Sort numberic in a string of text - sorting

Question

I tried some sort examble but can't find the way to solve this.I think i should find the right seperator and then sort it by numberic but it don't work as my desire.
This is my file:
abc_bla_bla_bla_reg0_bla_reg_1_0
abc_bla_bla_bla_reg0_bla_reg_5_0
abc_bla_bla_bla_reg0_bla_reg_2_0
abc_bla_bla_bla_reg0_bla_reg_10_0
abc_bla_bla_bla_reg0_bla_reg_15_0
abc_bla_bla_bla_reg2_bla_reg_15_0
abc_bla_bla_bla_reg2_bla_reg_9_0
abc_bla_bla_bla_reg2_bla_reg_7_0
abc_bla_bla_bla_reg3_bla_reg_26_0
abc_bla_bla_bla_reg3_bla_reg_3_0
abc_bla_bla_bla_reg3_bla_reg_5_0
And this is my desire result:
abc_bla_bla_bla_reg0_bla_reg_1_0
abc_bla_bla_bla_reg0_bla_reg_2_0
abc_bla_bla_bla_reg0_bla_reg_5_0
abc_bla_bla_bla_reg0_bla_reg_10_0
abc_bla_bla_bla_reg0_bla_reg_15_0
abc_bla_bla_bla_reg2_bla_reg_7_0
abc_bla_bla_bla_reg2_bla_reg_9_0
abc_bla_bla_bla_reg2_bla_reg_15_0
abc_bla_bla_bla_reg3_bla_reg_3_0
abc_bla_bla_bla_reg3_bla_reg_5_0
abc_bla_bla_bla_reg3_bla_reg_26_0

Answer 1

$ sort -t_ -k5,5 -k8,8n file
abc_bla_bla_bla_reg0_bla_reg_1_0
abc_bla_bla_bla_reg0_bla_reg_2_0
abc_bla_bla_bla_reg0_bla_reg_5_0
abc_bla_bla_bla_reg0_bla_reg_10_0
abc_bla_bla_bla_reg0_bla_reg_15_0
abc_bla_bla_bla_reg2_bla_reg_7_0
abc_bla_bla_bla_reg2_bla_reg_9_0
abc_bla_bla_bla_reg2_bla_reg_15_0
abc_bla_bla_bla_reg3_bla_reg_3_0
abc_bla_bla_bla_reg3_bla_reg_5_0
abc_bla_bla_bla_reg3_bla_reg_26_0
That may or may not produce the output you expect if the regN value in the 5th column can include 2-digit numbers.

Answer 2

Using awk
$awk -F"_" 'function print_array(arr,max){ for(i=1; i<=max; i++) if(a[i]){print a[i], a[i]="";} } key==$5{a[$8]=$0; key=$5; max=$8>max?$8:max} key!=$5{print_array(a,max); key=$5; a[$8]=$0; max=$8} END{print_array(a,max)}' file
Output:
abc_bla_bla_bla_reg0_bla_reg_1_0
abc_bla_bla_bla_reg0_bla_reg_2_0
abc_bla_bla_bla_reg0_bla_reg_5_0
abc_bla_bla_bla_reg0_bla_reg_10_0
abc_bla_bla_bla_reg0_bla_reg_15_0
abc_bla_bla_bla_reg2_bla_reg_7_0
abc_bla_bla_bla_reg2_bla_reg_9_0
abc_bla_bla_bla_reg2_bla_reg_15_0
abc_bla_bla_bla_reg3_bla_reg_3_0
abc_bla_bla_bla_reg3_bla_reg_5_0
abc_bla_bla_bla_reg3_bla_reg_26_0
Explanation:
awk -F"_" '
function print_array(arr,max) #Simply prints the hashed array from i=1 to max value array is holding
{
for(i=1; i<=max; i++)
if(a[i])
{print a[i], a[i]="";}
}
key==$5{a[$8]=$0; max=$8>max?$8:max} #Key here denotes the 5th field for eg. reg0 in line one. Initially key is null and it will satisfy the condition mentioned below i.e key!=$5. If the 5th field matches with the key set in previous line then push the record into array where the index in array will be the value at field 8 based on which you want to sort your results.
key!=$5{print_array(a,max); key=$5; a[$8]=$0; max=$8} #If key doesn't matches the 5th line it signifies we have a new record set and before proceeding further print the array we stored for previous record set based on 5th field.
END{print_array(a,max) #To print the last record set
}' file
key==$5{a[$8]=$0; max=$8>max?$8:max} : Key here denotes the 5th field for eg. reg0 in line one. Initially key is null and it will satisfy the condition mentioned below i.e key!=$5. If the 5th field $5 matches with the key set in previous line then push the record into array where the index in array will be the value at field 8 based on which you want to sort your results. This will work irrespective of the number of digits in $8.
key!=$5{print_array(a,max); key=$5; a[$8]=$0; max=$8} If key doesn't matches the 5th line it signifies we have a new record set and before proceeding further print the array we stored for previous record set based on 5th field.
END{print_array(a,max) Just to print the last record set

Answer 3

sort -V file
-V, --version-sort
natural sort of (version) numbers within text