Since I get files from the interenet, and have no idea of the CRLF format of
received files, I always use \R to detect any form of single or double
characters, and replace with \r\n so that I know that whatever form is in
the text, my replacement will be successful.
^!Replace "\R" >> "\r\n" ARTSW
provides that all forms of CRLF or CR or LF or LFCR are replaced with one
known entitiy. Then, throughout my libraries, I continue to use \R to detect
any form of line end. If I want to remove only excess blank lines, but
retain them between paragraphs, I use
^!Replace "\R\R+" >> "\r\n\r\n" ARTSW
> > Regarding your latest sample data, it appears to be dispensable
> > to search for complete IP-adresses. The only numbers at the start
> > of any line (or paragraph) are IP-adresses.
>Yes, and then the unexpected happens. I've found it pays to make
>searches as specific as reasonable possible. For me this means looking
>for four number groups because this is quick and easy and not checking
>the ranges for 254 or less, as that makes it too complicated to be worth
>it. No hard rules here, but more specific is better.
> Yes, I heartily concur! It is much easier to overkill the criteria and
> messages than debug obscure coding errors.
> > > To actually delete empty lines only you have to write...
> > > ^!If ^$GetLineSize$ > 0 Skip
> > > ^!DeleteLine
> >I wasn't sure if CR LF were counted as two characters or not and too
> >lazy to try. As they're not, you're right and this is an example of more
> >specific being better: A line with just one or two characters may be
> >improbable but it is possible.
> Actually, the CRLF sequence IS two characters which you can verify
> by creating a new doc with two or more blank lines and running this clip:
> ^!Jump Doc_Start
> ^!Select +2
> Only one line is selected. If the CRLF was only one character, 2 lines
> would be selected. If you try selecting just 1 character, nothing is
> Unsing the adage "If it works, don't fix it", I have been using this
> to delete an empty line:
> ^!Select +2
> ^!Replace "^P" >> "" HS
> But with the ideas brought out in this thread and my strong preference
> for non-crossing flow chart lines (a la Nassi-Shneiderman diagrams) and
> to make the code more readable, I will now use this snippet:
> ^!If ^$GetLineSize$ = 0 ^!DeleteLine
> Namaste', Art
> No virus found in this outgoing message.
> Checked by AVG - www.avg.com
> Version: 9.0.709 / Virus Database: 270.14.84/2530 - Release Date: 11/26/09
> [Non-text portions of this message have been removed]
[Non-text portions of this message have been removed]