“Updated: URL Activation with Regular Expressions”

Last week, a fairly well known, small software company agreed to use the regular expressions for URL activation that we use in Conversant. This morning they sent me a bug report, as they'd found a fairly common case that wasn't caught by the expressions. Specifically, it was skipping URL's with anchors (the part after the #) beginning with numbers.

I'd actually put that "bug" in there intentionally. Named links (anchors) in HTML are supposed to begin with letters (a-z or A-Z), never numbers... but the most common browser doesn't enforce that (of course), so lots of sites used numbers to name their anchors. Since it's so common, I fixed it. The expressions for finding and activating URLs in plain text have been updated to reflect this common (if incorrect) practice.