when not to script? (please forgive my noobishness)

[waffle247]

Hi there,

I'm new to the forums, been supporting various linux machines for a couple of years and usually use simple scripts for basic tasks. Recently I have been asked to create a script that goes and finds a whole bunch of .log and .txt files, searches them for a selection of values and returns an associated value (which has 3 possible conitations) which is then used to search on another file and return associated values from both searches into the same file. At the same time as that I need to get the extract part of the file path (/country/serialnumber) and use that as a label at the beggining of the line of returned values.

Does that make sense to anyone out there? The only reason a script is being requested I think is because someone doesn't want to go through the 250+ sub-folders (all of which contain 1 interface.log or interface.txt file and 1 config.log or config.txt file) and extract the information manually.

At the moment I can grep the files within the individual sub-folder and get the values I want out but it's very messy unfortunately:

grep -B 1 eth interface.* >interfacereport.txt

which gives me everything I want plus the above line for labels, plus a lot of gumf I don't need (which I haven't gone to the trouble of pasting in but will if someone thinks it might help) like so:

Name IP Address Zone
eth1 10.48.96.1/24 Trust

Anyway - I'm not sure if this will make sense to anyone else but I hope someone out there can make heads or tails of it!

Many thanks! (I apologise if this is a dumb question btw)

[waffle247]

hmm, maybe I phrased the question badly; ah well - I can't get my head around this

[wje_lf]

Make a couple of test data files with a few lines each.

Write a toy script which starts to do what you think you might want.

Look at the output and tell yourself, "No, that's not exactly what I had in mind."

Play with the script to make the output closer to what you want.

Introduce a few more lines of test data. The new test data should test unusual cases of data, if you can think of any.

Rinse. Repeat. Come back here if you have specific questions.

[waffle247]

thanks - I'll keep playing with my script in the way you suggest; I hope I get somewhere soon, my fingers are getting tired