Shell Programming and Scripting

I have a question about changing how parsing occurs currently for us:

input FILE123

Code:

TAGA01: 01
TAG02: daadsf
TAG03: adfasdf
TAGBBB04: 35
TAG05: asdfa
TAG07: adfd
TAG07: adfa3
TAG07: 234234
TAGCC08: 3525df
TAG09: adsfa
TAG10: 245
TAG11: nnnn
EOR:
TAGA01: 02
TAG02: abas
TAG03: asdfasd
TAGBBB04: E
TAG05: asdfasd
TAG07: acvasc
TAG07: czcvc
TAG07: 22
TAGCC08: adsfasd
TAG09: Y
TAG11: yyyy
EOR:
.
.
.

Note that some tags may not be in a record, and some tags may repeat in the same record.

I need to covert to the following inline format (limiter doesn't matter, and I can change it should the data include the limiter in other files) and trim it so the tag doesn't appear:

Code:

Format:
TAGA01 TAGCC08 TAGBBB04 TAG09 TAG11

output.file:
01 3535df 35 adsfa nnnn
02 adsfasd E Y yyyy
.
.
.

Here is what is used currently (from memory, so the syntax isn't correct but the idea is):

Code:

cat FILE123 | egrep "^TAGA01 ^TAGBBB04 ^TAGCC08 ^TAG09 ^TAG11" | awk -F. -f awkfile.awk > output.file

where awkfile.awk contains if statements and a printf output statement (again, syntax along with substring numbers are not correct - but the idea is there):

Code:

if ($1==TAGA01) {pTAGA01=substr($1,3)}
.
.
.
if ($1==TAG11) {
   pTAG11=substr($1,4)
   printf pTAGA01 ... pTAG11
}

I wanted to see different ideas for two reasons: one to see if this could be more efficient since every tag gets multiple ifs every time, and just to straight up learn something new.

Thanks for your time!

If you change EOR: to the blank line you can use perl in its paragraph mode:

Code:

sed 's/^EOR:$//' INPUTFILE | 
perl -00 -ne '/
TAGA01:\s+(.*?)\n
.*
TAGCC08:\s+(.*?)\n
# and so on
/xs && print "$1 $2\n"'

Thanks yazu...if you get a moment, could you explain a few things to me:
I understand what the sed command does, though not all of the syntax. From the man page: s/regexp/replacement. Why the ^ (was wondering about this character in the egrep command in the original script) and the $/ (is that the new line character)? And the third / indicates replace with a blank line since no other character is shown?

All of this is piped into the perl script.

Again, from the man page -00 will 'slurp' the stream in paragraph mode. -ne starts a while <> loop command. Not sure about perl command itself. Why the first /? The command searches for a tag:, but not sure what \s+(.*?) does, along with .* in the between tag searches. Not sure what /ns does, along with what the $1 and $2 refer to in the print line.

And I assume all of this can just be > into an output file, or is there a better way to do that with perl?

Thanks again!