Hi

I tried using the above command and all i get is K:\test>perl -pi -e 's/searchterm/replaceterm/' *.html
Can't open *.html: Invalid argument.

can you help

Thanks
ak
ajay.jindle@gmail.com

'Can't open *.html: Invalid argument'
..
means that it did not find any *.html files. Maybe you're in the wrong folder?

Hi Thanks but i wish it was as easy as that unfortunatly i was in the right folder. so if you have antother tip or suggestion would be much appericated

Hi
The above works fine if i supply a specific file name but not with a *.txt.

hereis what i did
Directory of C:\Documents and Settings\aajind\Desktop\battest

21/02/2006 12:28 <DIR> .
21/02/2006 12:28 <DIR> ..
14/02/2006 10:52 7,957 help.txt
16/02/2006 15:51 346 mynewsam.bat
21/02/2006 12:28 53 sample.txt
21/02/2006 12:28 53 sample.txt.bak
10/02/2006 13:22 33 sample2.txt
14/02/2006 08:23 38 sample3.txt
14/02/2006 10:33 692 testaj.bat
14/02/2006 10:33 692 testaj.vbs
8 File(s) 9,864 bytes
2 Dir(s) 32,098,480,128 bytes free

C:\Documents and Settings\aajind\Desktop\battest>perl -pi -i.bak -e 's/ma/sa/' *.txt
Can't open *.txt: Invalid argument.

C:\Documents and Settings\aajind\Desktop\battest>perl -pi -i.bak -e 's/ma/sa/' sample.txt

C:\Documents and Settings\aajind\Desktop\battest>

ajay.jindle@gmail.com if you can help
Thanks
aj

I have the same problem 'Can't open *.html: Invalid argument' . The script works in a directory under Linux, but in the same directory ('folder') under WinXP the script throws up the error message. In my Googling I've come across a suggestion that -i doesn't work, but not a clear explanation of why and what the workarounds are.

--
Zym

Found it.

Try perl -pi.bak -e 'BEGIN{@ARGV=<*.html>} s/searchterm/replaceterm/g'

--
Zym

Can anyone interested to answer me. I have a html file with some text written on it. Can i search for it and retrive more of the details from that html file?

sample perl code ---

--replace.pl---

#!/usr/bin/perl

if(-e $ARGV[0])
{
    $cmd = 'copy $ARGV[0] $ARGV[0].bak';
    `$cmd`;
}
else
{
    print 'File does not exist.\n';
    exit;
}

open(INPUT,'$ARGV[0].bak') or die 'Cannot open file: $!\n';
open(OUTPUT,'>$ARGV[0]');

while(<INPUT>){
    $_ =~ s/$ARGV[1]/$ARGV[2]/g;
    print $_;
    print OUTPUT $_;
}

close INPUT;
close OUTPUT;



---

call like

perl replace.pl file.html originaltext newtext


----

use at your own risk..

very useful

How to modify the given command if my Input String (Finding term) has / ..?

I tried for a set of text files it did not work.But for a single file it will work
Any suggestions??

-Tina

The way that Windows expands file names is different from UNIX, which is why you can't use *.html or *.txt with the in-place search and replace. Just use a command-line for loop:

for %x in (*.html) do perl -pi -i.bak 's/searchterm/replaceterm/' %x

cool, i just used this to remove some white spaces in a dhtml menu i copy pasted into vi:

sudo perl -pi -i.bak -e 's/\s{0,}(\<.*)/$1/g' dhtml_menu.dhtml

'just imagine an html markup file with 0 to many amount of whitespace before each tag'

kinda ugly but it works !! =)

[URL= http://rdoskfpe.com..xdxx[/URL] <a href=' http://dtzsmeog.com..</a> zowrtiax http://cpmgehvo.com wlfbbilg evioygjz

@Tina, basic search and replace script can be used for multiple files as well punching with a shell script.
using for loop of bash script