I recommend consulting my other remarks within this topic, but on this specific point: The title's inconsistency with the body text, i.e. "open tags" vs "openi

I recommend consulting my other remarks within this topic, but on this specific point: The title's inconsistency with the body text, i.e. "open tags" vs "opening tags", is especially and immediately notable because they are (in context) grammatically interchangeable but have dissimilar meanings. This is immediately suggestive of (but not diagnostic of) a writer revealing context and then switching to detail. As a longtime reader of requirements documents and S.O. questions, a mental flag to check the intended meaning of both is already raised at this point. The reference to XHTML is ambiguous, since it speaks to working around XHTML that is already present, rather than defining an input or output document, which means this is something thinking at the character level rather than in terms of DOM. This impression is verified by the comparison pairs in the question body, leaving the essential context question of what are the desired inputs and outputs?" glaringly unanswered. The early construct, "I need to ..." is a secondary flag that immediately reinforces the likelihood of a critical gap in context, since it does not explain why the need has arisen. To anyone familiar with the structure & semantics of the HTMLs, the omission of any mention of tag closing, having discussed "opening tags", boosts the sense that something relevant is missing from the question. The obviously flawed regular expression then attracts a "probable novice" qualifier, which is amplified by both the closing "what do you think?" and the original's emoticon. At this point, we're about ten seconds into the read-through and there's already a ton of labels pointing to a beginner who may be conflating dissimilar concepts and (through inexperience) choosing the wrong tool for the task, whatever that may be. Another few seconds to review comments, at which point it appears that author does understand the distinction between an "open tag" and an "opening tag", and this reweights the XHTML reference in the title toward being a need to output strict XHTML. Given their apparent beginner standing it's likely this is either their first S.O. post, or one of slightly too many, and a check in their profile for proximate questions immediately reveals the latter case. References to balancing of "dangling tags" abound, collapsing all preceding problem variants and likelihoods into the only focused understanding that fits: it's a PHP novice, bright but inexperienced, who wants to normalise a nonconforming document into XHTML by balancing the tags, and that the mostly likely use case is injecting an existing HTML fragment into a RSS XML feed. Elapsed time ca.90 seconds. Another minute to scroll through surrounding Q&A materials, to allow alternative options to appear (they don't). A few seconds for a chuckle at that old friend of an amusing answer, and recognise that the advice within is sound: what they really need is, indeed, a parser. Then, a pause to consider a contrasting notion, that what they're trying to do is write a parser. Consider evidence that author has taken at least one step towards inventing stacks on their own from first principles; prospect loses to existing top answer because the experience level required to write parsers cannot, ultimately, be acquired from a single S.O. question. Last but not least; the final boss: convincing the chorus of Hacker News to accept this interpretation. That takes much longer.