Rargs is kind of xargs + awk with pattern-matching support.

brew install rargs

nix-env -i rargs

(Currently available in unstable channel)

Download in the Release Page and put it in your PATH after uncompress.

cargo install --git https://github.com/lotabout/rargs.git

Suppose you have several backup files whose names match the pattern <scriptname>.sh.bak, and you want to map each filename back to <scriptname>.sh. We want to do it in a batch, so xargs is a natural choice, but how do we specify the name for each file? I believe there is no easy way.

With rargs, however, you are able to do:

ls *.bak | rargs -p '(.*)\.bak' mv {0} {1}

Here {0} refers to the whole input line, while {1} refers to the first group captured in the regular expression.

I had a bunch of URLs and their corresponding target filenames stored in a CSV file:

URL1,filename1
URL2,filename2

I hoped there was a simple way to download and save each URL with its specified filename. With rargs there is:

cat download-list.csv | rargs -p '(?P<url>.*),(?P<filename>.*)' wget {url} -O {filename}

Here (?P<group_name>...) assigns the name group_name to the captured group. This can then be referred to as {group_name} in the command.

Suppose you have an xSV file with lots of columns, and you only want to extract and format some of them, e.g.:

nobody:*:-2:-2:Unprivileged User:/var/empty:/usr/bin/false
root:*:0:0:System Administrator:/var/root:/bin/sh
daemon:*:1:1:System Services:/var/root:/usr/bin/false

Here's an example of how rargs can be used to process it:

$ cat /etc/passwd | rargs -d: echo -e 'id: "{1}"\t name: "{5}"\t rest: "{6..::}"'
id: "nobody"     name: "Unprivileged User"       rest: "/var/empty:/usr/bin/false"
id: "root"       name: "System Administrator"    rest: "/var/root:/bin/sh"
id: "daemon"     name: "System Services"         rest: "/var/root:/usr/bin/false"

rargs allow you to specify the delimiter (regex) to split the input on, and allows you to refer to the corresponding fields or field ranges. This allows it to be used as an AWK replacement for some simple but common cases.

1. receive the input on stdin and split it into lines
2. split (-d) or extract (-p) the input into named or numbered groups, with {0} matching the whole line
3. map the named and numbered groups into a command passed as the remaining arguments, and execute the command

rargs allows you to use any regular expression to match the input, and captures anything you are interested in. The syntax is the standard, mostly Perl-compatible Rust regex syntax used by tools such as ripgrep.

• positional (numbered) groups are captured with parentheses, e.g. '(\w+):(\d+)', and the corresponding groups are referred to by {1}, {2} etc. in the command
• named groups are captured with (?P<name>...) and referred to by {name} in the command

For simple usage, you might not want to write the whole regular expression to extract parts of the line. All you want is to split the groups by some delimiter. With rargs you can achieve this by using the -d (delimiter) option.

We already know how to refer to captures by number ({1}) or by name ({name}). There are also cases where you might want to substitute multiple fields at the same time. rargs also supports this with field-range expressions.

Suppose we have already captured 5 groups representing the strings 1, 2, 3, 4 and 5

• {..} gathers them all into 1 2 3 4 5 (note that they are separated by a space; this can be overridden by the -s option)
• {..3} results in 1 2 3
• {4..} results in 4 5
• {2..4} results in 2 3 4
• {3..3} results in 3

You can also specify a "local" separator (which will not affect the global setting):

• {..3:-} results in 1-2-3
• {..3:/} results in 1/2/3

Sometimes you may want to refer to the last few fields but have no way to predict the total number of fields of the input. rargs offer negative fields.

Suppose we have already captured 5 groups representing the strings 1, 2, 3, 4 and 5:

• {-1} results in 5
• {-5} results in 1
• {-6} results in nothing
• {-3..} results in 3 4 5

You can run commands in multiple threads to improve performance:

• -w <num> specifies the number of workers you want to run simultaneously
• -w 0 defaults the number of workers to the number of CPUs on your system

• {LINENUM} or {LN} to refer to current line number.

All feedback and PRs are welcome!