closes: https://github.com/TryGhost/Ghost/issues/14981
- Taxonomy-specific sitemaps were invalid xml when there was no data
- These invalid empty sitemaps were referenced in the index sitemap causing SEO tools to report errors
closes: https://github.com/TryGhost/Ghost/issues/15470
- When multiple browser tabs are open, each manipulate a different copy of ember data model, changes to the model in one tab are not reflected in the model of the other tab.
- When updating some settings, all current settings were sent to the API.
- As a result, when updating two different categories of settings (navigation/code inspection) in different tabs, the second update was overriding the first one.
- From a user perspective, this is not a natural behaviour. Only settings visible on-screen when clicking save should be modified.
Co-authored-by: Kevin Ansfield <kevin@lookingsideways.co.uk>
refs https://github.com/TryGhost/Toolbox/issues/441
- this is only v1 of the test I would like but it validates the keys on
a column definition are part of an allowlist
- this has already uncovered a bug with `maxLength` (vs `maxlength`)
fixes https://github.com/TryGhost/Team/issues/2054
This change adds the sentiment and positive_feedback counts to the posts models. This change isn't really ideal because there are some problems here:
- sentiment isn't really a count
- we don't need to include the sentiment and positive_feedback as a default for posts (but the same is true for attribution)
It would make sense to move this to separate endpoints that only fetch the analytics for a given post when the analytics page is opened. But for our initial skateboard version of audience feedback this should be a good start to already see the data.
refs https://github.com/TryGhost/Team/issues/2047
- We anticipate upcoming changes in the PUT /members/:id/subscriptions/:subscription_id endpoint , so covered it with a snapshot test to track the differences more precisely.
- Note, the test case contains a more explicit outgoing HTTP request mocking.
refs https://github.com/TryGhost/Team/issues/1967
This tests the full flow of publishing a newsletter, and then checking
that clicked links will increase the click count, generate events for
the member which clicked the link as well as the redirects contain the
correct query params.
refs https://github.com/TryGhost/Toolbox/issues/320
- Added more complex mobiledoc structure in the post.published test to check for correct transformation of special purpose `__GHOST_URL__`. The snapshot has a correct URL transformation, which gives confidence it works properly
refs https://github.com/TryGhost/Toolbox/issues/320
- There noe "roles" attached to the post's author when the 'post.added' event is fired. Webhooks function based of the model events and differ slightly with it's output comparing to the API response. For example, in case of Posts API, there'a an additional 'findOne' call (ref.: https://github.com/TryGhost/Ghost/blob/main/ghost/core/core/server/models/post.js#L1224-L1227) before returning the post to the endpoint handler and then passing that to the output serializer.
- If we want to have 1:1 copy of webhooks outputs and API outputs, we should rethink how we rely on model event data which is never the same as API controller level data.
refs a499f866f3
refs d817e5830d
- The user-agent used in outgoing Ghost requests (webhooks mostly) is dependent on the Ghost version - snapshots break if the matcher is not dynamic.
- There will be a few more webhooks tests coming soon, so makes sense to have this matcher moved to a common "framework matchers"
fixes https://github.com/TryGhost/Ghost/issues/14508
This change requires the frontend to send an explicit `emailType` when sending a magic link. We default to `subscribe` (`signin` for invite only sites) for now to remain compatible with the existing behaviour.
**Problem:**
When a member tries to login and that member doesn't exist, we created a new member in the past.
- This caused the creation of duplicate accounts when members were guessing the email address they used.
- This caused the creation of new accounts when using an old impersonation token, login link or email change link that was sent before member deletion.
**Fixed:**
- Trying to login with an email address that doesn't exist will throw an error now.
- Added new and separate rate limiting to login (to prevent user enumeration). This rate limiting has a higher default limit of 8. I think it needs a higher default limit (because it is rate limited on every call instead of per email address. And it should be configurable independent from administrator rate limiting. It also needs a lower lifetime value because it is never reset.
- Updated error responses in the `sendMagicLink` endpoint to use the default error encoding middleware.
- The type (`signin`, `signup`, `updateEmail` or `subscribe`) is now stored in the magic link. This is used to prevent signups with a sign in token.
**Notes:**
- Between tests, we truncate the database, but this is not enough for the rate limits to be truly reset. I had to add a method to the spam prevention service to reset all the instances between tests. Not resetting them caused random failures because every login in every test was hitting those spam prevention middlewares and somehow left a trace of that in those instances (even when the brute table is reset). Maybe those instances were doing some in memory caching.
fixes https://github.com/TryGhost/Ghost/issues/14508
This change requires the frontend to send an explicit `emailType` when sending a magic link. We default to `subscribe` (`signin` for invite only sites) for now to remain compatible with the existing behaviour.
**Problem:**
When a member tries to login and that member doesn't exist, we created a new member in the past.
- This caused the creation of duplicate accounts when members were guessing the email address they used.
- This caused the creation of new accounts when using an old impersonation token, login link or email change link that was sent before member deletion.
**Fixed:**
- Trying to login with an email address that doesn't exist will throw an error now.
- Added new and separate rate limiting to login (to prevent user enumeration). This rate limiting has a higher default limit of 8. I think it needs a higher default limit (because it is rate limited on every call instead of per email address. And it should be configurable independent from administrator rate limiting. It also needs a lower lifetime value because it is never reset.
