* Lazy load the pricing bom table when the "pricing" tab is selected
* Update django-debug-toolbar configuration
* Major refactoring for the 'can_build' function
- Use a single annotated query to the db, rather than a for loop (which is what a caveman would use)
- Query performance is greatly improved
- Also refactors existing variant-part-stock subquery code, to make it re-usable
* Use minified JS and CSS where possible
* Render a 'preview' version of each part image
- Saves load time when the image is quite large
- Adds a data migration to render out the new variation
* Adds 'preview' version of company images
* Defer loading of javascript files
Note: some cannot be deferred - jquery in particular
* Crucial bugfix for user roles context
- Previously was *not* being calculated correctly
- A non-superuser role would most likely display pages incorrectly
* Prevent loading of "about" on every page
- Load dynamically when requested
- Takes ~400ms!
- Cuts out a lot of fat
* Match displayed image size to preview image size
* Utilize caching framework for accessing user "role" information
- Reduces number of DB queries required by rendering framework
* Remove redundant query elements
* Remove 'stock' field from PartBrief serializer
- A calculated field on a serializer is a *bad idea* when that calculation requires a DB hit
* Query improvements for StockItem serializer
- Remove calculated fields
- Fix annotations
* Bug fixes
* Remove JS load test
- Loading of JS files is now deferred, so the unit test does not work as it used to
* Fix broken template for "maintenance" page
* Remove thumbnail generation migrations
- Already performed manually as part of ''invoke migrate"
- Running as a migration causes unit test problems
- Not sensible to run this as a data-migration anyway
* tweak for build table
* use shims for API view inheritation
* Add mixin for input sanitation
* fix clean operation to fix all string values
* Also clean up dicts
this is to future-proof this function
* Update docstirng
* proof custom methods against XSS through authenticated users
* Create custom ModelResource subclass
- Strips illegal starting characters from string cells
- Prevents formula injection
* Update all existing ModelResource classes to base off InvenTreeResource
* Handle more complex case where an illegal char is hidden behind another one
* Adds a Part API filter to limit query to valid conversion options for the specified part
* Refactor 'exclude_tree' filter to use django-filter framework
* Refactor the 'ancestor' filter
* Refactoring more API filtering fields:
- variant_of
- in_bom_for
* Adds API endpoint / view / serializer for converting a StockItem to variant
* stock item conversion now perfomed via the API
* Bump API version
* Add unit tests for new filtering option on the Part list API endpoint
* Adds unit test for "convert" API endpoint functionality
* Delete category via the API
* Delete StockLocation via the API
* Delete StockItem via the API
- Removes the final instance of AjaxDelete
* Remove URL path
* Add missing code
* Catch DoesNotExist error
* Move notificationtable function to js file
* Fix for custom metadata class
- Previously only worked if a POST or PUT action was available on the endpoint
- So, a ListAPIView endpoint would not actually work!
- Adding in a BulkDelete mixin to a ListAPIView caused failure
* Add unit test to ensure new OPTIONS metadata updates are checked
* Expand functionality of the existing BulkDelete mixin
- Allow deletion by custom filters
- Allow each implementing class to implement custom filters
- Adds more unit testing for BulkDelete mixin class
* Add bulk delete operation for Notification API
- Ensure users can only delete their *own* notifications
* Improve notification tables / buttons / etc
* Adds unit testing for bulk delete of notifications
- Fixed API permissions for notifications list endpoint
* Update BulkDelete operations for the StockItemTestResult table
* Use filters parameter in attachments table to ensure that only correct attachments are deleted
* JS linting
* Fixes for unit tests
* Introduces a BulkDelete API mixin class
- Allows deletion of multiple items against a single API request
* Bump API version
* Adds BulkDelete mixin to StockItemTestResult API class
* refactor "multi BOM Item delete" to use new approach
* Refactor various attachment API endpoints
* Refactor multi delete for StockItem
* Convert remaining enndpoints over
* Fix for API test code
* Fix "polarity" of modal form submit button
(cherry picked from commit 0e4550f28895af626715723e450792881f7fd882)
* Use existing API functionality to delete all test results for a particular StockItem
* Remove outdated forms / views
* Add new global setting to control auto-upload of test reports
* Adds callback to attach a copy of the test report when printing
* Fix for all attachment API endpoints
- The AttachmentMixin must come first!
- User was not being set, as the custom 'perform_create' function was never called
* Remove duplicated UserSerializer
* display uploading user in attachment table
* Add unit test to check the test report is automatically uploaded
- For some reason, the reverse URL lookups would fail for stock items which were not "in stock"
- e.g. detail pages for stock items would fail to render if the stock item was "destroyed" or "quarantined"
- changing the queryset to .none() fixed this
- I do not know why...
