nav[aria-label="Primary Navigation"] { padding: 0; & ul { list-style: none; width: 100%; display: flex; flex-direction: row; justify-content: start; align-items: start; gap: 30px; padding: 0; & li { margin: 0; } & ul li { list-style: none; } } }

search & replace in DCR

System

Hello,

Has anyone already written a script that parses all nodes of a DCR to replace a given string (I guess it should be a recursive script) ?
A simple substitution is not good enough because if the replacement string looks like one of the xml tag of the DCR it would mess up everything.
It should run on both "iwov" and "xml" DCR formats... so I guess the easiest is to use "TeamSite: Smiley Very Happy

CRnode".

I'm sorry for asking this but I'm very bad with recursive code.

Thank you all,
P-Y.

Find more posts tagged with

Comments

Johnny

Substitution should work... that's exactly what a script would do anyway.
Give an example of something that would not work.

John Cuiuli

Migrateduser

Hello,

Probably I didn't explain very well what I wanted to tell... so look at the below example, it's more clear.

Imagine you have this line in a DCR: <HelloWorld>HelloWorld</HelloWorld>

And you want to replace the value HelloWorld by HelloThere.

The Expected result is: <HelloWorld>HelloThere</HelloWorld> but not <HelloThere>HelloThere</HelloThere> !!!

That's why the DCR should be parsed to replace only the value and not the xml tags !

Johnny

Wouldn't something like this work?

s/(<.*?>)HelloWorld(<.*?>)/$1HelloThere$2/gs

If you have a text editor that supports directory search and replace as well as regex support (I use jedit) then you won't need to write a script if it's a one off.

I didn't give much thought to the regex, so it might have some holes to nut out.

John Cuiuli

jkm

Hi,
Normal substitution wont work properly becasue as he has given in his example there can be couple of tags or data in the VF fields as well which may be corrupted.
I have written such script once upon a time as user wants to rename couple of branches and then all the browse links we need to take care
XML::Twig will be the correct parser becasue we have faced some problems with XML: Smiley Very Happy

OM as it acts too smart sometime and DCR doesnt except tht.

Your script will depend upon the type of content,substitutions and amount of content you want to replace.

cheers !!!

nipper

Having written a lot of scripts lik ethis, I have found it easiest to use a regex as was
previously suggested. However I would strongly suggest you make a copy of templatedata
(or create a new workarea for testing) play with it on that and run. I also suggest that you
use grep (or search) to get a count of how many should be changed & verify that is the
number that you do change. There are some advantages to being paranoid.

Andy

Migrateduser

Check out this thread.

http://devnet.interwoven.com/forums/cgi-bin/showflat.pl?Cat=&Board=PRODUCTS_TEAMSITE&Number=70954&page=0&view=expanded&sb=5&o=0&part=1&vc=1

Windows 2000
Teamsite 6.5

Lal

Please check this attached file as an example.. It should help you to give a try..
Lal

<item name="Uploaded Files">
<value>
<item name="Document ID">
<value>1</value>
</item>

Here i am trying to replace the value of Document id with another values.

But if it is only a singe node, I guess the regex replace method suggested by
John Cuiuli is only required.

Both the method I tried and is working perfectly fine in our module..

Edited by Lal on 11/23/05 08:45 AM (server time).

rodigz

Substitution should work... that's exactly what a script would do anyway.
Give an example of something that would not work.

John Cuiuli

sample code.. it does not write the other xml..only writing the replace tag..
Any idea?

opendir(HOMEDIR, "$test_Dir") || die ("Unable to open directory");
while ($filename = readdir(HOMEDIR)) {
push(@testDir_files, $filename);
}
closedir(HOMEDIR);

foreach $testDir_file(@testDir_files) {

$test_Path = $test_Dir."/".$testDir_file;
#print "$test_Path";
print $test_Path ."\n\n\n";

open (FILE,"$test_Path")|| die ("Unable to open directory");
my @filecontent=; #Read entire file contents into variable
close(FILE); #Close file

open(OUTF,">$test_Path");
foreach $line (@filecontent) {
#print "$line";
if ($line =~s/find/replace/g) {
print "Searching in $line";
print OUTF "$line"; #Write modified variable into file
}
}
close(OUTF);

nipper

You did not say what was failing. I assume it is opening the files correctly.

