Replacing a string in between known strings over mutiple lines in Powershell - regex

Question

I'm entirely new to powershell and regex stuff, but i need help to do the following:
I need to replace 'VALUE' in a file given that i know the strings before and after it. It runs over multiple lines as well. eg:
<knownvalue1>
<knownvalue2>VALUE<knownvalue3>
knownvalue 2 and 3 are not unique so i need to include the knownvalue1 as the kind of 'identifier'
Also, To keep in a similar format as what has been previously done it needs to be along the lines of:
(gc $filename)-replace "(SEARCHPATTERN)","(REPLACEVAL)" | sc $filename
if you can't do it this way, then alternative ways will be okay.
I'm going crazy over this one so any help would be great.
Thanks.

Answer 1

Well, When you're importing the file and reading line for line you could split the line up like this:
Bassicly first we split it up on the delimiter > After doing so the string will be split up in these parts: <knownvalue2 and VALUE<knownvalue3 and (empty)
After that it would need to be split again on > however now that the string is in parts u need to split it on the parts. which makes it tricky.. and at the same time this is pretty dirty, atleast in my opinion..
$value = #()
$value += "<UniqueValue1>"
$value += "<knownvalue2>This Belongs to UniqueValue1<knownvalue3>"
$value += "</UniqueValue1>"
$value += "<UniqueValue2>"
$value += "<knownvalue2>This Belongs to UniqueValue2<knownvalue3>"
$value += "</UniqueValue2>"
[int]$count = 0
foreach($line in $value)
{
if($line.Equals("<UniqueValue1>"))
{
$parts = $value[($count + 1)].Split('>');
"Should be This Belongs to UnqiueValue1<knownvalue3"
$parts[1]
""
"Second Split"
$val = $parts[1].Split('<')
$val[0] + " #should be This belongs to UniqueValue1"
$val[1] + " #should be knownvalue3"
""
"Should be <knownvalue2"
$parts[0]
}
if($line.Equals("<UniqueValue2>"))
{
$parts = $value[($count + 1)].Split('>');
"Should be This belongs to UniqueValue2<knownvalue3"
$parts[1]
""
"Second Split"
$val = $parts[1].Split('<')
$val[0] + " #should be This Belongs to UniqueValue2"
$val[1] + " #should be knownvalue3"
""
"Should be <knownvalue2"
$parts[0]
}
"finished Reading line: $count"
$count++
}

Answer 2

Read the file as one string (gc outputs an array which is a very poor choice) and use a simple regex:
[IO.File]::ReadAllText($filename) -replace `
'(<knownvalue1>[\s\r\n]*<knownvalue2>)VALUE(<knownvalue3>)',
'$1(REPLACEVAL)$2'
Notes:
In PS3.0+ you can use (gc $filename -raw) instead of [IO.File]::ReadAllText($filename)
In case <knownvalue1>SKIP SOME OTHER TAGS<knownvalue2> to catch the SKIP section you can use <knownvalue1>(?s).*?<knownvalue2> to match any characters including newline (CR/LF).