Find non-standard URL

We have used the non-standard URLs in our book, we are trying to catch with grep, but can't backward from the center (/PDF/), kindly advice on this.

Here is the screenshots.

Trying ...

Need this...