- Updated error responses in the `sendMagicLink` endpoint to use the default error encoding middleware.
- The type (`signin`, `signup`, `updateEmail` or `subscribe`) is now stored in the magic link. This is used to prevent signups with a sign in token.
**Notes:**
- Between tests, we truncate the database, but this is not enough for the rate limits to be truly reset. I had to add a method to the spam prevention service to reset all the instances between tests. Not resetting them caused random failures because every login in every test was hitting those spam prevention middlewares and somehow left a trace of that in those instances (even when the brute table is reset). Maybe those instances were doing some in memory caching.
no issue
- All content-length snapshots should be using the same matcher for consistency - anyContentLength. It's more explicit about what the matcher is all about and might be useful to have content-length matchers in one place if it ever changes (the header value should be a damn digit after all, not a string!) (ref. https://www.rfc-editor.org/rfc/rfc7230#section-3.3.2)
refs https://github.com/TryGhost/Team/issues/2024
Without validation it was possible to send a string of comma separated
email addresses to the endpoint, and an email would be sent to each
address, bypassing any rate limiting.
This bug does not allow for an authentication bypass exploit. It is purely a
spam email concern.
Credit: Sandip Maity <maitysandip925@gmail.com>
refs https://github.com/TryGhost/Toolbox/issues/320
- The URL matcher is very likely to be reused in the future, so having it abstracted away gives two benefits:
1. Central place to document hacky behavior and easier future cleanup
2. The implementer of the e2e test does not have to see the "hacky note" and just concentrate on the implementation of the test
refs https://github.com/TryGhost/Toolbox/issues/320
- Header snapshot matching was missing from webhook e2e tests. With a bumped version of webhook-mock-receiver it's now possible to record and match webhook request headers.
closesTryGhost/Team#2007
- uses request context to add referrer source and medium for a new member
- uses integration name as referrer medium if exists
fixes https://github.com/TryGhost/Team/issues/2008
- New column that stores email click tracking at the time it was created
- Improved frontend side checks for when to show analytics
refs https://github.com/TryGhost/Toolbox/issues/425
refs https://github.com/TryGhost/Toolbox/issues/280
- The versioned API responses vary based on requested version (passed in request's 'accept-version' header). shared caches that sit between Ghost's origin server and the browser would be putting responses with same Vary into the same caching bucket, which is incorrect.
- This change makes response's Vary more granular and tells caching mechanisms to take 'Accept-Version' request header into account when caching.
- Informative read on the topic - https://www.fastly.com/blog/getting-most-out-vary-fastly
- renames `refSource`, `refMedium` and `refUrl` to `referrerSource`, `referrerMedium` and `referrerUrl` respectively for consistent naming across files and usages
- bumps member attribution script from alpha feature to now load for all sites. The script captures recent url history in localstorage to capture correct attribution for members.
- script is only loaded on the site if members is enabled
refs https://github.com/TryGhost/Team/issues/1967
- Test is good to test if the whole flow works as expected, and works together
- We can test independent parts in separate tests that have better coverage of more edge cases
- Adds a basic helper to get an agent for the frontend (spent too much time on a better solution so I decided to keep the existing supertest agent)
refs https://github.com/TryGhost/Ghost/pull/15471#discussion_r979902374
- the accent color value used by default content cta was copying the global site property which is redundant, and can be directly used
- originally, the accentColor property was extended to allow a fallback value for content ctas, but was later removed as we added default value to global site property directly
- the accentColor property is now deprecated and will be removed in next version, as existing themes might be relying on it for custom cta helpers
closes https://github.com/TryGhost/Team/issues/1898
- the default content cta always used the terminology as `post` when showing message that users don't have access to some content
- this caused confusion when users were looking at a page and message showed "This post is for subscribers only"
- updates the message to correctly reflect `page` vs `post` on the default cta
refs https://github.com/TryGhost/Toolbox/issues/410
- Private cache control was preventing browser or shared caches from storing Content APIs response. The type of data served through the Content API is very much of a "public" nature, so should be cacheable.
- Right now the 'max-age' value of 'cache-control' header is hardcoded to '0', without 'must-revalidate' value, to allow browsers to cache content slightly more aggressively. In the future the 'max-age' value will most-likely become configurable to allow even more aggressive HTTP caching.
refs https://github.com/TryGhost/Toolbox/issues/410
- The 'private' value in 'Cache-Control' response header for all errors made it impossible for shared caches (e.g.: Fastly, Cloudflare) to cache 404 responses efficiently.
- The change substitutes 'max-age=0' which should not effect the browser cache behavior but would allow shared caches to process such requests efficiently.
- A more loose caching logic only applies to 404 responses from GET requests that are not user-specific (non-authenticated, non-cookie containing requests)
closes https://github.com/TryGhost/Toolbox/issues/372
- The admin assets are served with a unique hash depending on the build with a year-long "max-age" value in the response cache-control header. The client browsers still do send 'If-None-Match' requests when there is a hard-refresh on the client side. There's no need for 'If-None-Match' requests though!
- With 'immutable' value in the cache-control header, the browser caches are treating responses as "hard-fresh" without sending redundant requests.
