Problem with "Auto-Tag from Amazon" in Japanese.

Post a reply

Smilies
:D :) :( :o :-? 8) :lol: :x :P :oops: :cry: :evil: :roll: :wink:

BBCode is ON
[img] is ON
[url] is ON
Smilies are ON

Topic review
   

Expand view Topic review: Problem with "Auto-Tag from Amazon" in Japanese.

Re: Problem with "Auto-Tag from Amazon" in Japanese.

by Prom » Sat Nov 29, 2008 5:03 am

Hi,

I have another problem with getting album information from amazon.co.jp.

When I am trying to tag an album from amazon.co.jp, every ID3 field is being displayed with the correct kanji except the title field. See here:
Image

Before it's all nonsense, and the tagging-suggestion is everything fine except the title information.
What could this be?

Thanks,

Prom

by REIKA » Sat Jan 19, 2008 2:49 am

I confirmed to work correctly in 3.0.2.1131.

by REIKA » Wed Jan 09, 2008 12:19 pm

http://www.oricon.co.jp/rank/ja/m/

The above is a sales ranking of 2007/12 in Japan.
I tried some album names and artist names in this.
The results were all "No result was found " .
Any Japanese CD information was not found though I tried others.
Therefore, the display of Japanese was not able to be verified.

English CD information on Amazon.com was able to be retrieved.
In other applications, Japanese CD information on Amazon.co.jp was able to be retrieved.
Therefore, I think that neither the network nor Amazon.co.jp are causes.

I tried in the following environments.
Win XP SP2 Japanese Ver.
MM3.0.2.1129 with/without Japanese language files(Latest Ver.)
Kaspersky = on/off

***********************
It is an additional information.
I manually inputted the word "friend" in listbox of "Auto-Tag from Web" for serch.

In case of MM2.5.5.998 , many results (Japanese informations were included) were found.
But Japanese informations were not displayed correctly.

In case of MM3.0.2.1129 , only one result (english) was found.

by jiri » Wed Jan 09, 2008 7:36 am

Re. what's new - it was taken from our internal bug-tracking system, that's why it probably wasn't as clear as it should be.

So, do you mean that it doesn't work well for you? We tested it even on Japanese version of Win XP and it worked fine. Do you have any simple example of what doesn't work (I suppose on Japanese Win XP)?

Jiri

Auto-tag from Amazon (Japanese) fails - [F4272]

by REIKA » Tue Jan 08, 2008 11:53 am

There is a line : "4237 Fixed remaining Japanese Localization Issues" in "MM3.0.2.1129 What's new".

But this issue does not seem to be included in them.
I want you to write in "What's new" concretely.

by REIKA » Fri Jan 04, 2008 3:37 pm

Is this fixed in the next release, too?
It is exciting ! Truthfully , I expected that it takes time to be fixed a little.
Thanks !

by jiri » Fri Jan 04, 2008 11:57 am

This should be fixed in the next release.

Jiri

by REIKA » Fri Jan 04, 2008 10:51 am

Because a large amount of strange character strings that cannot be made out at all are displayed,
the surprise when this bug is seen for the first time is large.
Perhaps, I think this bug is one of the big causes lowered the evaluation of MM in Japan.
Actually, many people have enumerated this problem as a dissatisfied point of MM in BBS,etc. of Japan.
Moreover, it is likely to become a differentiation that only has to be solved because Winamp has not
supported amazon.co.jp yet now.


I am using other convenient software for getting album art instead of MM for a long time.

http://team.thenexusnet.com/nexus/AAA/

This software can correctly get album arts of Japanese CDs from amazon.co.jp.
However, because the artist name and the album name are guessed based on the folder structure,
it is necessary to maintain the folder.

by Guest » Fri Dec 21, 2007 4:51 am

Has not this problem progressed at all?
MM3 is not improved at all either.
Cannot you solve it though think that the cause must be a code conversion from/to UTF8?

Me too

by fuzzynavel » Thu Nov 15, 2007 4:59 pm

I have the same problem (with MM3-RC2 on Vista). MM handles Japanese characters perfectly in everything but the Amazon.jp lookup. There it chokes, and the results that come back are blank. Any non-western characters in the list of results show as empty spaces.

I was hoping with the new version this would be fixed. Bummer.

Let me know if there is anything I can do to help test or solve this problem.

Exactly

by Andres_LO » Wed Feb 14, 2007 12:19 pm

Thank your for properly identifying this problem! I am a native English speaker, but I speak Japanese fluently and have a significant amount of Japanese music on my computer. You have no idea how hard it is to find a media player that properly recognizes (let alone can download), track information written in Japanese.

The strange thing was that for the first few weeks I had MM the Japanese recognition worked flawlessly. I could input Japanese (usually shift-js i believe) in the search box, get back properly encoded results (with track info) and update my ID3s. Now... nothing. I have no idea what happened. I am trying all combinations of regional options and MM settings.

One question. I thought that Japanese was non unicode. In fact I thought that ALL double byte character languages were non unicode. If that is the case how can amazon be returning unicode results? I thought just the opposite that I was sending out shift JS and MM was trying to convert everything to Unicode. especially since it does not support Japanese natively, but rather does so through windows libraries (is that the right term?). Am i totally wrong?

by Steegy » Wed Jun 14, 2006 1:09 pm

Message to Asian programmers visiting this page: please help the devs and your fellow countrymen/continentmen... 8)
You should know that it's a weird problem for all non-Asian people, so also not easy to solve.

by Guest » Wed Jun 14, 2006 12:42 pm

I also have this problem, also when I input Japanese characters it can't search at all, only if I input English characters.I don't know how to explain it in programming terms. Is there an easier way to help out?

As for my computer, I use Japanese as my secondary language on my English Version of Windows XP.
Image

by jiri » Thu May 25, 2006 10:37 am

Thanks for a detailed analysis, unfortunally I have no idea how should I properly work with Shift-JIS encoding.

To explain how it currently works: MM gets data from Amazon in UTF-8 encoding, decodes from UTF-8 to 16-bit wide Unicode (UTF-16) and shows it. It works well for all the languages I know. Do you have any programming reference explaining the problem of Japanese in the context of Win32 programming?

Jiri

by NiG » Thu May 25, 2006 6:08 am

screenshot : searching "My Little Lover".
Image
Japanese strings are not readable.

I'll show what happening in following picture.
Image
title of the album(large dotted) is "風と空のキリム".
2nd line is Hex-dump of the title encoded by Shift-JIS.
3rd lines shows same thing encoded by UTF-8.
And last line shows how looks when interpret (a) as Shift-JIS coded.
This is in agreement with a screenshot.

After all, it is thought that what is necessary is to make it deal with Japanese by UTF-8.

I found the topic
http://www.mediamonkey.com/forum/viewtopic.php?t=7324
I'm useing 2.5.2.951. is this version older?

Top