Add missing characters from dummy-index's adaptation, and adapt the z weight

xkcd-script/generator/pt5_svg_to_font.py

@@ -332,6 +332,14 @@ def charname(char):
     ('x',): psMat.translate(0, 20),
 }
 
+# Per-character weight nudge applied after the per-line scale correction.
+# Positive = thicker, negative = thinner.  Mirrors _GREEK_WEIGHT_NUDGE for the
+# main alphabet — used when an individual glyph's hand-drawn stroke is
+# noticeably heavier or lighter than its neighbours.
+_per_char_weight_nudge = {
+    ('z',): -15,  # z's diagonal strokes are heavier than the rest of the alphabet
+}
+
 # Pick out particular glyphs that are more pleasant than their latter alternatives.
 special_choices = {('C', ): dict(line=4),
                    ('G',): dict(line=4),
@@ -387,6 +395,15 @@ def charname(char):
                 c.transform(psMat.scale(_restore))
                 c.width = int(round(c.width * _restore))
 
+    # Per-character weight nudge: applied to source glyphs whose stroke weight
+    # is noticeably off from the rest of the alphabet.
+    _weight_nudge = _per_char_weight_nudge.get(chars)
+    if _weight_nudge:
+        c.correctDirection()
+        c.removeOverlap()
+        c.changeWeight(_weight_nudge)
+        c.simplify()
+
     # Per-character size adjustments: scale about the baseline (origin) to reduce
     # overall size while preserving stroke weight gained from changeWeight above.
     _operation_matrix = _per_char_operation.get(chars)

xkcd-script/generator/pt6_derived_chars.py

@@ -195,6 +195,26 @@ def _make_ogonek(font, cp, base_name):
     c.addReference('comma', psMat.compose(psMat.scale(-1, -1), psMat.translate(dx, dy)))
 
 
+def _make_spacing_modifier(font, cp, mark_name, y_bottom=None, lsb=40, rsb=40):
+    """Spacing modifier letter: reference to mark glyph, repositioned and given advance width.
+
+    Translates the mark so its bottom sits at y_bottom and its left ink edge is
+    at lsb.  Default y_bottom is 20 units above the x-height of 'a', placing
+    the modifier at a natural reading height as if hovering over a lowercase letter.
+    lsb/rsb: left/right side bearings of the resulting glyph.
+    """
+    if y_bottom is None:
+        y_bottom = int(round(font['a'].boundingBox()[3] + 20))
+    g = font.createMappedChar(cp)
+    g.clear()
+    mark_bb = font[mark_name].boundingBox()
+    dx = lsb - mark_bb[0]
+    dy = y_bottom - mark_bb[1]
+    g.addReference(mark_name, psMat.translate(dx, dy))
+    g.width = int(round(mark_bb[2] + dx + rsb))
+    return g
+
+
 # Codepoints that already exist as hand-drawn glyphs; skip to avoid overwriting.
 _SKIP_CPS = frozenset({
     0x00DC,  # Ü — hand-drawn
@@ -276,18 +296,44 @@ def _accented(cp, base_name, mark_name, gap=20, x_adj=0):
     (0x2033, 'quotedbl'),     # doubleprime
     (0x2035, 'quoteleft'),    # backprime
     # Dash aliases
+    (0x2010, 'hyphen'),       # Unicode hyphen
     (0x2011, 'hyphen'),       # non-breaking hyphen
     (0x00AD, 'hyphen'),       # soft hyphen
     (0x2212, 'endash'),       # minus sign
     (0x2012, 'endash'),       # figure dash
     (0x2015, 'emdash'),       # horizontal bar
+    # Modifier letter apostrophes (Polynesian, transliteration)
+    (0x02BB, 'quoteleft'),    # modifier letter turned comma ʻ
+    (0x02BC, 'quoteright'),   # modifier letter apostrophe ʼ
 ]
 for codepoint, source_name in ref_aliases:
     c = font.createMappedChar(codepoint)
     c.addReference(source_name)
     c.width = font[source_name].width
 
