Any ideas? I am attempting to write a script that uses sed.
If done this way it fails
- rmdec=“sed ‘s/…$//’”
- i1xmr=$(echo “$i1p/$apiresponse*1000” |bc -l |$rmdec)
But if i do it this way it works
- i1xmr=$(echo “$i1p/$apiresponse*1000” |bc -l | sed ‘s/…$//’)
You must log in or register to comment.
I have had some problems doing what you’re trying to do. Replacing the variable with a function would work.
rmdec() { sed ‘s/…$//’; }
Perfect. That did the trick. So when running a bash command like that putting it in a function keeps it from getting screwed up apparently
You’re assigning
rmdecto the output of sed. It should work if you wrap it as you did withi1xmr.Yeah, my goal is to shorten that sed command to that variable. It seems like it would work, but nope. It throws errors
It might be because it’s a single string, and might work if you store it or expand it as an array. I think it would in Zsh, anyway.
But the response to use a function instead is probably wiser.
Strings work fine, the problem is the (single) quotes:
~ $ foo="echo 'hello world'" ~ $ for x in $foo; do echo $x; done echo 'hello world' ~ $ $foo 'hello world' ~ $ eval "$foo" hello worldThe splitting is by whitespace, so the single quotes remain in the arguments. Using eval (and double quotes to preven splitting), it gets processed correctly. That said, don’t use eval; use functions or aliases instead.
Yep, the function did the trick. My guess is it was being misread at execution as a variable and thats why it was breaking
Your mistake is that after variable substitution bash does not handle quoted strings, i.e. it does not remove single quotes from
sedcommand line. If you really need this to happen, you have to useeval:i1xmr=$(echo "$i1p/$apiresponse*1000" | bc -l | eval $rmdec)However using functions is a better solution in general. But in this particular case, I guess, you only need to change the
bc’sscaleinstead of usingsed:i1xmr=$(echo "scale=17; $i1p/$apiresponse*1000" | bc -l)For better readability you may use heredoc instead of
echo:i1xmr=$( bc -l << EOF scale=17 $i1p/$apiresponse*1000 EOF )Oh dear Lord, and see, this is why I do not code for a living. What I ended up doing is using a function like this
- rmdec() { sed ‘s/…$//’; }
- i1xmr=$(echo “$i1p/$apiresponse*1000” |bc -l |rmdec)
Here, you guys can go laugh at my code now.
And again, using here-document greatly improves readability, like this.
Hmm. I had a look at the example given. I see the idea, but would cat be the thing to use or would it be echo <<-EOF > “$file”?
You need
catbecause it readsstdinand prints it tostdout.echodoes not readstdin, it prints its arguments.I have never used cat like that before. If you just ent cat abcd > file it says abcd doesnt exist but does create “file”. I know you can cat contents of a file into another file but why the <<-EOF > file works is a bit beyond me.
catdoes not createfile, your shell does when you redirect the standard output withfile.
Alright, I modified it and formatted it. However, for whatever reason, the output HTML in /var/www/html/index.html does not keep the formatting and is all just left aligned as before. That’s not really a problem, just more of a curiosity as to why it did not inherit the formatting of the input.
When using “<<-”, shell removes all tabs from the beginning of each line. So you have to use tabs for formatting inside your script and then spaces for HTML formatting, as in my example. Or use “<<” without dash to preserve tabs.
Again, you don’t need
sedfor this, simply setscale=2or how many digits after decimal point you need. Also you missed!in the shebang.So I need five digits after the decimal point, but then when I do the multiplication, I only want four digits in total. I did move “scale=” to five and that made the sed command in rmdec much shorter, but its still needed. I thought i could add “length=4” but that throws an error.
Multiply before you divide to keep precision:
1000 * $i1p / $apiresponseAh, that did the trick. Thanks
deleted by creator
