[SOLVED] Create new files in Linux based on conditions

Issue

I have a file in Linux. This file contains table names in it.

Now I want to check this file and create files based on condition.

table=${}
validateTable=$(hive --database "${hivedb}" -e "SHOW TABLES LIKE '$table'")
if [[ -z $validateTable ]]; then
    Add to file New_imports
else
     Add to file Already exists
fi

For example:

The file contains

table1 
table2
table3
table4

In the above tables table1 and table2 are already existing.

So I want two files

1) New_imports for tables that don't exist 2) Already exists for tables that exist

new_imports

table3
table4

already exists

table1
table2

How can I achieve my result

#!/bin/bash
while read table ; do
   table=${1:1:-1}
   validateTable=$(hive --database "${hivedb}" -e "SHOW TABLES LIKE '$table'")
   if [[ -z $validateTable ]]; then
      echo "$table" >> New_imports
   else
        echo "$table" >> Already_exists
   fi
done < tableFile

Solution

#!/bin/bash
while read table ; do
   table=$(echo "$table" | sed 's/[][]//g;s/'"'"'//g')
   validateTable=$(hive --database "${hivedb}" -e "SHOW TABLES LIKE '$table'")
   if [[ -z $validateTable ]]; then
      echo "$table" >> New_imports
   else
        echo "$table" >> Already_exists
   fi
done < tableFile

Should get your started. Making this bullet-proof and so it can accept arguments will take some more doing.

Managing the output files may take some doing. Note that with using >> means append, so each time you run this, you either need to delete those files, or manage them. I'd recommend keeping them with date-time stamps embedded in the file name like

 echo "$table" >> new_imports.$(/bin/date +%Y-%m-%d.%H:%M)

Also, I don't have anyway to test this with hive, so this is a general solution.

Answered By - shellter

Answer Checked By - Candace Johnson (WPSolving Volunteer)