SheetCompareData normalization
Transforming values into a consistent standard form, such as unifying date formats or trimming whitespace, so records can be compared or joined reliably.
Normalization, in the data-cleaning sense, means putting values into one canonical form so equivalent things compare as equal. That covers trimming whitespace, standardizing case, unifying date and number formats, and collapsing variants like $1,000 and 1000 to the same underlying value.
Without it, comparisons and joins break on cosmetic differences. 2024-01-05 and 01/05/2024 are the same date but won't match as text; yes and TRUE may mean the same thing to a human and nothing to a parser. Applying normalization rules before a diff, ignore case, trim whitespace, treat formatted and raw numbers as equal, is what separates real changes from formatting noise. (Note: 'normalization' also has a distinct meaning in relational database design, splitting data to reduce redundancy.)