UNIX for Dummies Questions & Answers

my file looks like this:
D8511049777 I 04/10/08 99817 1,923.33 927156 Kaufman
D8511049777I04100829981859003924884Martin


i'd like to look like this:
D8511049777I0410080000299817000000192333927156Kaufman
D8511049777I0410080000299818000000059003924884Martin

my spec sheet:

D: starting position at 1, and field lengh is 1
8511049777: starting position at 2, and field lengh is 10
I: starting position at 12, and field lengh is 1
04/10/08: starting position at 13, and field lengh is 6
0000299817: starting position at 19, and field lengh is 10
000000192333: starting position at 29, and field lengh is 12

Your posting history indicates that you should start to learn sed, awk and maybe shell.

hint: figure out what ${variable:n1:n2} does in bash

i used sed to get rid of / , . and space. how do i insert zero(0) to make my field lengh = 10 and other field lengh = 12

It's not clear what is the record format of your input file. Are all your records fields separated by a space or are all fixed in length?

Using sed to get rid of useless characters will be ok only in the second case, otherwise how could you then distinguish one field with the other if they aren't fixed in length?
"1,923.33" looks like an amount of money, which I will doubt it will be the same length for all the records!

If your input file is mixed up with both kind of records will be a lot harder or maybe impossible to parse and format the information you need.

after i used sed to get my file looks like this:
D8511049777 I 041008 99817 192333 927156 Kaufman

cat filename | awk '{print $5}' > filename2

and then i used command below to add my 0.

sed -e :a -e 's/^.\{1,10\}$/ &/;ta'
sed '1,$s/ /\0/g; s/\-\-/\-/g'

wow! this site is very very helpful.