fix: solve quantity false positives issue #2037

[CharlesNepote]

Description:

Change regexp to avoid many false positives.

Related issues and discussion: #2037

[VaiTon]

@CharlesNepote could you provide some cases that the old regex matched and this one does not? Thank you! :-)

[stephanegigandet]

The purpose of the more complex regexp was to flag things like this : "500 ge"

[stephanegigandet]

I think @VaiTon 's suggestion is a very good one: we should add test cases (both positives like "500 ge" and negatives like "1 litre") so that we don't break this later. There is already a test file : t/sitequality.t

[CharlesNepote]

Yes, you're both right. I'm not very comfortable with test cases but I know it's important, so I'll do it.

[teolemon]

rebased. @CharlesNepote can you write the test so we can merge

[teolemon]

Can we close @CharlesNepote ?

[codecov-commenter]

Codecov Report

Merging #2038 (d33f3d4) into main (e0f97b6) will increase coverage by 0.08%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##             main    #2038      +/-   ##
==========================================
+ Coverage   48.16%   48.25%   +0.08%     
==========================================
  Files         109      109              
  Lines       21015    21046      +31     
  Branches     4739     4739              
==========================================
+ Hits        10122    10155      +33     
+ Misses       9618     9617       -1     
+ Partials     1275     1274       -1     

Impacted Files	Coverage Δ
lib/ProductOpener/DataQualityFood.pm	59.94% <100.00%> (+0.53%)	⬆️
tests/unit/dataqualityfood.t	85.05% <100.00%> (+2.84%)	⬆️

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[sonarqubecloud]

Kudos, SonarCloud Quality Gate passed!   

0 Bugs
0 Vulnerabilities
0 Security Hotspots
0 Code Smells

No Coverage information
No Duplication information

[benbenben2]

@CharlesNepote , @stephanegigandet , @VaiTon , @teolemon

Tests have been added.
The regex (initially /(?:.*e$)|(?:[0-9]+\s*[kmc]?[gl]?\s*e)/i) have been updated.
Because it seems to be covered already by the second part of the regex, the first part of the regex (/(?:.*e$)|) has been removed.
Because this "e" letter is always (? could not find exception) preceded by "g" or "kg" or "l" or "ml", we can make [gl] as mandatory in the regex (by removing the question mark which makes it optional). This leads to: /(?:[0-9]+\s*[kmc]?[gl]\s*e)/i.

False positives such as "litre" or "verre" (ending by letter e) are not catch anymore with this regex.

[stephanegigandet]

Perfect, thank you!