Escaping whitespace within nested shell/perl scripts - perl

I'm trying to run a perl script from within a bash script (I'll change this design later on, but for now, bear with me). The bash script receives the argument that it will run. The argument to the script is as follows:
test.sh "myscript.pl -g \"Some Example\" -n 1 -p 45"
within the bash script, I simple run the argument that was passed:
#!/bin/sh
$1
However, in my perl script the -g argument only gets "Some (that's with the quotes), instead of the Some Example. Even if I quote it, it cuts off because of the whitespace.
I tried escaping the whitespace, but it doesn't work... any ideas?

To run it as posted test.sh "myscript.pl -g \"Some Example\" -n 1 -p 45" do this:
#!/bin/bash
eval "$1"
This causes the $1 argument to be parsed by the shell so the individual words will be broken up and the quotes removed.
Or if you want you could remove the quotes and run test.sh myscript.pl -g "Some Example" -n 1 -p 45 if you changed your script to:
#!/bin/bash
"$#"
The "$#" gets replaced by all the arguments $1, $2, etc., as many as were passed in on the command line.

Quoting is normally handled by the parser, which isn't seeing them when you substitute the value of $1 in your script.
You may have more luck with:
#!/bin/sh
eval "$1"
which gives:
$ sh test.sh 'perl -le "for (#ARGV) { print; }" "hello world" bye'
hello world
bye
Note that simply forcing the shell to interpret the quoting with "$1" won't work because then it tries to treat the first argument (i.e., the entire command) as the name of the command to be executed. You need the pass through eval to get proper quoting and then re-parsing of the command.
This approach is (obviously?) dangerous and fraught with security risks.

I would suggest you name the perl script in a separate word, then you can quote the parameters when referring to them, and still easily extract the script name without needing the shell to split the words, which is the fundamental problem you have.
test.sh myscript.pl "-g \"Some Example\" -n 1 -p 45"
and then
#!/bin/sh
$1 "$2"

if you really have to do this (for whatever reason) why not just do:
sh test.sh "'Some Example' -n 1 -p 45"
in:
test.sh
RUN=myscript.pl
echo `$RUN $1
(there should be backticks ` before $RUN and after $1)

Related

Bash adding single quotes in variable [duplicate]

