$declare -a inputs=("(1 3 4 8 6 2 7 0 5)" "(2 8 1 0 4 3 7 6 5)"
$ for i in ${inputs[#]}; do echo $i; done;
gives
(1
3
4
8
6
2
7
0
5)
(2
8
1
0
4
3
7
6
5)
I want each array in a row.
Use quotes:
for i in "${inputs[#]}"; do echo "$i"; done;
(1 3 4 8 6 2 7 0 5)
(2 8 1 0 4 3 7 6 5)
You need to use quotes. Say:
for i in "${inputs[#]}"; do echo $i; done
This would return:
(1 3 4 8 6 2 7 0 5)
(2 8 1 0 4 3 7 6 5)
Moreover, remove the ; after done unless it's the last line in your script!
Related
I have a .txt file with 9 lines:
1 2 3 4
1 2 3 5
1 2 3 6
1 2 3 4
1 2 3 5
1 2 3 6
1 2 3 4
1 2 3 5
1 2 3 6
I want to put the first 3 lines into one line, and the next three lines, and again the last three lines:
1 2 3 4 1 2 3 5 1 2 3 6
1 2 3 4 1 2 3 5 1 2 3 6
1 2 3 4 1 2 3 5 1 2 3 6
however it only gives me one consecutive line
I tried
cat old.txt | tr -d '\n' > new.txt
You can use paste to merge together lines.
paste -d " " - - - < input.txt
The -d " " uses a space to delimit between the lines being joined. Each - reads from stdin (and we're redirecting your input file to stdin). If you wanted to join more lines, just increase the number of - etc.
So I have a C program that outputs many numbers. I have to check them all. The problem is, each time I run my program, I need to change seeds. In order to do that, I've been doing it manually and was trying to make a shell script to work around this.
I've tried using sed but couldn't manage to do it.
I'm trying to get the output like this:
a=$(./algorithm < input.txt)
b=$(./algorithm2 < input.txt)
c=$(./algorithm3 < input.txt)
The output of each algorithm program is something like this:
12 13 315
1 2 3 4 5 6 7 8 10 2 8 9 1 0 0 2 3 4 5
So the variable a has all this output, and what I need is
variable a to contain this whole string
and variable a1 to contain only the third number, in this case, 315.
Another example:
2 3 712
1 23 15 12 31 23 3 2 5 6 6 1 2 3 5 51 2 3 21
echo $b should give this output:
2 3 712
1 23 15 12 31 23 3 2 5 6 6 1 2 3 5 51 2 3 21
and echo $b1 should give this output:
712
Thanks!
Not exactly what you are asking, but one way to do this would be to store the results of your algorithm in arrays, and then dereference the item of interest. You'd write something like:
a=( $(./algorithm < input.txt) )
b=( $(./algorithm2 < input.txt) )
c=( $(./algorithm3 < input.txt) )
Notice the extra () that encloses the statements. Now, a, b and c are arrays, and you can access the item of interest like ${a[0]} or $a[1].
For your particular case, since you want the 3rd element, that would have index = 2, hence:
a1=${a[2]}
b1=${b[2]}
c1=${c[2]}
Since you are using the Bash shell (see your tags), you can use Bash arrays to easily access the individual fields in your output strings. For example like so:
#!/bin/bash
# Your lines to gather the output:
# a=$(./algorithm < input.txt)
# b=$(./algorithm2 < input.txt)
# c=$(./algorithm3 < input.txt)
# Just to use your example output strings:
a="$(printf "12 13 315 \n 1 2 3 4 5 6 7 8 10 2 8 9 1 0 0 2 3 4 5")"
b="$(printf "2 3 712 \n 1 23 15 12 31 23 3 2 5 6 6 1 2 3 5 51 2 3 21")"
# Put the output in arrays.
a_array=($a)
b_array=($b)
# You can access the array elements individually.
# The array index starts from 0.
# (The names a1 and b1 for the third elements were your choice.)
a1="${a_array[2]}"
b1="${b_array[2]}"
