Shell Programming and Scripting

hi, all.

I got a question about awk:

If I have a lot of awk files for different tasks, but they'll share the same base directory, log setting, action lists, etc., is it possible to import them from a config file, so that I don't have to put them into each awk file?

For example, I need a variable LogFileHeader_StartServer="StartServerLog_".
For now, I have to write it into all my awk files.
And if someday I'll change it to another string, I'll have to modify all the files.

But If I can import it from a config file, the all I have to do is modify that config file.

Is this possible in awk?

Thanks.

Put your variables in some file, for example "vars.conf" defined as shell variables. Then create your awk scripts as follows:

Code:

#!/bin/bash
. vars.conf
awk -vLogFileHeader_StartServer=$LogFileHeader_StartServer -vsome_other_variable=$some_other_variable 'your
AWK
script'

Note: some awks needs a space between -v and the awk variable and the shell variables need to be quoted:

Code:

awk -v awkvar1="$shellvar" ...

Or use the variable behind your awk code

Code:

awk '{print var}' var="$shellvar" infile

The way awk reads files makes it easier to have a config file like

Code:

var1 value1
var2 value2

You could use it like

Code:

awk 'BEGIN {
while((getline<"/path/to/config") > 0)
{ C[$1]=$2 } }

...'

Then you could use the variables like C['var1']

If you put your awk code in program files and include them using the -f option, you can put the common code in a separate config file of awk commands and name it in the 1st -f option you pass to awk as in:

Code:

awk -f shared.awk -f application_specific.awk input_file...

Thanks!
I totally forgot that array can be used in such a case!

And thanks all other's help.
The -v is also good, just not suitable for my problem since I might have a long list of variables... @@