Is there a way to conditionally pipe the output of a command through sed, in a bash script? Depending upon a script option, I either want to pipe the output of a long pipe through sed, or omit the pipe through sed. Currently I'm doing
if [ $pipeit ]; then
sed_args='/omit this line/d'
else
sed_args='/$^/d' # pass-thru (what's a better sed pass thru?)
fi
some_cmd | sed "$sed_args"
I would keep it as simple as:
if [ $pipeit ]; then
some_cmd | sed '/omit this line/d'
else
some_cmd
fi
Why should you call sed if you don't need it? Just for your information, a possible sed command that does not change the input would be sed -n p
Btw, if some_cmd is kind of a large beast and you want to avoid duplicating it, wrap it into a function.
By default sed prints all lines:
if [ $pipeit ]; then
sed_args='/omit this line/d'
else
sed_args="" # pass-thru
fi
some_cmd | sed "${sed_args}"
There is this other tested solution:
some_cmd | if [ $pipeit ]; then
sed "/omit this line/d"
else
sed ""
fi
cat could be used instead of the sed ""
Finally, a string can be built and executed using eval.
some_cmd='printf "foo\n\nbar\n"'
if [ $pipeit ]; then
conditional_pipe='| sed "/foo/d"'
else
conditional_pipe=""
fi
eval "${some_cmd}" "${conditional_pipe}"
If some_cmd is complex it migth be tricky to build a string that would behave as expected with eval.
----
First solution for history
Using an impossible match with sed would make it print all lines to stdout:
$ printf "foo\n\nbar\n" | sed "/./{/^$/d}"
foo
bar
/./ selects a line with at least one char.
/^$/ selects an empty line.
Related
Why doesn't the follow bash script work? I would like it to output
two lines like this:
XXXXXXX
YYYYYYY
It works if I change the sed line to use a filename instead of the variable, but I want to use the variable.
#!/bin/bash
input=$(echo -e '=======\n-------\n')
for sym in = -; do
if [ "$sym" == '-' ]; then
replace=Y
else
replace=X
fi
printf "%s\n" "s/./$replace/g"
done | sed -f- <<<"$input"
The main problem is that you're giving sed two sources to read standard input from: the for loop that is fed through the pipe, and the variable coming through the here-string. As it turns out, the here-string gets precedence and sed complains that there are extra characters after a command (= is a command).
Instead of a here-string, you could use process substitution:
for sym in = -; do
if [ "$sym" == '-' ]; then
replace=Y
else
replace=X
fi
printf "%s\n" "s/./$replace/g"
done | sed -f- <(printf '%s\n' '=======' '-------')
You'll notice that the output isn't what you want, though, namely
YYYYYYY
YYYYYYY
This is because the sed script you end up with looks like this:
s/./X/g
s/./Y/g
No matter what you do first, the last command replaces everything with Y.
I'm trying to define a bash function, highlight, that I can use to highlight search terms in the output of a previous command. When I do this from the terminal, it works fine as follows:
# highlight all occurrences of bar in file foo
cat foo | egrep '(bar|$)'
Yes, catting is a simplified example, but it demonstrates how I can do this from the command line. I'd like to use this generically as: cat foo | hightlight bar
From what I've read, I can't simply pipe results to egrep like I hoped so I naively tried defining my bash function as:
highlight() {
while read line; do
pat="'("$1"|$)'"
echo \"$line\" | egrep $pat
done
}
However, this isn't working. Please advise.
Your quoting is just about entirely wrong.
pat="'("$1"|$)'"
you include literal single quotes in the pattern, and you're actually not quoting the function parameter.
echo \"$line\" | egrep $pat
You're including literal double quotes in the echo statement, and failing to quote both variables.
