I like it. The interface is clunky, but it's text based, so that's okay. Really, what a great tool. Regex expressions can be a real pain to build.
I think a lot of the comments are being overly critical. To take something as complicated to build as a regex and put a web interface on it is actually an amazing feat.
I would like to see the strings it generates as a single length of text, rather than concatenated variables. Of course I could do this myself, but it'll save countless minutes on all the programmers who use the tool, because ultimately, that's what they are going to do.
For many of the cases I was able to come up with (and express in this horrible, horrible interface), I was able to come up with much more concise and thus readable expressions that match the same thing.
Regexpes are very hard to read anyways, so please let's not use tools that make this even harder.
Can somebody explain to me how this manages to create regexes not confused by false positives? Like if you select a string that must be at the beginning of the input, it doesn't seem to enforce that.
Take a look at the "extract fields" feature in Splunk for an example of a much nicer interface for this kind of thing, although Splunk tends to make really ugly regexes.
Apart from the fact that it has been posted on HN an estimated 1000 times before, I find it strange that it does not even give you the actual RegExp. Instead it shows you the code that will generate the RegExp.