- For more about 'immutable' value read https://developer.mozilla.org/en-US/docs/Web/HTTP/Headers/Cache-Control#immutable
closes https://github.com/TryGhost/Team/issues/1942
- Added data fixtures for referrers
- Added new endpoint to fetch referrer stats for a given post: `/stats/referrers/posts/:id`
- Added new ReferrersStatsService, responsible for calculating referrer stats
refs TryGhost/Team#1931
- referrer source, medium and url will be stored in the events table along with rest of attribution data
- stores referrer information on two tables
- `members_created_events` for signups
- `members_subscription_created_events` for paid conversions
closes https://github.com/TryGhost/Team/issues/1933
- Added click_events to activity feed
- Added support for parsing click_events in the frontend
- Moved url parsing (transform ready) to model layer of LinkRedirect
- Moved `getEventTimeline` method to the top of the event repository
- Added description field to parsed events in the frontend (because we need a second line)
- Fixed: member email not returned in comment_event
refs d5f03ec0b1
- underlying error message varies across node versions so the content-length can't be fixed
- applied any-content-length matcher to the right test this time
no issue
- bumped `@tryghost/url-utils` to get access to the new lexical transform utilities
- updated the Post model's `parse()` and `formatOnWrite()` methods to transform the `lexical` field contents when reading/writing to ensure any links in content point at the correct place with `site.url` config changes
refs TryGhost/Team#1907
- calculates final attribution source and medium using captured referrer information in history
- adds new referrer-translator that goes through available history and based to determine most valid referrer info
- includes referrer url, source and medium in the attribution data for storage
no issue
- The explore endpoint needs to expose the total amount of published posts
- To be more consistent, this PR creates a PostStats class which is exposed as `stats` method within the PostService; just like it's done with the MemberService
- Moved existing method to return the date of the most recently published post into the stats service
- Updated the explore service test to reflect the new return property
no issue
- added `PostRevsion` model
- duplicated `mobiledoc_revision` creation routine in Post model's onSaving hook to create `post_revision` when model's `lexical` field has changed
- updated `mobiledoc_revision` creation to skip when `lexical` field is populated
no issue
- initially this will perform the same function as `mobiledoc_revisions` but storing `lexical` instead of `mobiledoc`
- naming is intentionally generic ready for later expansions
closes https://github.com/TryGhost/Team/issues/1908
### Problem:
- We need tracking on the paywall links in each email. (we cannot ignore them because those buttons are probably gonna have a higher paid conversion attribution than others).
- Currently we only add the paywall HTML to an email when processing each batch. So if we batch an email to 1.000 recipients per 100, we'll generate the paywall HTML 10 times.
- We cannot replace links in `renderEmailForSegment` because that methods will get called multiple times. We don't want to have multiple redirect instances created for the same link in the same email.
### Solution:
- Move the generation of the paywall to the `serialize` method of the post email serializer.
- Surround the generated paywall with HTML-comments so we can remove it if required in `renderEmailForSegment` depending on the member segment we are sending the email to.
---
### Before:
**Serialize output:**
```html
<html>
<body>
<h1>Generated email header</h1>
<p>Generated text</p>
<div>
<!-- POST CONTENT START -->
<h1>Post title</h1>
<p>Content visible for all members</p>
<!--members-only-->
<p>Content visible for paid members only</p>
<!-- POST CONTENT END -->
</div>
</body>
</html>
```
To be modified later by `renderEmailForSegment`:
**Paid members (nothing changed):**
```html
<html>
<body>
<h1>Generated email header</h1>
<p>Generated text</p>
<div>
<!-- POST CONTENT START -->
<h1>Post title</h1>
<p>Content visible for all members</p>
<!--members-only-->
<p>Content visible for paid members only</p>
<!-- POST CONTENT END -->
</div>
</body>
</html>
```
**Free members (paywall _added_):**
```html
<html>
<body>
<h1>Generated email header</h1>
<p>Generated text</p>
<div>
<!-- POST CONTENT START -->
<h1>Post title</h1>
<p>Content visible for all members</p>
<h2>Generated paywall here</h2>
<a href="https://subscribe.com">Subscribe to read the full post</a>
<!-- POST CONTENT END -->
</div>
</body>
</html>
```
### After this change:
**Serialize output:**
```html
<html>
<body>
<h1>Generated email header</h1>
<p>Generated text</p>
<div>
<!-- POST CONTENT START -->
<h1>Post title</h1>
<p>Content visible for all members</p>
<!--members-only-->
<p>Content visible for paid members only</p>
<!-- PAYWALL -->
<h2>Generated paywall here</h2>
<a href="https://subscribe.com/?tracked">Subscribe to read the full post</a>
<!-- POST CONTENT END -->
</div>
</body>
</html>
```
To be modified later by `renderEmailForSegment`:
**Paid members (paywall removed):**
```html
<html>
<body>
<h1>Generated email header</h1>
<p>Generated text</p>
<div>
<!-- POST CONTENT START -->
<h1>Post title</h1>
<p>Content visible for all members</p>
<!--members-only-->
<p>Content visible for paid members only</p>
<!-- POST CONTENT END -->
</div>
</body>
</html>
```
**Free members (members-only content removed):**
```html
<html>
<body>
<h1>Generated email header</h1>
<p>Generated text</p>
<div>
<!-- POST CONTENT START -->
<h1>Post title</h1>
<p>Content visible for all members</p>
<!-- PAYWALL -->
<h2>Generated paywall here</h2>
<a href="https://subscribe.com/?tracked">Subscribe to read the full post</a>
<!-- POST CONTENT END -->
</div>
</body>
</html>
```
no issue
- added `@tryghost/kg-lexical-html-renderer` dependency
- added `lexical` lib following the same pattern as our `mobiledoc` lib
- updated the Post model's `onSaving` hook to generate the `html` value from `lexical` when present
fixes https://github.com/TryGhost/Team/issues/1909
- The feature image caption is already escaped on the frontend
- Doing it again in the backend breaks the possibility to add links to the caption
- I checked and the `feature_image_alt` is not escaped in the frontend.
