Open
Conversation
I've renamed this to have a prefix of `pds` as we're going to be adding two more feature flags both related to PDS so it makes sense to group them together. Jira-Issue: MAV-2354
This replaces the `enqueue_bulk_updates` setting on `Settings.pds` with a feature flag which makes it easier to be enable/disable the feature without needing an infrastructure deploy. Jira-Issue: MAV-2354
This replaces `Settings.pds.enabled` with a feature flag that we can use to quickly enable or disable any PDS related functionality without needing to re-deploy the application. Jira-Issue: MAV-2354
This reduces the risk of supply chain attacks where the tag points to a different commit implicitly without us being aware. Jira-Issue: MAV-5897
…ic links on school page
This refactors how the options for the "who" step of the draft consent form are determined to simplify the code, and allow for a more complex set up around self-consent. Jira-Issue: MAV-5915
This adds a method on the `Patient` model that can be used to determine if a child can self-consent, following a competent Gillick assessment. Jira-Issue: MAV-5915
Gillick assessments are only considered valid for the day that they were recorded on. Outside of that, the child needs to be re-assessed before they can self-consent. Jira-Issue: MAV-5915
The national reporting teams uses the same vaccination table but requires more details on vaccination records like location and source, whereas full-fat Mavis doesn't need so much information. The new `show_details` arg controls whether to show or hide this info.
Shows the programme status (heading with colour, contextual detail text, vaccination outcomes table, and action link) for a patient on the session patient show page. This card links to the respective programme tab in the child’s record. Eventually this will be the only card on the page with a coloured heading so various statuses (consent, triage, etc) do not compete for attention and the programme status for the child is reflected right at the top. MAV-4846
When a triage has been performed, the programme card now shows a human-readable summary of the decision (e.g. "Joy, Nurse decided that Oliver is safe to vaccinate.") using the shared triage_summary helper. The delay vaccination case is updated to include the "until" date when present.
…undle-link-mismatch Suppress warning for `immunization.target` mismatch
…tions Improve date validations on imports
Only consider today’s Gillick assessments as valid
April design review
Fix session layout
We've seen some evidence that the batching queries generated by `PatientStatusUpdater#update_programme_statuses!` are (at least sometimes) slow in production, potentially because of a bunch of sequential scans relating to JOINs generated by the call to `ActiveRecord::QueryMethods#includes`. The call to `#includes` is necessary in order to trigger eager loading of associations to avoid N+1 queries when calling `Patient::ProgrammeStatus#assign`. However, this is only relevant for the batch queries which make the model instances available within each batch; not for the batching queries which just work out which model IDs to include in a given batch. Thus we can simplify the batching queries without changing the overall behaviour by using `ActiveRecord::Batches#in_batches` [1] instead of `#find_in_batches` and moving the call to `#includes` onto the batch relation. Note that as per the `ActiveRecord::QueryMethods#includes` docs [2]: > A separate query is performed for each association, unless a join is required > by conditions. This means that the JOINs were only being generated on the batching queries when the patient scope included a condition, e.g. when associated with an `ImmunisationImport`. Given the instance of `PatientStatusUpdaterJob` scheduled overnight [3] uses the default arguments, its batching query is *unscoped* and thus it may not directly benefit from any performance improvements made by the changes in this commit. [1]: https://api.rubyonrails.org/v8.1.3/classes/ActiveRecord/Batches.html#method-i-in_batches [2]: https://api.rubyonrails.org/v8.1.3/classes/ActiveRecord/QueryMethods.html#method-i-includes [3]: https://github.com/NHSDigital/manage-vaccinations-in-schools/blob/6895ee514a018aaa507a460e332a32b3d9e8b6a1/config/sidekiq.yml#L70-L73 Co-authored-by: Chris Lowis <chris.lowis@gofreerange.com> Co-authored-by: Chris Roos <chris.roos@gofreerange.com>
Refactor PDS feature flags
Add programme status card
…atient-status-updater Simplify batching query in PatientStatusUpdater
Pin all GitHub Actions to commit SHAs for security
Add mavis server
This adds the logic where if a child is eligible for the programme and hasn't been sent any consent requests and has parent contact details: - If part of a session where consent requests are scheduled to go out in the future, the programme status becomes `needs_consent_request_scheduled`. - If not assigned to a session with a request scheduled, the programme status becomes `needs_consent_request_not_scheduled`. Jira-Issue: MAV-5882
This allows us to run the end_to_end stack while running development locally.
