From: "phasis68 (Heesob Park)" <phasis@...>
Date: 2013-12-08T19:56:18+09:00
Subject: [ruby-core:58967] [ruby-trunk - Bug #9221] Time.parse performance	becomes exponentially worse as string length grows


Issue #9221 has been updated by phasis68 (Heesob Park).


It can be reproduced on ruby 2.0.0-p247 i386-mingw32.

I modified a test code to do not use extension library.

C:\work>ruby -v
ruby 2.0.0p247 (2013-06-27) [i386-mingw32]

C:\work>type test.rb
require 'benchmark'
str = "12:00 PM fffffffffffffffffffffffffffff"
pat=/((?:\d+\s*:\s*\d+(?:\s*:\s*\d+(?:[,.]\d*)?)?|\d+\s*h(?:\s*\d+m?(?:\s*\d+s?)
?)?)(?:\s*[ap](?:m\b|\.m\.))?|\d+\s*[ap](?:m\b|\.m\.))(?:\s*((?:gmt|utc?)?[-+]\d
+(?:[,.:]\d+(?::\d+)?)?|[[:alpha:].\s]+(?:standard|daylight)\stime\b|[[:alpha:]]
+(?:\sdst)?\b))?/i
p Benchmark.measure { pat.match(str) }

C:\work>ruby test.rb
#<Benchmark::Tms:0x24a3440 @label="", @real=3.303211, @cstime=0.0, @cutime=0.0,
@stime=0.0, @utime=3.2969999999999997, @total=3.2969999999999997>

Interestingly, this issue is not reproduced on irb session or without filename.

C:\work>ruby < test.rb
#<Benchmark::Tms:0x24cf4e0 @label="", @real=0.0, @cstime=0.0, @cutime=0.0, @stim
e=0.0, @utime=0.0, @total=0.0>

C:\work>irb
DL is deprecated, please use Fiddle
irb(main):001:0> require 'benchmark'
=> true
irb(main):002:0> str = "12:00 PM fffffffffffffffffffffffffffff"
=> "12:00 PM fffffffffffffffffffffffffffff"
<:].\s]+(?:standard|daylight)\stime\b|[[:alpha:]]+(?:\sdst)?\b))?/i
=> /((?:\d+\s*:\s*\d+(?:\s*:\s*\d+(?:[,.]\d*)?)?|\d+\s*h(?:\s*\d+m?(?:\s*\d+s?)?
)?)(?:\s*[ap](?:m\b|\.m\.))?|\d+\s*[ap](?:m\b|\.m\.))(?:\s*((?:gmt|utc?)?[-+]\d+
(?:[,.:]\d+(?::\d+)?)?|[[:alpha:].\s]+(?:standard|daylight)\stime\b|[[:alpha:]]+
(?:\sdst)?\b))?/i
irb(main):004:0> p Benchmark.measure { pat.match(str) }
#<Benchmark::Tms:0x27059a0 @label="", @real=0.0, @cstime=0.0, @cutime=0.0, @stim
e=0.0, @utime=0.0, @total=0.0>
=> #<Benchmark::Tms:0x27059a0 @label="", @real=0.0, @cstime=0.0, @cutime=0.0, @s
time=0.0, @utime=0.0, @total=0.0>


----------------------------------------
Bug #9221: Time.parse performance becomes exponentially worse as string length grows
https://bugs.ruby-lang.org/issues/9221#change-43517

Author: mpelzsherman (Michael Pelz-Sherman)
Status: Open
Priority: Normal
Assignee: 
Category: core
Target version: current: 2.1.0
ruby -v: ruby-2.0.0-p247
Backport: 1.9.3: UNKNOWN, 2.0.0: UNKNOWN


See attached script. Output:

parsing 12:00 PM fffffffffffffffff
2013-12-05 12:00:00 -0500
elapsed: 0.010443
parsing 12:00 PM ffffffffffffffffff
2013-12-05 12:00:00 -0500
elapsed: 0.017739
parsing 12:00 PM fffffffffffffffffff
2013-12-05 12:00:00 -0500
elapsed: 0.028127
parsing 12:00 PM ffffffffffffffffffff
2013-12-05 12:00:00 -0500
elapsed: 0.049885
parsing 12:00 PM fffffffffffffffffffff
2013-12-05 12:00:00 -0500
elapsed: 0.071379
parsing 12:00 PM ffffffffffffffffffffff
2013-12-05 12:00:00 -0500
elapsed: 0.112612
parsing 12:00 PM fffffffffffffffffffffff
2013-12-05 12:00:00 -0500
elapsed: 0.184517
parsing 12:00 PM ffffffffffffffffffffffff
2013-12-05 12:00:00 -0500
elapsed: 0.293784
parsing 12:00 PM fffffffffffffffffffffffff
2013-12-05 12:00:00 -0500
elapsed: 0.476253
parsing 12:00 PM ffffffffffffffffffffffffff
2013-12-05 12:00:00 -0500
elapsed: 0.786087
parsing 12:00 PM fffffffffffffffffffffffffff
2013-12-05 12:00:00 -0500
elapsed: 1.256976
parsing 12:00 PM ffffffffffffffffffffffffffff
2013-12-05 12:00:00 -0500
elapsed: 2.019426
parsing 12:00 PM fffffffffffffffffffffffffffff
2013-12-05 12:00:00 -0500
elapsed: 3.300646
parsing 12:00 PM ffffffffffffffffffffffffffffff
2013-12-05 12:00:00 -0500
elapsed: 5.37757
parsing 12:00 PM fffffffffffffffffffffffffffffff
2013-12-05 12:00:00 -0500
elapsed: 8.763601
parsing 12:00 PM ffffffffffffffffffffffffffffffff
2013-12-05 12:00:00 -0500
elapsed: 14.165842
parsing 12:00 PM fffffffffffffffffffffffffffffffff
2013-12-05 12:00:00 -0500
elapsed: 23.888907
...



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