join command option for generating tab-delimited file
When we want to join two files by a same key into one file, we can use join command. We can join lines by a first column value of tab-delimited files.
join -j 1 file1 file2
Indeed, join command itself is very useful, default output is not tab-delimited text, but space-delimited text. It is not good for us because most of data are tab-delimited. I just noticed an option for tab-delimited output.
join -j 1 -t "$(printf '\011')" file1 file2
It is required that files to join must be sorted. If not so, error will occur.
join: file1:7: is not sorted
Two files must be sorted like this.
sort -u file1 > file1s
sort -u file2 > file2s
By adding -u option, we can delete redundant lines.
Moreover, key for join must be unique. I could not overcome this restriction, and I wrote ad hoc Perl script for that.