* add a bespoke validation to the row parser
when buyer 1 is uploaded with working situation child, this should be validated with a custom message.
a test for this case was also created
* replicate the work from the last commit for the 2023 row parser and associated test file
* lint correction remove blank lines
- Previously, when attempting to import a log that already exists in the service, the importers would update its attributes.
- In most situations we don't want this as it could override changes that the user has made to the logs.
- This change adds an option to the log import services to allow this functionality to be toggled.
* Sends a correct email if there are only soft validations
* Add soft errors valid page
* Add soft errors confirm page
* Confirm the soft validations
* Reuse the how to fix template for check soft validations email
* Update email link
* Move soft validation confirmation to processor
* Correctly set the log status, remove redundant confirm_soft_validations
* Redirect successful upload to logs index and display a success banner
* Implement the soft validations only journey for sales logs
* Display the soft validation errors on the soft-errors-valid page
* Fix page alignment
* Fix tests by mapping housingneeds in csv hepler
* Add the sales soft validations to unpend_and_confirm_soft_validations
* Change naming
* Update method names
* refactor
* undo typo
* Only set the existing soft validations for correct types
* Fix path name
* Add missing error mappings for location fields
* Add missing tests and cancel button
* Change some wording
* Typos
# Context
- https://digital.dclg.gov.uk/jira/browse/CLDC-2344
- So the reported bug and these change do not match 100%
- As the reported bug should have already been resolved in an earlier change as the error message that was reported has already been removed from the system
- These changes remove the check from sales plus a few minor tidy ups
# Changes
- Remove dead code around error threshold which is no longer used or needed
- Inverted negative predicate to positive one, `incorrect_field_count` => `correct_field_count`
- Remove `validate_min_columns` from sales bulk upload
* feat: wip blank fields and dependent fields on upload tos ee if valid and can upload with missing info - this is not the exact ac on the ticket yet
* feat: update seed
* feat: wip commit
* feat: add postcodenk error so can clear on validation
* feat: add postcode validation back
* feat: move la vals to shared and move and add tests
* feat: add correct pluralisation to warning message
* feat: add blank compound invalid fields methods
* feat: update validations
* feat: update pluralisation
* refactor: lint
* feat: clear errors associated with blanked values so log status is set correctly on creation
* feat: validate instead
* feat: avoid duplicated errors
* feat: dont auto-refuse income, different to imports
* feat: validate after every blank method
* feat: delete la validator spec
* tests: update
* refactor: erblinting
* refactor: cleanup
* refactor: move pluralizer to helper
* feat: copy update
* feat: rename
* feat: refactor to avoid redundant re-validations and test
* refactor: rubocop
* test: update
* test: update
* test: update
* update sidekiq
* feat: clear errors
* feat: run clearing twice in case first clear creates different errors
* feat: remove moved file
* feat: undo validation file tweaks as shared/specific overlap could do with a more general refactor
* feat: update tests
# Context
- https://digital.dclg.gov.uk/jira/browse/CLDC-2316
- Implement bulk upload sales for new collection year 2023
- This is a first pass implementation and will probably have some bugs in it and we can address over time
# Changes
- Add CSV parser for sales 2023 to handle CSV structure
- Tweak collection window validation so error now contextual to year selected for upload
- Handle arbitrary ordering of CSV columns
- Fix ordering of errors in report and now ordered by cell
- Added `Upload your file again` link styled as button on error report to match lettings experience
- Update tooling to convert logs to 2023 csv rows with support for random column ordering
# Known issues
- There seem to be some issues with how UPRN is handled if the UPRN cannot be validated.