Let's say, you have a Bash alias like:
alias rxvt='urxvt'
which works fine.
However:
alias rxvt='urxvt -fg '#111111' -bg '#111111''
won't work, and neither will:
alias rxvt='urxvt -fg \'#111111\' -bg \'#111111\''
So how do you end up matching up opening and closing quotes inside a string once you have escaped quotes?
alias rxvt='urxvt -fg'\''#111111'\'' -bg '\''#111111'\''
seems ungainly although it would represent the same string if you're allowed to concatenate them like that.
If you really want to use single quotes in the outermost layer, remember that you can glue both kinds of quotation. Example:
alias rxvt='urxvt -fg '"'"'#111111'"'"' -bg '"'"'#111111'"'"
# ^^^^^ ^^^^^ ^^^^^ ^^^^
# 12345 12345 12345 1234
Explanation of how '"'"' is interpreted as just ':
' End first quotation which uses single quotes.
" Start second quotation, using double-quotes.
' Quoted character.
" End second quotation, using double-quotes.
' Start third quotation, using single quotes.
If you do not place any whitespaces between (1) and (2), or between (4) and (5), the shell will interpret that string as a one long word.
I always just replace each embedded single quote with the sequence: '\'' (that is: quote backslash quote quote) which closes the string, appends an escaped single quote and reopens the string.
I often whip up a "quotify" function in my Perl scripts to do this for me. The steps would be:
s/'/'\\''/g # Handle each embedded quote
$_ = qq['$_']; # Surround result with single quotes.
This pretty much takes care of all cases.
Life gets more fun when you introduce eval into your shell-scripts. You essentially have to re-quotify everything again!
For example, create a Perl script called quotify containing the above statements:
#!/usr/bin/perl -pl
s/'/'\\''/g;
$_ = qq['$_'];
then use it to generate a correctly-quoted string:
$ quotify
urxvt -fg '#111111' -bg '#111111'
result:
'urxvt -fg '\''#111111'\'' -bg '\''#111111'\'''
which can then be copy/pasted into the alias command:
alias rxvt='urxvt -fg '\''#111111'\'' -bg '\''#111111'\'''
(If you need to insert the command into an eval, run the quotify again:
$ quotify
alias rxvt='urxvt -fg '\''#111111'\'' -bg '\''#111111'\'''
result:
'alias rxvt='\''urxvt -fg '\''\'\'''\''#111111'\''\'\'''\'' -bg '\''\'\'''\''#111111'\''\'\'''\'''\'''
which can be copy/pasted into an eval:
eval 'alias rxvt='\''urxvt -fg '\''\'\'''\''#111111'\''\'\'''\'' -bg '\''\'\'''\''#111111'\''\'\'''\'''\'''
Since Bash 2.04 syntax $'string' allows a limit set of escapes.
Since Bash 4.4, $'string' also allows the full set of C-style escapes, making the behavior differ slightly in $'string' in previous versions. (Previously the $('string') form could be used.)
Simple example in Bash 2.04 and newer:
$> echo $'aa\'bb'
aa'bb
$> alias myvar=$'aa\'bb'
$> alias myvar
alias myvar='aa'\''bb'
In your case:
$> alias rxvt=$'urxvt -fg \'#111111\' -bg \'#111111\''
$> alias rxvt
alias rxvt='urxvt -fg '\''#111111'\'' -bg '\''#111111'\'''
Common escaping sequences works as expected:
\' single quote
\" double quote
\\ backslash
\n new line
\t horizontal tab
\r carriage return
Below is copy+pasted related documentation from man bash (version 4.4):
Words of the form $'string' are treated specially. The word expands to string, with backslash-escaped characters replaced as specified by the ANSI C standard. Backslash escape sequences, if present, are decoded as follows:
\a alert (bell)
\b backspace
\e
\E an escape character
\f form feed
\n new line
\r carriage return
\t horizontal tab
\v vertical tab
\\ backslash
\' single quote
\" double quote
\? question mark
\nnn the eight-bit character whose value is the octal
value nnn (one to three digits)
\xHH the eight-bit character whose value is the hexadecimal
value HH (one or two hex digits)
\uHHHH the Unicode (ISO/IEC 10646) character whose value is
the hexadecimal value HHHH (one to four hex digits)
\UHHHHHHHH the Unicode (ISO/IEC 10646) character whose value
is the hexadecimal value HHHHHHHH (one to eight
hex digits)
\cx a control-x character
The expanded result is single-quoted, as if the dollar sign had not been present.
See Quotes and escaping: ANSI C like strings on bash-hackers.org wiki for more details. Also note that "Bash Changes" file (overview here) mentions a lot for changes and bug fixes related to the $'string' quoting mechanism.
According to unix.stackexchange.com How to use a special character as a normal one? it should work (with some variations) in bash, zsh, mksh, ksh93 and FreeBSD and busybox sh.
I don't see the entry on his blog (link pls?) but according to the gnu reference manual:
Enclosing characters in single quotes
(‘'’) preserves the literal value of
each character within the quotes. A
single quote may not occur between
single quotes, even when preceded by a
backslash.
so bash won't understand:
alias x='y \'z '
however, you can do this if you surround with double quotes:
alias x="echo \'y "
> x
> 'y
I can confirm that using '\'' for a single quote inside a single-quoted string does work in Bash, and it can be explained in the same way as the "gluing" argument from earlier in the thread. Suppose we have a quoted string: 'A '\''B'\'' C' (all quotes here are single quotes). If it is passed to echo, it prints the following: A 'B' C.
In each '\'' the first quote closes the current single-quoted string, the following \' glues a single quote to the previous string (\' is a way to specify a single quote without starting a quoted string), and the last quote opens another single-quoted string.
Both versions are working, either with concatenation by using the escaped single quote character (\'), or with concatenation by enclosing the single quote character within double quotes ("'").
The author of the question did not notice that there was an extra single quote (') at the end of his last escaping attempt:
alias rxvt='urxvt -fg'\''#111111'\'' -bg '\''#111111'\''
│ │┊┊| │┊┊│ │┊┊│ │┊┊│
└─STRING──┘┊┊└─STRIN─┘┊┊└─STR─┘┊┊└─STRIN─┘┊┊│
┊┊ ┊┊ ┊┊ ┊┊│
┊┊ ┊┊ ┊┊ ┊┊│
└┴─────────┴┴───┰───┴┴─────────┴┘│
All escaped single quotes │
│
?
As you can see in the previous nice piece of ASCII/Unicode art, the last escaped single quote (\') is followed by an unnecessary single quote ('). Using a syntax-highlighter like the one present in Notepad++ can prove very helpful.
The same is true for another example like the following one:
alias rc='sed '"'"':a;N;$!ba;s/\n/, /g'"'"
alias rc='sed '\'':a;N;$!ba;s/\n/, /g'\'
These two beautiful instances of aliases show in a very intricate and obfuscated way how a file can be lined down. That is, from a file with a lot of lines you get only one line with commas and spaces between the contents of the previous lines. In order to make sense of the previous comment, the following is an example:
$ cat Little_Commas.TXT
201737194
201802699
201835214
$ rc Little_Commas.TXT
201737194, 201802699, 201835214
Simple example of escaping quotes in shell:
$ echo 'abc'\''abc'
abc'abc
$ echo "abc"\""abc"
abc"abc
It's done by finishing already opened one ('), placing escaped one (\'), then opening another one ('). This syntax works for all commands. It's very similar approach to the 1st answer.
I'm not specifically addressing the quoting issue because, well, sometimes, it's just reasonable to consider an alternative approach.
rxvt() { urxvt -fg "#${1:-000000}" -bg "#${2:-FFFFFF}"; }
which you can then call as:
rxvt 123456 654321
the idea being that you can now alias this without concern for quotes:
alias rxvt='rxvt 123456 654321'
or, if you need to include the # in all calls for some reason:
rxvt() { urxvt -fg "${1:-#000000}" -bg "${2:-#FFFFFF}"; }
which you can then call as:
rxvt '#123456' '#654321'
then, of course, an alias is:
alias rxvt="rxvt '#123456' '#654321'"
(oops, i guess i kind of did address the quoting :)
How to escape single quotes (') and double quotes (") with hex and octal chars
If using something like echo, I've had some really complicated and really weird and hard-to-escape (think: very nested) cases where the only thing I could get to work was using octal or hex codes!
Here are some basic examples just to demonstrate how it works:
1. Single quote example, where ' is escaped with hex \x27 or octal \047 (its corresponding ASCII code):
hex \x27
echo -e "Let\x27s get coding!"
# OR
echo -e 'Let\x27s get coding!'
Result:
Let's get coding!
octal \047
echo -e "Let\047s get coding!"
# OR
echo -e 'Let\047s get coding!'
Result:
Let's get coding!
2. Double quote example, where " is escaped with hex \x22 or octal \042 (its corresponding ASCII code).
Note: bash is nuts! Sometimes even the ! char has special meaning, and must either be removed from within the double quotes and then escaped "like this"\! or put entirely within single quotes 'like this!', rather than within double quotes.
# 1. hex; also escape `!` by removing it from within the double quotes
# and escaping it with `\!`
$ echo -e "She said, \x22Let\x27s get coding"\!"\x22"
She said, "Let's get coding!"
# OR put it all within single quotes:
$ echo -e 'She said, \x22Let\x27s get coding!\x22'
She said, "Let's get coding!"
# 2. octal; also escape `!` by removing it from within the double quotes
$ echo -e "She said, \042Let\047s get coding"\!"\042"
She said, "Let's get coding!"
# OR put it all within single quotes:
$ echo -e 'She said, \042Let\047s get coding!\042'
She said, "Let's get coding!"
# 3. mixed hex and octal, just for fun
# also escape `!` by removing it from within the double quotes when it is followed by
# another escape sequence
$ echo -e "She said, \x22Let\047s get coding! It\x27s waaay past time to begin"\!"\042"
She said, "Let's get coding! It's waaay past time to begin!"
# OR put it all within single quotes:
$ echo -e 'She said, \x22Let\047s get coding! It\x27s waaay past time to begin!\042'
She said, "Let's get coding! It's waaay past time to begin!"
Note that if you don't properly escape !, when needed, as I've shown two ways to do above, you'll get some weird errors, like this:
$ echo -e "She said, \x22Let\047s get coding! It\x27s waaay past time to begin!\042"
bash: !\042: event not found
OR:
$ echo -e "She said, \x22Let\x27s get coding!\x22"
bash: !\x22: event not found
One more alternative: this allows mixed expansion and non-expansion all within the same bash string
