[SOLVED] Retrieve default value with grep -e?

kristo5747 · 11-01-2011, 06:55 PM

I am parsing text files looking for specific entries like so

Code:

    grep -e 'Model' -e 'Manufacturer' -e 'Man Date' -e 'SW Version' -e' SW Name' -e 'HW Version' -e 'Receiver ID' JGMDTV356.HDD

This gives me an output like so

Code:

    Model         = HR24
    Manufacturer  = 100
    Man Date      = 04/14/2010
    SW Version    = 4D1
    HW Version    = 2.3
    Receiver ID   = 035635905389
    Model         = WDCWD5000AVVS-63M8B0 (Dragonfly-0)

The problem is that some files do not have the same number of fields. How can I do something like this?

Code:

   Model         = HR24
    Manufacturer  = 100
    Man Date      = N/A
    SW Version    = 4D1
    HW Version    = N/A
    Receiver ID   = N/A
    Model         = N/A

Could this be done?

syg00 · 11-01-2011, 09:07 PM

Not (easily) with grep - anything with associative arrays (bash, awk, perl, ...) you should be able to do something. Maybe populate all as "n/a" and only print the input if found, else the default. If you don't want to hard-code the fields, make up a file containing them, and read it in.

allend · 11-01-2011, 11:37 PM

An ugly hack using a bash script assuming that Model is always present and comes first and the other fields come in the same order if present.

Code:

#! /bin/bash

while read line
do
  if [[ $line =~ Model.* ]]; then
    echo $line
    read line 
    for i in "Manufacturer" "Man Date" "SW Version" "HW Version" "Receiver ID"
    do
      if [[ $line =~ $i.* ]]; then
        echo $line
        if [ "$i" != "Receiver ID" ]; then
          read line
        fi
      else
        echo $i" = N/A"
      fi
    done
  fi
done < test.txt

If test.txt contains:

Code:

    Model         = HR24
    Manufacturer  = 100
    Man Date      = 04/14/2010
    SW Version    = 4D1
    HW Version    = 2.3
    Receiver ID   = 035635905389
    Model         = WDCWD5000AVVS-63M8B0 (Dragonfly-0)
    Manufacturer  = 100
    SW Version    = 4D1

then the ouput is:

Quote:

Model = HR24
Manufacturer = 100
Man Date = 04/14/2010
SW Version = 4D1
HW Version = 2.3
Receiver ID = 035635905389
Model = WDCWD5000AVVS-63M8B0 (Dragonfly-0)
Manufacturer = 100
Man Date = N/A
SW Version = 4D1
HW Version = N/A
Receiver ID = N/A

kristo5747 · 11-02-2011, 11:27 AM

Quote:

Originally Posted by allend

An ugly hack using a bash script assuming that Model is always present and comes first and the other fields come in the same order if present.

Code:

#! /bin/bash

while read line
do
  if [[ $line =~ Model.* ]]; then
    echo $line
    read line 
    for i in "Manufacturer" "Man Date" "SW Version" "HW Version" "Receiver ID"
    do
      if [[ $line =~ $i.* ]]; then
        echo $line
        if [ "$i" != "Receiver ID" ]; then
          read line
        fi
      else
        echo $i" = N/A"
      fi
    done
  fi
done < test.txt

If test.txt contains:

Code:

    Model         = HR24
    Manufacturer  = 100
    Man Date      = 04/14/2010
    SW Version    = 4D1
    HW Version    = 2.3
    Receiver ID   = 035635905389
    Model         = WDCWD5000AVVS-63M8B0 (Dragonfly-0)
    Manufacturer  = 100
    SW Version    = 4D1

then the ouput is:

Ugly maybe, but this is a clever hack.

Unfortunately, this won't fly. Depending on the supplier that sends the file, Model may not even be present.

I need to devise a strategy where the relevant stuff is captured and the rest is rejected.

Thanks for your time.

chrism01 · 11-02-2011, 06:27 PM

In that case, I concur with syg00, grep is not the tool here, you need a proper prog lang (in my case Perl) that can handle arbitrary data layouts that may have missing elements.