csvgrep: Add support for multiple files

[jamieburchell]

Just like the regular grep command, it would be great if csvgrep supported multiple files. The implementation would need to take in to account the column headers present in each file.

Currently, I'm using cat to achieve what I need:

cat Files*.csv | csvgrep -c 4 -m "foo"

This could potentially match a column header in column 4 though, which wouldn't be the desired result.

[MrTrick]

I'd find this very useful as well, for running csvcut to extract the same columns from multiple files and appending them together.

Currently I have to use xargs, eg:

ls files/*_[my glob pattern].csv | xargs -n 1 csvcut -d ';' -c mycolumn

And it's very slow, I think in part because it has to spawn csvcut over and over again.

[dannguyen]

Is there any reason why you couldn't use csvstack for your situation? e.g.

csvstack *.csv | csvgrep -c 4 -m 'foo'

[jamieburchell]

That works perfectly. I didn't realise it existed!

Closing this, thanks.