remove all instances of string from comma seperated value

[iamtheky]

desired output (removing all instances of $name from $string)

 

du2: ad,43|gshy: 3d3,adf

 

You want more than just $name then, as 3 of the 4 pipes are gone as well?  so removes name and trailing separators first, and then the second pass cleans up the end where you want the preceding separator and name removed.

$string = "gsh: kae,as|du2: ad,43|gsh: kae,as|gshy: 3d3,adf|gsh: kae,as"
$name = "gsh: kae,as"
$sep="|"

$stripped = stringreplace($string , $name & $sep, "")
$stripped = stringreplace($stripped , $sep & $name , "")


msgbox(0, '' , $stripped)

Edited December 30, 2014 by boththose

[gcue]

 

If the separator is not the same, you should change it in the regex, or use a variable for this.

; $name = "gsh"
; $string = "gsh,as,43,abgsh,adf,gsh,gsh"
; $separator = ","

$name = "gsh: kae,as"
$string = "gsh: kae,as|du2: ad,43|gsh: kae,as|gshy: 3d3,adf|gsh: kae,as"
$separator = "|"

$final = StringRegExpReplace($string, "(?:(\Q" & $separator & "\E)(\Q" & $name & "\E)(?=(?1)|\Z))|(?:\A(?2)(?1)|\Z)", "")
MsgBox(0, "", $final)

works great!! thank you very much

[gcue]

 

You want more than just $name then, as 3 of the 4 pipes are gone as well?  so removes name and trailing separators first, and then the second pass cleans up the end where you want the preceding separator and name removed.

$string = "gsh: kae,as|du2: ad,43|gsh: kae,as|gshy: 3d3,adf|gsh: kae,as"
$name = "gsh: kae,as"
$sep="|"

$stripped = stringreplace($string , $name & $sep, "")
$stripped = stringreplace($stripped , $sep & $name , "")


msgbox(0, '' , $stripped)

only problem with that is if the name is slightly different at the beginning or end

[mikell]

BrewManNH extended 

$name = "gsh: kae,as"
$string = "gsh: kae,as|du2: ad,43|gsh: kae,as|gshy: 3d3,adf|gsh: kae,as"
$separator = "|"

$final = StringRegExpReplace($string, _ 
        '(\Q' & $name & $separator & '\E?)|(\Q' & $separator & '\E?\Q' & $name & '\E)', "")
MsgBox(0, "", $final)

Edited December 30, 2014 by mikell

[kylomas]

@mikell - That pattern fails on...

$name = "gsh: kae,as"
$string = "gsh: kae,as|du2: ad,43|gsh: kae,asx|gshy: 3d3,adf|gsh: kae,as"
$separator = "|"

$final = StringRegExpReplace($string, _
        '(\Q' & $name & $separator & '\E?)|(\Q' & $separator & '\E?\Q' & $name & '\E)', "")
MsgBox(0, "", $final)

@gcue - You're requirements are a long way from where we started, please define the complete strings that you are working with...

Edited December 30, 2014 by kylomas

[JLogan3o13]

@gcue - You're requirements are a long way from where we started, please define the complete strings that you are working with...

 

Amen, you have us floundering about here as your requirements change. Help us help you and provide exactly what you're looking for please!

[mikell]

@gcue - You're requirements are a long way from where we started, please define the complete strings that you are working with...

 

Sooo obvious... it's the usual trouble in most questions concerning regex 

Edit

BTW jguinch's regex in post #20 still works and seems the best (the only..) way

Edited December 30, 2014 by mikell

[kylomas]

BTW jguinch's regex in post #20 still works and seems the best (the only..) way

 

Yes, and now I'm going to spend the next couple hours trying to figure it out...

[iamtheky]

it may be the best, but it is not the parenthetical only.  #21 absolutely returns the desired output from the given input (that the input has undisclosed variations is not the scripts fault).

Edited December 30, 2014 by boththose

[mikell]

Yes, but it fails on kylomas' variation from #25

Anyway when it becomes really tricky with hazardous and variable requirements then the most reliable way remains StringSplit + ArrayDelete + ArrayToString  ^^

[iamtheky]

ahhh, i see now, variations in the string not the name.  And for sure if there were any amount of those I would totally go the stringsplit route.  Or go learn RegEx like JG.

[kylomas]

Or go learn RegEx like JG.

 

*sigh*  been try'in, been fail'in...

[mikell]

Keep fight'in 

[jguinch]

@kylomas : what is so hard to understand in the pattern ? Do you need explanations ?

Edit : here is a good way to understand it (I think) : https://regex101.com/r/iR0dK7/1

Edited December 30, 2014 by jguinch