JakeM

The world didn’t end and slimstats still rocks.

Waiting patiently for pages stats. ??

Your fix is working.

The problem mentioned in my last post is related to older IE versions. They don’t support utf-8 encoding in the URL.

Earlier, I found the same broken links in GA logs. So it’s not related to slimstat anymore.
In the meantime, we’re gonna have to wait for older IE versions to vanish. ??

Thanks for your support!

I’m gonna try to catch it through GA. I let you know.

Can you contact a mod to remove the links?

Thanks!

Camu,

Please remove the baidu links you posted earlier in this thread.

In the post that starts with:

If I put 聯 in the search field

and the next one:

your DB export tells me

I have encountered similar broken links in the last couple hours and I’m not sure If they are from baidu results or from people who are clicking those ones. ??

This referrer showed up in the last hour and the ip of the source was from Cambodia (different encoding?):

https://www.baidu.com/s?wd=%3F%3F弘%3F诀瘪

Got it installed a few hours ago, and it looks like you solved the bug.

I haven’t encountered a single corrupted referrer since.

Thanks!

P.S: I’ll change the status of this thread to resolved by tomorrow.

Let me know If you want me to test your bug fix. ??

WP SlimStat receives the URL encoded as EUC-CN, not UTF-8 (unless ie=ut8 is present).

I’m confused now, because If you take a look at the 2 last rows, the referrers contain the -utf-8 at the end, but still their search terms are corrupted (‘????????????’).

That’s correct.

The 1st link is what is sent as the referrer when you click one of the result in baidu. Then it is supposed to be translated back into Chinese characters (same as in the 2nd link) when received by wp and/or slimstat.

I’m not really sure which one of them two is in charge of the conversion.

Baidu works with UTF-8 only and convert big5 characters on the fly when you click search.

copy paste this 聯 in baidu and hit search. You’ll see this Traditional Chinese (big5) cchanging to 联 for Simplified Chinese (handled by UTF-8 ).

Check the last row and you’ll see that link is encoded with -utf-8 at it’s end but still doesn’t show up correctly in slimstat.

The problem doesn’t seem to occur on Baidu’ side.

https://www.j4.com.tw/big-gb/

Strictly UTF-8

I just send you some rows from wp_slim_stats on your email.

Thanks!

On your test website, you can try to create a page that contains big5 encoded paragraphs, and try to search some through baidu.com.

When the link works it shows up like that:

https://www.baidu.com/s?wd=%E5%B9%BF%E4%B8%9C%E6%98

And shows up like that when it doesn’t:

https://www.baidu.com/s?wd=%3F%3F%3F%3F%3F%3F%3F%3F</&gt;

https://www.baidu.com/s?wd=%3F%3F%3F%3F%3F%3F%3F%3F%3F?%3F%3F%3F%3F%3F%3F%3F?%3F%3F

%3F are in fact question marks (?)

The same referrer is sometimes correctly displayed. The reason is that it contains the utf-8 extension at the end of the link, for which I have no idea why.

https://www.baidu.com/baidu?word=公司&ie=utf-8

Hi Camu,

You can check the screenshots here:

1st: Right Now | 2nd: Spy View = https://imgur.com/a/L6kIk
3rd: Recent Search Terms = https://imgur.com/6rmIp
Let me know if you need me to help you debug and track down that issue.

PS: Thanks for your great plugin!

Just to add that Google Analytics display the referrers without a problem.