Trying out Awk,pls help!

[cyberray]

Hi guys i wanna append a sentence starting with @ to the previous line,used sed to try,but as the lines are too many..sed could not do what i wanna do,so i thought i try out awk..can anyone tell me how to do what i want with awk?

for example:

hey,how are u?
@ya i am fine!!

what i wanted:

hey,how are u?@ya i am fine!!

ya basically this is what i wanna..but as there is alot of lines i need to append to the previous line,can anyone suggest anything that awk can help with this requirement?thanks alot

[ghostdog74]

awk 'BEGIN{RS="\n@";ORS="@"}1' file

[cyberray]

hey thanks ghostdog74..but still got one problem..i tried out your command,it did append to the previous line,but there is a extra @ after the sentences

like this:
hey,how are u?@ya i am fine!!@

can tell me how to get rid of that?thanks alot!!

[ghostdog74]

RS means record separator. ORS means output record separarot. specifying RS="\n@" and ORS="@" means the awk will "change" every \n@ to @ in the output. The fact you have an extra @ at the back means there could be another record, something at the last line. Show your exact input file that you have.

[cyberray]

ok here the input file i have

@txndb@Command_History@oracle_history!> ----------------------
!> 01-10-07 08:17:18
!>
!> ----------------------
!> 01-10-07 11:01:35
!> \
!> ..."™.pwd
!> ..mkdir 20071001
!> .ftp 172.18.14.90
!> .ls -lrt
!> ..pwd
!> ..cd 20071001
!> ..ls
!> .ftp 172.18.14.90
!> .ls -lrt
!> ..pwd
!> ..cd @txo@app@enets2@monitoring
!> ..cd @txn
!> ..cd app
!> .cd enets2
!> ..cd monitoring
!> ..ls -l
!> ..rm aaa.txt
!> .clear
!> ..ls -l
!> ..cp -p @home@oracle@20071001@* .
!> ..ls -l
!> ..chmod 700 *
!> ..ls -l
!> ..more cron*
!> .crontab -e
!> .more cron*
!> .ls -l
!> ..vi query.sh
!> ..ls -l
!> ..vi query.sh
!> ..ls -l
!> ..more query.sh
!> ..vi batch_monitoring.sh
!> .ls -l
!> ..ftp 172.18.14.90
!> .ftp 172.18.14.90
!> .ls -l
!> ..vi query.sh
!> ..clear
!> ..ls -l
!> ..date
!> ..@batch_monitoring.sh
!> ..vi query.sql
!> .ls -l
!> ...@batch_monitoring.sh
!> ..ls -l
!> ..ftp 172.18.14.90
!> .cd @home
!> .ls
!> .more monitor.sh
!> ..more monitor.sh
!> ..cd @home
!> .ls
!> .ls -lt
!> .more monitor.sh
!> ..ftp 172.18.34.160
!> ..cd root
!> ..ls
!> .ls -lt
!> .ftp 172.18.34.160
!> ..ls -lt
!> .ls -lt > tmp.txt
!> .cd ..
!> ..cd @home
!> .ls
!> .ls -lt
!> .cd @txn@app
!> ..cd enets2
!> ..ls
!> .cd monitoring
!> ..ls
!> ..@sendmail_batchclose.sh
!> .cd @home
!> .ls -lt
!> .more monitor.sh
!> ..cd @txn@app@enets2@monitoring
!> ..ls
!> .more sendmail_batchclose.sh
!> ..ls
!> .clear
!> ..df -k
!> ..cat batch_close_status_montoring.txt | mail -s "test"
!> ..more sebdnmail*
!> ..more send*
!> .vi sendmail_bat*
!> ..@sendmail_batchclose.sh
!> .vi sendmail_b*
!> .ls -l
!> ...@sendmail_batchclose.sh
!> .cat batch_close_status_monitoring.txt | mail -s "test"
!> .ls -l
!> ..crontab -e
!> .vi @txn@app@enets2@monitoring@batch_monitoring.sh
!> ..ls
!> .more @etc@hosts
!> ..clear
!> ..df -k
!> ..df -k | mail -s "test"
!> .cd @txn@app@enets2@maint
!> .ls -lt
!> .chmod 700 *
!> ..ls -lt
!> .cd @
!> .clear
!> ..cd @txn@app@enets2@maint
!> .ls
!> .le -lt
!> .ls -lt
!> .cd ..
!> ..ls
!> .cd sql
!> .ls
!> .ls -lt
!> .cd ..
!> ..ls
!> .cd config
!> ..ls -lt
!> .cd ..
!> ..cd data
!> ..ls -lt
!> .cd ..
!> ..ls
!> .cd ftpScript
!> .ls -lt
!> .cd bk
!> ..ls -lt
!> .chmod 700 *
!> ..cd ..
!> ..cd lib
!> .ls -lt
!> .cd ..
!> ..ls
!> .ls -lt
!> .ls config
!> ..ls lib
!> .ls template
!> ..ls -lt bk
!> ..ls -lt
!> .ls -lt template
!> ..ls -lt log
!> .ls -lt
!> .ls -lt data
!> ..ls -lt log
!> .ls -lt sql
!> .cd sql
!> .chmod 700 *
!> ..ls -lt
!> .cd ..
!> ..ls -lt
!> .cd template
!> ..ls -lt
!> .chmod 700 *
!> ..ls -lt
!> .cd ..
!> ..ls -lt bk
!> ..ls -lt maint
!> .ls -lt template
!> ..ls -lt sql
!> ...#8.ls -t
!> ..pwd
!> ..ls -l
!> ..pwd
!> ..cd monitoring
!> ..ls -l
!> ..more sendmail_doub*
!> ..clear
!> ..ls
!> .ls -l
!> ..cat sendmail_doublebatch.sh


well...i just change the delimiter to !> and it work well too..just that behind will have a additonal !>..so i was wondering if u knw hw to solve this problem,thanks alot!

[ghostdog74]

try this then:

awk 'BEGIN{RS="\n!>";ORS="!>"}
RT{
gsub(/\\\n/,"")
print
}' "file"

[cyberray]

hmmm..alot of errors..i also dont knw why

errors:
'RT{' is not recognized as an internal or external command,
operable program or batch file.

'gsub' is not recognized as an internal or external command,
operable program or batch file.

[ghostdog74]

ok you don't have the GNU awk version. I am using GNU awk.
although i don't usually do this, but this may do what you want

awk 'BEGIN{RS="\n@";ORS="@"}1' file | sed '$d'

[Franklin52]

Try this:

awk '/^@/{if(NR > 1){print x $0}{x=$0}}' file

Regards

[cyberray]

haha..okok..thanks alot guys..i have one more question,cos that time i tried to use sed to do this problem,that is..i wanna match a certain word perhaps like /repdb/,so i do out a sed command which i want when match /repdb/ then it will be replace by the things inside line2.txt

this what i come up with:

sed "/repdb/r line2.txt" histtry2.txt>histtry3.txt

but like this way..it will get everything which have repdb inside..but what i want is /repdb/..with the / inside..anyone can help?can i use awk to do this?pls help..thanks..