Hi,

I have a problem I am trying to solve in the most effective manner possible - at least the most effective in a bash script. I realize that perl / etc would be better for this, but I wish to stay in shell for the time being.

The problem:

I have 3 files that I am working with, a "Translation" file (somewhat akin to how locale works), a "Settings" file, and various other files that I need to check against.

Sample "Translation" File:
alias1=realname1
alias2=realname2
alias3=realname3

Sample "Settings" File:
alias1=value1
alias2=value2
alias3=value3

Sample "file to check against"
realname1=value1
realname2=value2
realname3=value3

I think you can deduce what I am doing here. I am using the "Translation" file as an abstraction layer for my settings file and the files I want to check the values against.

Now the problem is...

The Settings file may have fewer elements than the Translation file. Therefore, I need to match the elements in field 1 of the settings to field 1 of the translation entries.

I am currently using arrays to do this. As you know, bash doesnt have associative arrays, just zero-indexed arrays. Thus, I need to create 2 arrays for each file, one containing field 1 and the other field 2. Works fine, but the problem is.. ITS SLOW AS HELL.

I am looking for ideas how to do this not using arrays. In summary: I need to match entries in settings and translation. Using those entries, I need to construct a list of realnames and their associated values. I then will use sed to parse files and overwrite the settings that do not match the ones in the list.

I hope this makes sense :)

perl