Hello,

I use UltraEdit often to parse through very large log files, and one task I'm constantly trying to streamline is the removal of uninteresting/repetitive lines.

Right now, I'm using the "Delete Lines Containing String" macro from the downloads section, but have found that it is often so slow that it's faster for me to simply copy the text, and replace all with the following RegEx: ^.*<PASTE>.*$

I end up doing this over and over again, and would like to streamline the Replace All process with a macro, using the currently selected text to delete all lines containing that text. However, when I try to use ^s in the macro along with the RegEx above, it never finds anything. I'm not sure what I'm missing. Current macro contents are below.

Of course, finding a way to speed up the script, which already works but is quite slow, would also fit my needs.
Thanks in advance for any tips.

^s and ^c are supported only when running a non regular expression replace or an UltraEdit regular expression replace. The UNIX and the Perl regexp engines don't support that UltraEdit special.

So you can use only

InsertMode
ColumnModeOff
HexOff
UltraEditReOn
Find RegExp AllFiles "%*^s*$"
Replace All ""

or to delete the (DOS/UNIX) lines completely

InsertMode
ColumnModeOff
HexOff
UltraEditReOn
Find RegExp AllFiles "%*^s*^r++^n"
Replace All ""

But please note that the last line of a file must have also a line termination to get it also deleted with that macro. And important to know is also that the selected string is interpreted as UltraEdit regular expression string. So if the selected string contains any UltraEdit regular expression character, the macro works different as you might expect. As long as you use the macro just for deleting lines containing a word(s), it will work for you.

Thanks so much yet again Mofi.

It's still not working for me for some reason.

I'm using the first code block you gave me:
InsertMode
ColumnModeOff
HexOff
UltraEditReOn
Find RegExp AllFiles "%*^s*$"
Replace All ""

But when I select the word Microsoft, it's still not removing any lines. If I change the code to include
Find RegExp AllFiles "Microsoft"
I get exactly what I expect the behavior to be when selecting and using ^s.

When I run the macro, the Cancel Operation box pops up very briefly, but nothing is ever removed.

What am I missing?

Thank you!

That difference in behavior is interesting. I have run some tests and could see that indeed the line with the selected text is not cleared with %*^s*$ while it is cleared with %?++^s?++$ and moving the caret to next line not cleared.

* is non-greedy while ?++ is greedy. But there should be no difference in this case. It should not matter if first * matches everything up to first occurrence of selected text and second * matches everything to end of line while first ?++ matches everything up to last occurrence of selected text and second ?++ matches everything to end of line.

I will report this to IDM by email. The IDM developers must evaluate it and decide if this difference in behavior is a bug or by design.

Edit on 2011-10-13: The issue with %*^s*$ not deleting also the line with the selected text was fixed with UE v17.30.0.1002 (first public release of UE v17.30).