I'm trying to evaluate a grep expression inside of a shell script, and that grep uses a literal asterisk (*), but that asterisk appears to be expanded by my bash instead of remaining a literal asterisk:
branch_description=$(git branch --list -vv | grep "^\*")
What can I do to run grep in this context and let it receive a literal asterisk in its PATTERN argument?
A solution is to use the ascii octal code :
branch_description=$(git branch --list -vv | grep "^\052")
See
man 7 ascii
The problem is that bash is interpreting the \ and stripping it away, because it's inside double quotes. Changing to
branch_description=$(git branch --list -vv | grep '^\*')
will do what you want. See the section on QUOTING in the bash manual.
You can use single quote in grep to avoid expansion of * by shell:
branch_description=$(git branch --list -vv | grep '^\*')
Related
I'm trying to inline a jq construct that itself requires pipes. I suspect I'm running into issues because bash is treating them as bash-level pipes, rather than part of the jq.
Testing at jqplay.org, this .[1] | [.timeEnded, .lifecycleState] | flatten gets me the result I need.
Trying to embed that in bash, I am trying to do something like:
status=$(curl -X GET <URL> | jq -r -c '.[1] | [.timeEnded, .lifecycleState] | flatten' | awk -F, '{print $2}' | sed 's/"//g')
With no escaping the pipes within the jq, I get
[.timeEnded,: command not found
I tried to escape those pipes as jq -r -c '.[1] \| [.timeEnded, .lifecycleState] \| flatten' but that gets me a jq syntax error:
jq: error: syntax error, unexpected INVALID_CHARACTER, expecting $end (Unix shell quoting issues?) at <top-level>, line 1:
.[1] \| [.timeEnded, .lifecycleState] \| flatten
jq: 1 compile error
Wrapping the entire jq command in double quotes (as well as the escape chars) gave me the same syntax error. I'm sure there's probably an easy answer here, but jq is new to me.
Any help would be appreciated.
I clearly suspect that you have an unbreakable space in this part:
jq -r -c '.[1] | [...
So, edit the line manually, and replace all spaces with real spaces (taking care to not type unbreakable spaces again with AltGr+space)
Embedding jq in bash - what needs to be escaped?
Using bash and bash-like shells, jq programs can often be specified quite simply on the command line using single-quoted strings, e.g.
$ jq -n '"abc"'
"abc"
However, using this technique, single quotes are a headache since bash
does not allow single quotes within ordinary single-quoted strings. The workaround is quite horrible:
$ jq -n '"a'"'"'b"'
"a'b"
So if the jq program does have embedded single-quotes, then
it's probably time to use the -f option, but if that is not
an option, then using the form $'STRING' should be considered.
In this case, though, there are two characters that can occur in jq programs and
that will require attention: single-quotes and backslashes
For example:
$ jq -n $'"a\\tb" | "\'\\(.)\'"'
"'a\tb'"
If I'm not mistaken, the required escaping can be done using:
sed -e $'s/\'/\\\'/g' -e $'s/\\\\/\\\\\\\\/g'
I have many text files and I want to list all files that contain foo.
grep -rc will display both matched and unmatched files, but i only want to see matched files. So I came up with this ugly pipeline:
grep -rc "foo" | grep -v ":0$"
The pipeline drops auto coloring.
Can I get the same result without a pipeline?
No, GNU grep doesn't provide such a feature. But you can tell grep to always color the output, then filter out lines ending in :<some ANSI escape codes>0 like this:
grep --color=always -rc 'foo' | grep -v $':\x1B\[m\x1B\[K0$'
($'...' syntax is a bash extension though.)
Below is a screenshot of the result on my terminal.
See ANSI-C Quoting, ANSI escape code.
I m using a script that run remote on server via ssh.
Inside the script I'm using this line from below:
ls | grep -oP "\d{4} -\d{2}-\d{2}"
On my local machine that run Ubuntu the script work fine.
But when I try to run it remote I got this
grep: invalid option -- 'P'
BusyBox v1.24.1 multi-call binary.
Usage: grep [-HhnlLoqvsriwFE] [-m N] [-A/B/C N] PATTERN/-e PATTERN/...-f file [FILE]...
The first thing I thought was an alias problem, i tryed
type grep
Output is: grep is /bin/grep I think this is ok.
What worries me is BusyBox (I do not know what it is) but i think this can be the problem ?
You may use [0-9] / [[:digit:]] instead of \d with POSIX BRE (no option) or ERE (-E option):
grep -o "[0-9]\{4\} -[0-9]\{2\}-[0-9]\{2\}"
grep -oE "[0-9]{4} -[0-9]{2}-[0-9]{2}"
Note that in the first command you need to escape the braces since unescaped { and } match literal brace symbols in a POSIX BRE regex. When escaped, they mean range (interval, limiting) quantifiers. And in the second command, POSIX ERE is enabled with -E, and the behavior is reverse: when the braces are escaped they are literal chars, else they are quantifiers.
I was under the impression that within single quotes, e.g. 'pattern', bash special characters are not interpolated, so one need only escape single quotes themselves.
Why then does echo "123" | grep '[0-9]+' output nothing, whereas echo "123" | grep '[0-9]\+' (plus sign escaped) output 123? (Likewise, echo "123" | grep '3|4' outputs nothing unless you escape the |.)
This is under bash 4.1.2 and grep 2.6.3 on CentOS 6.5.
grep uses Basic Regular Expressions, like sed and vi. In that you have to escape metacharacters, and it is tedious.
You probably want Extended Regular Expressions, so use egrep or grep -E (depending on the version in use). Check your man grep.
See also the GNU documentation for a full list of the characters involved.
Most languages use Extended Regular Expressions (EREs) these days, and they are much easier to use. Basic Regular Expressions (BREs) are really a throw-back.
That seems to be the regular expression engine that grep uses. If you use a different one, it works:
$ echo "123" | grep '[0-9]+'
$ echo "123" | grep -P '[0-9]+'
123
$ echo "123" | grep '3|4'
$ echo "123" | grep -P '3|4'
123
I am trying to run grep command involving regular expression '|' on a server using ssh.
ssh rpatil#192.168.1.5 grep -E "GapEvent|GapFilled" "$logFile" > $server-$testName.log
now '|' in the command is being treated as pipe and error "no command GapFilled" is being raised.
I tried 'GapEvent|GapFilled' or '(GapEvent|GapFilled)'
so how should regular expression "GapEvent|GapFilled" should be written so that | is not treated as pipe?
You need two levels of quotes since the command line is evaluated twice (once locally when ssh is executed and once when grep is executed on the remote side). You can use one of these patterns:
"'a|b'"
'"a|b"'
"\"a|b\""
Escape the | like this \|
grep -E "GapEvent\|GapFilled" "$logFile" file
Simply use two expressions:
grep -E -e "GapEvent" -e "GapFilled" "$logFile"
-E may no longer be needed here. -F may also be a preference.