 
+# ---------------------------------------------------------------------------
+# Inverted punctuation
+# ---------------------------------------------------------------------------
+
+# U+00A1 ¡ INVERTED EXCLAMATION MARK — ! flipped vertically.
+# U+00BF ¿ INVERTED QUESTION MARK — ? flipped vertically.
+# scale(1,-1) reflects about y=0; the translate re-maps [ymin,ymax] back to the
+# same vertical range so the glyph stays above the baseline.
+for _cp, _gname in [(0x00A1, 'exclam'), (0x00BF, 'question')]:
+    _src = font[_gname]
+    _src_bb = _src.boundingBox()
+    _g = font.createMappedChar(_cp)
+    _g.clear()
+    _layer = fontforge.layer()
+    for _c in _src.foreground:
+        _layer += _c
+    _g.foreground = _layer
+    _g.transform(psMat.compose(psMat.scale(1, -1), psMat.translate(0, _src_bb[1] + _src_bb[3])))
+    _g.correctDirection()
+    _g.width = _src.width
+
+
 # ---------------------------------------------------------------------------
 # Combining mark glyphs
 # ---------------------------------------------------------------------------
@@ -467,6 +513,24 @@ def _accented(cp, base_name, mark_name, gap=20, x_adj=0):
 _c0327.width = 0
 
 
+# ---------------------------------------------------------------------------
+# Spacing modifier letters (U+00A8, U+00AF, U+02C6–U+02DD)
+# Each is a reference to the corresponding combining mark, repositioned so its
+# bottom sits just above the x-height and given an advance width.
+# ---------------------------------------------------------------------------
+
+_sm_y = int(round(font['a'].boundingBox()[3] + 20))
+
+_make_spacing_modifier(font, 0x00A8, '_diaeresis_mark', _sm_y)   # ¨ spacing diaeresis
+_make_spacing_modifier(font, 0x00AF, '_macron_mark', _sm_y)      # ¯ spacing macron
+_make_spacing_modifier(font, 0x02C6, '_circumflex_mark', _sm_y)  # ˆ modifier letter circumflex accent
+_make_spacing_modifier(font, 0x02C7, '_caron_mark', _sm_y)       # ˇ modifier letter caron
+_make_spacing_modifier(font, 0x02D8, '_breve_mark', _sm_y)       # ˘ modifier letter breve
+_make_spacing_modifier(font, 0x02D9, '_dot_above_mark', _sm_y)   # ˙ modifier letter dot above
+_make_spacing_modifier(font, 0x02DC, '_tilde_mark', _sm_y)       # ˜ modifier letter small tilde
+_make_spacing_modifier(font, 0x02DD, '_double_acute_mark', _sm_y)  # ˝ modifier letter double acute accent
+
+
 # ---------------------------------------------------------------------------
 # Accented character tables
 # ---------------------------------------------------------------------------
@@ -1147,6 +1211,93 @@ def _greek_lc_to_uc(font, lc_cp, uc_cp, snap=True, weight_delta=0):
 _g.width = font[0x25BD].width
 