fixes https://github.com/TryGhost/Team/issues/1900
refs https://github.com/TryGhost/Team/issues/1901
- Defaults to the same value as the current email_track_opens setting for existing installations, otherwise defaults to true
- Had to use a custom migration because the `addSetting` helper doesn't support using an existing setting as current value
- Added a minimal UI to change the setting, but this still needs some design magic 🪄✨
- Link replacement is disabled if `email_track_clicks` is disabled. In the future we might consider to still do parial additions, such as source attribution and maybe redirects (to discuss).
refs https://github.com/TryGhost/Team/issues/1899
- Added `addEmailAttributionToUrl` method to MemberAttributionService. This adds both the source attribution (`rel=newsletter`) and member attribution (`?attribution_id=123&attribution_type=post`) to a URL.
- The URLHistory can now contain a new sort of items: `{type: 'post', id: 'post-id', time: 123}`.
- Updated frontend script to read `?attribution_id=123&attribution_type=post` from the URL and add it to the URLHistory + clear it from the URL.
- Wired up some external dependencies to LinkReplacementService and added some dummy code.
- Increased test coverage of attribution service
- Moved all logic that removes the subdirectory from a URL to the UrlTranslator instead of the AttributionBuilder
- The UrlTranslator now parses a URLHistoryItem to an object that can be used to build an Attribution instance
- Excluded sites with different domain from member id and attribution tracking
We're going to be adding more redirection logic into Ghost and it's
going to get confusing if we have names this generic. This makes it
clear which feature this service is related to.
Ideally in the future we can combine all of these into one redirects
service, but for now we will be running a specific service per feature
no issue
- fixed API returning "Invalid mobiledoc structure" errors when `mobiledoc:null` is sent in the payload alongside `lexical: '{...}'`
- updated Admin's `posts` and `pages` adapters to always add `?formats=mobiledoc,lexical` because the API doesn't return `lexical` by default
- added `lexical` attribute to Admin's Post model
- updated `lexical-editor` controller and related components to work with `lexical` always being a JSON string rather than a parsed object
- updated `<KoenigLexicalEditor>` to pass through the lexical state string as initial state and wired up the `onChange` prop
no issue
- bumped `@tryghost/admin-api-schema` to allow passthrough of the `lexical` property on post and page API endpoints
- prevented saving of blank document in the `mobiledoc` field if `lexical` is provided
- prevented API input containing both `mobiledoc` and `lexical` fields to avoid issues when both are present:
- not possible to know which content is latest/has precedence
- not possible to know which editor should be displayed in Admin
refs 9471384020
- previously added tests (any subsequent matcher updates) for browse endpoint were not using matchers that sufficiently covered the dynamic portions of the body
no issue
- left `mobiledoc` as the only default format added in the post/page input serializers for now to minimize API/test churn during these early stages of lexical development
- tested that the `lexical` field is not returned by default but can be requested via `?formats=lexical`
no issue
- similar to the `mobiledoc` field, the Content API should not return the source `lexical` field if requested via `?formats=`
- renamed `removeMobiledocFormat()` to `removeSourceFormats()` to better match it's behaviour
closes https://github.com/TryGhost/Team/issues/1884
- adds `post.lexical` ready for use by the lexical-powered editor re-write
- fulfils the same purpose as `posts.mobiledoc` so uses the same field properties
- added `lexical` to allowed formats in Post model so it won't be included by default in API responses meaning tests/snapshots don't need updating at present
closes https://github.com/TryGhost/Team/issues/1864
refs https://github.com/TryGhost/Team/issues/1881
- triggers free member email alert via event dispatch from member create method
- passes subscription/stripe data to member creation for paid members so free member alert can be ignored for them
- moves subscription created event being called from webhook controller to `linkSubscription`, allows creating subscription events for all new subscriptions instead of ones just via webhooks
refs https://github.com/TryGhost/Team/issues/1865
- refactors staff service to listen to member and subscription events
- triggers email alerts based on events instead of directly calling the service
- removes staff service dependency for members api
no issue
- Bumped into these tests when doing cleanup in the notifications service. Having full snapshot of requests is useful to have as a sanity check, so migrated this test suite quickly.
closes https://github.com/TryGhost/Team/issues/1772
- The user facing side of comments recently replaced `bio` with `expertise`.
- To remain consistent we replaced all the references of `bio` with `expertise` throughout the codebase.
- This includes a database column name changing migration, within the `members` table.
