add unicode-range test

[patrickkettner]

fixes #1058

This test embeds a base64'd truetype font with two characters, the <space> and m specifically, specifying the unicode-range to only use the <space> (U+002E) . It then creates four elements, two of each, one with the font and without, adds them to the DOM, and check their width. If the space changes width, and the m doesn't, then we have a winner.

[infojunkie]

Sounds great! I built your fork of Modernizr but when I run the tests through the web browser it only runs 97 tests, not including the new unicode-range one. How can I try it?

[patrickkettner]

Sorry about that, I left out a key bit.
Try it now

[infojunkie]

Thanks. I can see the unicoderange property but it returns false even under
Chrome which does support unicode-range.

I've copied your test to a JSFiddle to see it visually, but I can't tell
the difference between the spans: http://jsfiddle.net/infojunkie/VufQU/1/ -
maybe I misunderstand the test!

On Sat, Oct 19, 2013 at 12:27 AM, patrick kettner
notifications@github.comwrote:

Sorry about that, I left out a key bit.
Try it now

—
Reply to this email directly or view it on GitHubhttps://github.com//pull/1062#issuecomment-26645226
.

[infojunkie]

I've looked at http://css3test.com/ to see how they detect unicode-range support. Seems to be a very different test, based on creating a dummy element, assigning a test style and checking whether its properties match the given test value. Since I'm a n00b at browser testing, I am just reporting that this test seems to work :-)

[patrickkettner]

Man, im really sorry. This is embarrassing.
I had changed the test to the the . character, not the space.

updated test

That being said, after thinking about it at lunch, I was able to come up with a much simpler test that actually works.

I updated this pull accordingly.

RE: css3test, as it says up top

Caution: This test checks which CSS3 features the browser recognizes,
not whether they are implemented correctly.

I didn't actually dig into what they are doing, but I did see it fails in all IE, so its not 100%.

This test does seem to cover all browsers I can find accurately.

[infojunkie]

Thanks so much for updating the test! Works perfectly now.

On Sat, Oct 19, 2013 at 5:36 PM, patrick kettner
notifications@github.comwrote:

Man, im really sorry. This is embarrassing.
I had changed the test to the the . character, not the space.

updated test http://jsfiddle.net/VufQU/2/

That being said, after thinking about it at lunch, I was able to come up
wit ha much simpler test that actually works.

I updated this pull accordingly.

RE: css3test, as it says up top

Caution: This test checks which CSS3 features the browser recognizes,
not whether they are implemented correctly.

I didn't actually dig into what they are doing, but I did see it fails in
all IE, so its not 100%.

This test does seem to cover all browsers I can find accurately.

—
Reply to this email directly or view it on GitHubhttps://github.com//pull/1062#issuecomment-26662180
.

[patrickkettner]

@ryanseddon @stucox - how you livin?

[stucox]

LGTM.

I guess the only danger would be if Arial isn’t supported and it defaults to a monospace font (or at least one which has a . and m which are both the same width as Courier or whatever). Unlikely.