In a POSIX shell script, I'd like to use decimal numbers generated by seq -w both in string manipulations and in simple calculations using $(( ... )). This breaks because numbers with leading zeroes are interpreted as octal numbers. I have come up with something like this to actually get decimal numbers...
for n in $(seq -w 0 300); do
str="xxx $n"
dec=$(echo $n | sed -e 's,00*\(..*\)$,\1,')
...
done
... but I wonder if there's an easier / more obvious way to do this.
Use expr instead. eg instead of
for n in $(seq -w 0 10); do echo $n $((n + 4)); done
Use this:
for n in $(seq -w 0 10); do echo $n $(expr $n + 4); done
Rather than generate 0-padded strings with seq, generate decimal numbers and pad them with zeros with printf when necessary.
for n in $(seq 0 300); do
padded_n=$(printf '%03' "$n")
...
done
Hopefully, you know the upper-bound ahead of time. If not, you can use something uglier like
upper_bound=300
for n in $(seq 0 $upper_bound); do
padded_n=$(printf "%0${#upper_bound}d" "$n")
...
done