- Bumped up the comments-ui version to a new minor (0.10.x) as its a breaking change.
fixes https://github.com/TryGhost/Toolbox/issues/356
- this feature allows site Administrators to view a history log of staff
actions on their site so they can audit when and by whom that something happened
- this commit promotes the History log to GA
- this prevents the referrer/referer header being sent for requests that go to external domains
- this in turn prevents preview URLs from appearing in the analytics of sites that are linked to and clicked on from previews
- otherwise, preview URLs can be leaked to the owners of the linked and clicked sites
refs https://github.com/TryGhost/Team/issues/1795
- Snapshots help us detect unexpected changes in the `<head>` of all sites (e.g., newly introduced script tags)
- Added ghost_head tests for comment count helper
refs https://github.com/TryGhost/Team/issues/1871
This commit adds a test to the serialize method of `post-emaiserializer`. It checks whether the generated email HTML is valid and standard HTML5 and that all properties are escaped.
To do this validation, I depend on the new `html-validate` dev dependency. Just parsing the HTML with a HTML parser is not enough to guarantee that the HTML is okay.
Apart from that this fixes:
- Removed the sanitizeHTML method and replaced it with normal HTML escaping. We don't want to allow any HTML in the escaped fields. Whereas `sanitizeHTML` still allows valid HTML, but we don't want that and want the same behaviour as on the site. E.g., a post with a title `All your need to know about the <br /> tag` should actually render the same title and non-html content, being `All your need to know about the <br /> tag`
- The file, nft and audio card didn't (always) escape the injected HTML fields (new version @tryghost/kg-default-cards)
- `@tryghost/string` is bumped because it contains the new escapeHtml method
refs https://github.com/TryGhost/Ghost/pull/15375
- we currently pass all properties for the `tags` property of a
`page`/`post` body down further into Ghost, which is causing issues
because it's handling properties it doesn't expect
- this is showing up because it's triggering save history events for
tags when a post is edited
- this commit introduces a clean util which has an allowlist of
properties allows on tag relations
- this list was taken from the schema: 128f8fb006/packages/admin-api-schema/lib/schemas/posts.json (L214-L227)
refs https://github.com/TryGhost/Team/issues/1879
OpenSea updated their URL format for NFTs after adding support for Solana
which broke our regex, this updates to support the new format.
fixes https://github.com/TryGhost/Team/issues/1861
- Moved like and unlike endpoint handling to comments service and controller
- Moved small part of report logic to comments controller
- Added proper 401 authentication error when not authenticated as member
refs https://github.com/TryGhost/Toolbox/issues/384
- Existing adapter config was based on the notion there can only be one configuration per one adapter class. With adapter cache now allowing instantiating multiple adapter instances with the same base class it opened up a possibility to have shared configuration for a base class and then extend/override it in "feature" configurations (see tests in this commit for specific examples)
refs https://github.com/TryGhost/Team/issues/1875
- due to an misbehavior in our model layer, when `tiers` is set on a Post, it'll
trigger a save of the Tier, and this produces an extra event in the
`actions` table
- mapping the Tier(s) to just the ID prevents bookshelf-relations from
editing the Tier and thus prevents the extra event
- also fixed tests which were implicitly assuming supplying a slug to a
post would create the product
fixes https://github.com/TryGhost/Team/issues/1855
fixes https://github.com/TryGhost/Team/issues/1866
This commit moves all duplicate methods to get the support email address to a single location. Also methods to get the default email domain are moved.
For the location, I initially wanted to put it at the settings service. But that service doesn't feel like the right place. Instead I created a new settings helpers service. This service takes the settingsCache, urlUtils and config and calculates some special 'calculated' settings based on those:
- Support email methods
- Stripe (active) keys / stripe connected (also removed some duplicate code that calculated the keys in a couple of places)
- All the calculated settings are moved to the settings helpers
I'm not 100% confident in whether this is the right place to put the helpers. Suggestions are welcome.
fixes https://github.com/TryGhost/Team/issues/1870
Disables email sanitization that was enabled earlier because this bug is more important and urgent.
The recently introduced email sanitzation removes HTML comments from the post html.
- This breaks the email paid preview, because it depends on the `<!--members-only-->` comment.
- Breaks the Outlook comments `<!--[if !mso !vml]-->`
This commit reverts this change.
closes: https://github.com/TryGhost/Toolbox/issues/342
refs: 032a26f9f3
refs: 588c9d04e8
- Now that the old `users:no-owner` (now named 'users') is working correctly :)
- Was able to add loginAs[Role] methods for each staff role, so that it's possible to execute tests as that user and check permissions
- Refactored the email preview tests to use the new e2e framework and these methods, as an example
- This fixture is the main user fixture you'd want to use when testing staff roles
- At the moment it has a weird name that makes it less likely people will use it
- A tiny step in trying to make our fixture system make a tiny bit more sense
- This fixture would only work if the roles were inserted by the fixture system
- In most cases, this fixture was adding users without their associated roles
- Now we assume the roles exist already, and that we need to map users to each role
- This will allow us to more easily test user roles in e2e tests
refs https://github.com/TryGhost/Team/issues/1771
We don't have access to `req.brute.reset` due to the way the flow
works, we have one endpoint which sends an email with a magic link,
and another route which handles the login. We don't want to apply
brute force protection to both because our rate limiting is designed
for API requests not web page visits (which is how login is handled).
Because of this we require access to the underlying ExpressBrute
instance exposed by the spam-protection module, so that we can
perform the reset.
refs https://github.com/TryGhost/Team/issues/1074
Rather than relying on the global block to stop malicious actors from
enumerating email addresses to determine who is and isn't a user, we
want our user login brute force protection to be on an IP basis,
rather than tied to the username.
refs https://github.com/TryGhost/Toolbox/issues/389
- The e2e test suite log was full of ERR_NOCK_NO_MATCH warnings when the logging level was set to "warn". The cause of this warning was legit duplicated webhook trigger processing on test environment. Gah!
