Help with SED

[ferrad]

If anyone knows SED syntax, can they help? I have a bit of text:

Dev\TechRefactoring\BOPVT]

where there can be any number >=1 occurrences of the \ before \BOPVT. I am trying to remove the last \BOPVT to obtain only:

Dev\TechRefactoring

It's easy if there is only 1 \, but if the number is variable how do get all but the last removed?

[ferrad]

In case you were wondering:

sed -e "s/.*:\\[^\\]*\\\(\([^\\]*\\\)*\).*/\1/g" -e "s/\(.*\)\\$/\1/g"

works

[Squashman]

Any reason you need to use SED?

[ferrad]

I guess I have always used it and know the syntax pretty well. Is findstr similar or not as good?

[ferrad]

Here is what I need to do. I have:

31/03/2014 12:07 <JUNCTION> BOPVT [\??\d:\ReOTech\Dev\TechRefactoring\BOPVT]

I need to strip out everything before the 2nd backslash after d:, and everything after and including the last backslash, ie to get:

Dev\TechRefactoring

How do you do this in findstr?

[Squashman]

Here is what I need to do. I have:

31/03/2014 12:07 <JUNCTION> BOPVT [\??\d:\ReOTech\Dev\TechRefactoring\BOPVT]

I need to strip out everything before the 2nd backslash after d:, and everything after and including the last backslash, ie to get:

Dev\TechRefactoring

How do you do this in findstr?

So this example looks nothing like your first example. So first we need to understand how you are getting this output you have listed above.
I assume you are just using the DIR command. If you are, my question to you is why you are not using any of the switches to list only the bare format of the directory listing?

[ferrad]

Yes, straight from DIR. I need to determine if a directory is a JUNCTION or not. I don't know how to do this with any other DIR options.

If it is, then I need to find out where it is pointing to, eg in the example:

d:\ReOTech\Dev\TechRefactoring\BOPVT

Then, once I have done that I need to strip everything except:

Dev\TechRefactoring

[Squashman]

Code: Select all

 dir /AL /B /s

[Squashman]

Is this actually in your listing?
\??\

I don't see any of my junction points coming out like that at all.

[goldfish]

If anyone knows SED syntax, can they help? I have a bit of text:

Dev\TechRefactoring\BOPVT]

where there can be any number >=1 occurrences of the \ before \BOPVT. I am trying to remove the last \BOPVT to obtain only:

Dev\TechRefactoring

It's easy if there is only 1 \, but if the number is variable how do get all but the last removed?

what you need is gawk, not sed.

Code: Select all

echo "Dev\TechRefactoring\BOPVT" | gawk -F"\\" "$(NF)=\"\";1" OFS="\\"

[foxidrive]

This does it in GnuSED

Code: Select all

sed "s/.*:.[^\]*.\(.*\)\\.*/\1/"

[foxidrive]

what you need is gawk, not sed.

How are you Brian, it's been a while. How come you stopped using your other account?

[Squashman]

what you need is gawk, not sed.

How are you Brian, it's been a while. How come you stopped using your other account?

You beat me to it.

[ferrad]

Code: Select all

 dir /AL /B /s

This doesn't show where they are pointing to.

[ferrad]

Is this actually in your listing?
\??\

I don't see any of my junction points coming out like that at all.

Yes that's how it shows in Windows 7