If it is failing on the regex you may try this:



open(IN, "<$test_Path") || die " '$test_Path' ($!)";
{
        local $/ = undef;
        $text= <IN>;
}



print "Searching in $text";
if ($text =~s/find/replace/g) {

print "Replaced in $text \n";
print OUTF "$text"; #Write modified variable into file
}

rodigz

failing in this part

open(OUTF,">$test_Path");
foreach $line (@filecontent) {
#print "$line";
if ($line =~s/find/replace/g) {
print "Searching in $line";
print OUTF "$line"; #Write modified variable into file
}
}
close(OUTF);

rodigz

when writing the xml it does not write the whole xml it just write the result of the replace.

Adam Stoller

when writing the xml it does not write the whole xml it just write the result of the replace.

Well - that's the way it was written - if you want to write out all the XML (replaced or otherwise) remove the

if(...){...}

wrapping the code - so it looks more like:

...
open(OUTF,">$test_Path") || die " '$test_Path' ($!)";
foreach my $line (@filecontent) {
    $line =~s/find/replace/g;
    print OUTF $line; 
}
close(OUTF);

nipper

when writing the xml it does not write the whole xml it just write the result of the replace.

That happens because you only write for a line that has done the substitution:

if ($line =~s/find/replace/g) {
print OUTF "$line"; #Write modified variable into file

rodigz

it does not write the replace in the file but i can it shows in $line when I print it. hmm.
open(OUTF, "<$test_Path") || die " '$test_Path' ($!)";
{
local $/ = undef;
$line = "";
}
foreach my $line (@filecontent) {
$line =~s/find/replace/g;
print OUTF $line; #Write modified variable into file
print $line;

}
close(OUTF); #Close file

nipper

Change this

open(OUTF, "<$test_Path") || die " '$test_Path' ($!)"; # Open for read

to this

open(OUTF, ">$test_Path") || die " '$test_Path' ($!)"; # Open for write

You should have 2 opens, one for read & one for write.

rodigz

now getting error! (Invalid argument)

nipper

post your complete perl script

rodigz

see attached

nipper

Change to this:


open (FILE,"$test_Path")|| die ("Unable to open directory"); 
my @filecontent=; #Read entire file contents into variable
close(FILE); #Close file


open(OUTF, ">$test_Path") || die " '$test_Path' ($!)"; # Open for write

This was to slurp the complete file into a variable (you can run a single regex on it rather than looping), much faster.




open (FILE,"$test_Path")|| die ("Unable to open directory"); 
{
        local $/ = undef;
        $line = <FILE>;
}

$line =~ s/find/replace/g
open(OUTF, ">$test_Path") || die " '$test_Path' ($!)"; # Open for write
print OUTF $line;
close (OUTF);

also I fixed my previous post, some of my code was stripped out

Adam Stoller

see attached

While we're rewriting your code for you - for free - we might as well do the whole thing.

#!/usr/bin/perl
use strict;
use warnings;

my $test_Dir = '/default/main/Marketing/WORKAREA/Common/templatedata/Marketing/subcategory/data/Test';

opendir(HOMEDIR, "$test_Dir") ||  die ("Unable to open directory");
#----------------------------------------------------------------------
# Assumes (as original code appeared to) that there are no sub-directories
# within data directory
#----------------------------------------------------------------------
my @testDir_files = grep { ! -d "$test_Dir/$_" } readdir(HOMEDIR);
closedir(HOMEDIR);

foreach my $testDir_file (@testDir_files) {
my $test_Path = "$test_Dir/$testDir_file";

    open (FILE, "<$test_Path")|| die ("Unable to open file for reading");
my $filecontent = do { local $/; <FILE> };
    close(FILE); # Close file

my $count = ($filecontent =~ s/rebates/replace/gsm) || 0;
print "$test_Path: $count\n";
    open(OUTF, ">$test_Path") || die " '$test_Path' ($!)"; # Open for write
print OUTF $filecontent;
    close(OUTF); # Close file
}

Please pick up some books and learn how to write Perl code - you shouldn't need our help to write something like this.

rodigz

thanks..if the data has subdirectories?

nipper

Boy you are demanding.

Try replacing this:


opendir(HOMEDIR, "$test_Dir") ||  die ("Unable to open directory");
#----------------------------------------------------------------------
# Assumes (as original code appeared to) that there are no sub-directories
# within data directory
#----------------------------------------------------------------------
my @testDir_files = grep { ! -d "$test_Dir/$_" } readdir(HOMEDIR);
closedir(HOMEDIR);

with something like



use File::Find;
find(\&wanted, $test_Dir);



at the end of your script:

sub wanted {-f && push(@testDir_Files, $File::Find::name);}