The convention is capitalised variable names are intended for environment variables that are exported and passed between processes. This change makes it clear that only RAILS_ENV and REDIS_URL are, in fact, env vars.
…cheduled-not-scheduled-attempt-2 Implement logic needs consent status request scheduled/not scheduled
Use redis db 2 in end_to_end env
This helps prevent the E2E tests from being flaky. This has already been implemented for the national reporting `Vaccine`s.
Sort `Vaccine`s alphabetically in upload help test
Bumps [prettier](https://github.com/prettier/prettier) from 3.8.1 to 3.8.2. - [Release notes](https://github.com/prettier/prettier/releases) - [Changelog](https://github.com/prettier/prettier/blob/main/CHANGELOG.md) - [Commits](prettier/prettier@3.8.1...3.8.2) --- updated-dependencies: - dependency-name: prettier dependency-version: 3.8.2 dependency-type: direct:development update-type: version-update:semver-patch ... Signed-off-by: dependabot[bot] <support@github.com>
…/prettier-3.8.2 Bump prettier from 3.8.1 to 3.8.2
Bumps the bundler group with 1 update in the / directory: [yard](https://yardoc.org). Updates `yard` from 0.9.38 to 0.9.42 --- updated-dependencies: - dependency-name: yard dependency-version: 0.9.42 dependency-type: indirect dependency-group: bundler ... Signed-off-by: dependabot[bot] <support@github.com>
…134ee445d Bump yard from 0.9.38 to 0.9.42 in the bundler group across 1 directory
This rotates the API key we use in production as the old key was created 18 months ago and it's good practice to rotate regularly. Jira-Issue: MAV-6545
Since they are shared behaviour ... Jira-Issue: MAV-6062
Import::CSVData will be used to abstract certain operations out of CSVImportable, such as loading CSV data from user params. Jira-Issue: MAV-6063
It is already tested as part of CSVParser and the setup is confusing in these this tests. It needs to know that the encoding is either detected as part of `csv=` through the bom (which we don't use here), or in CSVParser, which is where this test really belongs anyway. Jira-Issue: MAV-6062
Recent changes rewired specs to test import by setting `csv_data` instead of `csv` to isolate what's being tested, but that left `csv=` largely being untested. Jira-Issue: MAV-6063
This changes import specs to not make test `subject` an action. This is stylistic, it can work either way, but considering how hard it's been to reason about the imports I thought it would be better to return to what's considered common convention to try to make the tests easier to understand. This became particularly apparent when a call to `parse_rows!` had to be added to tests (coming in a later change), meaning the `subject` would need a call to `parse_rows!`, or possibly `parse_rows!` added to the `before` block, but either way it was becoming un-idiomatic. Jira-Issue: MAV-6063
This is to make it easier to reason about the code: when parse_rows! is explicit it's clear what stage of the process is executing when. Jira-Issue: MAV-6063
The intention is to make the import processing simpler to understand. Since nearly all the processing happens in the sub-classes we just call that directly from the ProcessImportJob. Jira-Issue: MAV-6063
Rotate GOV.UK Notify API key
Refactor import specs
Refactor import process
Jira-issue: MAV-6746
thomasleese
approved these changes
Apr 20, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
9 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
We already have a status in
CSVImportablecalledprocessedso it's helper method processed? was being masked by this method. We also don't have to define aprocessed_at?as Rails has automatically generated one for use.Release version is tbd.
MAV-6746