re-format text from e.g. rainbowstream twitter output

twitter-format.pl

#!/usr/bin/perl

use strict;
use warnings;

$|++;

use Text::Format;
use IO::Select;
use Time::HiRes;
use Getopt::Long;

my $sleep_interval = 2;
my $slow = 0;
GetOptions(
  "sleep|s=i" => \$sleep_interval,
  "slow"      => \$slow,
) or die "Error parsing arguments";

my $s = IO::Select->new(\*STDIN);
my @input;
my $prev_tweet_head;
my @prev_tweet;
my $current_tweet_head;

while (1) {
  # read in available tweets
  while ($s->can_read(2)) {
    my $line = read_unbuffered_line();
    last unless defined $line;
    push @input, $line;
  }

  # sleep if nothing to read
  unless (@input) {
    # print the final tweet we may have buffered
    print_tweet($prev_tweet_head, @prev_tweet);
    undef $prev_tweet_head;
    undef @prev_tweet;
    sleep $sleep_interval;
    next;
  }

  # parse tweets
  while ($_ = shift(@input)) {
    # strip timestamp
    s/^.{29}//;

    # is this the start of new a tweet?
    if (m/^(\x1b\x5b\S+m|\s)+([^@]|.@.)+(\x1b\x5b\S+m|\s)+@\S+(\x1b\x5b\S+m|\s)+(just now|\S+ second.? ago|\S+ minute.? ago)(\x1b\x5b\S+m|\s)+$/) {
      s/(\S+just now|\S+ second.? ago|\S+ minute.? ago)(\x1b\x5b\S+m|\s)+$//;
      $current_tweet_head = "$_\n";

      # print the previous tweet
      print_tweet($prev_tweet_head, @prev_tweet);
      sleep $sleep_interval;
      undef @prev_tweet;
      next;
    }

    # skip line after start of new tweet
    if (defined $current_tweet_head) {
      $prev_tweet_head = $current_tweet_head;
      undef $current_tweet_head;
      next;
    }

    # queue the contents of the tweet for printing
    push @prev_tweet, $_;
  }

  # now loop to read new tweets as they come in
}

sub print_tweet {
  my $head = shift;
  return unless $head;
  my $t = Text::Format->new(columns=>`tput cols`,firstIndent=>4,rightMargin=>1,leftMargin=>1);
  my $text = $head . $t->paragraphs(@_);
  unless ($slow) {
    print $text, "\n";
    return;
  }
  my $b;
  foreach (split m/\n/, $text) {
    s!https?://\S+!!g;
    my $line = $_;
    $line =~ s/\e\[[0-9;]*m(?:\e\[K)?//g;
    if ($line =~ m/^\s+$/) {
      next if $b;
      $b = 1;
      print "\n";
      next;
    }
    undef $b;
    foreach (split m//) {
      print "$_";
      Time::HiRes::sleep(rand(1) / 5) if m/[A-Za-z]/;
    }
    print "\n";
  }
  print "\n" unless $b;

  return;
}

# https://stackoverflow.com/a/54816600/195656
sub read_unbuffered_line {
  my $line = "";
  while (sysread(STDIN, my $nextbyte, 1)) {
    $line .= $nextbyte;
    return $line if $nextbyte eq "\n";
  }
  return;
}

1;