cut, paste and join — Working with Columns
Asha's harvest file has four columns, but today she only needs the crop names. Tomorrow she'll want to line up two lists side by side, and after that match each picker with their garden plot. These are column jobs, and three tools cover them.
The rule
cut slices columns; paste glues lines; join matches on a key.
cut keeps only the columns you ask for. Use -d to give the separator and -f for the field numbers:
$ cut -d, -f2 harvest.csv
crop
tomato
basil
tomato
chilli
tomato
basil
$ cut -d, -f2,4 harvest.csv # crop and picker
$ cut -d, -f2- harvest.csv # field 2 to the end
$ cut -c1-10 harvest.csv # characters 1 to 10: the dateWithout -d, cut assumes the columns are separated by tabs.
paste glues files together line by line: line 1 with line 1, line 2 with line 2, and so on. Tabs go in between by default:
$ paste names.txt plots.txt
asha plot-1
ravi plot-2
meera plot-3
$ paste -d, names.txt plots.txt # comma instead of tab
$ paste -sd, names.txt # one file onto one line
asha,ravi,meerajoin matches lines from two files that share a key field, a bit like a database join. Both files must be sorted on that key. Here pickers.csv holds asha,plot-1, meera,plot-3 and ravi,plot-2, and totals.csv holds each picker's total kilograms:
$ join -t, pickers.csv totals.csv
asha,plot-1,4.9
meera,plot-3,4.1
ravi,plot-2,2.4-t, sets the separator. By default the key is the first field of each file.
(A related tool, split, does the opposite of gluing: split -l 1000 big.csv part- cuts a big file into pieces of 1000 lines each.)
The shortcut
Picture the file as a table, then ask which way you're working. Taking columns out means cut. Putting columns side by side by line position means paste. Putting them side by side by matching values means join, after sorting both files.
The data you want
The pickers are column 4 in a comma-separated file.
Two traps. cut -f4,2 still prints the fields in file order, field 2 and then field 4. cut never rearranges columns. And join on unsorted files quietly misses matches, so sort both files on the key first.
Drills
Answer fast — trust the rule.
Which command prints just the kg column (the 3rd) of harvest.csv?
What does paste a.txt b.txt do?
What must be true before join gives reliable results?
What does cut -d, -f4,2 harvest.csv print on its second line?
Remember
- cut slices columns, paste glues lines, and join matches on a key.
- Picture a table: taking columns out means cut, side by side by position means paste, and side by side by value means join.
- cut never reorders fields, cut splits on tabs unless you give
-d, and join needs both files sorted on the key.