Here is another demo of an alternative escaping technique.
First, read the main answer by #liori to see how the 2nd form below works. Now, read these two alternative ways of escaping characters. Both examples below are identical in their output:
CMD="gs_set_title"
# 1. 1st technique: escape the $ symbol with a backslash (\) so it doesn't
# run and expand the command following it
echo "$CMD '\$(basename \"\$(pwd)\")'"
# 2. 2nd technique (does the same thing in a different way): escape the
# $ symbol using single quotes around it, and the single quote (') symbol
# using double quotes around it
echo "$CMD ""'"'$(basename "$(pwd)")'"'"
Sample output:
gs_set_title '$(basename "$(pwd)")'
gs_set_title '$(basename "$(pwd)")'
Note: for my gs_set_title bash function, which I have in my ~/.bash_aliases file somewhere around here, see my other answer here.
References:
https://en.wikipedia.org/wiki/ASCII#Printable_characters
https://serverfault.com/questions/208265/what-is-bash-event-not-found/208266#208266
See also my other answer here: How do I write non-ASCII characters using echo?.
I just use shell codes.. e.g. \x27 or \\x22 as applicable. No hassle, ever really.
Since one cannot put single quotes within single quoted strings, the simplest and most readable option is to use a HEREDOC string
command=$(cat <<'COMMAND'
urxvt -fg '#111111' -bg '#111111'
COMMAND
)
alias rxvt=$command
In the code above, the HEREDOC is sent to the cat command and the output of that is assigned to a variable via the command substitution notation $(..)
Putting a single quote around the HEREDOC is needed since it is within a $()
IMHO the real answer is that you can't escape single-quotes within single-quoted strings.
Its impossible.
If we presume we are using bash.
From bash manual...
Enclosing characters in single quotes preserves the literal value of each
character within the quotes. A single quote may not occur
between single quotes, even when preceded by a backslash.
You need to use one of the other string escape mechanisms " or \
There is nothing magic about alias that demands it use single quotes.
Both the following work in bash.
alias rxvt="urxvt -fg '#111111' -bg '#111111'"
alias rxvt=urxvt\ -fg\ \'#111111\'\ -bg\ \'#111111\'
The latter is using \ to escape the space character.
There is also nothing magic about #111111 that requires single quotes.
The following options achieves the same result the other two options, in that the rxvt alias works as expected.
alias rxvt='urxvt -fg "#111111" -bg "#111111"'
alias rxvt="urxvt -fg \"#111111\" -bg \"#111111\""
You can also escape the troublesome # directly
alias rxvt="urxvt -fg \#111111 -bg \#111111"
A minimal answer is needed so that people can get going without spending a lot of time as I had to sifting through people waxing eloquent.
There is no way to escape single quotes or anything else within single quotes.
The following is, perhaps surprisingly, a complete command:
$ echo '\'
whose output is:
\
Backslashes, surprisingly to even long-time users of bash, have no meaning inside single quotes. Nor does anything else.
Most of these answers hit on the specific case you're asking about. There is a general approach that a friend and I have developed that allows for arbitrary quoting in case you need to quote bash commands through multiple layers of shell expansion, e.g., through ssh, su -c, bash -c, etc. There is one core primitive you need, here in native bash:
quote_args() {
local sq="'"
local dq='"'
local space=""
local arg
for arg; do
echo -n "$space'${arg//$sq/$sq$dq$sq$dq$sq}'"
space=" "
done
}
This does exactly what it says: it shell-quotes each argument individually (after bash expansion, of course):
$ quote_args foo bar
'foo' 'bar'
$ quote_args arg1 'arg2 arg2a' arg3
'arg1' 'arg2 arg2a' 'arg3'
$ quote_args dq'"'
'dq"'
$ quote_args dq'"' sq"'"
'dq"' 'sq'"'"''
$ quote_args "*"
'*'
$ quote_args /b*
'/bin' '/boot'
It does the obvious thing for one layer of expansion:
$ bash -c "$(quote_args echo a'"'b"'"c arg2)"
a"b'c arg2
(Note that the double quotes around $(quote_args ...) are necessary to make the result into a single argument to bash -c.) And it can be used more generally to quote properly through multiple layers of expansion:
$ bash -c "$(quote_args bash -c "$(quote_args echo a'"'b"'"c arg2)")"
a"b'c arg2
The above example:
shell-quotes each argument to the inner quote_args individually and then combines the resulting output into a single argument with the inner double quotes.
shell-quotes bash, -c, and the already once-quoted result from step 1, and then combines the result into a single argument with the outer double quotes.
sends that mess as the argument to the outer bash -c.
That's the idea in a nutshell. You can do some pretty complicated stuff with this, but you have to be careful about order of evaluation and about which substrings are quoted. For instance, the following do the wrong things (for some definition of "wrong"):
$ (cd /tmp; bash -c "$(quote_args cd /; pwd 1>&2)")
/tmp
$ (cd /tmp; bash -c "$(quote_args cd /; [ -e *sbin ] && echo success 1>&2 || echo failure 1>&2)")
failure
In the first example, bash immediately expands quote_args cd /; pwd 1>&2 into two separate commands, quote_args cd / and pwd 1>&2, so the CWD is still /tmp when the pwd command is executed. The second example illustrates a similar problem for globbing. Indeed, the same basic problem occurs with all bash expansions. The problem here is that a command substitution isn't a function call: it's literally evaluating one bash script and using its output as part of another bash script.
If you try to simply escape the shell operators, you'll fail because the resulting string passed to bash -c is just a sequence of individually-quoted strings that aren't then interpreted as operators, which is easy to see if you echo the string that would have been passed to bash:
$ (cd /tmp; echo "$(quote_args cd /\; pwd 1\>\&2)")
'cd' '/;' 'pwd' '1>&2'
$ (cd /tmp; echo "$(quote_args cd /\; \[ -e \*sbin \] \&\& echo success 1\>\&2 \|\| echo failure 1\>\&2)")
'cd' '/;' '[' '-e' '*sbin' ']' '&&' 'echo' 'success' '1>&2' '||' 'echo' 'failure' '1>&2'
The problem here is that you're over-quoting. What you need is for the operators to be unquoted as input to the enclosing bash -c, which means they need to be outside the $(quote_args ...) command substitution.
Consequently, what you need to do in the most general sense is to shell-quote each word of the command not intended to be expanded at the time of command substitution separately, and not apply any extra quoting to the shell operators:
$ (cd /tmp; echo "$(quote_args cd /); $(quote_args pwd) 1>&2")
'cd' '/'; 'pwd' 1>&2
$ (cd /tmp; bash -c "$(quote_args cd /); $(quote_args pwd) 1>&2")
/
$ (cd /tmp; echo "$(quote_args cd /); [ -e *$(quote_args sbin) ] && $(quote_args echo success) 1>&2 || $(quote_args echo failure) 1>&2")
'cd' '/'; [ -e *'sbin' ] && 'echo' 'success' 1>&2 || 'echo' 'failure' 1>&2
$ (cd /tmp; bash -c "$(quote_args cd /); [ -e *$(quote_args sbin) ] && $(quote_args echo success) 1>&2 || $(quote_args echo failure) 1>&2")
success
Once you've done this, the entire string is fair game for further quoting to arbitrary levels of evaluation:
$ bash -c "$(quote_args cd /tmp); $(quote_args bash -c "$(quote_args cd /); $(quote_args pwd) 1>&2")"
/
$ bash -c "$(quote_args bash -c "$(quote_args cd /tmp); $(quote_args bash -c "$(quote_args cd /); $(quote_args pwd) 1>&2")")"
/
$ bash -c "$(quote_args bash -c "$(quote_args bash -c "$(quote_args cd /tmp); $(quote_args bash -c "$(quote_args cd /); $(quote_args pwd) 1>&2")")")"
/
$ bash -c "$(quote_args cd /tmp); $(quote_args bash -c "$(quote_args cd /); [ -e *$(quote_args sbin) ] && $(quote_args echo success) 1>&2 || $(quote_args echo failure) 1>&2")"
success
$ bash -c "$(quote_args bash -c "$(quote_args cd /tmp); $(quote_args bash -c "$(quote_args cd /); [ -e *sbin ] && $(quote_args echo success) 1>&2 || $(quote_args echo failure) 1>&2")")"
success
$ bash -c "$(quote_args bash -c "$(quote_args bash -c "$(quote_args cd /tmp); $(quote_args bash -c "$(quote_args cd /); [ -e *$(quote_args sbin) ] && $(quote_args echo success) 1>&2 || $(quote_args echo failure) 1>&2")")")"
success
etc.
These examples may seem overwrought given that words like success, sbin, and pwd don't need to be shell-quoted, but the key point to remember when writing a script taking arbitrary input is that you want to quote everything you're not absolutely sure doesn't need quoting, because you never know when a user will throw in a Robert'; rm -rf /.
To better understand what is going on under the covers, you can play around with two small helper functions:
debug_args() {
for (( I=1; $I <= $#; I++ )); do
echo -n "$I:<${!I}> " 1>&2
done
echo 1>&2
}
debug_args_and_run() {
debug_args "$#"
"$#"
}
that will enumerate each argument to a command before executing it:
$ debug_args_and_run echo a'"'b"'"c arg2
1:<echo> 2:<a"b'c> 3:<arg2>
a"b'c arg2
$ bash -c "$(quote_args debug_args_and_run echo a'"'b"'"c arg2)"
1:<echo> 2:<a"b'c> 3:<arg2>
a"b'c arg2
$ bash -c "$(quote_args debug_args_and_run bash -c "$(quote_args debug_args_and_run echo a'"'b"'"c arg2)")"
1:<bash> 2:<-c> 3:<'debug_args_and_run' 'echo' 'a"b'"'"'c' 'arg2'>
1:<echo> 2:<a"b'c> 3:<arg2>
a"b'c arg2
$ bash -c "$(quote_args debug_args_and_run bash -c "$(quote_args debug_args_and_run bash -c "$(quote_args debug_args_and_run echo a'"'b"'"c arg2)")")"
1:<bash> 2:<-c> 3:<'debug_args_and_run' 'bash' '-c' ''"'"'debug_args_and_run'"'"' '"'"'echo'"'"' '"'"'a"b'"'"'"'"'"'"'"'"'c'"'"' '"'"'arg2'"'"''>
1:<bash> 2:<-c> 3:<'debug_args_and_run' 'echo' 'a"b'"'"'c' 'arg2'>
1:<echo> 2:<a"b'c> 3:<arg2>
a"b'c arg2
$ bash -c "$(quote_args debug_args_and_run bash -c "$(quote_args debug_args_and_run bash -c "$(quote_args debug_args_and_run bash -c "$(quote_args debug_args_and_run echo a'"'b"'"c arg2)")")")"
1:<bash> 2:<-c> 3:<'debug_args_and_run' 'bash' '-c' ''"'"'debug_args_and_run'"'"' '"'"'bash'"'"' '"'"'-c'"'"' '"'"''"'"'"'"'"'"'"'"'debug_args_and_run'"'"'"'"'"'"'"'"' '"'"'"'"'"'"'"'"'echo'"'"'"'"'"'"'"'"' '"'"'"'"'"'"'"'"'a"b'"'"'"'"'"'"'"'"'"'"'"'"'"'"'"'"'"'"'"'"'"'"'"'"'"'"'c'"'"'"'"'"'"'"'"' '"'"'"'"'"'"'"'"'arg2'"'"'"'"'"'"'"'"''"'"''>
1:<bash> 2:<-c> 3:<'debug_args_and_run' 'bash' '-c' ''"'"'debug_args_and_run'"'"' '"'"'echo'"'"' '"'"'a"b'"'"'"'"'"'"'"'"'c'"'"' '"'"'arg2'"'"''>
