From: Pete McBreen <pete@...>
Date: 2002-03-08T02:36:27+09:00
Subject: Re: file reading impossibly slow? 1.6.6 vs. 1.6.2 numbers

Peter Hickman suggested another test, so here are my results from Windows
NT4

D:\ruby\bin\test>\ruby162a\bin\ruby -v
ruby 1.6.2 (2000-12-18) [i386-cygwin]

D:\ruby\bin\test>\ruby162a\bin\ruby testfile.rb
1.672
1.572

D:\ruby\bin\test>ruby -v
ruby 1.6.6 (2001-12-26) [i586-mswin32]

D:\ruby\bin\test>ruby testfile.rb
30.414
31.736

Same script, but the 1.6.6 version is _18_ times slower than the old 1.6.2
cygwin version. And in case anyone is wondering about buffering, the times
are consistent when I alternate running the 1.6.6 and 1.6.2 versions.

I'd have to guess that there is something about the 1.6.6 mswin32 build
that is causing the grief, and/or the old cygwin build was really great:-)

The script is slightly different than Peter Hickman's, mainly because I
wanted to keep the file around. It is also only 100,000 records (2.4 MB)

#!ruby

tf = File.new("data.txt", "w+")
(1..100000).each {|i| tf.print "Here is a line of text\n" }
tf.close

tf = File.new("data.txt")
t1 = Time.now
tf.each_line {|line|}
puts Time.now - t1
tf.close

tf = File.new("data.txt")
t1 = Time.now
while(!tf.eof?)
        line = tf.gets
end
puts Time.now - t1
tf.close


I trust someone can figure this out,

Pete
----
Pete McBreen, McBreen.Consulting , Cochrane, AB
email: petemcbreen@acm.org    http://www.mcbreen.ab.ca/

Author, "Software Craftsmanship The New Imperative"
Addison-Wesley (C) 2002
http://www.amazon.com/exec/obidos/ASIN/0201733862
Finalist for the 2002 Jolt Awards