merge 2 files into third

[Meher1]

Hi, I have two files with the same format. Each of these files are | delimited and have a record identifier.
1 header, 2 detail, 3 trailer.

In the out put file:
I need to have the header of file a, merge the detail records of two files and sum up the corresponding trailer values of these two files excluding the identifier. Here is the format.

file a.
1|lkjh|ertwui|wweweqweefgh00|wqeqwe|sweqwe|
2|wert00|weqweqwe|weqweqwe|qweqweqwe|qewqe|qeqeqwe||qweqeqw|
2|wert00|weqweqwe|weqweqwe|qweqweqwe|qewqe|qeqeqwe||qweqeqw|
2|wert00|weqweqwe|weqweqwe|qweqweqwe|qewqe|qeqeqwe||qweqeqw|
2|wert00|weqweqwe|weqweqwe|qweqweqwe|qewqe|qeqeqwe||qweqeqw|
2|wert00|weqweqwe|weqweqwe|qweqweqwe|qewqe|qeqeqwe||qweqeqw|
3|1|2|3||5|

file b.
1|abcd|abcd|efgh00|bnm|sswee|
2|wert00|weqweqwe|weqweqwe|qweqweqwe|qewqe|qeqeqwe||qweqeqw|
2|wert00|weqweqwe|weqweqwe|qweqweqwe|qewqe|qeqeqwe||qweqeqw|
2|wert00|weqweqwe|weqweqwe|qweqweqwe|qewqe|qeqeqwe||qweqeqw|
3|1|2|3||3|

out put file c.
1|abcd|abcd|efgh00|bnm|sswee|
2|wert00|weqweqwe|weqweqwe|qweqweqwe|qewqe|qeqeqwe||qweqeqw|
2|wert00|weqweqwe|weqweqwe|qweqweqwe|qewqe|qeqeqwe||qweqeqw|
2|wert00|weqweqwe|weqweqwe|qweqweqwe|qewqe|qeqeqwe||qweqeqw|
2|wert00|weqweqwe|weqweqwe|qweqweqwe|qewqe|qeqeqwe||qweqeqw|
2|wert00|weqweqwe|weqweqwe|qweqweqwe|qewqe|qeqeqwe||qweqeqw|
2|wert00|weqweqwe|weqweqwe|qweqweqwe|qewqe|qeqeqwe||qweqeqw|
2|wert00|weqweqwe|weqweqwe|qweqweqwe|qewqe|qeqeqwe||qweqeqw|
2|wert00|weqweqwe|weqweqwe|qweqweqwe|qewqe|qeqeqwe||qweqeqw|
3|2|4|6||8|

**********************************************

Please help!

Thanks,
Meher.

 

[Ygor]

awk -F\| '
NR==1 || $1==2 {print}
$1==3 {for(i=2;i < NF; i++) a+=$i;max=i}
END {printf &quot;3&quot;;for(i=2; i < max; i++) printf &quot;|%d&quot;, a;printf &quot;|\n&quot;}
' a b > c

 

[Meher1]

Thanks Ygor for the quick reply. Sorry, I did not understand the solution. I copied my test files in to a nad b files. I tried executing the above code on command line but got this error.

**********************
&quot;}awk: There is a regular expression error.
?, *, or + not preceded by valid regular expression
The input line number is 1. The file is b.
The source line number is 1.
****************************************

Where did it go wrong? thx.

 

[Ygor]

Not sure why you are getting this error - the above code works on my version of awk. Maybe try....

awk -F\| '
NR==1 {print}
$1==2 {print}
$1==3 {for(i=2;i < NF; i++) a+=$i;max=i}
END {printf &quot;3&quot;;for(i=2; i < max; i++) printf &quot;|%d&quot;, a;printf &quot;|\n&quot;}
' a b > c