Shell Programming and Scripting

I have a text file that has blocks of text. Each block starts with ### and ends with End_###.
I wrote a perl script to search a string from line 2 (ignore any line starts with ###) of each block
if matched, need to print that whole block. According to the input file in below, it will print blocks #02, #04, #05.
Here is my question, I have no problems matching the SEARCH_STR, but once matched, how do I move back previous lines?
e.g. in block #02, matched in line 3, but I need to print lines 1 & 2.

Thanks in advance for any help!

MY SCRIPT:
=========
#!/46020/local/bin/perl -w
#
if ($#ARGV != 0) {
print "You must provide a search criteria.\n";
exit;
}
my $query = $ARGV[0];
my $INPUT = "/mypath/inputFILE.txt";
my $line;
my $tag = 0;
my @results = ();

open (FILE, "<$INPUT") or die "Cannot open '$INPUT' file: $!";
while ($line = <FILE>) {
next if ($line =~ /^#[0-9][0-9].* - / || $line =~ /^\s*$/);
if ($line =~ /$query/i) {
$tag = 1;
}
if ($line =~ /^End_#/ && $tag == 1) {
push (@results,$line) if ($tag == 1);
$tag = 0;
}
push (@results,$line) if ($tag == 1);
}
close (FILE);
print "@results\n";
exit;


INPUTFILE:
========
#01 - block start line here
some text here
more text here
more lines
more lines
more lines
End_#01

#02 - block start line here
some text here
this line contains the SEARCH_STR and something
more lines
more lines
End_#02

#03 - block start line here
some text here
more text here
more lines
more lines
more lines
End_#03

#04 - block start line here
some text here
more lines
more lines
the SEARCH_STR is here and something
some text here
more lines
more lines
more lines
End_#04

#05 - block start line here
some text here
more lines
the SEARCH_STR is here and something
more lines
End_#05


#06 - block start line here
some text here
more text here
more lines
more lines
more lines
End_#06

Try something like this:

Code:

#!/46020/local/bin/perl -w
$/="\n\n";
while (<STDIN>) { if (/$ARGV[0]/) { print $_ }}

./scriptname.pl SEARCH_STR < /mypath/inputFILE.txt

If Smiling_Dragons suggestion does not work, you can use the Tie::File module, which allows you to read and write to a file just like a perl array. See the Tie::File documentation for usage details. It's also a core module so you should have it installed already. If you get stuck, post back.

Just want to clarify, according to the inputFILE.txt, when run the script with the SEARC_STR it will print #02, #04, #05 blocks. e.g. scriptname search.
My code able to print matched plus following lines up to End_###, but I need a way to print lines before matched in that block.

Thank you.

Yes, I think we both understood that requirement, and both suggestions can be used to do what you want. If you are familiar with looping through perl arrays using subscripts: $array[n] then using Tie::File will be pretty straight forward once you read the documentation.

S_M's suggestion reads the file in chunks, not line by line, and looks like it should work although the last match might fail if there isn't two newlines at the end of the file. You will have to try it and see or maybe he can clear that up.

I think it'll be ok with a normal file termination - I tried it on a single newline and it looked alright but I havn't put a lot of testing into it for sure