Say I have the file data.log with the following contents:
[12/Mar/2015] /var/lib/file1.txt
[12/Mar/2015] /var/lib/file2.txt
[12/Mar/2015] /var/lib/file3.txt
How can I use awk and bash to store the contents of this file into an array where [12/Mar/2015] would be the element and /var/lib/fileN.txt its key?
bash:
# the data
$ cat data.log
[12/Mar/2015] /var/lib/file1.txt
[12/Mar/2015] /var/lib/file2.txt
[12/Mar/2015] /var/lib/file3.txt
# the associative array declaration
$ declare -A map
# read the data from the file into the array
$ while read -r date file; do map[$file]=$date; done < data.log
# iterate through the data
$ for key in "${!map[@]}"; do printf "%s => %s\n" "$key" "${map[$key]}"; done
/var/lib/file3.txt => [12/Mar/2015]
/var/lib/file2.txt => [12/Mar/2015]
/var/lib/file1.txt => [12/Mar/2015]
/var/lib/file2.txt => [12/Mar/2015] /var/lib/file1.txt => [12/Mar/2015] without the last line
In awk:
$ awk '{map[$2]=$1}END{for(key in map){printf "%s => %s\n",key,map[key]}}' infile
/var/lib/file2.txt => [12/Mar/2015]
/var/lib/file3.txt => [12/Mar/2015]
/var/lib/file1.txt => [12/Mar/2015]
Or, as a non-one-liner:
# For each line, use the second field as key and the first field as value
{
map[$2] = $1
}
END {
# Iterate over all keys
for (key in map) {
printf "%s => %s\n", key, map[key]
}
}
This goes into a file, for example script.awk, and is then called with
awk -f script.awk infile