AdWords is now Google Ads. Our new name reflects the full range of advertising options we offer across Search, Display, YouTube, and more. Learn more

Ads
3.7K members online now
3.7K members online now
For questions related to Google Shopping and Merchant Center. Learn to optimize your Shopping ads
Guide Me
star_border
Reply

Fixing / Detecting double-encoded utf-8

Visitor ✭ ✭ ✭
# 1
Visitor ✭ ✭ ✭

Great to reply to a 2 year old post.. but Google Product Feed is giving me a headache. 

You said: "later encoded as utf-8 -- that is not allowed"... ??? That does not make any sense. How come I cannot encode somethig 10 times back and forth utf-8 -> iso-8859-1 if I like. The result will be in correct format anyhow.

 

There is something really weird with Google Merchant center product feed import.

I have UTF-8 xml file which I'm trying to import there as swedish. It is really UTF-8. Every single character. And it has scandinavian characters in it. Those characters make merchant center to give me an error for product descriptions: 'double encoded utf-8'. I also (for fun) tried to import the xml with iso-8859-1 (yes, the xml was also formatted alike). It gave me the same error...??? duh.

 

Ok then I imported the same file without scandinavian characters (ä/ö) and all went well without errors.

 

I contacted support from merchant center. They informed me that "Swedish have different character than Finnish". What??? Ok, umm no. Swedish characters are really all alike Smiley Happy. When I asked more they said "we have already exhausted too much time for this". Thx.

 

The last straw. Rules in merchant center. Made a test to "find and replace" ä -> ä and ö -> ö. All well, the product feed went fine without errors. YES!. Next test to convert also capitals Ä -> Ä. Ok, rule's find and replace is damn CASE-INSENSITIVE. How can it be??

 

So i can get the file ok when I replace those characters which give me error but I cannot get character in correct case. Great.

 

If somebody can help I would really appreciate.

2 Expert replyverified_user

Fixing / Detecting double-encoded utf-8

Top Contributor
# 2
Top Contributor

If you like assistance from use and as we are simply community members and not part of Google, please provide samples of the data you submit.

 

Double encoding is an issue where a special character was decoded, than decoded again.

 

For example Ä in html entities is Ä but in your case could be &Auml

 

So this is not Google reading it incorrect, this is the file that is generated not being correctly setup. I have dealt with this allot and can understand the frustrations, but be ashured it's not Google, it is your file being wrong. Carefully inspect your coding and output and test to make sure there is no double encoding going on. Also make sure your mysql servers and php are all the same. For example if you send UTF-8 data to Mysql, while the mysql server is setup in latin. Than it will cause double encoding issues.

 

As there can be allot of reasons contributing to your issue, you need to make sure everything is setup from start to end. Contact your web developer and server admin to start resolving this issue.

 

Ultimately it is all down to the output written in your XML file.

 

Hope it helps.

 

Twitter | Linkedin | Community Profile | Shopping Feed Tips From FeedArmy
Did you find any helpful responses or answers to your query? If yes, please click on ‘Accept As Solution’

Fixing / Detecting double-encoded utf-8

Visitor ✭ ✭ ✭
# 3
Visitor ✭ ✭ ✭

Hi.

There is no difference if I use umlauts ä and ö to replace ä and ö.

 

Ill give you a very funny (kind of) example. Copypaste carefully Smiley Happy

 

This one will not go through. Gives double encoding utf-8 error.

<g:description><![CDATA[ääkköstestiä öökkösille
ääkköstestiä öökkösille. ääkköstestiä öökkösille
ääkköstestiä öökkösille. ääkköstestiä öökkösilleääkköstestiä öökkösille. ääkköstestiä öökkösille
abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract]]></g:description>

 

This one will not give error.

<g:description><![CDATA[ääkköstestiä öökkösille
ääkköstestiä öökkösille.
ääkköstestiä öökkösille. ääkköstestiä öökkösilleääkköstestiä öökkösille. ääkköstestiä öökkösille
abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract abstract]]></g:description>

 

There is not much difference. Please explain why other is fine and other is not.

Fixing / Detecting double-encoded utf-8

Top Contributor
# 4
Top Contributor

Copy pasting will not work, because what you are copying from what you see, which is already decoded by the browser. So that's why its the same. What you need to do is view the actual output, which you can do by viewing the browsers sourcecode.

Browser will make both look the same, however are actually different.

 

So please share the xml file link that is generated, even better a so called good one and bad one.

 

Thanks.

