feat: dq_category_specific_ingredient_percent_2

[benbenben2]

What

Add quality facets for jams having too small fruit quantity:
errors:

• en:specific-ingredient-fruit-quantity-is-below-the-minimum-value-of-35-for-category-jams
• en:specific-ingredient-fruit-quantity-is-below-the-minimum-value-of-25-for-category-redcurrants-jams
• ["en:specific-ingredient-<$specific_ingredient_id>-quantity-is-below-the-minimum-value-of-$quantity_threshold-for-category-<$category_id>]

info:

• en:missing-specific-ingredient-for-this-category

Following comment from @CharlesNepote and @aleene, I tried (maybe not perfectly done) to write the thresholds values directly in the taxonomy.

This is only for specific ingredients, only for lower than the value provided. If it works well, in future PR, it could be generalized for ingredients and nutriments and for maximal values (for labels it can be minimum (sugar, for example) or maximum (fibers, for example)). Also, to do eventually in the future, detect if jam/jelly should be extra-jam/extra-jelly.

Screenshot

Related issue(s) and discussion

Part of #1414

[sonarqubecloud]

Quality Gate passed

Kudos, no new issues were introduced!

0 New issues
0 Security Hotspots
No data about Coverage
No data about Duplication

See analysis details on SonarCloud

[codecov-commenter]

Codecov Report

Attention: 1 lines in your changes are missing coverage. Please review.

Comparison is base (5b33015) 49.28% compared to head (4a107f4) 49.31%.

Files	Patch %	Lines
lib/ProductOpener/DataQualityFood.pm	92.85%	0 Missing and 1 partial ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #9606      +/-   ##
==========================================
+ Coverage   49.28%   49.31%   +0.02%     
==========================================
  Files          66       66              
  Lines       20546    20560      +14     
  Branches     4946     4951       +5     
==========================================
+ Hits        10126    10139      +13     
  Misses       9132     9132              
- Partials     1288     1289       +1     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[CloCkWeRX]

My only suggestion would be seeing if there's an existing way to say "yes, this fails a data quality check for category, but no; it is not because the data is wrong".

This would cater for scenarios like https://www.npr.org/2020/10/01/919189045/for-subway-a-ruling-not-so-sweet-irish-court-says-its-bread-isnt-bread or https://www.delish.com/food-news/a49216/things-you-didnt-know-about-pringles/ (Potato Chips made from dried potato vs potato crisp vs potato chip)

• Average person and AI would call it... "bread" or "potato chip" when they are scanning it
• A regulator would call it something different

Otherwise, this somewhat implies that categories can only be derived from regulations

[stephanegigandet]

Looks good to me

[benbenben2]

My only suggestion would be seeing if there's an existing way to say "yes, this fails a data quality check for category, but no; it is not because the data is wrong".

This would cater for scenarios like https://www.npr.org/2020/10/01/919189045/for-subway-a-ruling-not-so-sweet-irish-court-says-its-bread-isnt-bread or https://www.delish.com/food-news/a49216/things-you-didnt-know-about-pringles/ (Potato Chips made from dried potato vs potato crisp vs potato chip)

* Average person and AI would call it... "bread" or "potato chip" when they are scanning it

* A regulator would call it something different

Otherwise, this somewhat implies that categories can only be derived from regulations

Yes, this is a good point. We really would like to have that. It is not only for this PR, but more generic. To be able to shutdown alerts that are false positives.