This is better:
highlight() {
while read -r line; do
pat="($1|$)"
echo "$line" | grep -E "$pat"
done
}
However, grep knows how to read from stdin, so simplify:
highlight() { grep -E "($1|$)"; }
I'm not sure what you read, but it is wrong. egrep stands for extended grep because it is using the extended POSIX regular expression syntax. It behaves like the standard grep -E
Read the man page:
egrep is the same as grep -E
It seems to me you should have changed this
pat="'("$1"|$)'"
To
pat="($1|\$)"
Also, I don't see the need of "$". In addition, I think it is better to move the initialization of "pat" out of the loop. Here is what I got (tested):
#!/bin/bash
function highlight {
pattern="$1"
while read line; do
echo "$line" | egrep --color "$pattern"
done
}
echo -e 'a\nb\nbar\nhibar' | highlight bar
I have written the following filter as a function in my ~/.bash_profile:
hilite() {
export REGEX_SED=$(echo $1 | sed "s/[|()]/\\\&/g")
while read line
do
echo $line | egrep "$1" | sed "s/$REGEX_SED/\x1b[7m&\x1b[0m/g"
done
exit 0
}
to find lines of anything piped into it matching a regular expression, and highlight matches using ANSI escape codes on a VT100-compatible terminal.
For example, the following finds and highlights the strings bin, U or 1 which are whole words in the last 10 lines of /etc/passwd:
tail /etc/passwd | hilite "\b(bin|[U1])\b"
However, the script runs very slowly as each line forks an echo, egrep and sed.
In this case, it would be more efficient to do egrep on the entire input, and then run sed on its output.
How can I modify my function to do this? I would prefer to not create any temporary files if possible.
P.S. Is there another way to find and highlight lines in a similar way?
sed can do a bit of grepping itself: if you give it the -n flag (or #n instruction in a script) it won't echo any output unless asked. So
while read line
do
echo $line | egrep "$1" | sed "s/$REGEX_SED/\x1b[7m&\x1b[0m/g"
done
could be simplified to
sed -n "s/$REGEX_SED/\x1b[7m&\x1b[0m/gp"
EDIT:
Here's the whole function:
hilite() {
REGEX_SED=$(echo $1 | sed "s/[|()]/\\\&/g");
sed -n "s/$REGEX_SED/\x1b[7m&\x1b[0m/gp"
}
That's all there is to it - no while loop, reading, grepping, etc.
If your egrep supports --color, just put this in .bash_profile:
hilite() { command egrep --color=auto "$#"; }
(Personally, I would name the function egrep; hence the usage of command).
I think you can replace the whole while loop with simply
sed -n "s/$REGEX_SED/\x1b[7m&\x1b[0m/gp"
because sed can read from stdin line-by-line so you don't need read
I'm not sure if running egrep and piping to sed is faster than using sed alone, but you can always compare using time.
Edit: added -n and p to sed to print only highlighted lines.
Well, you could simply do this:
egrep "$1" $line | sed "s/$REGEX_SED/\x1b[7m&\x1b[0m/g"
But I'm not sure that it'll be that much faster ; )
Just for the record, this is a method using a temporary file:
hilite() {
export REGEX_SED=$(echo $1 | sed "s/[|()]/\\\&/g")
export FILE=$2
if [ -z "$FILE" ]
then
export FILE=~/tmp
echo -n > $FILE
while read line
do
echo $line >> $FILE
done
fi
egrep "$1" $FILE | sed "s/$REGEX_SED/\x1b[7m&\x1b[0m/g"
return $?
}
which also takes a file/pathname as the second argument, for case like
cat /etc/passwd | hilite "\b(bin|[U1])\b"
my question seems to be general, but i can't find any answers.
In sed command, how can you replace the substitution pattern by a value returned by a simple bash function.
For instance, I created the following function :
function parseDates(){
#Some process here with $1 (the pattern found)
return "dateParsed;
}
and the folowing sed command :
myCatFile=`sed -e "s/[0-3][0-9]\/[0-1][0-9]\/[0-9][0-9]/& parseDates &\}/p" myfile`
I found that the caracter '&' represents the current pattern found, i'd like it to be passed to my bash function and the whole pattern to be substituted by the pattern found +dateParsed.
Does anybody have an idea ?
Thanks
you can use the "e" option in sed command like this:
cat t.sh
myecho() {
echo ">>hello,$1<<"
}
export -f myecho
sed -e "s/.*/myecho &/e" <<END
ni
END
you can see the result without "e":
cat t.sh
myecho() {
echo ">>hello,$1<<"
}
export -f myecho
sed -e "s/.*/myecho &/" <<END
ni
END
Agree with Glenn Jackman.
