38  Missing Unicode characters

Some characters might be missing in the template’s fonts. In HTML output the browser may find a fallback solution. In PDF output you get a LaTeX warning message such as:

[WARNING] Missing character: There is no ȧ (U+0227) (U+0227) in font STIXTwoText:mode=node;script=

There are three ways to solve this.

  1. The character is obtained by combining a letter and one or more diacritics (accents and the like). Then it can be declared in the template settings.
  2. The character cannot be recreated by combination but is available in the math font. Then it can be declared in the template settings.
  3. The character cannot be recreated by combination and is not available in the math font either. Then the template should be modified to use a different font for a Unicode range including this character.

Methods 1 and 2 are easy and should be tried first.

38.1 Adding a character obtained by combination

Requires: Pandokoma template.

Consider the character ‘à’, for instance. A LaTeX warning message would give its Unicode number in hexadecimal format: 00E0. (Hexadecimal means that A,B, … F are used in addition to digits 0…9. Unicode numbers are typically written U+xxx, i.e. U+00E0.) Search for “Unicode 00E0” online, and go to one of the websites that have a detailed page on each Unicode character. Look for a “decomposition mapping” such as:

Decomposition mapping:    a (U+0061)    ` (U+0300)

This shows that the Unicode character can be created by combining the a character with the character number 0300. 0300 is a special modifier or combining character, who combines with the preceding character. It should not be confused with the self-standing character ` (code 0060) that does not combine with the previous character.

In the template folder, open settings/common-latex.yaml. Search for the newunicodechar section:

# Unicode chars
newunicodechar:
- char: ǐ
  command: i\symbol{"030C}
- char: ȧ
  command: a\symbol{"0307}
- ...

Add an entry for your character. The char value should be the missing character itself (copy/paste it). The command value is the combination used to recreate it:

newUnicodechar:
- char: à
  command: a\symbol{"0300}
- char: ǐ
  command: i\symbol{"030C}
- ...

The combination should be entered in the same order as you found on the reference webpage: here we have a first then ```. (In Unicode combinations, the added diacritics come after the letter.)

The combining characters can be entered either directly (here, the a) or with \symbol{"...} command, where ‘…’ is the hexadecimal Unicode number (here, the modifier ` is entered with \symbol{"0300}.) So the above is equivalent to:

- char: à
  command: \symbol{"0061}\symbol{"0300}

because the Unicode number of a is 0061. You would normally enter modifier Unicode characters by number because they do not display well in text editors.

38.2 Adding a character from a math font

Requires: Pandokoma template.

For some characters, they are not available in the text font but they are available as math symbols. You can look them up in the Comprehensive LaTeX Symbol List (if you have a LaTeX installation, entering the command texdoc symbols in a terminal should display it; otherwise use the link).

Adding them to the template is similar to the combination method above. In the template, open settings/common-latex.yaml. Look for the unicodefrommath section. Here’s an example:

# Unicode symbols from math
unicodefrommath:
- char: '∴'
  is: \therefore

In the char field, I’ve entered the character. Instead of copying/ pasting it, I’ve used its HTML entity name, between quotation marks. (See section Chapter 13 on HTML entities. The quotation marks ensure that HTML entities are treated as such.) In the is field, I enter the LaTeX command that prints the desired characters.

The is field should be a LaTeX command that works in math mode. If it’s a text mode command, use \text{...}.

In HTML output, the LaTeX command is passed to MathJAX. Check your HTML output to see whether MathJAX displays it. If not, you’ll see red text with a yellow background where the character should be. Your only option then, the last resort, is to declare an alternative font for that character.

38.3 Declaring an alternative font for a certain Unicode range

This method requires changing both the LaTeX and HTML (CSS) templates and adding the relevant fonts to the template folder and website. You need to ask the template managed to do this.