Vista - Replace sed?

I'd like to process text in a pipeline in a way similar to sed with
multiple s commands. In particular, I'm trying to replace something
like this:

$P4CMD -p $P4PORT -P $P4PASSWD -u $USERNAME client -o $TEMPLATENAME \
| $SEDCMD -e "/^Client/s/$TEMPLATENAME/$CLIENTNAME/" \
-e "/^Host/s/:..*$/: $OWNERHOST/" \
-e "/^Owner/s/:..*$/: $USERNAME/" \
-e "/^Root/s/:..*$/: $CLIENTROOT/" \
-e "/^Options/s/ clobber/ noclobber/" \
-e "/^View/,$s/$TEMPLATENAME/$CLIENTNAME/" \
| $P4CMD -p $P4PORT -P $P4PASSWD -u $USERNAME client -i

with a PowerShell idiom.

That is, I want to pick out the lines that start with Host, Owner,
etc. through Root, and replace the substring from the first ":" to the
end of the line. For the Options I want to replace just clobber, if
it's there, with noclobber. For each of the lines from the one
starting with View until the end of the input, I want to replace the
template name with the client name.

I suspect that " $_ -replace" is my friend, but how do I restrict it
to matching lines? Especially, is there a way to get the behavior of "/
^View/,$" ?

"wscholine" <wscholine@gmail.com> wrote in message
news:1174428928.198766.221680@y66g2000hsf.googlegroups.com...
> I suspect that " $_ -replace" is my friend, but how do I restrict it
> to matching lines?

Split up the input by lines and then feed individual lines to $_.
Get-Content by default spits out one line at a time down the pipeline so if
you use it then you are set:

gc foo.txt | % { $_ -replace <pattern>, <replacement> }

If you are not reading a file (doesn't look like you are) then you can use
the string split operator: "foo`nbar".split("`n")

> Especially, is there a way to get the behavior of "/
> ^View/,$" ?

I'm not familiar with this sed regex syntax. Can you explain?

--
Keith

On Mar 21, 2:07 am, "Keith Hill" <r_keith_h...@mailhot.nospamIdotcom>
wrote:
> "wscholine" <wschol...@gmail.com> wrote in message
>
> news:1174428928.198766.221680@y66g2000hsf.googlegroups.com...
>
> > I suspect that " $_ -replace" is my friend, but how do I restrict it
> > to matching lines?
>
> Split up the input by lines and then feed individual lines to $_.
> Get-Content by default spits out one line at a time down the pipeline so if
> you use it then you are set:
>
> gc foo.txt | % { $_ -replace <pattern>, <replacement> }
>
> If you are not reading a file (doesn't look like you are) then you can use
> the string split operator: "foo`nbar".split("`n")
>
> > Especially, is there a way to get the behavior of "/
> > ^View/,$" ?
>
> I'm not familiar with this sed regex syntax. Can you explain?
>
> --
> Keith

It's an address range, it matches all of the lines beginning with one
that starts with "View" and continuing to the end of the input. I am
reading from a pipe (old style text pipe) in this case, but am
interested in processing files also.

> gc foo.txt | % { $_ -replace <pattern>, <replacement> }

If I have a bunch of patterns to replace, do I do the " | % { ... } "
multiple times?

"wscholine" <wscholine@gmail.com> wrote in message
news:1174485851.492476.306810@n76g2000hsh.googlegroups.com...
>> gc foo.txt | % { $_ -replace <pattern>, <replacement> }
>
> If I have a bunch of patterns to replace, do I do the " | % { ... } "
> multiple times?

You could create an array of patterns and replacement strings and do
something like this:

gc foo.txt | % { $line = $_; for ($i=0; $i -lt $patterns.Length; i++) {
$line = $line -replace $patterns[i], $replacement[i] }; $line}

Note: I haven't tested this code.

--
Keith