I have a web page with a link (anchor tag) to a text file that is in UTF-8 format. Linking to it from Safari on my Mac (Catalina) it looks fine, but linking to it from Safari on my iPhone does not display the non-ASCII characters correctly. In my HTML anchor (A) tag, how do I specify that the target file is UTF-8 so that Safari on my iPhone will display it correctly? ―Ken Nellis
iPhone SE, iOS 14
Bingo! Hadn't thought of that, but Googling Apache settings I got my answer: adding to .htaccess the following:
AddDefaultCharset UTF-8
Still, don't understand why HTML5 deprecated the charset attribute on the anchor (A) tag. My issue would seem to be an appropriate use of it. Thanx for the suggestion, Tom!
Ken Nellis wrote:
Could the respective sites themselves be providing charset hints to Safari?
I have not run into it for a long time, but the default of some web servers used to be to tell browsers to use Latin 1 encoding (regardless of what the page code may have stipulated). The fix was to modify some server setting. You might have a look at that. Do you see the same problem if you put a self-composed text with a bunch of utf8 accented characters in it on your server?
Apparently, why it works on my Mac is because I have Safari > Preferences > Advanced > Default encoding: Unicode (UTF-8). When I change it to Western (ISO Latin 1) it renders the file as does my iPhone. On my iPhone I looked in Settings > Safari, but didn't see a corresponding setting. Regardless, it seems my HTML anchor tag should be able to specify the file's encoding. It also seems that the iPhone's Safari should have that setting.
―Ken
Try making your text file UTF-16 instead of UTF-8.
Ken,
The HTML anchor allows you to specify a type, and for UTF-8 content, it would be the non-intuitive:
<a href="file:///Users/yourname/Desktop/exif.txt type="text/strings">exif.txt</a>
I have a text file on my Desktop that is a dump of EXIF data from an image. It makes no difference in its presentation on the web page whether I omit the anchor type clause above, and just let the <head> charset="UTF-8" control the display of the page.
Thanx, but that didn't work. I created a test page that shows the problem.
https://nellisks.com/tools/charsettest.html
From my Mac both links display correctly, but from my iPhone, the link to "My Copy" doesn't render correctly, even after I added type="text/strings". Did I do it wrong? Do you see the same results as I? ―Ken
Maybe I am going blind, even with corrective lenses, but when I review both of those links in Safari on a 12.9 inch iPad Pro running iPadOS 14.4.2, I see no difference. I am holding out for an iPhone 13, so don't have one running anything other than iOS 9.3.6.
Here are the IANA media types for that anchor tag type keyword.
Looking at the 3rd line, if I set Mac/Safari to Default encoding: Western (ISO Latin 1) then I see the following for "My File" on both my Mac and iPhone:
# © 2020 Unicode®, Inc.
But for "Unicode’s Copy" I see:
# © 2020 Unicode®, Inc.
So, if you're not blind, you're seeing different renderings than I am, so that's a whole new issue!
―Ken
Thanx, Tom, but the issue is how to coerce modifiable software correctly read an immutable file, in this case one that must remain an exact duplicate of the one on the Unicode site. The mystery remains: why does Unicode's copy of the file render correctly in Safari, but my exact duplicate renders incorrectly. Could the respective sites themselves be providing charset hints to Safari? (My site uses Apache. I have no idea what the Unicode site uses.)
It's good that Tom provided the synapse leading you to the .htaccess file entry.
In an HTML link to a text file, how to specify the charset of the file?