- For the above I think there is dependency on https://github.com/communitiesuk/submit-social-housing-lettings-and-sales-data/pull/1570 which should clear any errored fields so users can continue to create logs and fix within the service
# Context
- https://digital.dclg.gov.uk/jira/browse/CLDC-2287
- bulk upload errors on setup fields are not being categorised as setup errors
# Changes
- ensure errors on setup field are categorised as `setup` errors
- ordering of validations tweaked, so `validate_nulls` is further down the execution order. this is so any existing validation run first and `validate_nulls` only adds errors if there aren't any existing errors on a field
- removed old 16/60% tests are no longer required due to introduction of `how to fix` journey
* Clear invalid format previous postcode
* Fix import with income 2 outside of london validation
* Remove mscharge if it is under minimum
* Clear frombeds if it's out of range
* create a method on the FormHandler that returns the sales form questions for all years in the order that they appear in the form
* update csv email job to accomodate sales log export as well as lettings
add to tests to reflec the changes made
* write tests to cover the desired functionality of the SalesLogCsvService
* create the SalesLogCsvService
create a necessary method on the log to enable submission method to be included on the csv
derive values for the two halves of previous postcode for export
* add relevant links in the UI and pipe everything together in controllers
amend organisations controller to have flexibility to download logs of either type
add necessary methods to sales log controller, raising shared method to logs controller
update routing for amendments and additions
extract helper method to build urls for downloading logs within an organisation
* correct various linter complaints and tech review suggestions
* minor amendment to add old_id and reorder early columns
* undo my 'clever' refactor that broke things
* refactoring of csv service after some tech review and some UI testing in review app
* update tests to include a test of a full export and all values in teh csv
* correct minor routing error to ensure correct url is shown and tab selected after requesting csv email
* update organisations controller requests spec file to cover new functionality and make a minor amendment to authentication scope in the controller after error found in testing
* write request tests for the new functionality in the sales log controller, define authorisation in the controller
* minor correction after rubocop's kind suggestion'
* various corrections from first pass at PO, tech review, linter, etc
* refactor :ordered_sales_questions_for_all_years
* first pass at implementing flexible code-based form fixtures for testing
* second pass
* refactor all tests of :ordered_sales_questions_for_all_years to use new factories
* some refactoring in the testing of the csv service
* use that fact that params is always available in controllers and don't pass it around, inline some methods calls
* correct minor bug to ensure that "Return to logs" link returns to the correct index page
* remove reminder comments
* write further tests on the manipulation of questions into the csv headers, update factories of form constituents to allow the creation of forms with richer questions
* fix linter complaints
* minor alterations after rebase to account for changes made on other branches
* refactor after code review
* tweak fixtures after rebase containing alterations to the factory defaults
* Clear all the charges if the error is on tcharge
* Round earnings value upon import
* Rounds savings to the nearest 10
* Clear income on over_hard_max_for_london validation
* Refactor sale import validations to be consistent with lettings
* Extract charges attributes into a variable
* Add affected_question_ids to pregnancy check
* Update is_referrer_interruption_screen? check and naming
* Use interruption_screen_question_ids to set soft validation errors on relevant fields
* Add soft validations to sales bulk upload
* Add soft validations to lettings logs 23/24 bulk upload
* Add errors for optional soft validations
* Only add soft validations once
* Import helper methods
* Update test based on new validation messages
* Rebase fix
* [CLDC-2202] Allow coordinators to set created_by
* Scope user selection when data coordinator
* Remove managing_for_other_user_enabled
* Move sales created_by page and question out of common
* Address comments
- only select required users
- remove not needed CYA checks
# Context
- https://digital.dclg.gov.uk/jira/browse/CLDC-1888
- This is a continuation of https://github.com/communitiesuk/submit-social-housing-lettings-and-sales-data/pull/1277
- When bulk uploading we want to check users are not uploading data that already exists to prevent them submitting duplicate
# Changes
- This feature is behind a feature toggle. it has been disabled for staging for testing purposes but available in all other environments
- If a log already exists based off certain fields add errors to the associated fields
- We discount any hidden logs and only check "active" logs
- Added memoization to `#valid?` as an optimisation
* feat: add validation for number of field labels
* feat: check field labels are numbers
* feat: remove validations for number of columns
* refactor: rename non_blank_fields_count to valid_field_numbers_count
* test: add functions to generate custom field labels/values
* test: that extra invalid field labels don't cause issues
* test: that removing a valid field label reduces count by 1
* test: that adding a valid field label increases count by 1
* refactor: rename validate_fields_count and wrong_field_count
* feat: add validation that max col count not exceeded when no headers
* fix: convert numbers to strings in default_field_numbers
* feat: add leniency to max cols count (in case of 1 extra col)
* test: explicitly set year of bulk upload to be 2022
* test: add/update tests in validator_spec
* chore: lint
* test: remove tests from csv_parser_spec that were moved to validator_spec
* feat: update 2022 csv_parser to work with new validations
* refactor: define number of valid 2022 fields in one place
* refactor: remove redundant headers definition
* feat: update 2022 csv parser to have col flexibility like 2023
* test: for validator 2022 as well as 2023
* test: simplify 2022/2023 logic and improve layout
* chore: lint
* test: ensure context descriptions start with "when"
* refactor: check fields/columns count within csv parser, not validator
* test: update 2022 csv parser tests to work like 2023
* chore: lint
* feat: add validation to check UPRN exists if address doesn't
* test: add tests for UPRN/address fields
* test: make UPRN<=12 chars test more specific
* feat: show qu header on BU error template if check_answer_label missing
* test: that validate_nulls gets header if check_answer_label is missing
* fix: drop "known" from "You must answer UPRN known"
* fix: update uprn validation logic
* test: make UPRN being missing explicit
* test: improve test descriptions
* test: that errors added to address fields too when uprn & address fields all null
* refactor: standardise check for presence of needs type
* test: attempt to stub setup question to have nil check_answer_label
* Revert "test: attempt to stub setup question to have nil check_answer_label"
This reverts commit f5b3f6179b.
