Ordering Grep Output With Multiple Patterns - regex

Question

The file I'm searching (fruit.text) looks something like the below snippet, the data appears in random order that I cannot control.
....fruit=apple,...qty=3,...condition=bad,....
...qty=4,...condition=great,...fruit=orange,...
...condition=ok,...qty=2,...fruit=banana,...
My Grep command is: grep -Eo 'fruit.[^,]*'\|'qty.[^,]*'\|'condition.[^,]*' fruit.txt
This results in output like:
fruit=apple
qty=3
condition=bad
qty=4
condition=great
fruit=orange
condition=ok
qty=2
fruit=banana
Which is correct, however, I'm looking for the output to be ordered as I specified in the grep cmd. ie, exactly like the below:
fruit=apple
qty=3
condition=bad
fruit=orange
qty=4
condition=great
fruit=banana
qty=2
condition=ok

Answer 1

A solution with gawk:
first i added some extra ',' to the input:
....,fruit=apple,...,qty=3,...,condition=bad,....
...,qty=4,...,condition=great,...,fruit=orange,...
...,condition=ok,...,qty=2,...,fruit=banana,...
Then i wrote this awk script (fruit.awk):
{ fruit ="";
qty="";
condition="";
for (i = 1;i <= NF; i++){
delete a;
split($i,a,"=");
if (a[1]=="fruit" ) { fruit=a[2]; }
if (a[1]=="qty") { qty=a[2] }
if (a[1]=="condition") { condition=a[2] }
}
}
{ print "fruit=" fruit;
print "qty=" qty;
print "condition=" condition;
}
output of: gawk -F , -f fruit.awk fruit.txt:
fruit=apple
qty=3
condition=bad
fruit=orange
qty=4
condition=great
fruit=banana
qty=2
condition=ok

Answer 2

Using sed in some steps:
sed -E 's/^/,/;
s/(.*),(condition[^,]*)/\2\r,\1/;
s/(.*),(qty=[^,]*)/\2,\1/;
s/(.*),(fruit=[^,]*)/\2,\1/;
s/\r.*//;
s/,/\n/g' input.txt
I start with inserting a , for input where the interesting data starts in the first field.
After condition I add a \r, so I can remove the garbage after finding the fruit.