1:<bash> 2:<-c> 3:<'debug_args_and_run' 'echo' 'a"b'"'"'c' 'arg2'>
1:<echo> 2:<a"b'c> 3:<arg2>
a"b'c arg2
In the given example, simply used double quotes instead of single quotes as outer escape mechanism:
alias rxvt="urxvt -fg '#111111' -bg '#111111'"
This approach is suited for many cases where you just want to pass a fixed string to a command: Just check how the shell will interpret the double-quoted string through an echo, and escape characters with backslash if necessary.
In the example, you'd see that double quotes are sufficient to protect the string:
$ echo "urxvt -fg '#111111' -bg '#111111'"
urxvt -fg '#111111' -bg '#111111'
Here is an elaboration on The One True Answer referenced above:
Sometimes I will be downloading using rsync over ssh and have to escape a filename with a ' in it TWICE! (OMG!) Once for bash and once for ssh. The same principle of alternating quotation delimiters is at work here.
For example, let's say we want to get: Louis Theroux's LA Stories ...
First you enclose Louis Theroux in single quotes for bash and double quotes for ssh:
'"Louis Theroux"'
Then you use single quotes to escape a double quote '"'
The use double quotes to escape the apostrophe "'"
Then repeat #2, using single quotes to escape a double quote '"'
Then enclose LA Stories in single quotes for bash and double quotes for ssh: '"LA Stories"'
And behold! You wind up with this:
rsync -ave ssh '"Louis Theroux"''"'"'"'"''"s LA Stories"'
which is an awful lot of work for one little ' -- but there you go
Obviously, it would be easier simply to surround with double quotes, but where's the challenge in that? Here is the answer using only single quotes. I'm using a variable instead of alias so that's it's easier to print for proof, but it's the same as using alias.
$ rxvt='urxvt -fg '\''#111111'\'' -bg '\''#111111'\'
$ echo $rxvt
urxvt -fg '#111111' -bg '#111111'
Explanation
The key is that you can close the single quote and re-open it as many times as you want. For example foo='a''b' is the same as foo='ab'. So you can close the single quote, throw in a literal single quote \', then reopen the next single quote.
Breakdown diagram
This diagram makes it clear by using brackets to show where the single quotes are opened and closed. Quotes are not "nested" like parentheses can be. You can also pay attention to the color highlighting, which is correctly applied. The quoted strings are maroon, whereas the \' is black.
'urxvt -fg '\''#111111'\'' -bg '\''#111111'\' # original
[^^^^^^^^^^] ^[^^^^^^^] ^[^^^^^] ^[^^^^^^^] ^ # show open/close quotes
urxvt -fg ' #111111 ' -bg ' #111111 ' # literal characters remaining
(This is essentially the same answer as Adrian's, but I feel this explains it better. Also his answer has 2 superfluous single quotes at the end.)
in addition to #JasonWoof perfect answer i want to show how i solved related problem
in my case encoding single quotes with '\'' will not always be sufficient, for example if a string must quoted with single quotes, but the total count of quotes results in odd amount
#!/bin/bash
# no closing quote
string='alecxs\'solution'
# this works for string
string="alecxs'solution"
string=alecxs\'solution
string='alecxs'\''solution'
let's assume string is a file name and we need to save quoted file names in a list (like stat -c%N ./* > list)
echo "'$string'" > "$string"
cat "$string"
but processing this list will fail (depending on how many quotes the string does contain in total)
while read file
do
ls -l "$file"
eval ls -l "$file"
done < "$string"
workaround: encode quotes with string manipulation
string="${string//$'\047'/\'\$\'\\\\047\'\'}"
# result
echo "$string"
now it works because quotes are always balanced
echo "'$string'" > list
while read file
do
ls -l "$file"
eval ls -l "$file"
done < list
Hope this helps when facing similar problem
Another way to fix the problem of too many layers of nested quotation:
You are trying to cram too much into too tiny a space, so use a bash function.
The problem is you are trying to have too many levels of nesting, and the basic alias technology is not powerful enough to accommodate. Use a bash function like this to make it so the single, double quotes back ticks and passed in parameters are all handled normally as we would expect:
lets_do_some_stuff() {
tmp=$1 #keep a passed in parameter.
run_your_program $# #use all your passed parameters.
echo -e '\n-------------' #use your single quotes.
echo `date` #use your back ticks.
echo -e "\n-------------" #use your double quotes.
}
alias foobarbaz=lets_do_some_stuff
Then you can use your $1 and $2 variables and single, double quotes and back ticks without worrying about the alias function wrecking their integrity.
This program prints:
el#defiant ~/code $ foobarbaz alien Dyson ring detected #grid 10385
alien Dyson ring detected #grid 10385
-------------
Mon Oct 26 20:30:14 EDT 2015
-------------
shell_escape () {
printf '%s' "'${1//\'/\'\\\'\'}'"
}
Implementation explanation:
double quotes so we can easily output wrapping single quotes and use the ${...} syntax
bash's search and replace looks like: ${varname//search/replacement}
we're replacing ' with '\''
'\'' encodes a single ' like so:
' ends the single quoting
\' encodes a ' (the backslash is needed because we're not inside quotes)
' starts up single quoting again
bash automatically concatenates strings with no white space between
there's a \ before every \ and ' because that's the escaping rules for ${...//.../...} .
string="That's "'##$*&^`(##'
echo "original: $string"
echo "encoded: $(shell_escape "$string")"
echo "expanded: $(bash -c "echo $(shell_escape "$string")")"
P.S. Always encode to single quoted strings because they are way simpler than double quoted strings.
Here are my two cents -- in the case if one wants to be sh-portable, not just bash-specific ( the solution is not too efficient, though, as it starts an external program -- sed ):
put this in quote.sh ( or just quote ) somewhere on your PATH :
# this works with standard input (stdin)
quote() {
echo -n "'" ;
sed 's/\(['"'"']['"'"']*\)/'"'"'"\1"'"'"'/g' ;
echo -n "'"
}
case "$1" in
-) quote ;;
*) echo "usage: cat ... | quote - # single-quotes input for Bourne shell" 2>&1 ;;
esac
An example:
$ echo -n "G'day, mate!" | ./quote.sh -
'G'"'"'day, mate!'
And, of course, that converts back:
$ echo 'G'"'"'day, mate!'
G'day, mate!
Explanation: basically we have to enclose the input with quotes ', and then also replace any single quote within with this micro-monster: '"'"' ( end the opening quote with a pairing ', escape the found single quote by wrapping it with double quotes -- "'", and then finally issue a new opening single quote ', or in pseudo-notation : ' + "'" + ' == '"'"' )
One standard way to do that would be to use sed with the following substitution command:
s/\(['][']*\)/'"\1"'/g
One small problem, though, is that in order to use that in shell one needs to escape all these single quote characters in the sed expression itself -- what leads to something like
sed 's/\(['"'"']['"'"']*\)/'"'"'"\1"'"'"'/g'
( and one good way to build this result is to feed the original expression s/\(['][']*\)/'"\1"'/g to Kyle Rose' or George V. Reilly's scripts ).
Finally, it kind of makes sense to expect the input to come from stdin -- since passing it through command-line arguments could be already too much trouble.
( Oh, and may be we want to add a small help message so that the script does not hang when someone just runs it as ./quote.sh --help wondering what it does. )
If you're generating the shell string within Python 2 or Python 3, the following may help to quote the arguments:
#!/usr/bin/env python
from __future__ import print_function
try: # py3
from shlex import quote as shlex_quote
except ImportError: # py2
from pipes import quote as shlex_quote
s = """foo ain't "bad" so there!"""
print(s)
print(" ".join([shlex_quote(t) for t in s.split()]))
This will output:
foo ain't "bad" so there!
foo 'ain'"'"'t' '"bad"' so 'there!'
If you have GNU Parallel installed you can use its internal quoting:
$ parallel --shellquote
L's 12" record
<Ctrl-D>
'L'"'"'s 12" record'
$ echo 'L'"'"'s 12" record'
L's 12" record
From version 20190222 you can even --shellquote multiple times:
$ parallel --shellquote --shellquote --shellquote
L's 12" record
<Ctrl-D>
'"'"'"'"'"'"'L'"'"'"'"'"'"'"'"'"'"'"'"'"'"'"'"'"'"'"'"'"'"'"'"'"'"'s 12" record'"'"'"'"'"'"'
$ eval eval echo '"'"'"'"'"'"'L'"'"'"'"'"'"'"'"'"'"'"'"'"'"'"'"'"'"'"'"'"'"'"'"'"'"'s 12" record'"'"'"'"'"'"'
L's 12" record
It will quote the string in all supported shells (not only bash).
This function:
quote ()
{
local quoted=${1//\'/\'\\\'\'};
printf "'%s'" "$quoted"
}
allows quoting of ' inside '. Use as this:
$ quote "urxvt -fg '#111111' -bg '#111111'"
'urxvt -fg '\''#111111'\'' -bg '\''#111111'\'''
If the line to quote gets more complex, like double quotes mixed with single quotes, it may become quite tricky to get the string to quote inside a variable. When such cases show up, write the exact line that you need to quote inside an script (similar to this).
#!/bin/bash
quote ()
{
local quoted=${1//\'/\'\\\'\'};
printf "'%s'" "$quoted"
}
while read line; do
quote "$line"
done <<-\_lines_to_quote_
urxvt -fg '#111111' -bg '#111111'
Louis Theroux's LA Stories
'single quote phrase' "double quote phrase"
_lines_to_quote_
Will output:
'urxvt -fg '\''#111111'\'' -bg '\''#111111'\'''
'Louis Theroux'\''s LA Stories'
''\''single quote phrase'\'' "double quote phrase"'
All correctly quoted strings inside single quotes.
Here is another solution. This function will take a single argument and appropriately quote it using the single-quote character, just as the voted answer above explains:
single_quote() {
local quoted="'"
local i=0
while [ $i -lt ${#1} ]; do
local ch="${1:i:1}"
if [[ "$ch" != "'" ]]; then
quoted="$quoted$ch"
else
local single_quotes="'"
local j=1
while [ $j -lt ${#1} ] && [[ "${1:i+j:1}" == "'" ]]; do
single_quotes="$single_quotes'"
((j++))
done
quoted="$quoted'\"$single_quotes\"'"
((i+=j-1))
fi
((i++))
done
echo "$quoted'"
}
So, you can use it this way:
single_quote "1 2 '3'"
'1 2 '"'"'3'"'"''
x="this text is quoted: 'hello'"
eval "echo $(single_quote "$x")"
this text is quoted: 'hello'