* test: put test within context block for clarity
* test: fix typo in test description
* test: ensure one other non-blank field exists in row
* feat: output "this question" if qu has no header or check_answer_label
* chore: lint
* fix: add .presence to question.header and remove & before .downcase
* fix: add :after_log context to uprn validation
* rename variables in test file
* write tests for new validations
* validate that answers to numeric questions are given to the correct level of accuracy and in normal format with only digits
* add error messages for new validations
* add or alter step on numeric questions in sales, amend one test given step changes
* copy change
* remove validation that is specific to number format and fold that functionality into accuracy validation, rejig tests accordingly
* fix various tests
* remove reference to currency in validation message that concerns step only and doesn't check whether firled is currency related
* alter copy to avoid redundant repetition in validaiton message
* save changes made in rebase
* changes after PO review
two step values had been missed in previous work
in various places there were custom validations applied that duplicated the functionality of the generic min max validation that is applied to all numeric questions, in these situations sometimes the min and max on the question class was inconsistent with the limit that triggered validations
these have been corrected and made consistent
various tests were affected by this and have been amended accordingly
* remove tests introduced by mistake in rebase
* amend step value on earnings
# Context
- https://digital.dclg.gov.uk/jira/browse/CLDC-2235
- Support invalid options for bulk upload. one use case when user for a new collection year supplies a value only valid in the previous collection year
# Changes
- this validation works before `log.valid?` clears any fields
- as a result there is the potential to get 2 errors on a field for when it becomes blanked and invalid option occur together
- bulk upload validations are now split so that they run before or after `log.valid?`. this is due to the fact that `log.valid?` heavily mutates the `log` object. so we want to validate both before and after the data mutates depending on what needs to be checked
- errors must be duplicated and merged as calling `valid?` clears any existing errors on the object
- all validations are assigned a specific context otherwise they are added to the default context and will also be called when a context is given
# Context
- https://digital.dclg.gov.uk/jira/browse/CLDC-2279
# Changes
- split bulk upload feature toggle by log type ie lettings and sales
- this allows us to toggle with finer granularity
- moved `FeatureToggle` from initializer to service class. not sure why it was an initializer in the first place? this means its available to zeitwerk and will be reloaded when needed
* rename variables in test file
* write tests for new validations
* validate that answers to numeric questions are given to the correct level of accuracy and in normal format with only digits
* add error messages for new validations
* add or alter step on numeric questions in sales, amend one test given step changes
* copy change
* remove validation that is specific to number format and fold that functionality into accuracy validation, rejig tests accordingly
* fix various tests
* remove reference to currency in validation message that concerns step only and doesn't check whether firled is currency related
* alter copy to avoid redundant repetition in validaiton message
* save changes made in rebase
* changes after PO review
two step values had been missed in previous work
in various places there were custom validations applied that duplicated the functionality of the generic min max validation that is applied to all numeric questions, in these situations sometimes the min and max on the question class was inconsistent with the limit that triggered validations
these have been corrected and made consistent
various tests were affected by this and have been amended accordingly
* remove tests introduced by mistake in rebase
# Context
- https://digital.dclg.gov.uk/jira/browse/CLDC-2223
- Change how `renewal` works for bulk upload in 2023 compared to 2022
# Changes
- `renewal` is mandatory and must be filled in otherwise there is an error
- it is no longer inferred from another field and we just use the value from the user