- The source of duplicate webhook processing was duplication of event handlers. Event handlers were registered multiple times for same event because of the singleton nature of the "common/events" module - it remains the same instance and is not cleaned up between reboots. The deeper issue of events module initialization should be solved separately, this slightly hacky approach fixes the problem now and highlights it to be tackled in the future.
refs https://github.com/TryGhost/Team/issues/1074
Rather than relying on the global block to stop malicious actors from
enumerating email addresses to determine who is and isn't a user, we
want our user login brute force protection to be on an IP basis,
rather than tied to the username.
refs https://github.com/TryGhost/Toolbox/issues/389
- if we enable warning logs in E2E tests, we get a bunch of error
messages saying `ERROR Unhandled rejection: aborted` coming from the
SQLite DB reset code
- specifically, it's coming from the line that resets the DB by copying
the file
- this line was initially added because we would see random SQLite
"malformed database" errors
- I have a feeling that was due to something else, but I can't be sure
- I'm also not sure how else we should shut the DB connection, as this
is the recommended way but it throws an unhandled rejection
- this commit is a bit of a gamble because I'm not actually sure what
was causing the problem, but it gets rid of the errors locally and
doesn't regress on the random failures
fixes https://github.com/TryGhost/Team/issues/1860
**Problem:**
Members were not able to comment on a post that was only visible for members with a specific tier.
**Causes:**
Content gating was done on models with missing relations.
- The products relation was not loaded on the member when doing content gating
- The tiers relation was not loaded on the post when doing content gating
**Tests:**
- Added for tier-only posts
- Added for paid-only commenting
fixes https://github.com/TryGhost/Team/issues/1859
**Problem:**
When for some reason a member has an active subscription (or legacy comped subscription) for product A, and a comped subscription for product B. You cannot remove comped subscription B.
**Fixed by:**
Updating the API to allow more flexible product changes on members.
- Allow the removal of (comped) products on a member, as long as that product doesn't have a related subscription
- (still) allow the addition of comped products to a member, as long as that member doesn't have other active subscriptions. This matches the existing behaviour, but now this is only checked for added products.
- Includes tests for these edge cases
fixes https://github.com/TryGhost/Team/issues/1860
**Problem:**
Members were not able to comment on a post that was only visible for members with a specific tier.
**Causes:**
Content gating was done on models with missing relations.
- The products relation was not loaded on the member when doing content gating
- The tiers relation was not loaded on the post when doing content gating
**Tests:**
- Added for tier-only posts
- Added for paid-only commenting
fixes https://github.com/TryGhost/Team/issues/1859
**Problem:**
When for some reason a member has an active subscription (or legacy comped subscription) for product A, and a comped subscription for product B. You cannot remove comped subscription B.
**Fixed by:**
Updating the API to allow more flexible product changes on members.
- Allow the removal of (comped) products on a member, as long as that product doesn't have a related subscription
- (still) allow the addition of comped products to a member, as long as that member doesn't have other active subscriptions. This matches the existing behaviour, but now this is only checked for added products.
- Includes tests for these edge cases
fixes https://github.com/TryGhost/Team/issues/1679
These endpoints are safe to be removed, as they are only used by the admin app and usage has been removed over there. It is very unlikely that this endpoint has been used in a third party integration (in which case they will get a notification email).
no issue
- Return was missing for `res.end` if an invalid subscription_id was passed
- Added explicit `text/plain` `Content-Type` headers to error messages to avoid MIME sniffing
Signed-off-by: Elijah Conners <business@elijahpepe.com>
Co-authored-by: Simon Backx <simon@ghost.org>
In case there is an issue with the filtering of items in our client
side attribution script, we also check for and remove out of date
items here. This ensures that we do not erroneously attribute signups
or conversions to webpages from more than 24h ago.
refs TryGhost/Team#1826
- adds a method on user model which fetches all eligible users for a type of email alert
- restricts users to active `Owner` and `Administrators` with setting turned on
closes: https://github.com/TryGhost/Team/issues/1732
- adds a theme helper which outputs a working search button with a standard icon
- the icon adopts whatever the current color is from css, and has a set of default styles
- styles can be overridden with !important or the data attribute
- alternatively, any element in a theme may be turned into a search button by adding data-ghost-search
- this is meant to be a simple tool for non-theme-developers to easily add a search icon to their themes in a way that doesn't require css or html knowledge
refs: 203c8036fa
refs: 1fadbacdec
refs: 22fd7f289c
- There is something seriously weird about how content-length changes...
- It's different on CI to local sometimes...
- This particular test should not change IMO
refs https://github.com/TryGhost/Team/issues/1833
refs https://github.com/TryGhost/Team/issues/1834
We've added the attribution property to subscription and signup events when the
flag is enabled. The attributions resource is fetched by creating multiple relations
on the model, rather than polymorphic as we ran into issues with that as they can't
be nullable/optional.
The parse-member-event structure has been updated to make it easier to work with,
specifically `getObject` is only used when the event is clickable, and there is now a
join property which makes it easier to join the action and the object.
refs https://github.com/TryGhost/Team/issues/1833
refs https://github.com/TryGhost/Team/issues/1834
We've added the attribution property to subscription and signup events when the
flag is enabled. The attributions resource is fetched by creating multiple relations
on the model, rather than polymorphic as we ran into issues with that as they can't
be nullable/optional.
