Roster export cannot handle special characters

[GoogleCodeExporter]

What steps will reproduce the problem?
1. Use special characters, such as the Swedish 'å', 'ä', or 'ö' in team or
player names

What is the expected output? What do you see instead?
Expected output is that these characters are displayed as normal on
exported rosters. on XML, the entire name is remove, and in PDF the letters
(and some surrounding letters) are replaced by ?

What version of the product are you using?
r482 from SVN

Original issue reported on code.google.com by setomi...@gmail.com on 18 Dec 2009 at 3:20

[GoogleCodeExporter]

What is the result of viewing the BOTOCS roster?

Original comment by funnyfin...@hotmail.com on 18 Dec 2009 at 7:08

[GoogleCodeExporter]

I think this is the PDF roster? Right?

Original comment by Nimda...@gmail.com on 19 Dec 2009 at 1:48

[GoogleCodeExporter]

I have tried all three roster types and all work with both the above listed
characters and Danish æ, ø å.

Are the characters displayed properly on the HTML roster ie. the team's page?

Original comment by Nimda...@gmail.com on 20 Dec 2009 at 8:52

[GoogleCodeExporter]

Yes, they're correctly displayed on all HTML pages. As it works for you but not 
for 
me, could it be some kind of systems issue?

You can find my instance running at: 
http://www8.cs.umu.se/~danielh/obblm/obblm-read-
only/index.php?section=objhandler&type=1&obj=2&obj_id=1

I'm running the r482 version, as I have uncommitted changes and so haven't 
uploaded 
the changes to SVN

Original comment by setomi...@gmail.com on 21 Dec 2009 at 7:38

[GoogleCodeExporter]

Can you confirm that the export you are talking about is pdf?  I saw the ? in 
your
PDF as well.

Original comment by funnyfin...@hotmail.com on 21 Dec 2009 at 3:12

[GoogleCodeExporter]

I have no clue on why this is happening if it is displayed properly on the HTML 
page.
Evidently a charset issue, but I don't see how.

Original comment by Nimda...@gmail.com on 21 Dec 2009 at 4:00

[GoogleCodeExporter]

The problem is both in the XML export AND in the PDF one. I haven't tried the 
BOTOCS 
one (as I have that module disabled).

Also, there seem to be a bug in the current export code, I'll verify that it is 
also 
in the SVN and create a new issue for it.

Original comment by setomi...@gmail.com on 22 Dec 2009 at 7:24

[GoogleCodeExporter]

I have a clue on this: changin utf8_decode($p->name) to $p->name solves this 
issue 
for me.

Could this be because I am hosting this on a UTF-8 system? I verified that the 
result is the same regardless if I enter the name from Linux or Windows.

Original comment by setomi...@gmail.com on 22 Dec 2009 at 9:24

[GoogleCodeExporter]

That utf8_decode was to prevent "weird" swedish/danish/german characters from 
not
displaying correctly. :-/ Displays correctly in my (nowadays) old installation 
on
both windows and debian.

/Daniel

Original comment by blodae@gmail.com on 22 Dec 2009 at 11:30

[GoogleCodeExporter]

We'll have to ask Daniel S. this question - he wrote the roster.
Iøm pretty sure my system here is too all UTF8.

Original comment by Nimda...@gmail.com on 22 Dec 2009 at 11:36

[GoogleCodeExporter]

Does this help?

http://aspn.activestate.com/ASPN/Mail/Message/php-Dev/2082358

Original comment by funnyfin...@hotmail.com on 22 Dec 2009 at 2:25

[GoogleCodeExporter]

I talked to our system admin, and apparently we're NOT supporting UTF-8. Which
probably means that OBBLM currently is UTF-8 dependant.

Original comment by setomi...@gmail.com on 25 Jan 2010 at 8:23

[GoogleCodeExporter]

Thanks Will for the link.

Seems like the problem will persist even when changing utf8_decode to
mb_convert_encoding. :-(

Setomidor - you're using characterset ISO-8859-1 in both webserver and DB? 
(Since the
chars works when removing utf8_decode.)

/Daniel

Original comment by blodae@gmail.com on 25 Jan 2010 at 2:13

[GoogleCodeExporter]

It seems like a local charset issue? Any news?
I see that for some reason my browser reads your site as ISO-8859-1, when it 
should
be UTF8.

Original comment by Nimda...@gmail.com on 11 Feb 2010 at 8:28

[GoogleCodeExporter]

Well, the current status is that the original error was my own fault, as it 
turned
out my first hosting actually uses 8859-1, but that also means that OBBLM is 
UTF-8
dependant (which might or might not be a problem)

Original comment by setomi...@gmail.com on 13 Feb 2010 at 1:08

[GoogleCodeExporter]

If OBBLM is UTF-8 dependant then this is the default nature of PHP apps, I did 
not do
it. Your host should support this since it's the default PHP setting 
(apparently) to
be UTF8.

I'm closing this then.

Original comment by Nimda...@gmail.com on 13 Feb 2010 at 1:55

• Changed state: Invalid

[GoogleCodeExporter]

No, the UTF-8 dependency is because the utf8_decode() function is used, as 
mentioned
in comments 8 & 9 above.

Original comment by setomi...@gmail.com on 14 Feb 2010 at 5:20