CODEWITHSUNDEEP
rubycsv
dax
dax

Reputation: 11007

preserving whitespace when writing file to csv

This is a case of unfortunate input data, i think.

Given a file as can be found here, how can i preserve necessary whitespace as in the link? When I parse this using the code below, the final row gets compacted by being shifted to the left - and the dates get screwy (february gets 31 days, but december doesn't).

I know the code is doing exactly what I tell it. It is splitting on white space. It should maintain a certain number of rows, but since there are no delimiting characters, I'm not sure how to ask for what I want!

code is as follows: 

 #!/usr/bin/env ruby

 require 'open-uri'
 require 'csv'

 class MoonDataSeeder

   def initialize(year = nil)
     @year = year || Time.now.year
   end

   def seed
     convert_to_csv
   end

   private

   def convert_to_csv
     CSV.open('test_file', 'wb', :force_quotes => true, :skip_blanks => false) do |csv|
       feed_data.lines[-39..-7].each do |row|
         csv << row.split
       end
     end
   end

   def feed_data
     @feed_data ||= open(feed_uri).read
   end

   def feed_uri
     host = "http://aa.usno.navy.mil/cgi-bin/aa_moonill2.pl" 
     host + "?form=2&year=#{year}&task=00&tz=0&tz_sign=-1"
   end

   def year
     @year
   end

 end

Upvotes: 0

Views: 205

Answers (1)

Kristj&#225;n
Kristj&#225;n

Reputation: 18813

What you're really doing is parsing fixed-width data, rather than delimited data (well, maybe it used to be tabs, but now it's unhelpful spaces). Try the fixedwidth gem instead.

Or, I'd just do it manually. This works on the lines containing the data:

data = lines.map do |line|
  line.strip!
  [].tap do |pieces|
    pieces << line.slice!(0, 3)   # Day
    line.slice!(0, 4)             # Space
    until line.empty?
      pieces << line.slice!(0, 4) # Month
      line.slice!(0, 5)           # Space
    end
  end.map(&:strip)
end

And just for fun, here's a version using regular expressions

data = lines.map do |line|
  line.scan(/([\w. ]{4})( {4,5})?/).map(&:first)
end

Upvotes: 1

Related Questions