There's a (tiny) race condition where if an enclosing transaction
has not yet been committed when a user is created, and a sidekiq
worker begins find(:id) and fails, that the job is enqueued, started,
and skipped.
This changes the after_save hook on User to after_commit, ensuring the
user's assigned id is visible to other processes.
* Rename banned and comment_banned roles
* Add data update script to rename roles containing 'ban'
* Add named error for Suspended users
* Update unidiomatic method names
* Rename misc banned to suspended
* Apply suggestions from code review
Co-authored-by: Michael Kohl <me@citizen428.net>
* Add unit tests for suspended methods
This commit also adds TODO comments for removing banned and
comment_banned from the codebase after data update scripts have
successfully run on all of our Forems.
Co-authored-by: Michael Kohl <me@citizen428.net>
* User#followed_articles->Article.cached_tagged_with
This should improve performance on StoriesController#index
* Make cached_* work like their tagged_with cousins
* Swap to `cached_tagged_with_any`
This was using `tagged_with(tags, any: true)` before so this commit
swaps to using a scope that aligns with that method.
* Allow for searching by Tag models
* do the thing
* abracadabra run the build
* Add comment explaining Postgres regex anchors
* feat: add the two new models users_setting and users_notification_setting
* feat: add the settings and notification_settings table to the schema
* feat: add the user and notification models
* feat: add the user_id foreign key to the model
* chore: sneaky indent
* feat: add some fields from the profile attributes
* Revert "feat: add some fields from the profile attributes"
This reverts commit 376828746ded063a243505d317140fa5339227cf.
* chore: add some profile field attributes
* chore: remove language_settings
* chore: update indent
* chore: remove language_settings
* feat: changes to the tables
* chore: remove validation in favor of the foreign keys
* chore: add default for editor version
* Address PR review suggestions
* setting_spec.rb needs to be fixed; need help
* Working on PR review comments
* Continue with addressing PR review comments
* Remove normalize_config_values method; pass correct values from forms
* Address Travis failures
* revert some unnecessary changes in spec file
Co-authored-by: Arit Amana <msarit@gmail.com>
* Remove pro role on user, expose pro dashboard to all users as analytics
* Remove pro from Elasticsearch mappings
* Update user role docs to use :trusted over :pro
* Remove pro from Role model spec
* Remove more references to pro, as noted by @rhymes
* Add a failing test
This is failing because sponsorships are not removed by Users::Delete,
and raises a foreign key constraint error because user.destroy tries
to remove a referenced user.
* When destroying a user, destroy associated sponsorships
There is a `belongs_to` relationship in sponsorships, and a has_many
relationship in Organization, this adds a mirror relation on User with
the equivalent dependent destroy rules.
If an organization is removed, all sponsorships belonging to that
organization are removed too, and now, if a user is removed, all
sponsorships created by that user are removed.
Co-authored-by: djuber <djuber@forem.beta-reduction.com>
* Add PostgreSQL FTS for usernames
* Change profile_image_90 logic
* Make search_users private
* Add tsvector index on usernames
* Limit the number of search results
* Update index name
* Only check cache or database once
Since `@trusted` is a boolean true/false value, we can't use the
`@ivar || value` pattern, and want instead to check if the variable
has been defined.
Fixes#13006
* Add unit test around memoization behavior
In local testing I was able to get this to fail by changing the
2.times block to include `remove_instance_variable(:@trusted)` (so the
`allow...once` check is sufficient to ensure we are not calling the
cache more than once per instance when false).
When `acts-as-taggable-on`'s `.tagged_with()` is used with `any: true`,
the gem will use `SELECT *` regardless of any previous (or following) requests
of selecting a limited amount of columns.
Given that the `articles` table has [73 columns](https://dev.to/admin/blazer/queries/314-number-of-columns-in-all-tables)
that will amount to wasted RAM memory for columns we don't need.
By "unscoping" any previous `select()` we can optimize used memory.
Before:
```ruby
[24] pry(main)> Article.tagged_with([:ruby], any: true).select(:id, :name).to_sql
=> "SELECT \"articles\".*, \"articles\".\"id\", \"name\" FROM \"articles\" WHERE EXISTS (SELECT * FROM \"taggings\" WHERE \"taggings\".\"taggable_id\" = \"articles\".\"id\" AND \"taggings\".\"taggable_type\" = 'Article' AND \"taggings\".\"tag_id\" IN (SELECT \"tags\".\"id\" FROM \"tags\" WHERE (\"tags\".\"name\" LIKE 'ruby' ESCAPE '!')))"
