Ive got an alias to view my apache error log, which uses sed to format newline characters into newlines for better display. The following command works perfectly fine when entered via command line, but wont work as an alias. Every time I run this, I need to open my .profile, copy the contents of the alias, and paste it in command line.
tail -f /var/log/httpd/my-sandbox-error_log | sed -e 's/\\n/\n/g'
alias:
alias elog="tail -f /var/log/httpd/my-sandbox-error_log | sed -e 's/\\n/\n/g'"
Ive tried a number of approaches, have swapped the quotation characters, and experimented with escape characters a'plenty. It seems I can never get my alias to make use of the sed search/replace (error log is tailed, with no newline formatting). Im sure there is something trivial missing, to which I am naive. I am not a unix/linux expert.
Can anyone enlighten me as to why this does not work as an alias?
If you did set -x, echo alias elog=.., alias elog or otherwise got bash to write the result back to you, you'd see why it isn't working. \\ in double quotes becomes \.
The rule of thumb is that if you have to ask, you've exceeded the usefulness of an alias. Use a function instead:
elog() {
tail -f /var/log/httpd/my-sandbox-error_log | sed -e 's/\\n/\n/g'
}
This way you don't need any additional escaping.
Related
Let's say I have a line looking like this:
/Users/random/354765478/Tests/StoreTests/Base64Tests.swift
In this example, I would like to get the result:
Tests/StoreTests/Base64Tests.swift
How can I do if I want to get everything before the first pattern match (either Sources or Tests) using sed or awk?
I am using sed 's/^.*\(Tests.*\).*$/\1/' right now but it's falling:
echo '/Users/random/354765478/Tests/StoreTests/Base64Tests.swift' | sed 's/^.*\(Tests\)/\1/'
Tests.swift
Here's another example using Sources (which seems to work):
echo '/Users/random/741672469/Sources/Store/StoreDataSource.swift' | sed 's/^.*\(Sources\)/\1/'
Sources/Store/StoreDataSource.swift
I would like to get everything before the first, and not the last Sources or Tests pattern match.
Any help would be appreciated!
How can I do if I want to get everything before the first pattern match (either Sources or Tests).
Easier to use a grep -o here:
grep -Eo '(Sources|Tests)/.*' file
Tests/StoreTests/Base64Tests.swift
Sources/Store/StoreDataSource.swift
# where input file is
cat file
/Users/random/354765478/Tests/StoreTests/Base64Tests.swift
/Users/random/741672469/Sources/Store/StoreDataSource.swift
Breakdown:
Regex pattern (Sources|Tests)/.* match any text that starts with Sources/ or Tests/ until end of the line.
-E: enables extended regex mode
-o: prints only matched text instead of full line
Alternatively you may use this awk as well:
awk 'match($0, /(Sources|Tests)\/.*/) {
print substr($0, RSTART)
}' file
Tests/StoreTests/Base64Tests.swift
Sources/Store/StoreDataSource.swift
Or this sed:
sed -E 's~.*/((Sources|Tests)/.*)~\1~' file
Tests/StoreTests/Base64Tests.swift
Sources/Store/StoreDataSource.swift
With your shown samples please try following GNU grep. This will look for very first match of /Sources OR /Tests and then print values from these strings to till end of the value.
grep -oP '^.*?\/\K(Sources|Tests)\/.*' Input_file
Using sed
$ sed -E 's~([^/]*/)+((Tests|Sources).*)~\2~' input_file
Tests/StoreTests/Base64Tests.swift
would like to get everything before the first, and not the last
Sources or Tests pattern match.
