Remove Special Characters

In short Removing special characters strips symbols and punctuation and leaves the character classes you choose, which is usually letters and digits alone. The reason to do it is almost always that something downstream cannot cope: a file name that must avoid slashes and colons, a database field with a restricted character set, a URL slug, or a legacy system that predates Unicode. Choosing what to keep matters more than the stripping. Removing all punctuation from prose destroys sentence boundaries and makes the text unreadable and unsearchable; removing it from a list of identifiers is exactly right. The non-ASCII question is separate and worth thinking about before you answer it: stripping accented characters turns "Zoë" into "Zo", which is a corrupted name, while converting them to their plain equivalents gives "Zoe", which is a usable one. Use the accent remover for that case rather than this one. Stripping is irreversible, so keep the original — the removed characters cannot be inferred from what is left.