fys

perl output

pabs@halcyon:~/git/fys> perl perl.pl
host,time
blargh.org,"60d 3h"
blum.net,"40d 3h"
baz.biz,"32d 3h"

perl.pl

#!/usr/bin/perl

use warnings;
use strict;
use Text::CSV;

# create csv parser
my $csv = Text::CSV->new({
  binary => 1
}) or die "Cannot create CSV parser: " . Text::CSV->error_diag();

# read csv rows
my @rows;

# load input file
open my $fh, '<:encoding(utf8)', 'test.csv' 
  or die "Couldn't open test.csv: $!";

# read rows
while (my $row = $csv->getline($fh)) {
  # skip rows that do not have a status of "down"
  $row->[1] =~ /down/ or next;

  # get number of days
  my $days = ($row->[2] =~ /(\d+)d/) ? $1 : 0;
  
  # skip rows that are less than 30 days old
  next unless $days > 30;

  # add to output
  push @rows, {
    host => $row->[0],
    time => $row->[2],
    days => $days,
  };
}

# set line terminator
$csv->eol("\n");

# print csv header
$csv->print(*STDOUT, ['host', 'time']);

# write rows to output
$csv->print(*STDOUT, [$_->{host}, $_->{time}]) for (sort { 
  # sort rows in reverse order
  $b->{days} <=> $a->{days} 
} @rows);

python output

pabs@halcyon:~/git/fys> python python.py
host,time
blargh.org,60d 3h
blum.net,40d 3h
baz.biz,32d 3h

python.py

#!/usr/bin/python

# load libraries
import re, csv, sys

# cache down regex
DOWN_RE = re.compile(r"down", re.I)

def is_down(val):
  return re.match(DOWN_RE, val)

# cache days regex
DAYS_RE = re.compile(r"(\d+)d")

def to_days(val):
  # match days portion of time value
  md = re.match(DAYS_RE, val)

  # return 0 if regex didn't match
  if md is None:
    return 0

  # convert number of days to integer
  return int(md.group(1))

# open output csv and write header
out = csv.writer(sys.stdout)
out.writerow(["host", "time"])

# open input csv "test.csv"
with open('test.csv', 'rb') as src_fh:
  # convert to rows and filter on relevant rows
  rows = filter(lambda row: row["down"] and row["days"] > 30, [{
    "host": row[0], 
    "down": is_down(row[1]),
    "time": row[2],
    "days": to_days(row[2]),
  } for row in csv.reader(src_fh)])

  # sort rows by descending number of days
  rows.sort(lambda a, b: cmp(b["days"], a["days"]))

  # print rows to output csv
  for row in rows:
    out.writerow([row["host"], row["time"]])

ruby output

pabs@halcyon:~/git/fys> ruby ruby.rb
host,time
blargh.org,60d 3h
blum.net,40d 3h
baz.biz,32d 3h

ruby.rb

#!/usr/bin/env ruby

# load csv library
require 'csv'

# write to standard output
CSV($stdout) do |out| 
  # write output csv header
  out << %w{host time}

  # read from input csv "test.csv"
  CSV.read('test.csv').map do |row| 
    # match days portion of time column
    md = row[2].match(/(\d+)d/)

    # convert to integer, or zero if we couldn't match any day value
    days = md ? md[1].to_i : 0

    # map source row to hash of values
    { 
      host: row[0], 
      down: row[1] =~ /down/,
      time: row[2],
      days: days,
    }
  end.select do |row|
    # filter on rows that we care about
    row[:down] && row[:days] > 30
  end.sort do |a, b|
    # sort rows in reverse order
    b[:days] <=> a[:days]
  end.each do |row|
    # convert to csv row and add to output
    out << [row[:host], row[:time]]
  end
end

test.csv