upload or install via setFromZip, which is a method in the storage layer
These calls all have quite different logical flows at the moment, and need to be unified
For now, I've moved the existing "activate" function onto the bridge. This allows the theme service to be split from the frontend, and refactoring can start from there.
I hope to move this so there is less code in the actual bridge very soon, but my goal is not to require any server packages in the frontend part of this
I think ideally:
- all activation code, including validation, should probably be part of the theme engine
- the theme engine should offer 3 methods: getActive() canActivate() and activate()
- the theme service is then only responsible for loading themes in and out of storage, JSON responses for the API, and handing themes to the frontend via the bridge at the appropriate moment
- Modules in /shared are supposed to be standalone modules that can be required by the server or frontend
- As the server shouldn't require the frontend, and vice versa, shared modules should require neither
- Otherwise it just becomes a crutch for allowing cross-depenencies, and will create circular dependencies
The Bridge
- The bridge file is not meant to be a crutch sat allowing cross-dependencies, but rather a new component that manages the flow of data
- That data flows from the server/boot process TO the frontend, and should not flow in the other direction
- The management of that flow of data is necessarily hacky at the moment, but over time the architecture here should get clearer and better
- Still, for the time being it will need to handle requiring across components until that architecture matures
- Therefore, it should live in core root, not in core/shared