no issue
- if you start Ghost and you theme is invalid, you only get a warning, but no reason
- furthermore, if any error is thrown in Ghost, which is not a custom Ignition error, we take care that the error message to inherit from shows up
refs #7687
There are four main changes in this PR:
we have outsourced the base storage adapter to npm, because for storage developers it's annoying to inherit from a script within Ghost
we hacked theme storage handling into the default local storage adapter - this was reverted, instead we have added a static theme storage here
use classes instead of prototyping
optimise the storage adapter in general - everything is explained in each commit
----
* rename local-file-store to LocalFileStorage
I would like to keep the name pattern i have used for scheduling.
If a file is a class, the file name reflects the class name.
We can discuss this, if concerns are raised.
* Transform LocalFileStorage to class and inherit from new base
- inherit from npm ghost-storage-base
- rewrite to class
- no further refactoring, happens later
* Rename core/test/unit/storage/local-file-store_spec.js -> core/test/unit/storage/LocalFileStorage_spec.js
* Fix wrong require in core/test/unit/storage/LocalFileStorage_spec.js
* remove base storage and test
- see https://github.com/kirrg001/Ghost-Storage-Base
- the test has moved to this repo as well
* Use npm ghost-storage-base in storage/index.js
* remove the concept of getStorage('themes')
This concept was added when we added themes as a feature.
Back then, we have changed the local storage adapter to support images and themes.
This has added some hacks into the local storage adapters.
We want to revert this change and add a simple static theme storage.
Will adapt the api/themes layer in the next commits.
* Revert LocalFileStorage
- revert serve
- revert delete
* add storagePath as property to LocalFileStorage
- define one property which holds the storage path
- could be considered to pass from outside, but found that not helpful, as other storage adapters do not need this property
- IMPORTANT: save has no longer a targetDir option, because this was used to pass the alternative theme storage path
- IMPORTANT: exists has now an alternative targetDir, this makes sense, because
- you can either ask the storage exists('my-file') and it will look in the base storage path
- or you pass a specific path where to look exists('my-file', /path/to/dir)
* LocalFileStorage: get rid of store pattern
- getUniqueFileName(THIS)
- this doesn't make sense, instances always have access to this by default
* Add static theme storage
- inherits from the local file storage, because they both operate on the file system
- IMPORTANT: added a TODO to consider a merge of themes/loader and themes/storage
- but will be definitely not part of this PR
* Use new static theme storage in api/themes
- storage functions are simplified!
* Add https://github.com/kirrg001/Ghost-Storage-Base as dependency
- tarball for now, as i am still testing
- will release if PR review get's accepted
* Adapt tests and jscs/jshint
* 🐛 fix storage.read in favicon utility
- wrong implementation of error handling
* 🎨 optimise error messages for custom storage adapter errors
* little renaming in the storage utlity
- purpose is to have access to the custom storage instance and to the custom storage class
- see next commit why
* optimise instanceof base storage
- instanceof is always tricky in javascript
- if multiple modules exist, it can happen that instanceof is false
* fix getTargetDir
- the importer uses the `targetDir` option to ensure that images land in the correct folder
* ghost-storage-base@0.0.1 package.json dependency
no issue
🔥 Remove DIRTY HACK for API
- this is no longer needed, because themes get mounted in every case
✨ Switch to concept of 'mounted' theme
- check if active theme is mounted
- if not, mount it
- mounting is a function OF the active theme
🎨 Move theme middleware to theme module
🎨 Update theme middleware function names
- update the function names and comments to be more representative of their current functions
- this was pretty old and out of date!