The parse-member-event structure has been updated to make it easier to work with,
specifically `getObject` is only used when the event is clickable, and there is now a
join property which makes it easier to join the action and the object.
refs: https://github.com/TryGhost/Ghost/issues/14882
- I found a common pattern where catch predicates were being used to catch non-existent models in destroy methods, and sometimes elsewhere in the API endpoints
- The use of predicates is deprecated, and we're working to remove them from everywhere, so that we can remove bluebird
- In order to still handle these errors correctly, we needed a small change to mw-error-handler so that it can detect EmptyResponse errors from bookshelf, as well as 404s
Note: there is a small change as a result of this - the context on these errors now says "Resource not found" instead of "{ModelName} not found".
- I think this is acceptable for now, as we will be reviewing these errors in more depth later. It's quite easy to make changes, we just have to decide what with proper design input
refs https://github.com/TryGhost/Team/issues/1825
- adds 3 new columns to users table for storing email alert preferences for member signups/cancellation
- adds column for new member signup alert
- adds column for paid subscription started alert
- adds column for paid subscription canceled alert
- Updated default fixtures and tests for new columns
refs https://github.com/TryGhost/Toolbox/issues/356
- we have a very crude version of this before but it just wasn't
maintainable
- one of the first things I did here was to add `include=resource` on
the API call, so it returns the fields we need without extra API
requests
- after we have the id/slug, I could build a route and model array
dynamically, or return null if we can't redirect to the object (it
doesn't exist)
- copied over and rewrote the deletion test from the legacy file
- added a new test that checks that we get a 404 when attempting to delete an unknown post
- this is a guard to protect and futureproof the API whilst we do refactoring to improve 404 handling from bookshelf
- in turn this is aimed at helping to get rid of a bunch of catch predicates from the API
- I want to start rewriting the post tests using the new e2e framework, but it's quite a big task
- For now I have renamed the existing file, and will use the correct file name for writing modern versions of tests
- Note: I have a specific test that I'd like to add which is far easier to write in the new framework
- This change should facilitate moving forward more with the new framework
refs https://github.com/TryGhost/Toolbox/issues/356
- in order to show data that we might not necessarily still have around
(ie. when you delete a post, you might want the title), we're going to
start utilizing the `context` column
- right now, we store the `primary_name` for deleted events, and we also
store the `setting` `key` and `group` so we can reference it in the
audit log
no issue
- The site locale should be exposed within the public site config in order to handle i18n in third party apps
- Added the locale to Explore service to simplify fetching it when submitting a new site
fixes https://github.com/TryGhost/Team/issues/1829
- Remove the subdirectories when creating the Attribution instances
- URLs are now always stored relative to the subdirectory instead of the root directory (makes changing the subdirectory easier)
- Fixed returning absolute urls
- Added tests
- this endpoint returns the Ghost version, of which the minor just hit
double digits
- because of this, the content-length size changed, and the snapshot was
incorrect
- we've previously allowed overrides for the content-length to be any number (see
1fadbacdec)
- this commit allows the header to be any number so it doesn't fail when
the Ghost version is incremented
closes: https://github.com/TryGhost/Ghost/issues/15252
- comments are deleted when posts are deleted. Without cascade delete on parent_id, replies cannot be deleted
- this change means that deleting a post will delete all comments and replies without error
refs: https://github.com/TryGhost/Ghost/issues/15252
- all columns with a foreign key (references prop) must have a deletion strategy
- we just found a bug with this in the comments table - see referenced issue
- this fix adjusts the schema and migration for this change before its released so we don't have to write a horrible migration later
refs https://github.com/TryGhost/Team/issues/1727
- if feature flag is enabled, handles storing expiry date on complimentary subscriptions in `expiry_at` column of `members_products`
- updates the expiry value on both member edit or add with tiers
- expiry is passed as `expiry_at` in `tiers` list of a member
- includes `expiry_at` on tiers data of a member when flag is enabled
refs: 22fd7f289c
- in the mentioned commit I changed the tests so that we don't need to update snapshots for every labs flag change
- this commit does the same for content-length which didn't get picked up locally, but does on CI for some reason
- the goal is to allow the team to add and remove flags without needing to update a random snapshot
refs https://github.com/TryGhost/Team/issues/1822
Exposing the values through the API is restricted behind the alpha flag.
We're exposing the values by default when the flag is enabled for now,
but can reconsider that later.
refs https://github.com/TryGhost/Team/issues/1808
refs https://github.com/TryGhost/Team/issues/1809
refs https://github.com/TryGhost/Team/issues/1820
refs https://github.com/TryGhost/Team/issues/1814
### Changes in `member-events` package
- Added MemberCreatedEvent (event, not model)
- Added SubscriptionCreatedEvent (event, not model)
### Added `member-attribution` package (new)
- Added the AttributionBuilder class which is able to convert a url history to an attribution object (exposed as getAttribution on the service itself, which handles the dependencies)
```
[{
"path": "/",
"time": 123
}]
```
to
```
{
"url": "/",
"id": null,
"type": "url"
}
```
- event handler listens for MemberCreatedEvent and SubscriptionCreatedEvent and creates the corresponding models in the database.
