Check 049: Limit to official whitespace characters?

Observed behaviour

Check 049 treats these Unicode values as whitespace:

  WHITESPACE_CHARACTERS = [
      0x0009, 0x000A, 0x000B, 0x000C, 0x000D, 0x0020, 0x0085, 0x00A0, 0x1680,
      0x2000, 0x2001, 0x2002, 0x2003, 0x2004, 0x2005, 0x2006, 0x2007, 0x2008,
      0x2009, 0x200A, 0x2028, 0x2029, 0x202F, 0x205F, 0x3000, 0x180E, 0x200B,
      0x2060, 0xFEFF
  ]

Expected behaviour

The Unicode 11.0 property list attached a whitespace property to these Unicode values:

0009..000D    ; White_Space # Cc   [5] <control-0009>..<control-000D>
0020          ; White_Space # Zs       SPACE
0085          ; White_Space # Cc       <control-0085>
00A0          ; White_Space # Zs       NO-BREAK SPACE
1680          ; White_Space # Zs       OGHAM SPACE MARK
2000..200A    ; White_Space # Zs  [11] EN QUAD..HAIR SPACE
2028          ; White_Space # Zl       LINE SEPARATOR
2029          ; White_Space # Zp       PARAGRAPH SEPARATOR
202F          ; White_Space # Zs       NARROW NO-BREAK SPACE
205F          ; White_Space # Zs       MEDIUM MATHEMATICAL SPACE
3000          ; White_Space # Zs       IDEOGRAPHIC SPACE

The difference isn’t large, but maybe we should only check for official whitespace characters? Can’t seem to access that information from Python’s unicodedata module though 😕

[anthrotype]

maybe we can add to our fontTools.unicodedata module. we already can parse these kind of files https://github.com/fonttools/fonttools/blob/b38e2bd8acbebe98980f63a5bb490010e0c22134/MetaTools/buildUCD.py#L64

we use that to build the Scripts.txt, ScriptExtensions.txt, and Blocks.txt, which have the same format as PropList.txt.

PR? 😉

[drj11]

This check is now com.google.fonts/check/whitespace_ink

The list includes whitespace that should have drawings, and also incudes non-whitespace that should not have drawings.

Some observations: The “extras” are U+180E MONGOLIAN VOWEL SEPARATOR U+200B ZERO-WIDTH SPACE U+2060 WORD JOINER U+FEFF ZERO WIDTH NO-BREAK SPACE (but Byte-Order-Mark in actual use)

These are not whitespace (in the properties sense) but should not have drawings (I’m making an assumption about MONGOLIAN VOWEL SEPARATOR because I don’t know about it). So they are appropriate for this test. It’s entirely possible there are other char codes like this, that are not whitespace, but have no drawing.

OGHAM SPACE MARK is the other way around. It is a whitespace character that is supposed to have a drawing (at least as far as I understand it, I don’t know much about the script).

This test is really about glyphs with no drawings, which may or not be whitespace.

In conclusion:

• Some non-whitespace glyphs should not have drawings (and I think we could test for that, as we do now);
• Some (one!) whitespace glyphs should have drawings. So we should remove OGHAM SPACE MARK from this list.