🚨 Fixup tests for middleware
- ensure the objects match what we expect
- based partially on theme docs
Update TODO
📡 Add debug for the 3 theme activation methods
There are 3 different ways that a theme can be activated in Ghost:
A. On boot: we load the active theme from the file system, according to the `activeTheme` setting
B. On API "activate": when an /activate/ request is triggered for a theme, we validate & change the `activeTheme` setting
C. On API "override": if uploading a theme with the same name, we override. Using a dirty hack to make this work.
A: setting is done, should load & validate + next request does mounting
B: load is done, should validate & change setting + next request does mounting
C: load, validate & setting are all done + a hack is needed to ensure the next request does mounting
✨ Validate w/ gscan when theme activating on boot
- use the new gscan validation validate.check() method when activating on boot
✨ New concept of active theme
- add ActiveTheme class
- make it possible to set a theme to be active, and to get the active theme
- call the new themes.activate() method in all 3 cases where we activate a theme
🎨 Use new activeTheme to simplify theme code
- make use of the new concept where we can, to reduce & simplify code
- use new hasPartials() method so we don't have to do file lookups
- use path & name getters to reduce use of getContentPath etc
- remove requirement on req.app.get('activeTheme') from static-theme middleware (more on this soon)
🚨 Improve theme unit tests (TODO: fix inter-dep)
- The theme unit tests are borked! They all pass because they don't test the right things.
- This improves them, but they are still dependent on each-other
- configHbsForContext tests don't pass if the activateTheme tests aren't run first
- I will fix this in a later PR
no issue
🎨 simplify loader - use loadOneTheme for init
- use loadOneTheme for init
- move updateThemeList to the one place that it is used
- this just reduces the surface area of the loader
🎨 Move init up to index temporarily
- need to figure out what stuff goes in here as well as loading themes
- will move it again later once I've got it figured out
🎨 Reorder & cleanup theme middleware
- move the order in blog/app.js so that theme middleware isn't called for shared assets
- add comments & cleanup in the middleware itself, for clarity
🎨 Simplify the logic in themes middleware
- Separate out config dependent on settings changing and config dependent on request
- Move blogApp.set('views') - no reason why this isn't in the theme activation method as
it's actually simpler if it is there, we already know the active theme exists & can remove the if-guard
🎨 Improve error handling for missing theme
- ensure we display a warning
- don't have complex logic for handling errors
- move loading of an empty hbs object into the error-handler as this will support more cases
🐛 Fix assetHash clearing bug on theme switch
- asset hash wasn't correctly being set on theme switch
🎨 Remove themes.read & test loader instead
- Previously, we've simplified loader & improved error handling
- We are now able to completely remove theme.read as it's nothing more than a wrapper for package.read
- This also means we can change our tests from testing the theme reader to loader
refs #8093✨ Add activate theme permission
- add permission to activate themes
- update tests
- also: update tests for invites
TODO: change how the active theme setting is updated to reduce extra permissions
✨ Move theme validation to gscan
- add a new gscan validation method and use it for upload
- update activate endpoint to do validation also using gscan
- change to using SettingsModel instead of API so that we don't call validation or permissions on the settings API
- remove validation from the settings model
- remove the old validation function
- add new invalid theme message to translations & remove a bunch of theme validation related unused keys
📖 Planned changes
🚨 Tests for theme activation API endpoint
🐛 Don't allow deleting the active theme
🚫 Prevent activeTheme being set via settings API
- We want to control how this happens in future.
- We still want to store the information in settings, via the model.
- We just don't want to be able to change this info via the settings edit endpoint
🐛✨ Fix warnings for uploads & add for activations
- warnings for uploads were broken in f8b498d
- fix the response + adds tests to cover that warnings are correctly returned
- add the same response to activations + more tests
- activations now return a single theme object - the theme that was activated + any warnings
🎨 Improve how we generate theme API responses
- remove the requirement to pass in the active theme!
- move this to a specialist function, away from the list
🎨 Do not load gscan on boot
no issue
🎨 Switch themes API to use config.availableThemes
- this gets rid of the only places where settings.availableThemes are used
🔥 Get rid of settings.availableThemes
- this is no longer used anywhere
- also get rid of every related call to updateSettingsCache
🔥 Replace config.availableThemes with theme cache
- Creates a tailor-made in-memory cache for themes inside the theme module
- Add methods for getting & setting items on the cache
- Move all references to config.availableThemes to use the new cache
- This can be abstracted later to support other kinds of caches?
🎨 Start improving theme lib's API
Still TODO: simplifying/clarifying:
- what is the structure of the internal list
- what is the difference between a package list, and a theme list?
- what is the difference between reading a theme and loading it?