If you want to use bash function in sed, something like this :
sed -rn 's/^([[:digit:].]+)/`date -d #&`/p' file |
while read -r line; do
eval echo "$line"
done
My file here begins with a unix timestamp (e.g. 1362407133.936).
Bash function inside sed (maybe for other purposes):
multi_stdin(){ #Makes function accepet variable or stdin (via pipe)
[[ -n "$1" ]] && echo "$*" || cat -
}
sans_accent(){
multi_stdin "$#" | sed '
y/àáâãäåèéêëìíîïòóôõöùúûü/aaaaaaeeeeiiiiooooouuuu/
y/ÀÁÂÃÄÅÈÉÊËÌÍÎÏÒÓÔÕÖÙÚÛÜ/AAAAAAEEEEIIIIOOOOOUUUU/
y/çÇñÑߢÐð£Øø§µÝý¥¹²³ªº/cCnNBcDdLOoSuYyY123ao/
'
}
eval $(echo "Rogério Madureira" | sed -n 's#.*#echo & | sans_accent#p')
or
eval $(echo "Rogério Madureira" | sed -n 's#.*#sans_accent &#p')
Rogerio
And if you need to keep the output into a variable:
VAR=$( eval $(echo "Rogério Madureira" | sed -n 's#.*#echo & | desacentua#p') )
echo "$VAR"
do it step by step. (also you could use an alternate delimiter , such as "|" instead of "/"
function parseDates(){
#Some process here with $1 (the pattern found)
return "dateParsed;
}
value=$(parseDates)
sed -n "s|[0-3][0-9]/[0-1][0-9]/[0-9][0-9]|& $value &|p" myfile
Note the use of double quotes instead of single quotes, so that $value can be interpolated
I'd like to know if there's a way to do this too. However, for this particular problem you don't need it. If you surround the different components of the date with ()s, you can back reference them with \1 \2 etc and reformat however you want.
For instance, let's reverse 03/04/1973:
echo 03/04/1973 | sed -e 's/\([0-9][0-9]\)\/\([0-9][0-9]\)\/\([0-9][0-9][0-9][0-9]\)/\3\/\2\/\1/g'
sed -e 's#[0-3][0-9]/[0-1][0-9]/[0-9][0-9]#& $(parseDates &)#' myfile |
while read -r line; do
eval echo "$line"
done
You can glue together a sed-command by ending a single-quoted section, and reopening it again.
sed -n 's|[0-3][0-9]/[0-1][0-9]/[0-9][0-9]|& '$(parseDates)' &|p' datefile
However, in contrast to other examples, a function in bash can't return strings, only put them out:
function parseDates(){
# Some process here with $1 (the pattern found)
echo dateParsed
}
I have little problem with specifying my variable. I have a file with normal text and somewhere in it there are brackets [ ] (only 1 pair of brackets in whole file), and some text between them. I need to capture the text within these brackets in a shell (bash) variable. How can I do that, please?
Bash/sed:
VARIABLE=$(tr -d '\n' filename | sed -n -e '/\[[^]]/s/^[^[]*\[\([^]]*\)].*$/\1/p')
If that is unreadable, here's a bit of an explanation:
VARIABLE=`subexpression` Assigns the variable VARIABLE to the output of the subexpression.
tr -d '\n' filename Reads filename, deletes newline characters, and prints the result to sed's input
sed -n -e 'command' Executes the sed command without printing any lines
/\[[^]]/ Execute the command only on lines which contain [some text]
s/ Substitute
^[^[]* Match any non-[ text
\[ Match [
\([^]]*\) Match any non-] text into group 1
] Match ]
.*$ Match any text
/\1/ Replaces the line with group 1
p Prints the line
May I point out that while most of the suggested solutions might work, there is absolutely no reason why you should fork another shell, and spawn several processes to do such a simple task.
