Shell Programming and Scripting

Hi,

I need to read each line in all files .If last character in each line is not ;(semicolon) then i need to append next line also.Here each line may consists of multiple semicolons(.but i need to read only last character.I need to achieve this using shell script.And i need to remove tht last semicolon also.I need to achieve this in all files in a directory and i need to redirect modified files to the same directory.
How to achieve this?

For example my file consists of
asdfgh;adffg;afdfg;
adfsffvfgggg;
aedfdsfs
fsff;

Here i need output as
asdfgh;adffg;afdfg
adfsffvfgggg
aedfdsfs fsff

How to achieve this?

if you have GNU sed:

Code:

sed -i 's/;$//' file*

or using Perl:

Code:

perl -pi -e 's/;$//' file*

Hi,

If there is no semicolon its not appending next line with present line.And also its not working if my file contains pipe line characters like '|'.

For example if file contains

user|find|ghjk|gg|;
sufgh|fghj|hgj
Dhj|hjk|jkl;
its not removing semicolon
and its not appending 3red line to 2nd line

try this

sed -e :a -e '/;$/!N;s/\n//; ta' -e 's/;$//' file

Code:

awk 'ORS=/;$/?RS:FS{sub(/;$/,"")}1' filename

Use nawk or /usr/xpg4/bin/awk on Solaris.

As far as the entire task is concerned:

Code:

awk 'FNR == 1 {
  system("[ -f \""fnt"\" ]&&mv "fnt" "fn)
  fn = FILENAME
  fnt = FILENAME"__"++c
}
ORS = /;$/ ? RS : FS {
  sub(/;$/,"")
  print > fnt
} END {
  system("[ -f \""fnt"\" ]&&mv "fnt" "fn)
}' filenames*

Use nawk or /usr/xpg4/bin/awk on Solaris.

Hi quintet

using this sed -e :a -e '/;$/!N;s/\n//; ta' -e 's/;$//' file its not redirecting to my directory again