fix(numbers): correctly determine malformed decimals

[olunusib]

Fixes #928

Problem:

The current code was primarily designed for numbers like 3,400.00, where the r-stripped value (3,400.) matches the expected proper (3,400) + . (3,400.) format. It doesn't work for numbers like 3,400.60 since 3,400.6 (the value after r-stripping) is incorrectly perceived as not equal to the expected format, which is the proper (3,400.6) + . (3,400.6.), thereby leading to its erroneous detection as improperly formatted.

Solution:

Added a more robust method for stripping trailing zeros, specifically targeting only those zeros that appear after the last non-zero digit following the decimal symbol.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Comparison is base (aca7663) 91.43% compared to head (e89a6c9) 91.44%.

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #1042      +/-   ##
==========================================
+ Coverage   91.43%   91.44%   +0.01%     
==========================================
  Files          26       26              
  Lines        4436     4444       +8     
==========================================
+ Hits         4056     4064       +8     
  Misses        380      380              

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

[akx]

Thank you! I took the liberty of rebasing this, cleaning up the logic a bit and adding a few more test cases for good measure.