convert the Twitter archive csv file into something useful sheesh

csv-to-array.awk

#!/usr/local/bin/gawk -f

# I can't get multiline regular expressions to work well outside of
# TextMate, so before running this on tweets.csv I opened it in
# TextMate and replaced:

# ([^"])\n
# for
# $1\\n

# I had to run it twice (!?) to turn all the multiline tweets into
# single-line tweets with '\n' instead of actual line breaks.  Then
# you can run this:

BEGIN {
  FS="\",\"";
  print "module.exports = [";
  ORS = ",\n";
}
{
  # http://www.linuxquestions.org/questions/programming-9/removing-first-line-with-awk-470129/#post2362671
  if (NR != 1) {
    if (substr($6,0,2) != "RT"){
      tweet = gensub("'","\\\\'","g",$6);
      # lol just keep adding backslashes until it works
      tweet = gensub(/\\([^'n])/,"\\\\\\\\\\1","g",tweet);
      print "'"tweet"'";
    }
  }
}
END {
  ORS = "\n";
  print "];";
}

# POST PROCESSING

# Life is too short to figure out why awk sucks so much, so I just
# went back to TextMate and substituted:

# \\\',$
# for
# \'',

# to close some strings.  And manually deleted the trailing comma on
# the last array element.