xargs pass multiple arguments to perl subroutine?

I know how to pipe multiple arguments with xargs:
echo a b | xargs -l bash -c '1:$0 2:$1'
and I know how to pass the array of arguments to my perl module's subroutine from xargs:
echo a b | xargs --replace={} perl -I/home/me/module.pm -Mme -e 'me::someSub("{}")'
But I can't seem to get multiple individual arguments passed to perl using those dollar references (to satisfy the me::someSub signature):
echo a b | xargs -l perl -e 'print("$0 $1")'
Just prints:
-e
So how do I get the shell arguments: $0, $1 passed to my perl module's subroutine?
I know I could just delimit a;b so that the xarg {} could be processed by perl splitting it to get individual arguments), but I could also just completely process all STDIN with perl. Instead, my objective is to use perl -e so that I can explicitly call the subroutine I want (rather than having some pre-process in the script that figures out what subroutine to call and what arguments to use based on STDIN, to avoid script maintenance costs).
While bash's argument are available as $# and $0, $1, $2, etc, Perl's arguments are available via #ARGV. This means that the Perl equivalent of
echo a b | xargs -l bash -c 'echo "1:$0 2:$1"'
is
echo a b | xargs -l perl -e'CORE::say "1:$ARGV[0] 2:$ARGV[1]"'
That said, it doesn't make sense to use xargs in this way because there's no way to predict how many times it will call perl, and there's no way to predict how many arguments it will pass to perl each time. You have an XY Problem, and you haven't provided any information to help us. Maybe you're looking for
perl -e'CORE::say "1:$ARGV[0] 2:$ARGV[1]"' $( echo a b )
I am not sure about the details of your design, so I take it that you need a Perl one-liner to use shell's variables that are seen in the scope in which it's called.
A perl -e'...' executes a Perl program given under ''. For any variables from the environment where this program runs -- a pipeline, or a shell script -- to be available to the program their values need be passed to it. Ways to do this with a one-liner are spelled out in this post, and here is a summary.
A Perl program receives arguments passed to it on the command-line in #ARGV array. So you can invoke it in a pipeline as
... | perl -e'($v1, $v2) = #ARGV; ...' "$0" "$1"
or as
... | xargs -l perl -e'($v1, $v2) = #ARGV; ...'
if xargs is indeed used to feed the Perl program its input. In the first example the variables are quoted to protect possible interesting characters in them (spaces, *, etc) from being interpreted by the shell that sets up and runs the perl program.
If input contains multiple lines to process and the one-liner uses -n or -p for it then unpack arguments in a BEGIN block
... | perl -ne'BEGIN { ($v1, $v2) = splice(#ARGV,0,2) }; ...' "$0" "$1" ...
which runs at compile time, so before the loop over input lines provided by -n/-p. The arguments other than filenames are now removed from #ARGV, so to leave only the filenames there for -n/-p, in case input comes from files.
There is also a rudimentary mechanism for command-line switches in a one-liner, via the -s switch. Please see the link above for details; I'd recommend #ARGV over this.
Finally, your calling code could set up environment variables which are then available to the Perl progam in %ENV. However, that doesn't seem to be suitable to what you seem to want.
Also see this post for another example.

