- As of redis 7, a list is added to the result of fetching the module list
- This list cannot be "decoded",so use `frappe.as_unicode` that handles bytes as well as other types
- Test to block Item fields (which aren’t in Website Item) in E Commerce Settings as filters
- Removed unnecessary function and setup in E Commerce Settings test
- Removed commented useless test
- Test to check custom field as filter
- tweak `filters.py` to correctly query filter field values from Website Item
- Use Website Item for filter field options in Settings and Item Group Field Filter table
- Separate Item group and Item autocomplete dict definition
- Add payload along with Item group, containing namke and route
- Pass weightage while defining item group autocomplete dict (auto sort)
- Use payload while getting results for categories in search
- Remove check to show categories, always show
- Search fields mandatory if reidsearch enabled
- Code separation (rough)
- bind the right classes to the filter lookup field
- make class names more descriptive
- make filter lookup field more visible with white bg and border
- bind lookup input field js in `views.js`
- make filter lookup field functioning for atribute filters too
- added placeholder to lookup field
- Use absolute route even 3-4 sub-category levels down
- Remove scroll from category pills due to accessibility issues
- Arrange sub-category pills alphabetically
- Added 'Include descendants' checkbox, which will pull child item group products too
- Build item group filters in query engine file
- Include logic in filter engine
- Clean up Website section of Item Group page (UX)
- Add util to fetch child item groups including self
- Multiple Item Variants creation fails due to extra spaces in attributes from popup. Clean them before passing to server side
- Mention explicit method to build variants cache to avoid ambiguity between old method path (pre-refactor)
- If the number of variants are large (almost 2lakhs), the query to get variants and attribute data takes time
- If the no.of disabled attributes is large as well, the list comprehension weeding out disabled variants takes forever
- We dont need to loop over the variants data so many times
- Avoid any `if a in list(b)` is best when the iterables have tremendous data
