03-15-2018
Without more detailed, representative info on input files' and variables' contents it is very difficult for me - and mayhap others in here - to reproduce and understand (and even less comment on possible improvements of) the problem.
There's no obvious bug in your code snippet, so only generic suspicions can be uttered: non-*nix text files (DOS line terminators, missing terminator at end-of-file), unexpected behaviour of the for loops, regex special characters in the respective shell variables, ...
10 More Discussions You Might Find Interesting
1. UNIX for Advanced & Expert Users
Hi,
I'm trying to compare Actual.html with a baseline.html
However, everytime it fails b'coz of the timestamp differences
between the two. So, thought of stripping off the timestamp
from both the *html files before comparing using below sed
command over Solaris Unix platform:... (3 Replies)
Discussion started by: elearn.latha
3 Replies
2. Shell Programming and Scripting
Hi,
I need convert a dump file in the following format : (please note that line numbers are provided for easy look)
Original file:
1 2007-10-2482.90 No trade 0 0.00 100000.00
2 100000.00
3 0.00
4 HOLD
5 2007-10-2589.75 Bought 1114 1114 100000.00 0.00
... (5 Replies)
Discussion started by: sabyasm
5 Replies
3. Shell Programming and Scripting
cat file.txt
fvnuiehuewf
ruevhxncvkjrh
zxjvurhfuwe
jkhvBEGINvfnvf
ijrgioe
Trying to delete a line that has the pattern "BEGIN"
cat sedtest
filename=file.txt
pattern=BEGIN
sed "/^$pattern/d" "$filename" (9 Replies)
Discussion started by: cola
9 Replies
4. Shell Programming and Scripting
Can anyone help me get this small sed script to work in shell on the command line?
I need it in a one liner really as i want to edit many scripts in a for loop and dont want to have to invoke a separate script each time.
#!/bin/sh
sed '/mailx\ -s.*$ {
i\
#Comment above mailx line ... (5 Replies)
Discussion started by: lavascript
5 Replies
5. Shell Programming and Scripting
Hi
Is it possible to do the following in a single command
/usr/xpg4/bin/sed -e '/rows selected/d' /aemu/CALLAUTO/callauto.txt > /aemu/CALLAUTO/callautonew.txt
/usr/xpg4/bin/sed -e '/^$/d' /aemu/CALLAUTO/callautonew.txt > /aemu/CALLAUTO/callauto_new.txt
exit (1 Reply)
Discussion started by: aemunathan
1 Replies
6. Shell Programming and Scripting
Hi there,
I'd like to delete the beginning of a line up until it finds a certain word or character string: in this case, I'd like to delete each line up to the word "mounting".
Thanks ;)
Susan (12 Replies)
Discussion started by: kitykity
12 Replies
7. Shell Programming and Scripting
Hi all,
I ahve a program which has to delete a line in a file... if i run the sed command through shell prompt it works fine. But if run it using code its throwing error. May i know where i am doing wrong.
the file has 3 lines
# cat /root/.ssh/known_hosts... (4 Replies)
Discussion started by: vivek d r
4 Replies
8. Shell Programming and Scripting
sed /'1-2'/&^/ filename
suppose there is a file containing three lines , how do we do delete the word from each line?
hyter efr frf
rerfer efe ewd
cdcf evrfgf erfv
the output has to look like
frf
ewd
erfv (2 Replies)
Discussion started by: Rajeev Nukala
2 Replies
9. Shell Programming and Scripting
It looks like if matching and deleting the last line confuses 'sed' so it does not recognize '$' address. Consider:
sed -e '/^3/d' -e '$ a text'
supposed to delete a line starting with '3' and then append 'text' after the last line of input. But, if it is the last line of input which starts... (2 Replies)
Discussion started by: msz59
2 Replies
10. Shell Programming and Scripting
I have a simple task to replace unix line feed end of line characters with carriage returns.
When I run the following “change file in place” sed instruction from the command line all the Line feeds are successfully replaced with Carriage returns.
sed -i 's/$/\r/' lf_file.txt
But that same... (1 Reply)
Discussion started by: hawkman2k
1 Replies
cr(1) General Commands Manual cr(1)
NAME
cr - converts text files between nix EOL and dos EOL
SYNOPSIS
cr - | + <input file> <output file>
DESCRIPTION
Text files, such as tle files, that come from a dos source usualy have the ^M symbol at the end of every line. Cr converts files between
the dos newline format and the normal *nix newline format by stripping the ^M to convert dos to *nix, using the '-' option, or adding ^M to
a *nix file to create the proper dos file when the '+' option is used.
Although this extra character is not often a problem, programs like seesat5, which are data driven will encounter parsing problems when the
extra character is present. It is these problems that cr is intended to repair.
Options
- | + One or the other of these options is required. The '-' option is used to remove ^M from all newlines found in the dos file. The '+'
option is used to add ^M to every newline found in a *nix file.
input file
Fully delineated path to the input file. As this program is used in the dos environment as well, standard input is not used.
output file
Fully delineated path to the output file. As this program is used in the dos environment as well, standart output is not used.
SEE ALSO
seesat5(1), seesat5(7), SEESAT5.INI(5), tle(5)
BUGS
Cr is not an inteligent program. It methodicaly replaces/removes the offending character when it finds it in the correct context. Newline
sequences found in contexts other than 'newline' will be replaced/removed just like those found in the proper context. Passing a binary
file through cr is not advised, for this reason. Send all inqueries to Dale Scheetz <dwarf@polaris.net>.
Debian Linux 2 April 96 cr(1)