Skip to content
Redfin Solutions logoRedfin Solutions logoContact
character set

Handling High Unicode Character Sets in Web Fonts

When odd characters are displayed instead of the intended text, it’s often due to the font not supporting “high unicode” characters like the marcon character over the letter “a”: ā. Checking the google font @import url in a SASS file showed that the latin extended character set was not included. To do this:

  1. Find your font at google fonts using the search bar in the upper right and click the font name
  2. Click the “SELECT THIS FONT” link in the upper right
  3. Click the “Family Selected” black bar that appears at the bottom right of the page to expand the url builder window
  4. Click customize
  5. Scroll to the bottom and select “Latin Extended”
  6. Go back to the “Embed” tab and you will see that the urls in the STANDARD and @IMPORT tabs have been updated with “subset=latin-ext”
  7. Paste the embed code of your choice into your style sheets or sass files and voila! Your special characters should render.

In a recent case they did not render because the google font being used, “Lato,” was actually out of date despite there being a latin extended option (your font will not render if the font author has not created a glyph for that character). The updated Lato font files were on the font author’s website and downloaded to a font folder within the Drupal theme.

The author had conveniently included a CSS file with an @font-face block linking to each typeface file in the download. Enter the below link in your browser to see the data returned:

https://fonts.googleapis.com/css?family=Lato:300,300i,400,400i&subset=latin-ext

Each @font-face block can contain a number of attributes such as font-weight, font-style, font-family, and src. The author had chosen to name each typeface differently and set font-weight to normal for each despite having light, normal, and bold etc. character sets in the download.

The Google font API implements a different approach, however. In the above example all @font-face blocks return with a font-family identifier of “Lato” as opposed to a unique identifier per block like “Lato-Light” or “LatoRegular” as the font author had done. Google also explicitly sets a numerical value in the font-weight attribute as opposed to a consistent font-weight value of “normal” for all.

You can use either approach, but if you set up your @font-face rules like Google does with consistent font-family names and numerical font-weights just make sure you specify the font-weight declaration in addition to the font-family name when building the declarations that apply to your markup so that the correct set of glyphs can be found and applied.