Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.lang.java.programmer > #28409 > unrolled thread

byte stream vs char stream buffer

Started byRoedy Green <see_website@mindprod.com.invalid>
First post2014-05-06 23:48 -0700
Last post2014-05-18 23:54 +0200
Articles 2 on this page of 22 — 6 participants

Back to article view | Back to comp.lang.java.programmer

Contents

  byte stream vs char stream buffer Roedy Green <see_website@mindprod.com.invalid> - 2014-05-06 23:48 -0700
    Re: byte stream vs char stream buffer Andreas Leitgeb <avl@auth.logic.tuwien.ac.at> - 2014-05-08 13:30 +0000
      Re: byte stream vs char stream buffer Roedy Green <see_website@mindprod.com.invalid> - 2014-05-09 09:34 -0700
        Re: byte stream vs char stream buffer Roedy Green <see_website@mindprod.com.invalid> - 2014-05-10 13:18 -0700
          Re: byte stream vs char stream buffer Robert Klemme <shortcutter@googlemail.com> - 2014-05-10 23:24 +0200
            Re: byte stream vs char stream buffer Roedy Green <see_website@mindprod.com.invalid> - 2014-05-10 16:08 -0700
              Re: byte stream vs char stream buffer Robert Klemme <shortcutter@googlemail.com> - 2014-05-11 16:02 +0200
                Re: byte stream vs char stream buffer markspace <markspace@nospam.nospam> - 2014-05-11 11:50 -0700
                  Re: byte stream vs char stream buffer markspace <markspace@nospam.nospam> - 2014-05-11 12:10 -0700
                  Re: byte stream vs char stream buffer Robert Klemme <shortcutter@googlemail.com> - 2014-05-11 21:34 +0200
                    Re: byte stream vs char stream buffer markspace <markspace@nospam.nospam> - 2014-05-11 12:55 -0700
                  Re: byte stream vs char stream buffer Roedy Green <see_website@mindprod.com.invalid> - 2014-05-16 16:55 -0700
                    Re: byte stream vs char stream buffer markspace <markspace@nospam.nospam> - 2014-05-16 17:36 -0700
                      Re: byte stream vs char stream buffer Roedy Green <see_website@mindprod.com.invalid> - 2014-05-16 19:52 -0700
                Re: byte stream vs char stream buffer Roedy Green <see_website@mindprod.com.invalid> - 2014-05-16 16:37 -0700
            Re: byte stream vs char stream buffer Roedy Green <see_website@mindprod.com.invalid> - 2014-05-10 17:21 -0700
              Re: byte stream vs char stream buffer Martin Gregorie <martin@address-in-sig.invalid> - 2014-05-11 09:05 +0000
                Re: byte stream vs char stream buffer Roedy Green <see_website@mindprod.com.invalid> - 2014-05-12 02:07 -0700
                  Re: byte stream vs char stream buffer Silvio <silvio@internet.com> - 2014-05-12 12:20 +0200
                    Re: byte stream vs char stream buffer Roedy Green <see_website@mindprod.com.invalid> - 2014-05-16 20:19 -0700
                    Re: byte stream vs char stream buffer Roedy Green <see_website@mindprod.com.invalid> - 2014-05-18 05:28 -0700
                      Re: byte stream vs char stream buffer Silvio <silvio@internet.com> - 2014-05-18 23:54 +0200

Page 2 of 2 — ← Prev page 1 [2]

#28532

FromRoedy Green <see_website@mindprod.com.invalid>
Date2014-05-18 05:28 -0700
Message-ID<hq9hn9lv7n1kfdcd3dfa3ur33d6dirui5v@4ax.com>
In reply to#28446
On Mon, 12 May 2014 12:20:11 +0200, Silvio <silvio@internet.com>
wrote, quoted or indirectly quoted someone who said :

>It would be interesting to repeat this experiment with different 
>character encodings and see if using UTF-16 versus UTF-8 makes a 
>difference here.

UTF-16 ends up reading twice as many bytes.  However the magic 50%
ratio is still optimal.