First thing is to understand reason of that, you are using
sed 's/^.*\(Tests.*\).*$/\1/'
observe that * is greedy, i.e. it will match as much as possible, therefore it will always pick last Tests, if it would be non-greedy it would find first Tests but sed does not support this, if you are using linux there is good chance that you have perl command which does support that, let file.txt content be
/Users/random/354765478/Tests/StoreTests/Base64Tests.swift
then
perl -p -e 's/^.*?(Tests.*)$/\1/' file.txt
gives output
Tests/StoreTests/Base64Tests.swift
Explanation: -p -e means engage sed-like mode, alterations in regular expression made: brackets no longer require escapes, first .* (greedy) changed to .*? (non-greedy), last .* deleted as superfluous (observe that capturing group will always extended to end of line)
(tested in perl 5, version 30, subversion 0)
Each file's name starts with "input". One example of the files look like:
0.0005
lii_bk_new
traj_new.xyz
0
73001
146300
I want to delete the lines which only includes '0' and the expected output is:
0.0005
lii_bk_new
traj_new.xyz
73001
146300
I have tried with
sed -i 's/^0\n//g' input_*
and
grep -RiIl '^0\n' input_* | xargs sed -i 's/^0\n//g'
but neither works.
Please give some suggestions.
Could you please try changing your attempted code to following, run it on a single Input_file once.
sed 's/^0$//' Input_file
OR as per OP's comment to delete null lines:
sed 's/^0$//;/^$/d' Input_file
I have intentionally not put -i option here first test this in a single file of output looks good then only run with -i option on multiple files.
Also problem in your attempt was, you are putting \n in regex of sed which is default separator of line, we need to put $ in it to tell sed delete those lines which starts and ends with 0.
In case you want to take backup of files(considering that you have enough space available in your file system) you could use -i.bak option of sed too which will take backup of each file before editing(this isn't necessary but for safer side you have this option too).
$ sed '/^0$/d' file
0.0005
lii_bk_new
traj_new.xyz
73001
146300
In your regexp you were confusing \n (the literal LineFeed character which will not be present in the string sed is analyzing since sed reads one \n-separated line at a time) with $ (the end-of-string regexp metacharacter which represents end-of-line when the string being parsed is a line as is done with sed by default).
The other mistake in your script was replacing 0 with null in the matching line instead of just deleting the matching line.
Please give some suggestions.
I would use GNU awk -i inplace for that following way:
awk -i inplace '!/^0$/' input_*
This simply will preserve all lines which do not match ^0$ i.e. (start of line)0(end of line). If you want to know more about -i inplace I suggest reading this tutorial.
In the following file I want to replace all the ; by , with the exception that, when there is a string (delimited with two "), it should not replace the ; inside it.
Example:
Input
A;B;C;D
5cc0714b9b69581f14f6427f;5cc0714b9b69581f14f6428e;1;"5cc0714b9b69581f14f6427f;16a4fba8d13";xpto;
5cc0723b9b69581f14f64285;5cc0723b9b69581f14f64294;2;"5cc0723b9b69581f14f64285;16a4fbe3855";xpto;
5cc072579b69581f14f6428a;5cc072579b69581f14f64299;3;"5cc072579b69581f14f6428a;16a4fbea632";xpto;
output
A,B,C,D
5cc0714b9b69581f14f6427f,5cc0714b9b69581f14f6428e,1,"5cc0714b9b69581f14f6427f;16a4fba8d13",xpto,
5cc0723b9b69581f14f64285,5cc0723b9b69581f14f64294,2,"5cc0723b9b69581f14f64285;16a4fbe3855",xpto,
5cc072579b69581f14f6428a,5cc072579b69581f14f64299,3,"5cc072579b69581f14f6428a;16a4fbea632",xpto,
For sed I have: sed 's/;/,/g' input.txt > output.txt but this would replace everything.
The regex for the " delimited string: \".*;.*\" .
(A regex for hexadecimal would be better -- something like: [0-9a-fA-F]+)
My problem is combining it all to make a grep -o / sed that replaces everything except for that pattern.
The file size is in the order of two digit Gb (max 99Gb), so performance is important. Relevant.
Any ideas are appreciated.
sed is for doing simple s/old/new on individual strings. grep is for doing g/re/p. You're not trying to do either of those tasks so you shouldn't be considering either of those tools. That leaves the other standard UNIX tool for manipulating text - awk.
