I'm trying to extract the db element's value attributes from an xml file using xmllint
My XPath Query needs to navigate to the correct dbtype, which I wish to store in an environment variable
Without an environment variable (hard coded), this command works
echo 'cat //rdbmsinfo/dbtype[#value="sqlserver"]/db/#value' | xmllint --shell "config.xml" | grep -v ">" | cut -f 2 -d "=" | tr -d \ | sed 's/"//g'
I am trying with below command:
echo 'cat //rdbmsinfo/dbtype[#value="$ldb_source_typ"]/db/#value' | xmllint --shell "config.xml" | grep -v ">" | cut -f 2 -d "=" | tr -d \ | sed 's/"//g'
where ldb_source_typ is a variable to get the value from shell parameter as "sqlserver". But the above syntax is not giving any output.
Why isn't this working, and how can I fix it?
My config.xml XML looks like this:
<?xml version="1.0"?>
<rdbmsinfo>
<dbtype value="sqlserver">
<db value="sqlsrv1">
<dbhostip>192.168.0.1</dbhostip>
<dbhostportno>2000</dbhostportno>
<dbusername>sample</dbusername>
<dbpassword>sample</dbpassword>
</db>
<db value="sqlsrv2">
<dbhostip>192.168.0.2</dbhostip>
<dbhostportno>2000</dbhostportno>
<dbusername>sample</dbusername>
<dbpassword>sample</dbpassword>
</db>
</dbtype>
<dbtype value="postgresql">
<db value="postsql1">
</db>
</dbtype>
</rdbmsinfo>
As I just found out, you need to set the environment variable between single quotes ( ' ).
So the following should work:
echo 'cat //rdbmsinfo/dbtype[#value="'$ldb_source_typ'"]/db/#value' | xmllint --shell "config.xml" | grep -v ">" | cut -f 2 -d "=" | tr -d \ | sed 's/"//g'
I was able to accomplish this with the following:
echo "cat //rdbmsinfo/dbtype[#value="ldb_source_type"]/db/#value" | sed 's/ldb_source_type/"'$SQLSERVER'"/g' | xmllint --shell config.xml| grep -v ">" | cut -f 2 -d "=" | tr -d \ | sed 's/"//g'
Which produces the following output:
sqlsrv1
-------
sqlsrv2
I didn't care enough to figure out how to make the value placeholder use an $ symbol, I may come back and edit it eventually.
Edit: here's one with a $, but you need to escape it with a :
echo "cat //rdbmsinfo/dbtype[#value="\$ldb_source_type"]/db/#value" | sed 's/\$ldb_source_type/"'$SQLSERVER'"/g'| xmllint --shell config.xml| grep -v ">" | cut -f 2 -d "=" | tr -d \ | sed 's/"//g'
Related
I am trying to write a simple shell script that takes a text file as input and checks all non-punctuated words against a dictionary (english.txt). It should return all non-matching (misspelled) words. I am using grep but it does not seem to successfully match all the lines in english.txt. I have included my code below.
#!/bin/bash
cat $1 |
tr ' \t' '\n\n' |
sed -e "/'/d" |
tr -d '[:punct:]' |
tr -cd '[:alpha:]\n' |
sed -e "/^$/d" |
grep -v -i -w -f english.txt
I have the following command:
httpd.conf | grep AuthUserFile | cut -d" " -f4 | sed -e 's|["'\'']||g'
the output of this is:
/etc/httpd/secure/htpasswd.training
I did:
httpd.conf | grep AuthUserFile | cut -d" " -f4 | sed -e 's|["'\'']||g'| cat
However this just returned:
/etc/httpd/secure/htpasswd.training
I want to cat the contents of the file. How do I do this?
Piping to xargs cat will pass stdin as an argument to cat, printing the file.
Alternatively try: cat $( some command printing a filename ).
You could try surrounding your initial command with backticks as the argument to cat, like so:
cat `httpd.conf | grep AuthUserFile | cut -d" " -f4 | sed -e 's|["'\'']||g'`
This short script scrapes some log files daily to create a simple extract. It works from the command line and when I echo the $cmd and copy/paste, it also works. But it will breaks when I try to execute from the script itself.
I know this is a nightmare of patterns that I could probably improve, but am I missing something simple to just execute this correctly?
#!/bin/bash
priorday=$(date --date yesterday +"%Y-%m-%d")
outputfile="/home/CCHCS/da14/$priorday""_PROD_message_processing_times.txt"
cmd="grep 'Processed inbound' /home/rules/care/logs/RootLog* | cut -f5,6,12,16,18 -d\" \" | grep '^"$priorday"' | sed 's/\,/\./' | sed 's/ /\t/g' | sed -r 's/([0-9]+\-[0-9]+\-[0-9]+)\t/\1 /' | sed 's/ / /g' | sort >$outputfile"
printf "command to execute:\n"
echo $cmd
printf "\n"
$cmd
ouput:
./make_log_extract.sh command to execute: grep 'Processed inbound' /home/rules/care/logs/RootLog.log /home/rules/care/logs/RootLog.log.1
/home/rules/care/logs/RootLog.log.10
/home/rules/care/logs/RootLog.log.11
/home/rules/care/logs/RootLog.log.12
/home/rules/care/logs/RootLog.log.2
/home/rules/care/logs/RootLog.log.3
/home/rules/care/logs/RootLog.log.4
/home/rules/care/logs/RootLog.log.5
/home/rules/care/logs/RootLog.log.6
/home/rules/care/logs/RootLog.log.7
/home/rules/care/logs/RootLog.log.8
/home/rules/care/logs/RootLog.log.9 | cut -f5,6,12,16,18 -d" " | grep
'^2014-01-30' | sed 's/\,/./' | sed 's/ /\t/g' | sed -r
's/([0-9]+-[0-9]+-[0-9]+)\t/\1 /' | sed 's/ / /g' | sort
/home/CCHCS/da14/2014-01-30_PROD_message_processing_times.txt
grep: 5,6,12,16,18: No such file or directory
As grebneke comments, do not store the command and then execute it.