```
Note, how the SQL query is `articles.*, articles.column_a`
After:
[25] pry(main)> Article.tagged_with([:ruby], any: true).unscope(:select).select(:id, :name).to_sql
=> "SELECT \"articles\".\"id\", \"name\" FROM \"articles\" WHERE EXISTS (SELECT * FROM \"taggings\" WHERE \"taggings\".\"taggable_id\" = \"articles\".\"id\" AND \"taggings\".\"taggable_type\" = 'Article' AND \"taggings\".\"tag_id\" IN (SELECT \"tags\".\"id\" FROM \"tags\" WHERE (\"tags\".\"name\" LIKE 'ruby' ESCAPE '!')))"
```
`articles.*` is gone :-)
- https://github.com/mbleigh/acts-as-taggable-on/issues/936
- 47da5036de/lib/acts_as_taggable_on/taggable/tagged_with_query/any_tags_query.rb (L2-L8)
* Use proper Mailchimp newsletter IDs respective to methods
* Add guard clauses when Mailchimp newsletter isn't enabled
* Rubocop correction
* Only unsub after destroy if Mailchimp newsletter is enabled
* Refactor a bit
* Fix typo
* Fix boolean logic for guard clauses
* Update tests to account for new SiteConfig logic
* Break guard clause into two clauses
* Add script to remove feature flag
* Remove :feeds_import feature flag and RssReader and related classes
* Change Feeds::ImportArticlesWorker signature to support Sidekiq Cron serialization
* Replace RssReader::Assembler with Feeds::AssembleArticleMarkdown
* Removing Assembler
* Add validation for payment_pointer and clean its space before save
* Add test for payment_pointer
* fix payment pointer regexp
* fix test
* Remove space with strip
Co-authored-by: Jacob Herrington <jacobherringtondeveloper@gmail.com>
* Remove redundant test
* Refactor clean_payment_pointer to strip_payment_pointer
- Move callback from before_save to before_validation
* Refactor PAYMENT_POINTER_REGEXP
* style: remove last comment in regexp
* fix typo
Co-authored-by: Jacob Herrington <jacobherringtondeveloper@gmail.com>
* Prepare to drop profile columns from user
* Update code and factory
* Also remove unused constant
* Move validation from user to profile
* Remove Profiles::ExtractData service object
* Add more comments
* Simplify sameAs attribute generation
* Obey me machine, I am your master
* Fix condition order in guard clause
* Temporarily disable callback
* Fix specs
* Reduce usage of Profile#refresh_attributes!
* Remove leftover comment
* Handle social media links differently
* More spec fixes
* Fix specs for admin profile fields controller
* Fix specs after merge
* Fix remaining specs
* Update user show request spec
* Add comment for follow_hiring_tag
* Only save profile when user is valid
* Fix seeds.rb for profile fields
* Switch from before_save to after_save
* Undo unrelated formattin change
* Update spec/fixtures/files/profile_fields.csv
Co-authored-by: Molly Struve <mollylbs@gmail.com>
* Remove data update script and spec
* Fix spec
* Fix typo in comment
* Fix typo in comment
* Move article resave logic to service object
* Move profile field creation to before(:suite)
* Refactor error handling in Profiles::Update
* Fix Profiles::Update specs and refactor
* Temporarily disable spec
* Add ProfileValidator
* Clean up
* Move DB ready check into app/lib
* Refresh attributes after importing from CSV
* Fix specs
* Remove unused file
* A girl has no name. A profile neither.
* Fix specs
* Add responds_to? check
* Spec fix
* Add name to user fields in profile settings page
Co-authored-by: Molly Struve <mollylbs@gmail.com>
* Add podcast_appearances joined table (forem#82)
* Add podcast appearances keys unique constraint and creator property (forem#82)
* Add podcast_appearance role validation forem(#82)
* Add spec test for podcast_appearances model (forem#82)
* Small refactoring - place podcast appearance association by alphabetical order
* Adapt m2m association to pass rspec test (forem#82)
* Rename podcast appearance model to podcast episode appearance and remove fk indexing
* Rename podcast appearance model to podcast episode appearance and remove fk indexing
* Rename podcast episode appearance models and spec models. Update podcast episode and user spec tests (forem#82)
* Rename podcast episode appearances composite index to follow Rails naming pattern (forem#82)
* Remove ddl_transaction disabling and conurrent index adding from appearance migration
* Add role validation in model spec (forem#82)
* Add explanatory comment to how RssReaderFetchUserWorker works
* Add Feeds::ImportWorker worker
* Enable Feeds::Import for single user feed fetching behind a feature flag
* Add Feeds::ValidateUrl service
* Add feature flag to user's validate feed URL feature
* Remove todo notices
* Add feature flag to fetch_all_rss task
* Add feature flag to RssReaderWorker
* Bring back short-circuit for Articles::RssReaderWorker