# Print output strings.
# (The newlines in $a and $b are gobbled by echo, since they are not quoted.)
echo "Output a:" $a
echo "Output b:" $b
# Print third elements.
echo "3rd from a: $a1"
echo "3rd from b: $b1"
This script outputs
Output a: 12 13 315 1 2 3 4 5 6 7 8 10 2 8 9 1 0 0 2 3 4 5
Output b: 2 3 712 1 23 15 12 31 23 3 2 5 6 6 1 2 3 5 51 2 3 21
3rd from a: 315
3rd from b: 712
Explanation:
The trick here is that array constants (literals) in Bash have the form
(<space_separated_list_of_elements>)
for example
(1 2 3 4 a b c nearly_any_string 99)
Any variable that gets such an array assigned, automatically becomes an array variable. In the script above, this is what happens in a_array=($a): Bash expands the $a to the <space_separated_list_of_elements> and reads the whole expression again interpreting it as an array constant.
Individual elements in such arrays can be referenced like variables by using expressions of the form
<array_name>[<idx>]
like a variable name. Therein, <array_name>is the name of the array and <idx> is an integer that references the individual element. For arrays that are represented by array constants, the index counts elements continuously starting from zero. Therefore, in the script, ${a_array[2]} expands to the third element in the array a_array. If the array would have less elements, a_array[2] would be considered unset.
You can output all elements in the array a_array, the corresponding index array, and the number of elements in the array respectively by
echo "${a_array[#]}"
echo "${!a_array[#]}"
echo "${#a_array[#]}"
These commands can be used to track down the fate of the newline: Given the script above, it is still in $a, as can be seen by (watch the quotes)
echo "$a"
which yields
12 13 315
1 2 3 4 5 6 7 8 10 2 8 9 1 0 0 2 3 4 5
But the newline did not make it into the array a_array. This is because Bash considers it as part of the whitespace that separates the third and the fourth element in the array assignment. The same applies if there are no extra spaces around the newline, like here:
12 13 315\n1 2 3 4 5 6 7 8 10 2 8 9 1 0 0 2 3 4 5
I actually assume that the output of your C program comes in this form.
This will store the full string in a[0] and the individual fields in a[1-N]:
$ tmp=$(printf '12 13 315\n1 2 3 4 5 6 7 8 10 2 8 9 1 0 0 2 3 4 5\n')
$ a=( $(printf '_ %s\n' "$tmp") )
$ a[0]="$tmp"
$ echo "${a[0]}"
12 13 315
1 2 3 4 5 6 7 8 10 2 8 9 1 0 0 2 3 4 5
$ echo "${a[3]}"
315
Obviously replace $(printf '12 13 315\n1 2 3 4 5 6 7 8 10 2 8 9 1 0 0 2 3 4 5\n') with $(./algorithm < input.txt) in your real code.
I've been trying to figure this silly thing with awk in the last hours but no luck so far.
I understand how to plot every second line, for example:
awk 'NR%2' file
and I also understand how to print a column based file if one column is within a specific range, for example:
awk '{if ($1 > 'yourvalue') print}' file
What I don't quite get is how to combine the two.
In practize, if I have a file organized as:
1 3 6 8
2 8 4 5
3 9 8 7
4 7 3 5
5 7 3 6
6 2 4 6
7 1 4 7
8 3 2 1
9 7 5 3
10 4 5 6
11 8 2 5
how can I get, for example:
1 3 6 8
3 9 8 7
5 7 3 6
7 1 4 7
8 3 2 1
9 7 5 3
10 4 5 6
11 8 2 5
so return every two lines if column 1 is smaller than 7 and print normally the rest.
I tried to combine everything in one single line but I always get errors.
You can reverse the 2nd condition and use OR condition to combine them:
awk 'NR%2 || $1>=7' file
1 3 6 8
3 9 8 7
5 7 3 6
7 1 4 7
8 3 2 1
9 7 5 3
10 4 5 6
11 8 2 5
You can combine conditions using && (and) and || (or).
You can use parentheses for nesting conditions.
For example:
awk 'cond1 && (cond2 || cond3)' file
This:
awk '{if ($1 > 7) print}' file
... is equivalent to this:
awk '$1 > 7 { print }' file
... because you can write conditions outside of the {...} to use as filters.
... which is equivalent to:
awk '$1 > 7' file
... because the default action is to print.
My colleague has given me a file, in which half of the lines are made of 8 columns of info and the other half are made of the 9th column of info. They are always next to each other, e.g.
