I have the following extract from a script that fetches weather information from accuweather:
wget -O ./weather_raw $address
if [[ -s ./weather_raw ]]; then
egrep 'Currently|Forecast<\/title>|_31x31.gif' ./weather_raw > ./weather
sed -i '/AccuWeather\|Currently in/d' ./weather
sed -i -e 's/^[ \t]*//g' -e 's/<title>\|<\/title>\|<description>\|<\/description>//g' ./weather
sed -i -e 's/<img src="/\n/g' ./weather
sed -i '/^$/d' ./weather
sed -i -e 's/_31x31.*$//g' -e 's/^.*\/icons\///g' ./weather
sed -i -e '1s/.$//' -e '3s/.$//' -e '6s/.$//' ./weather
for (( i=2; i<=8; i+=3 ))
do
im=$(sed -n ${i}p ./weather)
sed -i $i"s/^.*$/$(test_image $im)/" ./weather
done
fi
The command that triggers the code above is in a conkyrc file and its ~/.conkyblue/accu_weather/rss/acc_rss. When I run the conkyrc script from the prompt, I get an error
sed: can't read /home/me/weather: No such file or directory
And indeed when I check, the "weather" file is not created. However if run the command ~/.conkyblue/accu_weather/rss/acc_rss directly from the prompt, it works as expected and create and puts content into the /home/me/weather file.
I don't know anything about the sed command although I'm trying to learn it as a result of this bother.
What could be the problem with the code. I don't think its a permission issue since the folder its writing into is my home folder and I of-course own it.
Thanks
It should have been created by egrep.
When you run your script, the weather directory will be created in the pwd of the script process.
Check and see why egrep does not create the file, or in which directory it created it.
Related
I'm having trouble in getting sed to remove just the specific line I want. Let's say I have a file that looks like this:
testfile
testfile.txt
testfile2
Currently I'm using this to remove the line I want:
sed -i "/$1/d" file
The issue is that with this if I were to give testfile as input it would delete all three lines but I want it to only remove the first line. How do I do this?
With grep
grep -x -F -v -- "$1" file
# or
grep -xFv -- "$1" file
-F is for "fixed strings" -- turns off regex engine.
-x is to match entire line.
-v is for "everything but" the matched line(s).
-- to signal the end of options, in case $1 starts with a hyphen.
To save the file
grep -xFv -- "$1" file | sponge file # `moreutils` package
# or
tmp=$(mktemp)
grep -xFv -- "$1" file > "$tmp" && mv "$tmp" file
So match the whole line.
var=testfile
sed -i '/^'"$var"'$/d' file
# or with " quoting
sed -i "/^$var\$/d" file
You can learn regex with fun online with regex crosswords.
I want to add below line to build.sh file to a line number 26
sed -i 's/-DskipTests //' dev/make-distribution.sh
I tried with this command
sed "26 a sed -i 's/-DskipTests //' dev/make-distribution.sh" build.sh
But this is giving error
sed: 1: "26 a sed -i 's/-DskipTe ...": command a expects \ followed by text`
Try putting it in as if it were multiline add as a workaround. Does this work for you?
sed "26 a\\sed -i 's/-DskipTests //' dev/make-distribution.sh" build.sh
You need to quote your quote since you are compounding quote-types.
c.f. the manual, though - it ought to work as it is.
I need to run one input file after another (same folder) and have previously used something along the lines of:
sed -i -e 's/file1/file1/g'
sh run file 1
sed -i -e 's/file1/file2/g'
sh run file 2
However, I recently tried to use this and it wouldn't work. The run file works so I assume I'm just using the sed -i -e command incorrectly?
I'm requesting help with a very simple script...
#!/usr/bin/sed -f
sed '/11,yahoo/d'
sed '/2506,stackover flow/d'
sed '/2536,reddit/d'
Just need it to remove three matches that account for 18408 in my file, data.csv
% sed -f remove.sed < data.csv
sed: 3: remove.sed: unterminated substitute pattern
Doing these same lines individually is no problem at all, so what am I doing wrong with this?
Using freeBSD 10.1 and its implementation of sed, if that matters.
This, being a sed script, should not have "sed" at each line.
Either change it to:
#!/usr/bin/sed -f
/11,yahoo/d
/2506,stackover flow/d
/2536,reddit/d
Or to
#!/bin/sh
sed -e /11,yahoo/d \
-e /2506,stackover flow/d \
-e /2536,reddit/d
I'm trying to run this sed script on all the files in a directory:
sed.s:
/<constants>/a\
<const type="profElem" name="mission_description" value="NCEP and NCAR Reanalysis Monthly Means and Other Derived Variables"/>
but whenever I run:
find . -exec sed -f sed.s -i {} \;
I get the error:
sed: -i may not be used with stdin
How do I get this to work?
It appears that your version of sed requires you to pass an extension for backups to the -i option. If you feel pretty confident in your command you could try to give it a zero-length extension like so:
find . -exec sed -f sed.s -i '' {} \;