Shell Programming and Scripting

Hello!

I need to come back one line in awk when i detect a regular expression, i'm working with this:

Code:

/<POLITIC>/,/<\/POLITIC>/

when i found the line that matches with <\/POLITIC> i need to come back one line, the opposite of getline that takes a new line and continue executing the code in the same line, i need to come back one line and continue executing the code with this new line.

Someone knows how can i do?

Thanks!

What you're working with (i.e. "I'm working with this") doesn't tie in with the question you're asking.

Code:

/<POLITIC>/,/<\/POLITIC>/

This doesn't match a regular expression. It matches a range of lines between two "regular expressions".

You want to "go back" from what after matching "<\POLITIC>"? Back from "<\POLITIC>" or back from the line before "<\POLITIC>"?

More useful is if you show what your input is and what your output should be, as well as what code you've written so far.

Yes, sorry, it matches a range, i didn't express correctly.

>You want to "go back" from what after matching "<\POLITIC>"? Back from "<\POLITIC>" or back from the line before "<\POLITIC>"?

I want to go back from the line before "<\POLITIC>".

When it arrives to the end of the range: "<\POLITIC>" i wanna go back one line.

Well, there is no "rewind" function in awk, so:

Code:

$ cat Test
awk '
  NR == 1 { LINE = $0; next }
  /<POLITIC>/,/<\/POLITIC>/ { if ( $0 ~ /<\/POLITIC>/ ) print "POLITIC:  goes here" }
  { print LINE; LINE = $0 }
' file1


$ cat file1
1
2
<POLITIC>
blah 1
blah 2
blah 3
</POLITIC>
blah
more blah
and still
more blah


$ ./Test
1
2
<POLITIC>
blah 1
blah 2
POLITIC:  goes here
blah 3
</POLITIC>
blah
more blah
and still

My approach when I need to do this is to pass the file name twice to the script. On the first pass write the line numbers you are interested in to an array. On the second pass do your processing.