Community

Can I use Regex to delete all lines in a text file that end with a specific string, in this case -a---(plus the EOL character)?

That's 'minus,a,minus,minus,minus,EOL'

Search | Replace... (<F8>):

Find what: .*-a---\n
Replace with: [nothing]

[X] Regular expression

Replace All

Thanks a bunch, appreciate the speedy reply. I was plainly over-thinking it, as it looks so obvious now!

In fact, I often do this sort of thing in two steps:

1. Mark all the lines to be deleted:

Search | Find... (<F5>):

Find what: -a---$

[X] Regular expression

Mark All

2. Delete the marked lines:

Edit | Delete | Bookmarked Lines

This gives me a better chance of avoiding doing something stupid (and the regex is slightly simpler).

As per intention both approaches differ in that the former will not delete a last line in the file and the latter does.

They only differ if the last line is not terminated with a newline (with TextPad 8.15.1 and 9.0.1).

Yes, that's the point. If the requirement includes CR or LF then your approach would still delete a potential last line of the file that wouldn't have a CR or LF, which would be wrong.

Pedantically, yes. But the omission of a newline at the end of a text file is usually an error.

That could be fixed by adding a positive lookahead for the line break ...

ben_josephs wrote:the omission of a newline at the end of a text file is usually an error

Only in POSIX and Unix environments. However, TextPad usually lives in... well... none of both.

No, I've learnt otherwise. Always implying a newline at the end is an important detail/requirement and is different from dealing with reality. Just because it should be like that does by far not mean it is done like that by everyone. And Windows is a whole different environment on top of that.

Thank you gentlemen! You've added precious time back to my days!