taxonomy: Added unknown Croatian ingredients to the taxonomy (part 9)

[benbenben2]

What

• added HR in taxonomy

• seems to me that we could make difference between:

  • filling (garniture in FR) that is inside a preparation (like in doughnut)
  • coating (nappage in FR) that is around a preparation

  I tried to apply changes accordingly

• Also wondering if we could merge preparation and filling? Can we say that preparations are fillings? As it is not clear to me (there is term en:compound as well), I made all preparations as children of en:preparation (although preparation alone does not exists).

Related issue(s) and discussion

• Fixes #-none-

[stephanegigandet]

I think this entry should be "Homogenized", not "Homogenization"

[stephanegigandet]

not related to this change, but "caramel chocolate preparation" should not be under E150 I think

[stephanegigandet]

• seems to me that we could make difference between:

  • filling (garniture in FR) that is inside a preparation (like in doughnut)
  • coating (nappage in FR) that is around a preparation

  I tried to apply changes accordingly

• Also wondering if we could merge preparation and filling? Can we say that preparations are fillings? As it is not clear to me (there is term en:compound as well), I made all preparations as children of en:preparation (although preparation alone does not exists).

Another option woul be to make "preparation", "filling" and "coating" ingredients processing instead. That way we don't have to add entries for basically all fruits, vegetables etc. ("strawberry preparation" and so on) and their translations in all languages.

We will have to be careful when counting % of fruits and vegetables (e.g. "strawberry preparation 50% (strawberries, water, sugar)" should not count as 50% of fruits), but it can be done easily.

[aleene]

In principle I am not in favour to put it in processing, as the idea was that a processing step would not influence the characteristics of an ingredient. For filling this is not the case (strawberry and strawberry filling are only related on one ingredient.
However we are talking here about compounds: an X filling can have multiple ingredients. So the question is more what to do with compounds? As an ingredient this is unimportant. For percentage determination maybe. For analysis purpose we do not care where it is a filling or whatever. It messes up the ingredients taxonomy and could just be replaced by the word compound. It belongs more to a list of "ingredients" to be ignored.

[benbenben2]

Good point from @aleene
Would that be an option to create a new .txt file for compounds?

"strawberry filling" may be related to more than one ingredient. For example, this "apple filling" (Punjenje od jabuke) contains more than apple: https://hr.openfoodfacts.org/product/3856021206184/pita-of-jabuka-s-budget

[stephanegigandet]

Some queries that are useful to see what kind of fillings etc. we have in ingredients lists:

https://uk.openfoodfacts.org/ingredients?filter=filling

https://fr.openfoodfacts.org/ingredients?filter=preparation

[stephanegigandet]

We could have a compound taxonomy that we use for ingredient parsing. Whenever we see "[something] preparation" (or filling, cover etc.), we map it to an ingredient id "preparation". Same for "préparation [something]" in French etc.

Today we have things like:

{
id: "fr:preparation-a-base-de-poivron",
ingredients: [
],
percent: 5.1,
percent_estimate: 5.1,
percent_max: 5.1,
percent_min: 5.1,
text: "préparation à base de poivron"
},

And we would replace them with:

{
id: "en:preparation",
type: "compound",
ingredients: [
],
percent: 5.1,
percent_estimate: 5.1,
percent_max: 5.1,
percent_min: 5.1,
text: "préparation à base de poivron"
},

[codecov-commenter]

Codecov Report

All modified and coverable lines are covered by tests ✅

Comparison is base (8d8be8e) 48.68% compared to head (f872c66) 48.68%.

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #9236   +/-   ##
=======================================
  Coverage   48.68%   48.68%           
=======================================
  Files          65       65           
  Lines       20268    20268           
  Branches     4896     4896           
=======================================
  Hits         9867     9867           
  Misses       9141     9141           
  Partials     1260     1260           

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

[benbenben2]

This is a good idea.

[stephanegigandet]

@benbenben2 I think we can merge as-is (maybe change "homogenization" to "homogenized" in the ingredients processing first).

I'm filing a bug for the proposal to deal with compounds.

[stephanegigandet]

New issue for the parsing of "xyz preparation": #9345

[stephanegigandet]

Thank you!

[sonarcloud]

Kudos, SonarCloud Quality Gate passed!   

0 Bugs
0 Vulnerabilities
0 Security Hotspots
0 Code Smells

No Coverage information
No Duplication information