How to use LSB_JOBINDEX in bsub array job arguments in Platform LSF?

I would like to pass LSB_JOBINDEX to as an argument to my script instead of using an environment variable.
This makes my script more LSF agnostic and avoids creating a helper script that uses the environment variable.
However, I was not able to use LSB_JOBINDEX in arguments: it only works as part of the initial command string.
For example, from a bash shell, I use the test command:
bsub -J 'myjobname[1-4]' -o bsub%I.log \
'echo $LSB_JOBINDEX' \
'$LSB_JOBINDEX' \
\$LSB_JOBINDEX \
'$LSB_JOBINDEX' \
"\$LSB_JOBINDEX"
and the output of say bsub2.log is:
2 $LSB_JOBINDEX $LSB_JOBINDEX $LSB_JOBINDEX $LSB_JOBINDEX
So in this case, only the first $LSB_JOBINDEX got expanded, but not any of the following ones.
But I would rather not pass the entire command as a single huge string as the 'echo $LSB_JOBINDEX' in this example. I would prefer to just use separate arguments as in a regular bash command.
I've also tried to play around with %I but it only works for -o and related bsub options, not for the command itself.
Related: Referencing job index in LSF job array
Tested in LSF 10.1.0. Related documentation: https://www.ibm.com/support/knowledgecenter/en/SSWRJV_10.1.0/lsf_admin/job_array_cl_args.html
bsub will add single quotes around the arguments if the argument starts with $. For example. If the bsub command line is
bsub command -a $ARG1 -b $ARG2
Then bsub will add quotes to the arguments to the 2nd and 4th parameters. The command is stored like this
command -a '$ARG1' -b '$ARG2'
One way to prevent this is to put the commands in a script. Like this:
$ cat cmd
echo $LSB_JOBINDEX
echo "line 2"
echo $LSB_JOBINDEX
Then run your job like this:
$ bsub -I < cmd
Job <2669> is submitted to default queue <normal>.
<<Waiting for dispatch ...>>
<<Starting on hostA>>
0
line 2
0
Note that the -I is not needed. Its just so you can see the job output on the bsub's stdout.
EDIT
OK. Looks like this works. But its not really a serious answer since it's so ugly. The thing is that bsub will surround the argument with single quotes if the argument starts with $. So the strategy is to find some way to make sure that the first character in the argument isn't a $. One way is to put any character other than $ as the first character of the argument. Follow it by a backspace literal, followed by the $. Note that it needs to be the actual backspace character, not ^ followed by H. Use ctrl-v followed by a ctrl-h to get the literal appended to the command line.
$ bsub -I echo "x^H\$LSB_JOBINDEX" "x^H\$LSB_JOBINDEX"
Job <2686> is submitted to default queue <normal>.
<<Waiting for dispatch ...>>
<<Starting on hostA>>
0 0
EDIT2
A tab literal also works. Not that its much better.
$ bsub -I echo " \$LSB_JOBINDEX" " \$LSB_JOBINDEX"
Job <2687> is submitted to default queue <normal>.
<<Waiting for dispatch ...>>
<<Starting on hostA>>
0 0