1 2 3 4 5 6 7 8
1.1
2 3 4 5 6 7 8 9
1.2
...
a b c d e f g h
abcd
I know how to paste every two lines as one and print them out in Python. But I was wondering if it's possible to do that even more conveniently in BASH?
Thanks guys!
You could use sed or awk, as other answers have mentioned. Those answers are all good.
You could also do this easily in pure shell.
$ while read line1; do read line2; echo "$line1 $line2"; done < input.txt
1 2 3 4 5 6 7 8 1.1
2 3 4 5 6 7 8 9 1.2
Note that whitespace is not preserved.
There's another tool available on most unix-like systems called paste:
$ paste - - < input.txt
1 2 3 4 5 6 7 8 1.1
2 3 4 5 6 7 8 9 1.2
In this case, there's a big space in the first line because paste separates columns using tabs, by default, and the trailing space in the first line of input.txt caused the separating tab to be offset to the next column. You can read paste's man page for options to control this.
Another awk
awk '{f=$0;getline;print f,$0}' file
1 2 3 4 5 6 7 8 1.1
2 3 4 5 6 7 8 9 1.2
And just for the fun of it a gnu awk
awk -v RS="[0-9][.][0-9]" '{$1=$1;print $0,RT}' file
1 2 3 4 5 6 7 8 1.1
2 3 4 5 6 7 8 9 1.2
Here is set the Record Separator to the value in line two.
Then the RT will have the actual separator stored.
try:
awk '{printf "%s%s",$0,(NR%2?FS:RS)}' file
or:
awk 'NR%2{printf "%s ",$0;next}7' file
test:
kent$ echo "1 2 3 4 5 6 7 8
1.1
2 3 4 5 6 7 8 9
1.2"|awk '{printf "%s%s",$0,(NR%2?FS:RS)}'
1 2 3 4 5 6 7 8 1.1
2 3 4 5 6 7 8 9 1.2
kent$ echo "1 2 3 4 5 6 7 8
1.1
2 3 4 5 6 7 8 9
1.2"|awk 'NR%2{printf "%s ",$0;next}7'
1 2 3 4 5 6 7 8 1.1
2 3 4 5 6 7 8 9 1.2
You can sed:
sed 'N;s/\n/ /' file
or awk:
awk 'NF==1{print $0}{printf "%s ",$0}' file
I need to combine columns of different lengths into one column using BASH. Here is an example input file:
11 1 2 3 4 5 6 7 8
12 1 2 3 4 5 6 7 8
13 1 2 3 4 5 6 7 8
14 1 2 5 6 7 8
15 1 2 7 8
And my desired output:
1
1
1
1
1
3
3
3
5
5
5
5
7
7
7
7
7
The input data is pairs of columns as shown. Each pair is separated from another by a fixed number of spaces. Values within a pair of columns are separated by one space. Thanks in advance!
Using GNU awk for fixed width field handling:
$ cat file
1 2 3 4 5 6 7 8
1 2 3 4 5 6 7 8
1 2 3 4 5 6 7 8
1 2 5 6 7 8
1 2 7 8
$ cat tst.awk
BEGIN{ FIELDWIDTHS="1 1 1 3 1 1 1 3 1 1 1 3 1 1 1" }
{
for (i=1;i<=NF;i++) {
a[NR,i] = $i
}
}
END {
for (i=1;i<=NF;i+=4)
for (j=1;j<=NR;j++)
if ( a[j,i] != " " )
print a[j,i]
}
$ gawk -f tst.awk file
1
1
1
1
1
3
3
3
5
5
5
5
7
7
7
7
7
You may try the following:
awk -f ext.awk input.txt
where input.txt is your input data file and ext.awk is:
BEGIN {
ncols=4 # number of columns
nspc=3 # number of spaces that separates the columns
}
{
str=$0;
for (i=1; i<=ncols; i++) {
pos=match(str,/^([0-9]+) ([0-9]+)/,a)
if (pos>0) {
b[NR,i]=a[1]
if (NR==1) colw[i]=RLENGTH; #assume col width are given as in first row
}
str=substr(str,colw[i]+1+nspc);
}
}
END {
for (i=1;i<=ncols;i++)
for (j=1;j<=NR;j++) {
if (b[j,i]) print b[j,i];
}
}