repeat

(c) 2005-2012 Evan Moran http://github.com/evanmoran/repeat

Repeat is freely distributable under the MIT license. http://opensource.org/licenses/mit

How does it work?

Repeat allows commands to be automated with automagic substitution.

Imagine you have a comma separated file called data.csv:

Evan,     3.14,    Batman
Laura,    19,      James Bond
Sarah,    42,      Wolverine

Here is a simple example:

repeat '#1 loves #3' data.csv --comma

output:

Evan loves Batman
Laura loves James Bond
Sarah loves Wolverine

Repeat can handle any type of separator: --tab, --whitespace, --comma or you can provide your own type of separator with a regular expression.

Now let's craft a more complex command:

repeat 'echo "#1 loves #3" > #1.txt ' data.csv --comma

output:

echo "Evan loves Batman" > Evan.txt
echo "Laura loves James Bond" > Laura.txt
echo "Sarah loves Wolverine" > Sarah.txt

By default repeat simply prints the commands and does not run them. To execute the commands use the execute option with -x or --execute:

repeat 'echo "#1 loves #3" > #1.txt ' data.csv --execute --comma

This outputs nothing but creates three files:

Evan.txt
Laura.txt
Sarah.txt

Use verbose mode (-v or --verbose) to print commands as they are being run:

repeat 'cat #1.txt' data.csv --execute --verbose --comma

output:

[cat Evan.txt]
Evan loves Batman
[cat Laura.txt]
Laura loves James Bond
[cat Sarah.txt]
Sarah loves Wolverine

Finally clean everything up:

repeat 'rm #1.txt' data.csv --execute --verbose --comma

output:

[rm Evan.txt]
[rm Laura.txt]
[rm Sarah.txt]

Usage

repeat [options] format [files...]

format                  The format string of the command to execute
                        Arguments are passed with the dollar sign:

                          #1   First argument
                          #2   Second argument
                               (and so on...)

options

  -h, --help            Help on usage with examples
  -x, --execute         Execute the command instead of printing it
  -c, --comment <c>     Regexp to detect comment line (default: '(//|\#)')
  -s, --separator <c>   Regexp to split content on (default: '\t')
  -m, --marker <c>      Regexp to find argument in the format (default: '\#')

                        Note: Initially `$` seemed a better choice for the marker
                        but in practice it required too much character escaping.

  --comma               Alias for --separator ','
  --tab                 Alias for --separator '\t'
  --tabs                Alias for --separator '\t+'
  --spaces              Alias for --separator ' +'
  --whitespace          Alias for --separator '\s+'

  --strict              Prevent commands from running if any #args are missing
                        This is useful if your data is irregular and missing
                        arguments could lead to bad commands (default: off)

files                   List of files to use as arguments to the format string

                        Each line corresponds to a single format execution.
                        The --separator defines what the file is split on,
                        where the first part becomes #1, the second #2, etc.

Format Strings

Format strings act much like regular expression and printf formats. They can reorder items and place them around other characters.

#1          First argument
#2          Second argument
...         ...

Advanced formatting: (all on first argument)

Format strings can also come with printf-like formatting. The first argument is the argument number as shown above. The second part after the command is the printf format:

#{1,i}      Convert to an integer (truncated precision)
#{1,b}      Convert to a binary number
#{1,o}      Convert to an octal number
#{1,c}      Convert to an ascii character
#{1,d}      Convert to a signed integer
#{1,u}      Convert to an unsigned integer
#{1,e}      Convert to floating point in scientific notation
#{1,x}      Convert to hex
#{1,X}      Convert to hex with CAPITAL LETTERS

#{1,6s}     Convert to a string with six spaces or more
            ('hi' => '    hi')

#{1,-6s}    Convert to a string with six spaces left justified
            ('hi' => 'hi    ')

#{1,6i}     Convert to an integer with six spaces
            ('3.14159' => '     3')

#{1,0.4f}   Convert to a float rounded to 4 decimal places
            ('3.14159' => '3.1416')

#{1,6.4i}   Float with minimum of six spaces and 4 decimal places
            ('3.14159' => ' 3.1416')