Using a random sample data file of 209,715,200 chars 419,430,402
bytes.
Using aggregate buffersize of 65,536 bytes.
Using charset UTF-16
end read using HunkIO 2.65 seconds
end read using using BufferedReader backed with BufferedInputStream
ratio 0.10 buffsize 65536 bytes  2.70 seconds
end read using using BufferedReader backed with BufferedInputStream
ratio 0.20 buffsize 65536 bytes  2.68 seconds
end read using using BufferedReader backed with BufferedInputStream
ratio 0.30 buffsize 65536 bytes  2.67 seconds
end read using using BufferedReader backed with BufferedInputStream
ratio 0.40 buffsize 65536 bytes  2.68 seconds
end read using using BufferedReader backed with BufferedInputStream
ratio 0.50 buffsize 65536 bytes  2.61 seconds <<<
end read using using BufferedReader backed with BufferedInputStream
ratio 0.60 buffsize 65536 bytes  2.72 seconds
end read using using BufferedReader backed with BufferedInputStream
ratio 0.70 buffsize 65536 bytes  2.73 seconds
end read using using BufferedReader backed with BufferedInputStream
ratio 0.80 buffsize 65536 bytes  2.78 seconds
end read using using BufferedReader backed with BufferedInputStream
ratio 0.90 buffsize 65536 bytes  2.72 seconds






-- 
Roedy Green Canadian Mind Products http://mindprod.com
Young man, in mathematics you don’t understand things.
You just get used to them. 
~ John von Neumann (born: 1903-12-28 died: 1957-02-08 at age: 53)

[toc] | [prev] | [next] | [standalone]

#28534

FromSilvio <silvio@internet.com>
Date2014-05-18 23:54 +0200
Message-ID<53792c2f$0$2957$e4fe514c@news2.news.xs4all.nl>
In reply to#28532
On 05/18/2014 02:28 PM, Roedy Green wrote:
> On Mon, 12 May 2014 12:20:11 +0200, Silvio <silvio@internet.com>
> wrote, quoted or indirectly quoted someone who said :
>
>> It would be interesting to repeat this experiment with different
>> character encodings and see if using UTF-16 versus UTF-8 makes a
>> difference here.
>
> UTF-16 ends up reading twice as many bytes.  However the magic 50%
> ratio is still optimal.
>
>
> Using a random sample data file of 209,715,200 chars 419,430,402
> bytes.
> Using aggregate buffersize of 65,536 bytes.
> Using charset UTF-16
> end read using HunkIO 2.65 seconds
> end read using using BufferedReader backed with BufferedInputStream
> ratio 0.10 buffsize 65536 bytes  2.70 seconds
> end read using using BufferedReader backed with BufferedInputStream
> ratio 0.20 buffsize 65536 bytes  2.68 seconds
> end read using using BufferedReader backed with BufferedInputStream
> ratio 0.30 buffsize 65536 bytes  2.67 seconds
> end read using using BufferedReader backed with BufferedInputStream
> ratio 0.40 buffsize 65536 bytes  2.68 seconds
> end read using using BufferedReader backed with BufferedInputStream
> ratio 0.50 buffsize 65536 bytes  2.61 seconds <<<
> end read using using BufferedReader backed with BufferedInputStream
> ratio 0.60 buffsize 65536 bytes  2.72 seconds
> end read using using BufferedReader backed with BufferedInputStream
> ratio 0.70 buffsize 65536 bytes  2.73 seconds
> end read using using BufferedReader backed with BufferedInputStream
> ratio 0.80 buffsize 65536 bytes  2.78 seconds
> end read using using BufferedReader backed with BufferedInputStream
> ratio 0.90 buffsize 65536 bytes  2.72 seconds
>
>
>
>
>
>

Thank you Roedy. Interesting results. Kind of surprising as well.

Cheers,

Silvio

[toc] | [prev] | [standalone]

Page 2 of 2 — ← Prev page 1 [2]

Back to top | Article view | comp.lang.java.programmer

csiph-web