Shows using the spreadsheet gem to read a file that is flawed. Worksheets and dimensions are correct, but cell data cannot be read. File that causes the issue can be found at: http://dl.dropbox.com/u/83477489/sample.xls

read_flawed_spreadsheet.rb

#!/usr/bin/env ruby

wb = Spreadsheet.open('sample.xls')

puts "Opened Spreadsheet, #{wb.sheet_count} sheets encountered"
(0..wb.sheet_count-1).each do |sheet_number|
  ws = wb.worksheet(sheet_number)
  puts "\tWorksheet #{ws.name} has #{ws.row_count} rows and #{ws.column_count} columns}; and Dimensions: #{ws.dimensions}"
end

puts "Opened Spreadsheet, #{wb.sheet_count} sheets encountered"

ws = wb.worksheet(0)
puts "\tWorksheet '#{ws.name}' has #{ws.row_count} rows and #{ws.column_count} columns}; and Dimensions: #{ws.dimensions}"

row_index = 5
row = ws.row row_index
puts "\tWorksheet '#{ws.name}' row #{row_index}: #{row.to_s}"

puts "\tWorksheet '#{ws.name}' cell(#{row_index}, 0): #{ws.cell(row_index, 0)}"

Does the solution of Javix work for you?

require 'spreadsheet'

book = Spreadsheet.open('sample.xls')
puts "file found" if book
puts "Sheets found: #{book.worksheets.size}"
book.worksheets.each do |sheet|
puts "Got sheet: #{sheet.name}"
end

sheet1 = book.worksheets[0]
puts "Got the first sheet: #{sheet1.name}"
puts "cell 1: #{sheet1[2,1]}"

I can't figure out out why the standard way to read rows values like that:

sheet1.rows each { | row| puts row }

also see: https://groups.google.com/d/msg/rubyspreadsheet/vkHOulBnjMc/r792XlGEWLsJ

read_flawed_spreadsheet.rb
results in:
in `

': uninitialized constant Spreadsheet (NameError)

Where as the solution of Javix results in:

file found
Sheets found: 7
Got sheet: Summary Statistics
Got sheet: Patient Sample
Got sheet: Local Weights
Got sheet: Age.Gender Dist
Got sheet: Total Cost Risk Dist
Got sheet: Hospitalization Risk Dist
Got sheet: Build Options
Got the first sheet: Summary Statistics
cell 1:

Yes, that's the same thing that I have been getting. You can read the sheets and their titles, but you cannot see any of the data.

So, no, this is still not working for me.