[ruby-core:39649] [Ruby 1.9 - Bug #4576] Range#step miss the last value, if end-exclusive and has float number

Issue #4576 has been updated by Yui NARUSE.


I made a patch which fixes following 3 cases:
* the error of loop count
* duplicated values from underflow
* excess of the end value

diff --git a/ChangeLog b/ChangeLog
index 0abb211..0f0c28b 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,16 @@
+Wed Sep 21 11:17:22 2011  NARUSE, Yui  <[e-mail]>
+
+	* numeric.c (ruby_float_step): improve floating point calculations.
+	  [ruby-core:35753] [Bug #4576]
+
+	* numeric.c (ruby_float_step): correct the error of floating point
+	  numbers on the excluding case.
+	  patched by Masahiro Tanaka [ruby-core:39608]
+
+	* numeric.c (ruby_float_step): use the end value when the current
+	  value is greater than or equal to the end value.
+	  patched by Akira Tanaka [ruby-core:39612]
+
 Tue Sep 20 18:08:51 2011  Nobuyoshi Nakada  <[e-mail]>
 
 	* vm_insnhelper.c (vm_get_cvar_base): reduce duplicated checks and
diff --git a/numeric.c b/numeric.c
index 18f5e1c..973da1f 100644
--- a/numeric.c
+++ b/numeric.c
@@ -1689,11 +1689,27 @@ ruby_float_step(VALUE from, VALUE to, VALUE step, int excl)
 	    if (unit > 0 ? beg <= end : beg >= end) rb_yield(DBL2NUM(beg));
 	}
 	else {
+	    double prev = beg == 0 ? -1 : 0;
 	    if (err>0.5) err=0.5;
-	    n = floor(n + err);
-	    if (!excl || ((long)n)*unit+beg < end) n++;
+	    if (excl) {
+		if (n>0) {
+		    if (n<err)
+			n = 1;
+		    else
+			n = floor(n - err) + 1;
+		}
+	    } else {
+		n = floor(n + err) + 1;
+	    }
 	    for (i=0; i<n; i++) {
-		rb_yield(DBL2NUM(i*unit+beg));
+		double d = i*unit+beg;
+		if (d == prev) continue;
+		if (d >= end) {
+		    if (!excl) rb_yield(DBL2NUM(end));
+		    break;
+		}
+		prev = d;
+		rb_yield(DBL2NUM(d));
 	    }
 	}
 	return TRUE;
diff --git a/test/ruby/test_float.rb b/test/ruby/test_float.rb
index e77b9e6..4fc8a6b 100644
--- a/test/ruby/test_float.rb
+++ b/test/ruby/test_float.rb
@@ -508,4 +508,39 @@ class TestFloat < Test::Unit::TestCase
       sleep(0.1+0.1+0.1+0.1+0.1+0.1+0.1+0.1+0.1+0.1)
     end
   end
+
+  def test_step
+    1000.times do
+      a = rand
+      b = a+rand*1000
+      s = (b - a) / 10
+      assert_equal(11, (a..b).step(s).to_a.length)
+    end
+
+    prev = 0
+    (1.0..(1.0+1E-15)).step(1E-16) do |current|
+      assert_not_equal(prev, current)
+      prev = current
+    end
+
+    (1.0..12.7).step(1.3).each do |n|
+      assert_operator(n, :<=, 12.7)
+    end
+  end
+
+  def test_step_excl
+    1000.times do
+      a = rand
+      b = a+rand*1000
+      s = (b - a) / 10
+      assert_equal(10, (a...b).step(s).to_a.length)
+    end
+
+    assert_equal([1.0, 2.9, 4.8, 6.699999999999999], (1.0...6.8).step(1.9).to_a)
+
+    e = 1+1E-12
+    (1.0 ... e).step(1E-16) do |n|
+      assert_operator(n, :<, e)
+    end
+  end
 end
----------------------------------------
Bug #4576: Range#step miss the last value, if end-exclusive and has float number
http://redmine.ruby-lang.org/issues/4576

Author: Joey Zhou
Status: Open
Priority: Normal
Assignee: 
Category: core
Target version: 1.9.4
ruby -v: -


=begin
Hi, I find that:

* if: range.exclude_end? == true
* and: any one in [begin_obj, end_obj, step] is a true Float(f.to_i != f)
* and: unless begin_obj + step*int == end_obj
* then: the result will miss the last value.

for example:

 p (1...6.3).step.to_a # => [1.0, 2.0, 3.0, 4.0, 5.0], no 6.0
 p (1.1...6).step.to_a # => [1.1, 2.1, 3.1, 4.1], no 5.1
 p (1...6).step(1.1).to_a # => [1.0, 2.1, 3.2, 4.300000000000001], no 5.4

 p (1.0...6.6).step(1.9).to_a # => [1.0, 2.9], no 4.8
 p (1.0...6.7).step(1.9).to_a # => [1.0, 2.9, 4.8]
 p (1.0...6.8).step(1.9).to_a # => [1.0, 2.9, 4.8], no 6.7

Maybe the #step is ok on integers, but there's something wrong if the range is end-exclusive and contain float numbers.
=end



-- 
http://redmine.ruby-lang.org