Replacing new line sign ("\n") when the line after starts with ";"

piotrtorchala

Hi,

I have a 600k rows file, and I have to delete all the new line signs (“/n”) that are followed by a new line beginning with “;”.

Is there anyone that can help with it?

Best,
Piotr

Alan Kilborn

@piotrtorchala

Please show a sample of your data; best to follow the instructions HERE for this type of question.

piotrtorchala

Here is the data sample:

;01;ABC;ABC/n
;01;A/n
BC;ABC/n
;01;ABC;ABC/n

And I have to replace “/n” by " " in all lines not beginning with “;”. In this example: to replace “/n” in row 2 by " ".

PeterJones

@piotrtorchala,

First, thanks for trying to put in into a text box in your second post. For future reference, the syntax is ``` , not ''' – I used moderator power to change it for you this time. (If you have trouble remembering that, then just use the </> button on the toolbar to insert the two lines of that sequence to start and end your example text.)

And now, on to your question:

And I have to replace “/n” by " " in all lines not beginning with “;”. In this example: to replace “/n” in row 2 by " ".

Your subject says Replacing new line sign ("\n") when the line after starts with ";" but your example text says /n . I believe you intended to say \n even in your example text.

You are also not clear whether you have a literal backslash followed by a literal n before the actual newline character(s), or whether you were just trying to indicate that you have a newline sequence there.

The other question becomes, do you really have just a unix-style LF (\n), or do you really have a Windows-style CR LF (\r\n) at the end of each line? Because it will change the results. (You can tell by View > Show Symbol… > Show End of Line, or just by looking near the lower-right of your Notepad++ status bar and see whether it says “Windows (CRLF)” or “Unix (LF)” down there.

If the /n in your example text really means “I have a windows-style end-of-line sequence at the end of each line”, and you want to change any “end-of-line sequence followed by a semicolon” with just a space, then you could use

FIND = \r\n;
REPLACE = \x20 (or type a space in the replacement box)
SEARCH MODE = Regular Expression

If you truly have just a unix-LF-style line ending, then instead use:

FIND = \n;
REPLACE = \x20 (or type a space in the replacement box)
SEARCH MODE = Regular Expression

-—

Useful References

piotrtorchala

Thank you @PeterJones for your support and all the tips.

Answering your question:

@PeterJones said in Replacing new line sign ("\n") when the line after starts with ";":

or do you really have a Windows-style CR LF (\r\n)

I have a Windows-style (CRLF).

Unfortunately, the guidance given is not solving my issue, maybe a description of how I identified the lines that have to be merged with previous lines would be helpful, accompanied by a screenshot example.

To identify all lines that I would like to delete the \n from the line before I used the following search expression:

FIND = ^(?!;)

Example on how it looks in the data base:

N++.PNG

Unfortunately, there is 10k+ of such lines within 600k+ lines file.

PeterJones

@piotrtorchala ,

FIND = ^(?!;)

This is why it’s always a good idea when asking for help to share what you tried: I had misinterpreted your original statement that you wanted to replace the newlines before a semicolon at the start of the line. But it now that you showed that, I think that you really wanted to replace the newlines that don’t have a semicolon at the start of the next line.

You actually had all the pieces that you needed: you just needed to replace the ; in mine with the (?!;) from yours, and it would have worked as you wanted (if I’m understanding)

FIND = \r\n(?!;)
REPLACE = \x20 (or type a space in the replacement box)
SEARCH MODE = Regular Expression

If this still isn’t right, you will need to show both “before” and “after” data.

piotrtorchala

@PeterJones it is absolutely it, it works perfectly! Thank you so much!

I’m a little bit embarrassed that I haven’t connected the dots…

Thanks once again!