The shell provides you with all the tools you need:
$ var='foo[bar] pinch'
$ var=${var#*[}; var=${var%%]*}
$ echo "$var"
bar
See: http://mywiki.wooledge.org/BashFAQ/073
Sed is not necessary:
var=`egrep -o '\[.*\]' FILENAME | tr -d ][`
But it's only works with single line matches.
Using Bash builtin regex matching seems like yet another way of doing it:
var='foo[bar] pinch'
[[ "$var" =~ [^\]\[]*\[([^\[]*)\].* ]] # Bash 3.0
var="${BASH_REMATCH[1]}"
echo "$var"
Assuming you are asking about bash variable:
$ export YOUR_VAR=$(perl -ne'print $1 if /\[(.*?)\]/' your_file.txt)
The above works if brackets are on the same line.
What about:
shell_variable=$(sed -ne '/\[/,/\]/{s/^.*\[//;s/\].*//;p;}' $file)
Worked for me on Solaris 10 under Korn shell; should work with Bash too. Replace '$(...)' with back-ticks in Bourne shell.
Edit: worked when given [ on one line and ] on another. For the single line case as well, use:
shell_variable=$(sed -n -e '/\[[^]]*$/,/\]/{s/^.*\[//;s/\].*//;p;}' \
-e '/\[.*\]/s/^.*\[\([^]]*\)\].*$/\1/p' $file)
The first '-e' deals with the multi-line spread; the second '-e' deals with the single-line case. The first '-e' says:
From the line containing an open bracket [ not followed by a close bracket ] on the same line
Until the line containing close bracket ],
substitute anything up to and including the open bracket with an empty string,
substitute anything from the close bracket onwards with an empty string, and
print the result
The second '-e' says:
For any line containing both open bracket and close bracket
Substitute the pattern consisting of 'characters up to and including open bracket', 'characters up to but excluding close bracket' (and remember this), 'stuff from close bracket onwards' with the remembered characters in the middle, and
print the result
For the multi-line case:
$ file=xxx
$ cat xxx
sdsajdlajsdl
asdajsdkjsaldjsal
sdasdsad [aaaa
bbbbbbb
cccc] asdjsalkdjsaldjlsaj
asdjsalkdjlksjdlaj
asdasjdlkjsaldja
$ shell_variable=$(sed -n -e '/\[[^]]*$/,/\]/{s/^.*\[//;s/\].*//;p;}' \
-e '/\[.*\]/s/^.*\[\([^]]*\)\].*$/\1/p' $file)
$ echo $shell_variable
aaaa bbbbbbb cccc
$
And for the single-line case:
$ cat xxx
sdsajdlajsdl
asdajsdkjsaldjsal
sdasdsad [aaaa bbbbbbb cccc] asdjsalkdjsaldjlsaj
asdjsalkdjlksjdlaj
asdasjdlkjsaldja
$
$ shell_variable=$(sed -n -e '/\[[^]]*$/,/\]/{s/^.*\[//;s/\].*//;p;}' \
-e '/\[.*\]/s/^.*\[\([^]]*\)\].*$/\1/p' $file)
$ echo $shell_variable
aaaa bbbbbbb cccc
$
Somewhere about here, it becomes simpler to do the whole job in Perl, slurping the file and editing the result string in two multi-line substitute operations.
var=`grep -e '\[.*\]' test.txt | sed -e 's/.*\[\(.*\)\].*/\1/' infile.txt`
Thanks to everyone, i used Strager's version and works perfectly, thanks alot once again...
var=`grep -e '\[.*\]' test.txt | sed -e 's/.*\[\(.*\)\].*/\1/' infile.txt`
Backslashes (BSL) got munched up ... :
var='foo[bar] pinch'
[[ "$var" =~ [^\]\[]*\[([^\[]*)\].* ]] # Bash 3.0
# Just in case ...:
[[ "$var" =~ [^BSL]BSL[]*BSL[([^BSL[]*)BSL].* ]] # Bash 3.0
var="${BASH_REMATCH[1]}"
echo "$var"
2 simple steps to extract the text.
split var at [ and get the right part
split var at ] and get the left part
cb0$ var='foo[bar] pinch'
cb0$ var=${var#*[}
cb0$ var=${var%]*} && echo $var
bar