cmontr:
Let me see if I can rephrase your question in a way that we can all understand unambiguously. Your ldap database contains records named 'cquestion', which have some values which may or may not be unique. You want the corresponding 'dn' records to be reported, but only for those 'cquestion' records which are unique. You do not want any of the 'cquestion' records reported. Does this accurately express your requirements?

Can someone who understands the ldap database confirm that the records in the database are organized in groups starting with a 'dn' record, and ending with a 'cquestion' record? The snippet of text given in the example would seem to indicate this, and if it is the case, it should simplify the problem.

--- rod.

Hi everyone...first thank you very much for helping out. I got this running but need to iterate now:

#
# LQcmontr.awk
#
BEGIN{ headingFound = 0; }
/^cquestion: What's your name/ { headingFound = 1; }
/^dn:/ {
if( headingFound == 1 ){
print $2;
headingFound = 0;
}
}

# awk -f LQcmontr.awk bigFile.txt


#######################

Ok this runs ok. Now I want to print this in another format - in another words I want to give an input as cquestion when I run the script. When I ran the script is should ask me something like : "Enter your cquestion: " And when I enter, it should do the printing for its dn value which is working. I appreciate if anyone is familiar with it and can help me. Thanks million again.

add this in the BEGIN
then use choice in your code.

In awk, how does one match against the value of a variable?
If I do
awk will not interpolate the variable 'choice', but will simply take it as the literal word. Perl makes this kind of trick easy.
--- rod.

Probably because Perl takes the best of sed, c, and awk, lump them together into a powerful language to make it easier for people to program with..

Are you running on a Unix or BSD machine. Maybe the -B (as in before) option doesn't exist in your version of grep. My example doesn't place the dn: after cquestion:. To do that you could use sed, using the hold register. This is where you would be better off writing a sed script and running the command like
sed -f dn_cquestion.sed ldapjunk.txt
Where dn_cquestion is a sed script. Also if there are dn's without corresponding cquestion's then you may need to branch out when another dn is read.

The exceptions are what can make using sed very complicated. Often when you have patterns over a range of files, you need to use the advanced options and build up lines in the input register or use the hold register.

regarding to the same issue, I need to be able to print the whol record for the cquestion. Could anyone help me how ? Thanks very much.

in another words, how can I print this from a "cquestion: what is xyz? " -- please let me know if anyone can help...thanks in advance...

and for every "what is xyz" print this:

dn: uid=poi.1111,ou=abc,o=cc.com
authpassword;orclcommonpwd: {SHA}adhnbckhsypzx=
businesscategory: abc
createtimestamp: 20050709198316x
creatorsname: cn=bulkload
modifiersname: cn=bulkload
modifytimestamp: 20050709198316x
objectclass: top
objectclass: cuser
cquestion: xyf88561
dn: uid=poi.2222,ou=def,o=dd.com
authpassword;orclcommonpwd: {SHA}adhnbckhsypzx=
businesscategory: abc
createtimestamp: 20050709198316x
creatorsname: cn=bulkload
modifiersname: cn=bulkload
modifytimestamp: 20050709198316x
objectclass: top
objectclass: cuser
cquestion: what is xyz?

I've read the whole thread and still don't fully comprehend what you're
after ... does this help at all?


I assume that your file is proper LDIF?

Cheers,
Tink

========================

Hi - What I am trying to print out every field for the particular question in the file...earlier I was printing only dn value for the asked question in this file...I tried your script but did not work. Could you please take a look again? thank you very much. This was the old code to print dn from the question in the txt file.


#!/bin/sh
# ctest.awk
#
BEGIN{ headingFound = 0; }
/^cquestion:/ { headingFound = 1; }
/^dn:/ {
if( headingFound == 1 ){
print $2 $3;
headingFound = 0;
}
}

# awk -f test2.awk ab.txt > test2.txt

What did it give you? And where did you create the file? Is it using
Unix or DOS line-ends? If it's from DOS my thing won't work - convert
it to Unix first ...


Cheers,
Tink

================

It is in unix - solaris os.

i put that in a script

#!/usr/bin/awk -f

awk 'BEGIN{RS=ORS="\n\n"; FS=OFS="\n"} /cquestion: What is your name?/ {print $0}'

and ran as:

awk -f test.awk a.txt > test.txt where a.txt contains the data file and test.txt is to be the extracted file

error I got was

awk: syntax error near line 3
awk: bailing out near line 3


I appreciate for the help...

OK ... use the GNU awk instead of the Solaris default one - if it's not installed
(yet) you can grab if from sun's download site. The thing that comes with Solaris
by default is cr*p.


Cheers,
Tink

==========

I am sorry but I only have user account on the server as oracle

is there any other way to resolve this?

Thanks much for the help

thanks for the time....I got it fixed...here it worked as ...

BEGIN { RS = ""; FS = "\n"; IGNORECASE=1 }

/cquestion: What?{ print $0, "\n" }