Metatag and html tags encoding

I think this is because strip_tags() is ran on the values, and then they're ran through check_plain() again before they're finally output.

Does this resolve the problem?

I am still encountering this issue even with the patch in #5 applied. Example output:

<meta name="description" content="This rich yet simple chocolate cake gets a little kick from a&nbsp;touch of smoked Spanish paprika and espelette. Espelette is a smoky&nbsp;chile powder from France.">

@markdorison: If you see " " in a field it comes from the source field's data. That said, because it's a known problem we can fix it.

@DamienMcKenna: I applied the patch in #10 and I was still seeing the issue. When I var_dump() $value before your changes, I am not seeing &nbsp but instead the a-circumflex character:

"This rich yet simple chocolate cake gets a little kick from a touch of smoked Spanish paprika and espelette. Espelette is a smoky chile powder from France."

Some searching (https://www.google.com/search?q=nbsp+a-circumflex&oq=nbsp+a-circumflex) describes this as a character encoding issue. It seems to be converting successfully back to &nbsp by the time it makes it to the browser, but at the point you are trying to replace &nbsp, there isn't one there to replace.

There's a very fine line between correcting encoding errors and screwing up a non-English site. That's one reason that delving further into this makes me nervous. Out of interest, for the node with the a-circumflex problem, do you also see that character in the node display itself?

@DamienMcKenna: By the time the text makes it into markup it seems to have reverted back to the expected &nbsp. The only place I have seen the a-circumflex is when I var_dump() the value in tidyValue().

How does this work?

@DamienMcKenna #14 works for me.

@markdorison: Thanks for letting me know.

I've committed this.

@DamnienMcKenna: Absolutely; thank you for the fix!

I'm confused. Why wouldn't decode_entities() call itself convert ' ' to a space? That's what it's designed to do.

$ php -r 'print ord(html_entity_decode(" ")) . "\n";'
160
$ php -r 'print ord(html_entity_decode(" ")) . "\n";'
32

This difference is why.