OK, so here is a story from what happened at work a couple of weeks ago.
A customer was having difficulty with importing some data into a program we supply - he was getting an error and the import failed.
So I asked him to email the 'class' to me so I could try in-house.
When I imported it, the import was successful - but that was no surprise since the customer is not using the newest version of the software, and I found that this was the result that had been fixed.
Anyway, I saved it as a new class of a different type: the original data was in a format that is typically used for polygons (such as police beat boundaries) but the data was 'point' data - in this case, the locations of bridges.
So I emailed the data back to him in this new format (CSV) with a covering email, only to have a bounce from their corporate mail marshal telling me in no uncertain terms that profanity was not allowed in email communication.
I wrote nothing that could even be vaguely construed as being profane, rude, or anything of the sort: to do so would be highly unprofessional to say the least.
So I opened up the CSV I had tried to send and did a search for a possible Scunthorpe Problem
... I'm glad that nobody appeared behind me as I did Control-F searches for rude words!
Anyway. Finally found it. Pissmire Road Bridge. Could hardly believe my eyes.
In this day and age, surely regular expressions could get rid of this problem once and for all? All you have to do is check for letters before or after the perceived profanity, and if there are some then it's part of a valid word. Surely there is no need for such broken behaviour any more?
I've heard cases where the word "assassinate" has been changed by word replacement to "buttbuttinate" and I doubt the chemical arsenic would be permitted either.....