This Regex Generator site works as follows, you provide text examples by highlighting the strings to extract; the strings to not extract are the surrounding text. The Regex Generator finds out a regular expression that complies with provided examples and performs the desired text extraction task.

This is a research project of the Machine Learning Lab http://machinelearning.inginf.units.it/

The engine has also been released opensource here: https://github.com/MaLeLabTs/RegexGenerator

There is an interesting discussion about here: https://news.ycombinator.com/item?id=9946681

I've used this almost everytime I make a regex, simply because it is easier than learning regex all over again..

Edit: actually I also use Regex101.com, but not as often as regex gen ++

I've obtained a regex like this: \d\w\w\w\d++(?=\w)|(?<=\d[^\d]\w)\d++|(?<=\w\w\w\d\w\w\w\w\w\w\w\w)\w

for me this regex is bullshit! totally useless!

In order to obtain a good solution, it is important to provide representative examples; it is also important that such examples provide surrounding text that has not to be extracted. We have observed that poor solutions often appear when users provide little variety in examples and no negative examples at all. I suggest to improve the provided examples ad re-execute the solution search.

I'm not affiliated in any way to this project, but there's no need to be a dick about a poor result you got, mostly because you are clearly showing that you didn't read the instructions given on that same page at all.

The website authors recently published the details behind their tool: http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=7...