- how do we update the theme list (add/remove)
- how do we refresh the theme list? (hot reload?!)
- how do we get from an internal list, to one that is sent as part of the API?
- how are we going to handle theme storage: read/write, such that the path is configurable
🎨 Use themeList consistently
🎨 Update list after storage
no issue
🔥 remove unused loadThemes API method
🚨 Add tests for themes.readOne
🔥 Don't update settings cache for imports
- this isn't needed as of #8057
- settings.edit fires an event, that will result in the update happening automatically
🎨 Move validation to themes
- slowly collecting all theme-related code together
🔥 Reduce DEBUG output
- all this info is a bit tooooo much!
no issue
* ✨ Add new server start & stop events
* 🔥 Get rid of unused availableApps concept
- when we need an API endpoint for a list of apps, we'll build one 😝
* ✨ Move theme loading into a module
- move loading from API method to a module method and use as needed
- wire up read one vs read all as per LTS
- read one (the active theme) on boot, and read the rest after
- fudge validation - this isn't all that helpful
* Settings API tests need to preload themes
- this used to automatically happen as part of loading settings
- now we need to trigger this to happen specifically for this test
closes#4172, closes#6948, refs #7491, refs #7488, refs #7542, refs #7484
* 🎨 Co-locate all admin-related code in /admin
- move all the admin related code from controllers, routes and helpers into a single location
- add error handling middleware explicitly to adminApp
- re-order blogApp middleware to ensure the shared middleware is mounted after the adminApp
- TODO: rethink the structure of /admin, this should probably be an internal app
* 💄 Group global middleware together
- There are only a few pieces of middleware which are "global"
- These are needed for the admin, blog and api
- Everything else is only needed in one or two places
* ✨ Introduce a separate blogApp
- create a brand-new blogApp
- mount all blog/theme only middleware etc onto blogApp
- mount error handling on blogApp only
* 🎨 Separate error handling for HTML & API JSON
- split JSON and HTML error handling into separate functions
- re-introduce a way to not output the stack for certain errors
- add more tests around errors & an assertion framework for checking JSON Errors
- TODO: better 404 handling for static assets
Rationale:
The API is very different to the blog/admin panel:
- It is intended to only ever serve JSON, never HTML responses
- It is intended to always serve JSON
Meanwhile the blog and admin panel have no need for JSON errors,
when an error happens on those pages, we should serve HTML pages
which are nicely formatted with the error & using the correct template
* 🐛 Fix checkSSL to work for subapps
- in order to make this work on a sub app we need to use the pattern `req.originalUrl || req.url`
* 🔥 Get rid of decide-is-admin (part 1/2)
- delete decide-is-admin & tests
- add two small functions to apiApp and adminApp to set res.isAdmin
- mount checkSSL on all the apps
- TODO: deduplicate the calls to checkSSL by making blogApp a subApp :D
- PART 2/2: finish cleaning this up by removing it from where it's not needed and giving it a more specific name
Rationale:
Now that we have both an adminApp and an apiApp,
we can temporarily replace this weird path-matching middleware
with middleware that sets res.isAdmin for api & admin
* 🎨 Wire up prettyURLs on all Apps
- prettyURLs is needed for all requests
- it cannot be global because it has to live after asset middleware, and before routing
- this does not result in duplicate redirects, but does result in duplicate checks
- TODO: resolve extra middleware in stack by making blogApp a sub app
* ⏱ Add debug to API setup
* 🎨 Rename blogApp -> parentApp in middleware
* 🎨 Co-locate all blog-related code in /blog
- Move all of the blogApp code from middleware/index.js to blog/app.js
- Move routes/frontend.js to blog/routes.js
- Remove the routes/index.js and routes folder, this is empty now!
- @TODO is blog the best name for this? 🤔
- @TODO sort out the big hunk of asset-related mess
- @TODO also separate out the concept of theme from blog
* 🎉 Replace middleware index with server/app.js
- The final piece of the puzzle! 🎉🎈🎂
- We no longer have our horrendous middleware/index.js
- Instead, we have a set of app.js files, which all use a familiar pattern
* 💄 Error handling fixups
