I have a simple bash function dividing two numbers:
echo "750/12.5" | bc I'd like to take the output from bc and append /24 and pipe said result to another instance of bc.
Something like:
echo "750/12.5" | bc | echo $1 + "/24" | bc Where $1 is the piped result.
P.S. I realize I could just do echo "750/12.5/24" | bc my question is more in regards to the appending of text to a pipe result.
I achieved this using sed and replacing the end of the line:
echo "my text" | sed 's/$/ more text/' Returns:
my text more text Your example:
echo "750/12.5" | bc | sed 's/$/\/24/' | bc Alternative approach:
TL;DR: Use cat to concatenate with stdin (-) as an argument. Note that it'll newline-separate your concatenated streams; remove if undesired.
Unfortunately this approach is a bit cumbersome for your particular use case since bc fails if there's a newline in the middle of that expression:
$ echo "750/12.5" | bc | cat - <(echo "/24") | bc 60 (standard_in) 2: parse error You can work around that by deleting the newline from the original bc result:
$ echo "750/12.5" | bc | tr -d '\n' | cat - <(echo "/24") | bc 2 However, I came upon this question by googling "bash append to pipe", and in my case I actually do want that newline between the concatenations. So for those here for the same reason, use | cat - <(...):
$ echo foo | cat - <(echo bar) foo bar $ echo foo | cat - <(echo bar) | sort | cat - <(echo last_line) bar foo last_line $ echo foo | cat - <(echo bar) | sort | cat <(echo header) - <(echo last_line) header bar foo last_line 
echo "first part..." (which adds a newline), use printf "%s" "first part...." (or printf "first part..." if there are no % in it) which will not put a newline, and thus makes appending the following of the line of computation easier) In the simplest of the options, this does append to the pipe stream:
$ echo "750/12.5" | { bc; echo "/24"; } 60 /24 However that has an unexpected newline, to avoid that you need to either use tr:
$ echo "750/12.5" | { bc | tr -d '\n' ; echo "/24"; } 60/24 Or, given the fact that a command expansion removes trailing newlines:
$ printf '%s' $( echo "750/12.5" | bc ); echo "/24" 60/24 But probably, the correct way should be similar to:
$ echo "$(echo "750/12.5" | bc )/24" 60/24 Which, to be used in bc, could be written as this:
$ bc <<<"$(bc <<<"750/12.5")/24" 2 Which, to get a reasonable floating number precision should be something like:
$ bc <<<"scale=10;$(bc <<<"scale=5;750/12.5")/24" 2.5000000000 Note the need of two scale, as there are two instances of bc.
Of course, one instance of bc needs only one scale:
$ bc <<<"scale=5;750/12.5/24" In fact, what you should be thinking about is in terms of an string:
$ a=$(echo "750/12.5") # capture first string. $ echo "$a/24" | bc # extend the string 2 The comment about scale from above is still valid here.
{…} is simple. The answer walks toward showing the user a "command substitution" as the best solution. Read again. Something like this seems to work using xargs:
$ echo "750/12.5" | bc |xargs -I{} echo "{}+20" |bc 80 Or even:
$ echo "750/12.5" | bc |echo "$(</dev/stdin)+20" |bc 80 And why not :
$ bc <<<20+$(bc <<<"750/12.5") 80 You could use awk to insert/append anything to the pipe:
$ date | awk 'BEGIN{print "prefix line"} {print} END{print "suffix line"}' prefix line Mo 21. Aug 15:24:30 CEST 2023 suffix line where
date - is your input pipe (current date for example here)BEGIN{print "prefix line"} - this block will insert output before main{print} - prints input pipe's contents; one could do some processing here if neededEND{print "suffix line"} - this appends output after mainTo do same in single line, just use printf:
$ date | awk 'BEGIN{printf("prefix line >> ")} {printf($0)} END{printf(" << suffix line")}' prefix line >> Mo 21. Aug 15:32:08 CEST 2023 << suffix line/upload/sc505enb/tmp> $ echo "750/12.5" | bc | awk '{printf($0)} END{printf("/24\n")}' 60/24 another possible BASH way - create a helper function which processes the stdin with read command and then prints extra line(s):
$ function append_lines(){ local data; while read -r data; do echo -en "$data"; done; echo "$*"; } $ date | append_lines " +suffice_;line1" " ++line2" " +++line_3" [Mon Aug 21 16:00:37 CEST 2023][ +suffice_;line1 ++line2 +++line_3] $ echo "750/12.5" | bc | append_lines "/24" | bc 2 where
while read -r data; do echo -en "$data"; done; - just passes the stdin (pipe) through. remove the -en modifier to print separate input lines.echo "$*"; - prints all arguments of append_lines() to stdout
echo $(echo "750/12.5" | bc )/24" | bc?vipeas an interactive on-the-pipevisession.