* Remove outdated forms / views for stock item tracking
- No longer used anywhere
- Manual stock tracking entries are no longer supported anyway
* Futher cleanup:
- Delete unused template
- Remove URL which pointed to a nonexistent template
- Remove button for manually adding stock tracking entries
* Adds ability to return item into stock via the API
* Remove old server-side form / view for returning stock from a customer
* Add unit tests for new API endpoint
* Add pre-commit to the stack
* exclude static
* Add locales to excludes
* fix style errors
* rename pipeline steps
* also wait on precommit
* make template matching simpler
* Use the same code for python setup everywhere
* use step and cache for python setup
* move regular settings up into general envs
* just use full update
* Use invoke instead of static references
* make setup actions more similar
* use python3
* refactor names to be similar
* fix runner version
* fix references
* remove incidential change
* use matrix for os
* Github can't do this right now
* ignore docstyle errors
* Add seperate docstring test
* update flake call
* do not fail on docstring
* refactor setup into workflow
* update reference
* switch to action
* resturcture
* add bash statements
* remove os from cache
* update input checks
* make code cleaner
* fix boolean
* no relative paths
* install wheel by python
* switch to install
* revert back to simple wheel
* refactor import export tests
* move setup keys back to not disturbe tests
* remove docstyle till that is fixed
* update references
* continue on error
* add docstring test
* use relativ action references
* Change step / job docstrings
* update to merge
* reformat comments 1
* fix docstrings 2
* fix docstrings 3
* fix docstrings 4
* fix docstrings 5
* fix docstrings 6
* fix docstrings 7
* fix docstrings 8
* fix docstirns 9
* fix docstrings 10
* docstring adjustments
* update the remaining docstrings
* small docstring changes
* fix function name
* update support files for docstrings
* Add missing args to docstrings
* Remove outdated function
* Add docstrings for the 'build' app
* Make API code cleaner
* add more docstrings for plugin app
* Remove dead code for plugin settings
No idea what that was even intended for
* ignore __init__ files for docstrings
* More docstrings
* Update docstrings for the 'part' directory
* Fixes for related_part functionality
* Fix removed stuff from merge 99676ee
* make more consistent
* Show statistics for docstrings
* add more docstrings
* move specific register statements to make them clearer to understant
* More docstrings for common
* and more docstrings
* and more
* simpler call
* docstrings for notifications
* docstrings for common/tests
* Add docs for common/models
* Revert "move specific register statements to make them clearer to understant"
This reverts commit ca9665462202c2d63f34b4fd920013b1457bbb6d.
* use typing here
* Revert "Make API code cleaner"
This reverts commit 24fb68bd3e1ccfea2ee398c9e18afb01eb340fee.
* docstring updates for the 'users' app
* Add generic Meta info to simple Meta classes
* remove unneeded unique_together statements
* More simple metas
* Remove unnecessary format specifier
* Remove extra json format specifiers
* Add docstrings for the 'plugin' app
* Docstrings for the 'label' app
* Add missing docstrings for the 'report' app
* Fix build test regression
* Fix top-level files
* docstrings for InvenTree/InvenTree
* reduce unneeded code
* add docstrings
* and more docstrings
* more docstrings
* more docstrings for stock
* more docstrings
* docstrings for order/views
* Docstrings for various files in the 'order' app
* Docstrings for order/test_api.py
* Docstrings for order/serializers.py
* Docstrings for order/admin.py
* More docstrings for the order app
* Add docstrings for the 'company' app
* Add unit tests for rebuilding the reference fields
* Prune out some more dead code
* remove more dead code
Co-authored-by: Oliver Walters <oliver.henry.walters@gmail.com>
* Add pre-commit to the stack
* exclude static
* Add locales to excludes
* fix style errors
* rename pipeline steps
* also wait on precommit
* make template matching simpler
* Use the same code for python setup everywhere
* use step and cache for python setup
* move regular settings up into general envs
* just use full update
* Use invoke instead of static references
* make setup actions more similar
* use python3
* refactor names to be similar
* fix runner version
* fix references
* remove incidential change
* use matrix for os
* Github can't do this right now
* ignore docstyle errors
* Add seperate docstring test
* update flake call
* do not fail on docstring
* refactor setup into workflow
* update reference
* switch to action
* resturcture
* add bash statements
* remove os from cache
* update input checks
* make code cleaner
* fix boolean
* no relative paths
* install wheel by python
* switch to install
* revert back to simple wheel
* refactor import export tests
* move setup keys back to not disturbe tests
* remove docstyle till that is fixed
* update references
* continue on error
* use relativ action references
* Change step / job docstrings
* Refactor 'DeleteManufacturerPart' form
- Remove duplicated forms
- Update style to match other forms
- Block on each deletion before progressing to the next one
* PEP fix
* Adds deleteSupplierParts function
* Refactor all instances of supplier part deletion
* Refactor tractor : use convenience function for bootstraptable.getSelections
* Add deleter for manufacturerpartparameter.
Refactor existing code into a single function
* Refactor deletion for stock items
* JS linting
* Fix display of stock labels
- If 'shipped' or 'installed', don't display 'allocated' flag
* Switch stock item data around
* Add 'available' and 'allocation' information to the StockItem detail page
- Cache some context data to the view renderer
* Stock table now also displays allocation informatoin
- If the stock item had been created as part of a Build Order, and subsequently "converted" to a variant part, the conversion operation will fail
- Patch allows the build reference to be linked based on either the base part, or any conversion options