Twitter | Linkedin | Community Profile | Shopping Feed Tips From FeedArmy
Did you find any helpful responses or answers to your query? If yes, please click on ‘Accept As Solution’

Fixing / Detecting double-encoded utf-8

Visitor ✭ ✭ ✭
# 5
Visitor ✭ ✭ ✭

Umm.. could you please test first 2 cdata’s. I thought at first this is crazy but those handwritten texts (as utf-8) work differently, even they are just slightly different. Other cdata has one sentence deleted.

Fixing / Detecting double-encoded utf-8

[ Edited ]
Visitor ✭ ✭ ✭
# 6
Visitor ✭ ✭ ✭

So yeah.. i was on a sickleave as my kid was sick. But really, I'm beginning to call this a weird bug. Even I can see you are not believing me when I say copypasting in this case really works and would like to see you trying this.


These two sentences of rubbish work differently when copypasted into the XML as cdata description.

If you compare option 1 and 2 you can see below there is one sentence missing in option 2.

There should not be any reason for product feed import to give an error of double encoded utf8 for the option 1 and NOT for the option 2.

 

1) This one DOES NOT work (gives double encoded utf-8 error):

---

ääkköstestiä öökkösille
ääkköstestiä öökkösille. ääkköstestiä öökkösille
ääkköstestiä öökkösille. ääkköstestiä öökkösilleääkköstestiä öökkösille. ääkköstestiä öökkösille

---

XML here: https://drive.google.com/file/d/10rw8kGH3IH2LQxdTGHDrtyYLgt5akvKm/view?usp=sharing

 

2) This one works without (NO double encoded utf8 error):

---

ääkköstestiä öökkösille
ääkköstestiä öökkösille.
ääkköstestiä öökkösille. ääkköstestiä öökkösilleääkköstestiä öökkösille. ääkköstestiä öökkösille

---

XML here: https://drive.google.com/file/d/1Nu17r_8c10mJ-RLOrKoIo-Fhl8KGh1ds/view?usp=sharing

 

If you replace ä / ö with umlauts it does not make any difference.

Fixing / Detecting double-encoded utf-8

Top Contributor
# 7
Top Contributor

I did not check the data, because I want to check the original file. As I am volunteering and to be as helpful as possible checking from the source is required. Otherwise we are going around in circles. If providing the exact URL you are submitting to Google Merchant is not possible, than I'm not able to assist.

 

I have been doing this for many years and I want to help you, but you do need to provide a real world example and not just a copy paste from somewhere with random text.

 

Hence why I asked, please provide the XML file you are submitting to Google and not some random text that is not a product.

 

Alternatively if this is not possible, please contact Google directly : https://support.google.com/adwords/answer/7218750

 

Please note this is a member to member forum, meaning we are volunteers trying to help other people and are not part of Google. Due to the technical issue you have, for me to help you, I require real data to analyze.

 

I hope you understand this is not a simply issue you are having.

Twitter | Linkedin | Community Profile | Shopping Feed Tips From FeedArmy
Did you find any helpful responses or answers to your query? If yes, please click on ‘Accept As Solution’

Fixing / Detecting double-encoded utf-8

[ Edited ]
Visitor ✭ ✭ ✭
# 8
Visitor ✭ ✭ ✭

Hi,

I already posted links.. but somehow my post didnt appear.. I hope it gets there now.

And yes I've been working with different character sets also for over 2 decades Smiley Happy But this is something new eh.

Fixing / Detecting double-encoded utf-8

Visitor ✭ ✭ ✭
# 9
Visitor ✭ ✭ ✭

Hi, my links appeared in my post above now.

And thanks for answering this issue. I thought at first that you could just paste the text to some existing product feed xml Smiley Happy

 

So there is now 2 issues:

1) double encoded utf-8 error

2) merchant center "Rules" - find and replace is doing it case-insensitive.

Fixing / Detecting double-encoded utf-8

Top Contributor
# 10
Top Contributor

Thanks for the links, so your issue stems from two special characters sitting next to each other. It seems that Google is reporting this as double encoding. As swedish is not my native language, do you have words where two identical special character letters sit next to each other? If yes, than you need to report this to Google as this is a technical issue on there side. If no, then of course do not submit these kind of words.

 

Regardless as you have reported the issue, I'm guessing you do have words with two identical special characters, so you need to submit feedback to Google by going to the top right corner, click on feedback, and give them clear explenation that it is due to what I explained. Than give them 1 word example. That triggers the issue.

 

However this can take a few months for them to fix, so I recommend removing words that have two identical special characters next to each other. Than resubmit the feed and you will be fine.

 

So instead of submitting ääkköstestiä submit äkköstestiä 

 

Hope it helps.

Twitter | Linkedin | Community Profile | Shopping Feed Tips From FeedArmy
Did you find any helpful responses or answers to your query? If yes, please click on ‘Accept As Solution’