What you can do is to execute it but firstly print it: Bash: Print each command before executing?
priorday=$(date --date yesterday +"%Y-%m-%d")
outputfile="/home/CCHCS/da14/$priorday""_PROD_message_processing_times.txt"
set -o xtrace # <-- set printing mode "on"
grep 'Processed inbound' /home/rules/care/logs/RootLog* | cut -f5,6,12,16,18 -d\" \" | grep '^"$priorday"' | sed 's/\,/\./' | sed 's/ /\t/g' | sed -r 's/([0-9]+\-[0-9]+\-[0-9]+)\t/\1 /' | sed 's/ / /g' | sort >$outputfile"
set +o xtrace # <-- revert to normal
I get the ip address like that :
Ip=`ifconfig | grep inet | grep -v -E 'inet6|127.0.0.1' | \
tr -d [:alpha:] | tr -s [:space:] | cut -d: -f2`
I have an ip like this for instance : 10.1.0.76
I want to make a new variable with the Ip variable to have another ip, for instance my new variable will return : 10.1.0.178
Just the last number change, so I want to get just a part of Ip variable (10.1.0.) and add another number to the end.
I tried with sed but I always have mistakes like "there's no file call'd ..."
Can you help me ?
You can use parameter expansion: It's simply: ${Ip%.*}.178
${Ip%.*} is the ip with the last dot and everything after it removed. The .178 is what you want to append after that.
Here it is in context:
# Your original expression
Ip=`ifconfig | grep inet | grep -v -E 'inet6|127.0.0.1' | \
tr -d [:alpha:] | tr -s [:space:] | cut -d: -f2`
# assign a new variable with the ip with different end octet
newIp=${Ip%.*}.178
# Show new ip
echo "$newIp"
Well, given that you have IP in a format x.y.z.w, you can use perl regex:
$ echo "120.20.31.78" | perl -pe 's/(.*)\..*/$1\.123/'
120.20.31.123
This will repace last number ("78") with "123".
So, in your case (assuming your "Ip" variable is set correctly), it would be:
Ip=ifconfig | grep inet | grep -v -E 'inet6|127.0.0.1' | tr -d [:alpha:] | tr -s [:space:] | cut -d: -f2 | perl -pe 's/(.*)\..*/$1\.123/'
see this, I hope it is what you want:
kent$ echo $ip
10.1.0.76
kent$ echo $part
178
kent$ sed -r "s/(.*\.).*/\1$part/" <<< $ip
10.1.0.178
to set $ip with new value:
kent$ ip=$(sed -r "s/(.*\.).*/\1$part/" <<< $ip)
kent$ echo $ip
10.1.0.178
I am trying to build a Bash CGI Script that takes in coordinates as parameters from the url and uses osmosis to extract the map and the splitter and mkgmap to make the map so that it can be opened with Qlandkarte. My problem being is that when i type wget localhost/cgi-bin/script.pl?top=42&left=10&bottom=39&right=9&file=map.osm the linux terminal reads the file with the coordinates. How can I make wget just activate the script so it takes the coordinates and executes the commands. And also when the map is created at the end how can a return the file that was created by the script.
Thanks
#!/bin/bash
TOP=`echo "$QUERY_STRING" | grep -oE "(^|[?&])top=[0.0-9.0]+" | cut -f 2 -d "=" | head -n1`
LEFT=`echo "$QUERY_STRING" | grep -oE "(^|[?&])left=[0.0-9.0]+" | cut -f 2 -d "=" | head -n1`
BOTTOM=`echo "$QUERY_STRING" | grep -oE "(^|[?&])bottom=[0.0-9.0]+" | cut -f 2 -d "=" | head -n1`
RIGHT=`echo "$QUERY_STRING" | grep -oE "(^|[?&])right=[0.0-9.0]+" | cut -f 2 -d "=" | head -n1`
FILE=`echo "$QUERY_STRING" | grep -oE "(^|[?&])file=[^&]+" | sed "s/%20/ /g" | cut -f 2 -d "="`
$(sudo osmosis --read-xml file=bulgaria.osm --bounding-box top=$TOP left=$LEFT bottom=$BOTTOM right=$RIGHT --write-xml file=$FILE)
$(sudo java -Xmx900m -jar splitter.jar --max-nodes=110000 $FILE)
$(sudo java -ea -Xmx900m -jar mkgmap.jar --tdbfile --route -c template.args)
echo "Content-type: text/html"
echo ""