Search Results

You can make function to your .profile to overwrite external or builtin command, no need to setup PATH and create scriptfile.

Insert into .profile:

vi()
{
for f in $@
do
[ -s "$f"...

I'm not sure what all you need, but hopefully this examples give some ideas for your needs.

1st I made some example file. I hope that if you use some input files, you give those also. Much faster...

Alias works also in in noninteractive scripts.

Make own rc-file if not like add this to default rc-file:

$HOME/.xxrc file:

alias nawk='echo No nawk'
for awk in /usr/bin/gawk...

Multiple file testing like rbatte1 tried to tell, if you need only know if some while include something.

found=0
for f in A*.rej
do
[ ! -s "$f" ] && continue
found=1
break
done
[...

Shortly:
You have csv file which is something like:

00001,0012
...

And when you open in Excel by clicking your email attachment, the looks like:

1,12
...


It's Excel import default. I...

Using ksh93:

#!/usr/bin/ksh
# 2016-12-30 16:20:11
LOG_TME="12302016162011"
NLOG_TME="${LOG_TME:4:4}-${LOG_TME:0:2}-${LOG_TME:2:2} ${LOG_TME:8:2}:${LOG_TME:10:2}:${LOG_TME:12:2}"
# or...

Ksh, bash, all posix-sh you can use builtin substr:

logmm=${NLOG_TME:0:2}
logdd=${NLOG_TME:2:2}
# ...
# no need echo, simply:
NLOG_TME="$logyy-$logmm-$logdd $logth:$logtm:$logts"

In this forum have been lot of questions parse xml using awk.

I have published some solution (https://github.com/kshji/awk/tree/master/Xml) in one location. Sort toolset to parse xml using awk....

Here is example to calculate diff between login and session time ... easy to modify ..
using ksh93

#!/usr/bin/ksh
prev=""
name="someuser"

for logu in /tmp/$lognam/wtmp*
do
last -f...

Give example data which you have tried to parse.

Env problem ?
Add line

env > /tmp/sapenv.tmp

to the script and run manually and via cron. Compare env.

Make your env file

cd /usr/sap/saprouter
env > myenv


Edit myenv if needed.
...

Maybe this example helps use to make comparing little easier in shell. You can add breaks as you like.

for file in *.txt
do
stat="DEPTH"
while IFS="," read f1 f2 DEPT SDEPT rest
do...

Works fine in posix shells: dash, ksh, zsh, bash, ...

# this solution no problem even input lines include some special chars, $IFS chars and so on.
while read file
do
first_line=$(head -1...

Have tested:

# add to to the pipe : delete spaces
# | tr -d " "
# and then make string comparing
while [ "$process_count" != 0 ]

ksh93 using builtin time conversion and calculation:

#!/usr/bin/ksh93

X1=02 ### Hour
Y1=20160405 ## Date
Z1=4 ## I want to Back 4 Hour

day=$(printf "%(%Y-%m-%d)T" $Y1)
timestamp="$day...

oifs="$IFS"
while IFS="|" read county_code county_name usr_nm passwd domain xstr
do
IFS="$oifs"
echo "Connecting to ${domain}"

$ORACLE_HOME/bin/sqlplus...

Here is two examples. Also read -p is usable.

Using read -n 1.

#!/bin/ksh
# bash/ksh93/zsh/... but not dash (=posix)

# if read key ENTER, bash return newline and ksh return carriage return...

Most of syntax between ksh93 (https://github.com/att/ast) and bash is same. All previous scripts works fine in bash (http://www.gnu.org/software/bash/manual/bashref.html) and ksh93.
Both support...

Look system cmd (http://php.net/manual/en/function.system.php), but http-server has not usually enough privileges to do ex. rm some user directory.

Need sudo...

awk '
BEGIN {
FS=":"
}

{ # default for all lines
color="yellow"
}

$1 == "red" {
color=$1
}

{ # all lines
...

Example forever loop and if status is 0, then break the loop. Case is nice command to use string comparing.


#!/bin/bash
# or ksh
while true
do
# your db code here
# after that:...

Usually CGI need to return content type. And using HERE method it's easier to write web page with variable and command expand. Rest is html/javascript problem, not scripting.


#!/bin/bash
cat...

for i in server1*
do
[ "$i" = "server1*" ] && continue # there wasn't any files
# remove echo if looks good
echo mv "$i" "${i/server1/server2}"
done

Please: give the full if - then fi block. Easier to say.

If you need really put something to do in background, then do something (parallel) and after that wait the child process to end, solution could be something like:


child()
{
parentfile=$1

...