Regex to identify invalid URL slugs

Question

I am struggling to find a regex that will identify invalid URL slugs. I often have revisions tasks where the translators add apostrophes or accented letters in the URL slugs and I keep removing them manually. I&rsquo;d like to improve my workflow and spot the erroneous characters automatically. 
 
 A valid URL slug is composed of number, letters and hyphens (or underscore, depending on the chosen variant). Nothing else. Thus the regex for valid URL slugs is fairly simple: ^[a-z0-9]+(?:-[a-z0-9]+)*$ 
 
 Now, where I am struggling is in negating this expression. I tried to use negative lookahead but it did not work. Any hint? 
 
 Philippe

Paul Filkin · Answer

Philippe Noth 
 I think a better approach might be to create a QA check to look for segments where the source was correct but the target was not... like this: 
 
 This way you can use the expression you have validated is always going to identify a correct URL Slug and test against the source and target. If the source matches but the target doesn't then it will report an error... like this for example: 
 
 Would that work?

Paul Filkin · Answer

Philippe Noth 
 My first thought is you should encourage your client to include this setting in the project for their own benefit. 
 My second is even simpler and might be better anyway! Use the Advanced display filter like this: 
 
 And then just click on "Reverse" next to "Apply Filter" which gets you this: 
 
 Just one the ones you need to correct.

Philippe Noth · Answer

Thanks Paul, of course the Reverse option does the trick!

Trados Studio > 5. Regex and XPath

Regex to identify invalid URL slugs