Perl Search and Replace Command Not Working When Called from Inside another Perl Script

I am a beginner with Perl. I am using the Below Perl Command To Search and Replace "/$" Sequence in my tcl Script. This works well When used on the linux Command Line directly.
perl -p -i -e 's/\/\$/\/\\\$/g' sed_list.tcl
I am calling to Call the above Perl One liner in another Perl script using System Command and only with " ` " Back Tick.
system(`perl -p -i -e 's/\/\$/\/\\\$/g' sed_list.tcl`);
`perl -p -i -e 's/\/\$/\/\\\$/g' sed_list.tcl`;
I am getting the Below error. Please Help With this issue.
Bareword found where operator expected at -e line 1, near "$/g"
(Missing operator before g?)
Final $ should be \$ or $name at -e line 1, within string
syntax error at -e line 1, near "s//$/"
Execution of -e aborted due to compilation errors.
I Dont Know if I Can use any other Separation Operator like % and # just like SED command but, When I used '%' operator for separation, I didn't see error but job is not done.
`perl -p -i -e 's%\/\$%\/\\\$%g' sed_list.tcl`;
I couldn't find sufficient results for this particular issue of '$' variable on the web. Any help is appreciated.
Some one here Suggested that I should Escape all Back Slashes while using System Command or calling another command using BackTicks from inside a perl script. But later they have deleted their answer. It worked for me. I would like to thank every one for taking effort and helping me out in solving my question.
Here is the correct working code.
`perl -p -i -e 's/\\\/\\\$/\\\/\\\\\\\$/g' sed_list_gen.tcl`;
or Use System function as shown Below
system("perl -p -i -e 's/\\\/\\\$/\\\/\\\\\\\$/g' sed_list_gen.tcl");
Thanks once again for the community for helping me out. . .
You can execute an external command by passing the command to a system function or by using backticks(``) operator. Please pass the command to the system() function as a string:
system(q{perl -p -i -e 's/\/\$/\/\\\$/g' sed_list.tcl})
or use backticks as:
`perl -p -i -e 's/\/\$/\/\\\$/g' sed_list_gen.tcl`
Edit:
As suggested by Paul in the comments.

