Participating Frequently

Answered

Need Regex Help

Forum|Forum|6 years ago
July 16, 2019
5 replies
516 views

Question: How do I get a regex wildcard to select only between the two nearest elements instead of going on to further similar elements?

I have a VERY long list of names that need to have the last name separated from the rest by a tab character. In order to accomplish this, I've inserted a placeholder character (#) just before the return of each line, and replaced all spaces in the names with a different placeholder character (@) for my regex to reference. These names are all first, middle initial, then last, having a space on each side of the middle initial. So, with my placeholder characters inserted, the list looks like this (short sample):

Edward@C.@Sellner#

James@J.@Megivern#

J.@Philip@Newell#

I need to replace the @ placeholder just before the last name with a tab. When I try and do that, my regex is selecting everything between the # placeholder and the first @ placeholder, instead of stopping at the seconds @ placeholder just before the last name. Here's what I'm using in my search:

@([\s\S]*?)#

which I would replace with:

(tab)$1#

after which I would convert all placeholders back to their original state.

The problem is that this regex is going to the first @ placeholder instead of stopping at the one just before the last name.

What should I be doing differently here?

Thanks.

regex

This topic has been closed for replies.

Correct answer WolfShade

Cayce, was this of any assistance to you? Just curious.

V/r,

^ _ ^

WolfShadeCorrect answer

Legend

Cayce, was this of any assistance to you? Just curious.

V/r,

^ _ ^

C

CayceAuthor

Participating Frequently

Thanks, WolfShade! Great help!

WolfShade

Legend

You're welcome. And thank you for marking my answer as correct. I do appreciate it.

V/r,

^ _ ^

WolfShade

Legend

Okay, I've got it. But I also just discovered a bug in how Dw implements the replace.

IF YOU USE REPLACE ALL, IT WORKS. If you hit 'Replace' for just a single instance and do it repeatedly, it will skip lines. ???

Anyhoo.. my final solution:

Edward@C.@Sellner#
James@J.@Megivern#
J.@Philip@Newell#
FIND using REGEX:  ([^@]{1,})(@)([^@]{1,})(@)([^#]{1,}#)
REPLACE:  $1$2$3    $5  (This is with four spaces, as a tab cannot be used as part of the replace.)
REPLACE ALL command:
Edward@C.    Sellner#
James@J.    Megivern#
J.@Philip    Newell#
REPLACE one at a time:
Edward@C.    Sellner#
James@J.@Megivern#
J.@Philip    Newell#

HTH,

^ _ ^

WolfShade

Legend

Actually, I did not quite think my solution all the way through. The backreference should not be around the @. There should be two of them around everything else, and manually insert the tab.

I will play with it some more, and post my final suggestion.

V/r,

^ _ ^

WolfShade

Legend

Okay.. I was trying to use lookbehinds, but couldn't get it to work.

If you are trying to do this in Dw, use the following:

(@)[^@]*#[\r\n]

Looking for an @ followed by one or more of anything that isn't @ followed by # with a line feed. The parenthesis around the first @ is a backreference ($1) that can be replaced with the tab that you want.

HTH,

^ _ ^