refs https://github.com/TryGhost/Ghost/issues/12150
- `destroy` method was using incorrect cancel subscriptions method - stripe.cancelStripeSubscriptions - which doesn't exist
- Fixes call with intended method - `stripe.cancelAllSubscriptions` - to cancel all subscriptions
refs TryGhost/Ghost#12127
- Adds new `updateSubscription` method to members-api which allows updating individual subscription for a member
- New method only allows toggling of cancellation at period end for a subscription at the moment
no issue
- When items are parsed from CSV empty values were interpreted as empty strings - ''. Empty strings are always transformed into 'null' values in Ghost's model layer and are much more problematic to validate comparing to plain `null`. Specifically validation was imossible for 'format: date-time' with JSON schema validation through ajv when the value of date property was an empty string
- This behavior resemples one present in Ghost's model layer - 95880dddeb
- When testing performance overhead for this change did not spot any statistically significant change in performance (tested set was 50K rows)
no-issue
Using models internally and in the exported API means that we avoid expensive
`toJSON` calls, which affects performance when looping through large lists of
members. It also allows us to take advantage of the new relations used in the
models.
The addition of "ByID" methods for linking stripe customers and setting
complimentary subscriptions allows bulk imports to avoid the overhead of creating
a model for each members, instead passing an id string. n.b. currently the impl
_does_ still create models, but it makes it easier to optimise and refactor in the
future.
no-issue
Previously we would blindly put subscriptions into the database when we
received a webhook, which could result in orphaned rows that were not
linked to a customer (and by extension a member)
This updates the logic so that we will only add subscriptions if we have
a record of their customer.
Customers are only added during a checkout.session.completed webhook, at
which point a member is guarunteed, but for formailty and safety against
changes in the flow, the logic has been applied to inserting customers
too.
refs https://github.com/TryGhost/Ghost/issues/11557
If a subscription failed to delete, we would error and bailout of the
process, this updates it to log the error so that site owners have a
record of the error in the logs, but also to continue through the rest
of the subscriptions.
no-issue
Destroy is terminology we usually use for the model layer and was a
little confusing without context, this method is used in one place so
it's a low effort cleanup with minimal repercussions
refs https://github.com/TryGhost/Team/issues/342
- Send magic link middleware was not using custom status code from error and sending 500
- Updates error code to be picked from err object if present, or fallback to 500 as before otherwise
refs https://github.com/TryGhost/Ghost/issues/12074
Some sites may have had duplicate webhooks created due to a race
condition. This updates the members-api to cleanup _all_ webhooks before
starting, allowing it to create webhooks on a fresh slate, and removing
possible causes of 401 errors due to incorrect webhook secrets.
refs https://github.com/TryGhost/Ghost/issues/12065
This protects us against multiple instances of the members-api being
started simultaneously and race conditions where inbetween the initial
"GET" of a plan which returns empty, and the "POST" of a plan to create
it, another instance has already created it.
no-issue
This fixes a problem when subscribing to a Plan (Price) with a default
trial period. We also add logging to add a little more information about
which flow we're entering.
Subscriptions that are started with a trial have a `setup_intent`
present on the Checkout Session object, which was incorrectly causing us
to determine that we are in a "setup" flow and attempt to update a
customers card details.
We now use the `mode` property of the Checkout Session to determine
whether we are handling a new Subscription, or if we are in a "setup"
flow and should update the Customer's card details.
no-issue
This fixes a problem when subscribing to a Plan (Price) with a default
trial period. We also add logging to add a little more information about
which flow we're entering.
Subscriptions that are started with a trial have a `setup_intent`
present on the Checkout Session object, which was incorrectly causing us
to determine that we are in a "setup" flow and attempt to update a
customers card details.
We now use the `mode` property of the Checkout Session to determine
whether we are handling a new Subscription, or if we are in a "setup"
flow and should update the Customer's card details.
refs https://github.com/TryGhost/Ghost/issues/12061
Due to a bug in Ghost webhooks are now created with a trailing "/" which
meant that the previous webhooks to that (without a slash) was never
removed.
This results in users receiving emails from stripe about failed webhook
delivery, which is not good at all.
This fix lists out the webhooks and finds (if present) the webhook which
matches the current URL, minus the trailing slash. If found it will then
attempt to delete that webhook thus stopping the emails from Stripe.
I've added a note to remove this code as it should only ever need to run
once, and can be removed for the Ghost release after these changes.
no-issue
* Refactored model dependencies
This groups all of the model depenencies into a single models object,
and renames the models with more concise identifiers
* Fixed spacing
* Added webhook support to metadata
* Refactored stripe configure to have better logging
* Refactored webhook creation to reuse existing webhook
* Installed @types/stripe
no issue
- The extra payload added to magic link token included `name`, `labels` and `oldEmail`
- Refactor in commit [here](bf63ffe424 (diff-9f9ef757543bb9a90baba0d3bea76a83L157-R169)) changed the `body` variable assignment causing the payload objection creation to not include the extra data from request body
- Updates `body` to `req.body` to use correct data from request
refs 5c46786ebc
- This is continuation of work removing csv-parser as main CSV handling library with more suitable papaparse library
- Referenced commit introduced papaparse as a library to serialize JSON to CSV, this changeset takes it a step further and replaces CSV to JSON seriazliation logic
no issue
- When processing CSV files `parse` function now allows for the client to specify "mapping" parameter in format of a hash as follows:
{ destination_property_name: 'source_column_name'}
e.g.:
{
name: 'weird_name_column',
email: 'email_column'
}
- It is done so to allow for the end user to provide exact mapping of the fields to be transformed into JSON.
