I somewhat disagree that the famous answer is an instance of cargo-culting. Even if not for entertainment it's probably more valuable than literally answering the question posed. Let's say that you now have a method that extracts tags in the manner prescribed... now what? I'd bet the OP was only trying to write that method as a subprocedure for some larger task much more efficiently dealt with another way. In fact if you look at that poster's question history around that time, it's apparent that they were trying to render potentially-malformed XHTML.

regardless of cargo-culting or not, the rebuttal explains why the famous answer is utterly wrong: the problem can be solved by regular expressions and it cannot be solved using an xml parser. While the answer is funny and probably good general advice, it is incorrect.

You can't really argue with a mathematical theorem.

But you can argue against someone misunderstanding what a theorem says and applying it in the wrong context.

The question is about tokenization, not parsing.

They aren’t asking to parse all valid xml. They wanted to parse a small known subset. Regex can be an excellent tool in that situation.

> They aren’t asking to parse all valid xml. They wanted to parse a small known subset. Regex can be an excellent tool in that situation.

Rarely better than using an existing general XML parser and then validating the additional constraints, especially in terms of implementation and maintenance cost.

The question is not about parsing, it is about tokenization.