Awk inside of qsub

I have a bash script in which I have a few qsubs. Each of them are waiting for a preivous qsub to be done before starting.
My first qsub consist of sending files in a certain directory to a perl program and having the outfiles printed in a new directory. At the end, I echo the array with all my jobs names. This script works as intented.
mkdir -p /perl_files_dir
for ID_FILES in `ls Infiles_dir/*.txt`;
do
JOB_ID=`echo "perl perl_scirpt.pl $ID_FILES" | qsub -j oe `
JOB_ID_ARRAY="${JOB_ID_ARRAY}:$JOB_ID"
done
echo $JOB_ID_ARRAY
My second qsub is meant to sort all my previous files made with my perl script in a new outfile and to start after all these jobs are done (about 100 jobs) with depend=afterany. Again, this part is working fine.
SORT_JOB=`echo "sort -m -n perl_files_dir/*.txt >>sorted_file.txt" | qsub -j oe -W depend=afterany$JOB_ID_ARRAY`
SORT_ARRAY="${SORT_ARRAY}:$SORT_JOB"
My issue is that in my sorted file, I have a few columns I wish to remove (2 to 6), so I came up with this last line using awk piped to sed with another depend=afterany
SED=`echo "awk '{\$2="";\$3="";\$4="";\$5="";\$6=""; print \$0}' sorted_file.txt \
| sed 's/ //g' >final_file.txt" | qsub -j oe -W depend=afterany$SORT_ARRAY`
This last step creates final_file.txt, but leaves it empty. I added SED= before my echo because it would otherwise give me Command not found.
I tried without the pipe so it would just print everything. Unfortunately it prints nothing.
I assume it is not opening my sorted file and this is why my final file is empty after my sed. If it's the case, then why won't awk read it?
In my script, I am using variables to define my directories and files (with the correct path). I know my issue is not about find my files or directories since they are perfectly defined at the beginning and used throughout the script. I tried to write the whole path instead of a variable and I get the same results.
for ID_FILES in `ls Infiles_dir/*.txt`
Simplify this to
for ID_FILES in Infiles_dir/*.txt
ls lists the files you pass it (except when you pass it directories, then it lists their content). Rather than telling it to display a list of files and parse the output, use the list of files you already have! This is more reliable (parsing the output of ls will fail if the file names contain whitespace or wildcard characters), clearer and faster. Don't parse the output of ls.
SORT_JOB=`echo "sort -m -n perl_files_dir/*.txt >>sorted_file.txt" | qsub -j oe -W depend=afterany$JOB_ID_ARRAY`
You'd make your life simpler if you used the right form of quoting in the right place. Don't use backquotes, because it's difficult to know how to quote things inside. Use $(…) instead, it's exactly equivalent except that it is parsed in a sane way.
I recommend using a here document for the shell snippet that you're feeding to qsub. You have fewer quoting issues to worry about, and it's more readable.
While we're at it, always put double quotes around variable substitutions and command substitutions: "$some_variable", "$(some_command)". Annoyingly, $var in shell syntax doesn't mean “take the value of the variable var”, it means “take the value of the variable var, parse it as a list of wildcard patterns, and replace each pattern by the list of matching files if there are matching files”. This extra stuff is turned off if the substitution happens inside double quotes (or in a here document, by the way): "$var" means “take the value of the variable var”.
SORT_JOB=$(qsub -j oe -W depend="afterany$JOB_ID_ARRAY" <<'EOF'
sort -m -n perl_files_dir/*.txt >>sorted_file.txt
EOF
)
We now get to the snippet where the quoting was actually causing a problem.
SED=`echo "awk '{\$2="";\$3="";\$4="";\$5="";\$6=""; print \$0}' sorted_file.txt \
| sed 's/ //g' >final_file.txt" | qsub -j oe -W depend=afterany$SORT_ARRAY`
The string that becomes the argument to the echo command is:
awk '{$2=;$3=;$4=;$5=;$6=; print $0}' sorted_file.txt | sed 's/ //g' >final_file.txt
This is syntactically incorrect, and that's why you're not getting any output.
You didn't escape the double quotes inside what was meant to be the awk snippet. It's a lot clearer if you use a here document. Also, you don't need the SED= part. You added it because you had a command substitution (a command between …), which substitutes the output of a command. But since you aren't interested in the output of the qsub command, don't take its output, just execute it.
qsub -j oe -W depend="afterany$SORT_ARRAY" <<'EOF'
awk '{$2="";$3="";$4="";$5="";$6=""; print $0}' sorted_file.txt |
sed 's/ //g' >final_file.txt
EOF
I'm not familiar with qsub, but presumably there's a way to get the error output and the return status of the commands it runs. Inspect that error output, you should have seen the errors from awk.
The version of awk that I am using, does not like the character escapes
awk --version
GNU Awk 3.1.7
spuder#cent64$ awk '{\$2="";\$3="";\$4=""; print \$0}' foo.txt
awk: {\$2="";\$3="";\$4=""; print \$0}
awk: ^ backslash not last character on line
Try the following syntax
awk '{for(i=2;i<=7;i++) $i="";print}' foo.txt
As a side note, if you are using Torque 4.x you may not be able to use a comma separated list of jobs with -W depend=, instead you may need to create a new PBS declarative (-W) for each job.
eg...
#Invalid syntax in newer versions of torque
qsub -W depend=foo,bar
Resources
backslash in gawk fields
Print all but the first three columns
http://docs.adaptivecomputing.com/torque/help.htm#topics/commands/qsub.htm#-W