Not a bug Problem with Greek board names

[Salamanca]

Affected version

2.2

Hi,

There's a problem as it seems, with the conversion of Greek board names with the "Romanize titles in URLs" function

example:
greek word μαθηματικά will be converted to "mathimatik" in the URL.

The problem is with letters with accents. If it was "μαθηματικα" it would be converted fine, but it's totally incorrect not to have accents in Greek language.

Thanks
A.

 

[Salamanca]

this also applies to thread names, and everything actually.

 

[Ozzy47]

ACP search: romanize

Uncheck the option.

 

[Salamanca]

ACP search: romanize

Uncheck the option.

Hi, unfortunately, this option is not good either. the reason is that greek characters on the URL is bad for SEO.

 

[Mike]

Generally speaking, we consider Romanization a "best reasonable effort" option, but we can't guarantee that every character from every language may be converted. (The process is essentially a massive map of all the characters that need to be converted. It's a library that we don't commonly apply changes to.)

The context is slightly different, but you can see a discussion of the same issue here:

Not a bug - Canonical url of tags is not correct (missing characters)

Hello all! :) I use xF2.1 in a Greek language board. I have the following issue: words with Greek characters which have "tono" (τόνος in Greek is a small line above one vowel of each word - all greek words have one τόνο). Examples (word - canonical url) : κάπνισμα - /index.php?tags/kpnisma/...

xenforo.com

If this is significant to you, then disabling the Romanization option would be necessary (or making custom changes for your specific needs). It's worth noting that essentially all browsers correctly display non-ASCII characters in the URL so I don't see why Google wouldn't recognize the characters from a search perspective (though the characters would also be in the title and h1 tags on the page as well).

 

[Salamanca]

Generally speaking, we consider Romanization a "best reasonable effort" option, but we can't guarantee that every character from every language may be converted. (The process is essentially a massive map of all the characters that need to be converted. It's a library that we don't commonly apply changes to.)

The context is slightly different, but you can see a discussion of the same issue here:

Not a bug - Canonical url of tags is not correct (missing characters)

Hello all! :) I use xF2.1 in a Greek language board. I have the following issue: words with Greek characters which have "tono" (τόνος in Greek is a small line above one vowel of each word - all greek words have one τόνο). Examples (word - canonical url) : κάπνισμα - /index.php?tags/kpnisma/...

xenforo.com

If this is significant to you, then disabling the Romanization option would be necessary (or making custom changes for your specific needs). It's worth noting that essentially all browsers correctly display non-ASCII characters in the URL so I don't see why Google wouldn't recognize the characters from a search perspective (though the characters would also be in the title and h1 tags on the page as well).

Hi Mike,

Thanks for the information. Unfortunately, using non-ascii characters in the URL is a problem for several reasons, it's a not an option of us.
However, if this just an issue with mapping, I can help you with that. It seems that the mapping misses the accented characters like ά ό ή and so on. This can be easily fixed if we just add those characters.
If you want to include this on the next update, then great! If for some reason you dont want to do this, at least let me know where the mapping is so I may fix this for our installation.

Thanks.

 

[Salamanca]

Generally speaking, we consider Romanization a "best reasonable effort" option, but we can't guarantee that every character from every language may be converted. (The process is essentially a massive map of all the characters that need to be converted. It's a library that we don't commonly apply changes to.)

The context is slightly different, but you can see a discussion of the same issue here:

Not a bug - Canonical url of tags is not correct (missing characters)

Hello all! :) I use xF2.1 in a Greek language board. I have the following issue: words with Greek characters which have "tono" (τόνος in Greek is a small line above one vowel of each word - all greek words have one τόνο). Examples (word - canonical url) : κάπνισμα - /index.php?tags/kpnisma/...

xenforo.com

If this is significant to you, then disabling the Romanization option would be necessary (or making custom changes for your specific needs). It's worth noting that essentially all browsers correctly display non-ASCII characters in the URL so I don't see why Google wouldn't recognize the characters from a search perspective (though the characters would also be in the title and h1 tags on the page as well).

Hello,

I would really appreciate it if you can give me directions to find those mappings so that I can update them and include the greek accented characters.

Thanks
Alex

 

[Salamanca]

Hello,

I've updated the file, and it works now. The file is at:

\src\vendor\dokuwiki\utf8\utf8.php

The missing ones are:

// Missing accented Greek
  'ά'=>'a','Ά'=>'A','έ'=>'e','Έ'=>'E','ί'=>'I','Ί'=>'I','ή'=>'i','Ή'=>'I','ύ'=>'y','Ύ'=>'Y','ώ'=>'o','Ώ'=>'O','ό'=>'o','Ό'=>'O',