#4555 – 404 errors when DOCman document has only UTF-8 characters Edit

Hi

This might be an issue with the URL length but I am a bit surprised as 30 characters times 4 bytes in only about 120 chars or so. As the limit is 255 chars, that should not be an issue.
On my dev sites, I don't have greek URLs but I do have hebrew URLs which are much longer and works very fine.

1 - Does this only happens with DocMan? please do a test with an standard article. It might absolutely be an issue in DocMan itself.
2 - Does that URL looks good and is complete when you search for it in the URL manager?
3 - What happens if you customize it to be a bit shorter? does it save properly and is redisplayed normally in the URL manager, without some characters missing?
4 - Please update to the latest dev version (from development versions download area). It should not have any effect on your problem, but will provide a level field for investigation.

Rgds

Hi

Despite what they say, it's most likely a bug in DOCMan (I am sure you realize the fact it happens only with DOCMan and not all your other content is important here).

The only way to find out is that you take a full backup of the site using Akeeba backup. I can then download this backup, install on my local machine and then actually debug the PHP code in DOCMan to see where it fails.

1 - Can you confirm DocMan is set to "Use Joomla router" under the "By component" tab of sh404SEF configuration?

2 - After doing the backup, please provide superadmin access to the site backend, so that I can download the backup and proceed with the debugging.

Rgds

Hi

1 - Please upload the jpa file to this online private folder.

2 - Yes, "Use Joomla router" bypass sh404SEF for a component, so pretty much nothing can be done with URls.

3 - As this is a greek site, I will have difficulties navigating it. Can you provide one or two full links to pages with this issue?

Rgds

Hi

JPA file was just uploaded (using e-mail xxxx).

Thanks, I will work on that today. You actually do not need to be logged in, the access is free when one knows the link.

As for the navigation, within the demo site I created I left the english-speaking default data and I created a docman main menu item. By entering docman from the front-end, I created two documents - one with long and one with short alias.

Thanks, I'll get back to you.

Rgds

Hi

Well, I installed the test site locally, and my findings are a bit weird. The main point is that I can indeed recreate 404s, but I can do that also when sh404SEF is disabled!

Without sh404SEF:

[xxxx]

and

[xxxx]

With sh404SEF - Use joomla router:

Exact same result: 404 on first link, success with second link

I found the issue: it is linked to the size of the database field for the URLs, which is exceeded with those long URLs (because they are stored URL encoded. I will produce a version with longer fields for you to test (I made the changes manually to the DB and I get the same results as with the other two options).

Rgds

Hi

Well, I was a bit confused when I recapped my tests above I think. I did all the changes to the database, etc so that sh404SEF now works with longer URLs (2048). This was a long needed change anyway. So it's done, I installed on the test site and I basically I get the same results as above. That is:

1 - link to PDF with a latin char alias work fine in all cases
2 - Link with (long) greek characters alias never works. By that I mean that they do not work without sh404SEF at all. I'm not sure why I did not stress that out before, as I said before, I got confused when testing the 2 documents links that you put in the test site.

Can you please confirm: simply set "Enable URL optimization" to No, which entirely disable sh404SEF, and test the 2 links: the second one works, while the first one fails.

Looks like a DOCMAn bug to me.

Rgds

Hi

Please mind that I have changed the component sh404SEF configuration to Joomla Router!.

Yes, the problem only occurs when using default handler.

[EDIT] And yes, on my local machine, only one link works, the one with latin characters. With sh404SEF or without.

Rgds

Hi again,

I have uploaded to the online private folder the fixed sh404SEF version. Please install and test again, with and without sh404SEF enabled.

Please note that you should probably delete the faulty URLs in the URL manager, so that new ones (full length) can be created again.

Rgds

Hi

That's not something sh404SEF would have anything to do with. We do not set page title for DOCman - unless you have manually set a page title for a specific DOCMan URL, from the URL manager?

E.g.: the URL "/pdfs/40-rfs-isafe-prospect/file" shows "file" on the browser window, whereas "/pdfs/41-rfs-isafe-διαδικασια-εγκαταστασησ/file" shows the Docman document title.

I suspect this is standard DOCMan behavior?

Rgds

Hi

Don't know what to say.. I can only verify that there are no URL customization from within the sh404SEF component..

Again, I can only comment that the page title (for DOCMan) is not something sh404SEF is setting. The page title is set by DOCMan.

That said, I'm not sure what you mean by that exactly. The xxxxx/file URL is that of the PDF file, it's not an HTML document. So it does not have a title. If the PDF is displayed in your browser (and not downloaded) then the title displayed depends on the browser you use and several other things. Again, it's definitely something sh404SEF can have an impact on: we can only modify HTML pages on your website, not PDF files.

Rgds

Hi

The titles I mention do show within the same browser (e.g. latest Mozilla Firefox) when the PDF file is opened within the browser window. In some of the browser windows the title appears in others not.

That's just because you use your browser to view PDF. But this has nothing to do with an HTML page "page title" and is not something sh404SEF or Joomla will have an effect on.

I think it's just the file name actually, so yes, maybe DOCMan can do something about it.

Closing this ticket now, feel free to open a new one as needed. If you do so, please mention this ticket number in the new one.

If you created any superadmin account for us, be sure to delete or block it now to avoid unnecessary risk in the future.

Be sure to also check out wbAMP, our new Accelerated Mobiles pages plugin for Joomla - the next big thing is SEO, direct from Google themselves!



Rgds