Approve, improve or reject question edits with significant code beautifying

Question

Editing questions, apparently a lot of contributors improve readability of code samples, not only correcting whitespace. Their edits make use of their domain knowledge. E.g. upper and lower case in Apex and SOQL like here:

While I agree it improves readability, I tend to dislike such measures. (I'm currently skipping such cases.) It makes it more likely to introduce errors. It is often inconsistent.

In addition to that the proofreader changes the code necessarily without explanation, if she/he changes the code at all. The original author might not understand this change.

On the other hand I've modified cited code myself. E.g. I've added "S" at the beginning of a code example, because due to copy/paste sloppiness it started "tring s = 'something'".

Where do you draw the line?

Answer 1

I've never consulted anyone, nor seen an SFSE "standard," so I just take it on a case-by-case basis.

I, myself, do like those measures, so if I came across your example while reviewing, I would "accept the challenge" and would look at it very closely - in order to make sure that it did not seem to introduce any errors as you mentioned. If it did not, then I would approve it.

But there are other types/categories of edits about which I feel similar to the way you feel about this type/category. When I come across those, I do what you are doing with this one: Skip. That way someone who has been doing this longer or feels more comfortable making a decision can approve or reject.

So I think you are doing the right thing - if you don't know where the line is, skip it, and let others make that call.

The good thing is, the author can always post a comment asking about an edit, or just roll it back if they think it made significant changes to their original intent.