Split a text file in ruby - ruby

Question

I have a text file with several different sections. Each section has a header followed by the actual data. For example:
Header1
x,y,z
x,y,z
x,y,z
Header2
a,b,c
a,b,c
a,b,c
I want to read through the file in one pass and do different things with the data present under each section. I know how to parse the data, but I'm having trouble figuring out how to code the logic for "Do this until hitting Header2, then do something else until Header3, etc."
I'm using ruby, and I haven't really come across any examples of doing this. Any suggestions?

Answer 1

At the simplest you could do something like this:
# Process lines for header1
def do_header1(line)
puts line.split(/,/).join("|")
end
# Process lines for header2
def do_header2(line)
puts line.split(/,/).map{ |e| e.upcase}.join(",")
end
header1 = false
header2 = false
# Main loop
File.open("file.txt").each_line do |line|
if line.chomp == 'Header1' # or whatever match for header1
header1 = true
header2 = false
next
end
if line.chomp == 'Header2' # or whatever match for header2
header1 = false
header2 = true
next
end
do_header1(line) && next if header1
do_header2(line) && next if header2
end
If the number of headers becomes too high, you can start tracking headers with an integer:
header = -1
# Main loop
File.open("file.txt").each_line do |line|
if line.chomp == 'Header1' # or whatever match for header1
header = 1
next
end
if line.chomp == 'Header2' # or whatever match for header2
header = 2
next
end
do_header1(line) && next if header == 1
do_header2(line) && next if header == 2
end

Answer 2

A solution using objects. For each line you ask each parser if a new section has started that the parser can parse.
class Section1Parser
def section? potential_header
potential_header.chomp == 'Header1'
end
def parse line
puts "Section 1: #{line.split(/,/).join("|")}"
end
end
class Section2Parser
def section? potential_header
potential_header.chomp == 'Header2'
end
def parse line
puts "Section 2: #{line.split(/,/).join("|")}"
end
end
parsers = [Section1Parser.new, Section2Parser.new]
selected_parser = nil
File.open("c:\\temp\\file.txt").each_line do |line|
if new_parser_detected = parsers.detect {|p| p.section? line }
selected_parser = new_parser_detected
next # skip header
end
selected_parser.parse line if selected_parser
end

Answer 3

Would something like this work?
File.open('datafile').each_line do |s|
if s =~ /^headerpattern$/
#Start a new parsing block
...
else
#Parse data
...
end
end

Answer 4

In my case 'Header' was in form of following string OBJECT ObjectType ObjectNumber ObjectName
if File.exist?("all.txt") then
object_file = File
File.open("all.txt").each_line do |line|
file_name = case
when line.match('^OBJECT Table.*')
"TAB" + line.split[2] + ".TXT"
when line.match('^OBJECT Form.*')
"FOR" + line.split[2] + ".TXT"
when line.match('^OBJECT Report.*')
"REP" + line.split[2] + ".TXT"
when line.match('^OBJECT Dataport.*')
"DAT" + line.split[2] + ".TXT"
when line.match('^OBJECT XMLPort.*')
"XML" + line.split[2] + ".TXT"
when line.match('^OBJECT Codeunit.*')
"COD" + line.split[2] + ".TXT"
when line.match("^OBJECT MenuSuite.*")
"MEN" + line.split[2] + ".TXT"
when line.match('^OBJECT Page.*')
"PAG" + line.split[2] + ".TXT"
when line.match('^OBJECT Query.*')
"QUE" + line.split[2] + ".TXT"
end
unless file_name.nil?
File.exist?(file_name) { File.delete(file_name) }
object_file = File.open(file_name,"w")
end
object_file.write(line)
end
end
But there are some prerequisites: I'm always sure that first line of the file will contain a header. I'm also not closing file (this will definitely draw my karma to the zero one day).