### Changes in `members-api` package
- Added urlHistory to `sendMagicLink` endpoint body + convert the urlHistory to an attribution object that is stored in the tokenData of the magic link (sent by Portal in this PR: https://github.com/TryGhost/Portal/pull/256).
- Added urlHistory to `createCheckoutSession` endpoint + convert the urlHistory to attribution keys that are saved in the Stripe Session metadata (sent by Portal in this PR: https://github.com/TryGhost/Portal/pull/256).
- Added attribution data property to member repository's create method (when a member is created)
- Dispatch MemberCreatedEvent with attribution
### Changes in `members-stripe-service` package (`ghost/stripe`)
- Dispatch SubscriptionCreatedEvent in WebhookController on subscription checkout (with attribution from session metadata)
refs https://github.com/TryGhost/Toolbox/issues/364
refs 147ec91162
- This looks like a subtle bug that has gone unnoticed for years. Have checked if we rely on the logic anywhere (mostly used in image-dimensions frontend helper) - we don't access the "url" directly.
- There is no reasoning attached behind why the cached size was stored as a url (see refed commit)
- WHY is this even being fixed? Caches can store anything... does not mean we should! Inconsistent data becomes a real PITA if the cache is persisted and is hard to repopulate (e.g. to migrate the cached data format).
refs https://github.com/TryGhost/Toolbox/issues/364
- The "new Map()" cache was a "hidden cache" that did not follow any specific pattern. Following the cache adapter pattern here makes it possible swapping out the cache for alternative implementations - e.g. Redis storage
refs https://github.com/TryGhost/Toolbox/issues/364
- The InMemoryCache is an implementation of the cache adapter interface and allows to test cache in the works which is "close to the real world". Being able to do so in tests for image sizes cache manager proves we can use other cache adapters such as Redis based ones.
refs https://github.com/TryGhost/Toolbox/issues/364
- Doing the `.catch(errors.NotFoundError...` was throwing another error as this syntax did not work with native promises. Checking `instanceof` works 100% and is way more explicit/readable way to handle this type of error differently
refs https://github.com/TryGhost/Toolbox/issues/364
- It was using an outdated syntax and relied on Bluebird depencency. Updated the syntax to async/await and dropped the Bluebird dependency.
refs https://github.com/TryGhost/Team/issues/1799
Rather than using the `adminAuthAssets` config which is not updated to
be aware of running in a different directory to the cwd, we use the
getContentPath method which handles all of the directory checking.
Without this, we were unable to serve the admin-auth iframe, as the
directory was incorrect for self hosters.
- we've had an optimization in CI that copies a fresh SQLite DB to a
file, and copies it back when we need to do a DB reset
- I originally only let this run in CI but we've had it around for a
while so we should GA it to run on local machines
- there may be edge cases, but we should fix them instead of letting
tests run slower for development
- this also makes sure we clean up any existing files before
initializing the DB
- prior to this commit, if you add or remove a faeture flag, you also have to update the snapshots for the settings tests
- feature flags are intended to be very easy to add and remove, and so this extra step doesn't fit with our needs
- it's also unnecessary, we don't need to verify the exact contents of the labs setting
- added core and builtin integrations to test fixtures
- allowed passing a custom api key id to generate JWT
- updated admin key auth test to make successful request with a `core` integration, which doesn't work atm because relations are not returned
refs https://github.com/TryGhost/Toolbox/issues/363
- this API framework is standalone and should be pulled out into a
separate package so we can define its boundaries more clearly, and
promote better testing of smaller parts
refs https://github.com/TryGhost/Toolbox/issues/363
- AFAICT, this is an outdated and unused concept from when subscribers
were a thing, but members are now how we do things, and this is causing code
to linger around that we don't need
refs https://github.com/TryGhost/Toolbox/issues/363
- this middleware is standalone and I suspect we're going to be touching
it further when we work on Ghost's caching in the near future
refs https://github.com/TryGhost/Team/issues/1726
- free trial offers don't need a stripe coupon created for them
- checkout sessions for free trial offers ignore stripe coupon and directly pass the trial days value
- trial days of an offer take precedence over trial days added as default to a tier
refs https://github.com/TryGhost/Toolbox/issues/363
- this commit pulls all code involving the Mailgun client SDK into one
new package called `mailgun-client`
- this means we should be able to replace `mailgun-js` (deprecated) with
`mailgun.js` (the new, official one) without editing code all over the
place
- this also lays some groundwork for better testing of smaller
components
refs https://github.com/TryGhost/Team/issues/1723
- Added count.replies to comments
- Added replies endpoint
- Limited returned replies to 3.
- Replaced likes_count with count.likes in comments
- Instead of fetching all the likes of a comment to determine the total count, we'll now use count.likes
- Instead of fetching all the likes of a comment to determine whether a member liked a comment, we'll now use count.liked (which returns the amount of likes of the current member, being 0 or 1). This is mapped to `liked` to make it more natural to work with.
The `members.test.snap` file changed because we no longer include `liked: false` if we didn't fetch the liked relation. And in the comments events of the activity feed the liked property is therefore removed.
These changes requires an update to the `bookshelf-include-count` plugin:
- Updated to also work for nested relations
- This moves the count queries from the `bookshelf-include-count` plugin to the `countRelations` method of each model.
- Updated to keep the counts after saving a model (crud.edit didn't return the counts before)