You have a ;-separated CSV that you want to make ,-separated. That's simply:
$ awk -v FPAT='[^;]*|"[^"]+"' -v OFS=',' '{$1=$1}1' file
A,B,C,D
5cc0714b9b69581f14f6427f,5cc0714b9b69581f14f6428e,1,"5cc0714b9b69581f14f6427f;16a4fba8d13",xpto,
5cc0723b9b69581f14f64285,5cc0723b9b69581f14f64294,2,"5cc0723b9b69581f14f64285;16a4fbe3855",xpto,
5cc072579b69581f14f6428a,5cc072579b69581f14f64299,3,"5cc072579b69581f14f6428a;16a4fbea632",xpto,
The above uses GNU awk for FPAT. See What's the most robust way to efficiently parse CSV using awk? for more details on parsing CSVs with awk.
If I get correctly your requirements, one option would be to make a three pass thing.
From your comment about hex, I'll consider nothing like # will come in the input so you can do (using GNU sed) :
sed -E 's/("[^"]+);([^"]+")/\1#\2/g' original > transformed
sed -i 's/;/,/g' transformed
sed -i 's/#/;/g' transformed
The idea being to replace the ; when within quotes by something else and write it to a new file, then replace all ; by , and then set back the ; in place within the same file (-i flag of sed).
The three pass can be combined in a single command with
sed -E 's/("[^"]+);([^"]+")/\1#\2/g;s/;/,/g;s/#/;/g' original > transformed
That said, there's probably a bunch of csv parser witch already handle quoted fields that you can probably use in the final use case as I bet this is just an intermediary step for something else later in the chain.
From Ed Morton's comment: if you do it in one pass, you can use \n as replacement separator as there can't be a newline in the text considered line by line.
This might work for you (GNU sed):
sed -E ':a;s/^([^"]*("[^"]*"[^"]*)*"[^";]*);/\1\n/;ta;y/;/,/;y/\n/;/' file
Replace ;'s inside double quotes with newlines, transpose ;'s to ,'s and then transpose newlines to ;'s.
I want to open a text file that has a list of 500 IP addresses. I want to make the following changes to one of the lines and save the file. Is it possible to do that with awk or sed?
current line :
100.72.78.46:1900
changes :
100.72.78.46:1800
You can achieve that with the following:
sed -ie 's/100.72.78.46:1900/100.72.78.46:1800/' file.txt
The i option will update the original file, and a backup file will be created. This will edit only the first occurrence of the pattern. If you want to replace all matching patterns, add a g after the last /
This solution, however (as point out on the comments) fails in many other instances, such as 72100372578146:190032, which would transform into 72100.72.78.46:180032.
To circumvent that, you'd have to do an exact match, and also not treat the . as special character (see here):
sed -ie 's/\<100\.72\.78\.46:1900\>/100.72.78.46:1800/g' file.txt
note the \. and the \<...\> "word boundary" notation for the exact match. This solution worked for me on a Linux machine, but not on a MAC. For that, you would have to use a slightly different syntax (see here):
sed -ie 's/[[:<:]]100\.72\.78\.46:1900[[:>:]]/100.72.78.46:1800/g' file.txt
where the [[:<:]]...[[:>:]] would give you the exact match.
finally, I also realized that, if you have only one IP address per line, you could also use the special characters ^$ for the beginning and end of line, preventing the erroneous replacement:
sed -ie 's/^100\.72\.78\.46:1900$/100.72.78.46:1800/g' file.txt
I have a bash script that I use to setup a simple php script on my server. I am stuck with how to correctly change a variable with sed with the script. Here is what I have tried:
echo "Enter Portal Password:"
read PORTPASS;
sed -i 's/$ppass =".*"/$ppass ="$PORTPASS"/' includes/config.php
The above changes the variable in the config file but it only changes it to $PORTPASS it does not change it to what I input in the script.
I also tried this and it does change the $PORTPASS correctly, but it remove the " " around the variable in the file.
sed -i 's/$ppass ='".*"'/$ppass ='"$PORTPASS"';/' includes/config.php
Here is what I'm trying to change in the conf.php file: $ppass ="password";
Try:
sed -i "s/\$ppass =\".*\"/\$ppass =\"$PORTPASS\"/" includes/config.php
You have to use double quotes (") around the command so that the shell will evaluate $PORTPASS before passing it to sed, so then you have to "escape" all of the double quotes within the command.