Shell Programming and Scripting

Hi all,

I know how to count lines, count byte of a file, but really dont know how to compare line , I am newbie and hope you can help me to learn more about UNIX.
Here is my problem.
====
How :

1. File A (Flatfile with NAME|DATE|ID1|ID2|ID3|ID4|ID5)
CJKA|2005-12-10-08.01.30.000000|1111111111|ECI|1112221111|1113331111|1114441111
DMCH|2005-12-10-08.28.00.000000|1111113333|CCCOSSR|1112223333|1113333333|1114443333
DLMQ|2005-12-11-01.00.00.000000|1111117777|STLITE|1112227777|1113337777|1114447777
DMCH|2005-12-10-08.20.00.000000|1111114444|MTRRD|1112224444|1113334444|1114444444
CJKA|2005-12-10-08.10.00.000000|1111112222|OMCRTVI|1112222222|1113332222|1114442222
R1L7|2005-12-10-08.11.00.000000|1111115555|CLAIMS|1112225555||
DLMQ|2005-12-10-08.00.00.000000|1111116666|QAS1|1112226666|1113336666|1114446666
DLMQ|2005-12-11-01.10.00.000000|1111118888|SVCPLN|1112228888|1113338888|1114448888
CJKA|2005-12-10-08.13.30.000000|1111119999|OMCRTVI|1112229999|1113339999|1114449999

2. File B ( 4 lines)
records='9'
bytes='724'
CJKA|2005-12-10-08.01.30.000000|1111111111|ECI|1112221111|1113331111|1114441111
CJKA|2005-12-10-08.13.30.000000|1111119999|OMCRTVI|1112229999|1113339999|1114449999

How we write a ksh script to compare and give another output file

1. a. Check record count in file A is equal 9 in file B.
b. Check byte count in file A is equal line 2 file B 724
c. First record in File A is line 4 in file B
d. last record in file A is line 4 in File B

If it did not match then give the error do not match . Which one is not match

2. If it is match then give the out put
File C name is yyyymmdd.dat in 2nd field of file A , like 20051210.dat , it will be a flatfile
NAME|DATE||||ID4|ID3
====

Thanks for sharing your knowledge .

look at man:

join
sort
uniq

I really dont see sort, uniq in here

Here is what I did for 1c and 1d
====

$AFIRST=(head -1 A)
$BTHREE=(cat B | sed -n 3p)
if [[ $AFIRST = $BTHREE ]];
then
print "they match"
else
print "they don't match"
fi


ALAST=$(tail -1 A)
BFOUR=$(cat B | sed -n 4p)
if [[ $ALAST = $BFOUR ]]
then
print "they match"
else
print "they don't match"
fi
=======

I got this part for 1a and 1b
ALINES=$(cat A | wc -l)
BFIRST=$(head -1 B)

how do you compare ALINE=9 and BFRIST=records'9' ; how to have only 9 = 9

i guess i'm kinda lost in what you're trying to do. bigearsbilly's sugestion of using uniq as an option actually sounds like a pretty good suggestion to me.