Using $ in search string 1

[wgmallard]

I have a file format that has blank lines within the record, and uses $$$$ as a record end.
Thus \n$$$$\n is the end of the record, the the next character starts the next record.
I have made several attempts to parse this with awk,
RS = "\n$$$$\n" does not work
RS = /\n\$\$\$\$\n/ does not work
and several other options do not work.
Now the thing that makes this strange is that if I go into the file and do an edit
$$$$ -> ####, i.e. changing the 4 $ to 4 #, the following value for RS works perfectly:
RS="\n####\n"

It would appear that the fact that the $ is the end of match marker is getting in the way here, but it would seem that it should be escaped with the \$ and the character just looked at as $, but it is not.
In one sense I have solved the problem by editing the file first and then processing it, but I would like to know if anyone has an idea of what is going on and what RS value I should be using to get the desired result.
Thanks.
Gary

 

[feherke]

Hi

awk -vRS='\n\\$\\$\\$\\$\n' 1 /input/file

[gray]# or[/gray]

awk 'BEGIN{RS="\n\\$\\$\\$\\$\n"}1' /input/file

You assign a string to [tt]RS[/tt], which will be interpreted as regular expression. In regular expression you have to escape the literal dollar sign ( $ ) with a backslash ( \ ). In string you have to escape regular backslash with a backslash.

Feherke.
[link feherke.github.com/]

http://feherke.github.com/

[/url]

 

[wgmallard]

Feherke,
Thanks for the solution, it worked perfectly and more importantly thank you for the discussion of what the issue in the problem was. I had not been aware that the RS string was interpreted as a regular expression when given as a string. I had even tried RS = /\n\$\$\$\$\n/ with no luck - which I would have thought was the same, but it to did not work, but probably would have had it been double escaped. I should note that this is being done in an awk script since subsequent processing is more complex, but this was the key.
Thanks again.
Gary