 
+# ---------------------------------------------------------------------------
+# Additional composites
+# ---------------------------------------------------------------------------
+
+# U with diaeresis + tone mark: Ǖ Ǘ Ǚ Ǜ / ǖ ǘ ǚ ǜ  (Pinyin)
+# Ü / ü are used as bases so the new mark floats above the existing dots.
+_Udi = font[0x00DC].glyphname  # Ü hand-drawn
+_udi = font[0x00FC].glyphname  # ü derived
+_accented(0x01D5, _Udi, '_macron_mark')        # Ǖ
+_accented(0x01D6, _udi, '_macron_mark', gap=8)  # ǖ
+_accented(0x01D7, _Udi, '_acute_mark')         # Ǘ
+_accented(0x01D8, _udi, '_acute_mark', gap=8)   # ǘ
+_accented(0x01D9, _Udi, '_caron_mark')         # Ǚ
+_accented(0x01DA, _udi, '_caron_mark', gap=8)   # ǚ
+_accented(0x01DB, _Udi, '_grave_mark')         # Ǜ
+_accented(0x01DC, _udi, '_grave_mark', gap=8)   # ǜ
+
+# A/a with diaeresis + macron: Ǟ ǟ  (Livonian, Skolt Sámi)
+# gap=30/20: the base's bb top already includes the diaeresis dots; extra
+# clearance prevents the macron from pressing against the dots.
+_Adi = font[0x00C4].glyphname  # Ä
+_adi = font[0x00E4].glyphname  # ä
+_accented(0x01DE, _Adi, '_macron_mark', gap=30)  # Ǟ
+_accented(0x01DF, _adi, '_macron_mark', gap=20)  # ǟ
+
+# AE/ae + macron: Ǣ ǣ  (Old Norse transliteration)
+# x_adj shifts the macron rightward: Æ/æ's geometric centre sits left of the
+# visual centre (E side is heavier than A side).
+_Ae_name = font[0x00C6].glyphname  # Æ
+_ae_name = font[0x00E6].glyphname  # æ
+_accented(0x01E2, _Ae_name, '_macron_mark', x_adj=100)        # Ǣ
+_accented(0x01E3, _ae_name, '_macron_mark', gap=25, x_adj=50)  # ǣ
+
+# G/g + caron: Ǧ ǧ  (Skolt Sámi, some transliteration systems)
+# x_adj for ǧ: g's full bounding box includes a wide descender that pulls
+# its geometric centre left of the bowl centre.
+_accented(0x01E6, 'G', '_caron_mark')               # Ǧ
+_accented(0x01E7, 'g', '_caron_mark', gap=8, x_adj=80)  # ǧ
+
+# K/k + caron: Ǩ ǩ  (Skolt Sámi)
+_accented(0x01E8, 'K', '_caron_mark')         # Ǩ
+_accented(0x01E9, 'k', '_caron_mark', gap=8)   # ǩ
+
+# G/g + acute: Ǵ ǵ  (Upper Sorbian)
+# x_adj for ǵ: same descender-bowl offset as ǧ.
+_accented(0x01F4, 'G', '_acute_mark')                    # Ǵ
+_accented(0x01F5, 'g', '_acute_mark', gap=20, x_adj=80)  # ǵ
+
+# A-ring + acute: Ǻ ǻ  (Northern Sámi)
+# The acute floats above the existing ring; base bb already includes the ring.
+_Aring_name = font[0x00C5].glyphname  # Å
+_aring_name = font[0x00E5].glyphname  # å
+_accented(0x01FA, _Aring_name, '_acute_mark')        # Ǻ
+_accented(0x01FB, _aring_name, '_acute_mark', gap=8)  # ǻ
+
+# AE/ae + acute: Ǽ ǽ  (Faroese, Danish orthography)
+_accented(0x01FC, _Ae_name, '_acute_mark')        # Ǽ
+_accented(0x01FD, _ae_name, '_acute_mark', gap=8)  # ǽ
+
+# O-stroke + acute: Ǿ ǿ  (Faroese, Danish orthography)
+# Ø/ø are derived earlier in this script; their bounding boxes include the stroke.
+_Ost_name = font[0x00D8].glyphname  # Ø
+_ost_name = font[0x00F8].glyphname  # ø
+_accented(0x01FE, _Ost_name, '_acute_mark')        # Ǿ
+_accented(0x01FF, _ost_name, '_acute_mark', gap=8)  # ǿ
+
+# DZ/dz digraph ligatures: DZ Dz dz and DŽ/Dž/dž: DŽ Dž dž
+# Letters are placed ink-edge to ink-edge with a small gap, like IJ above.
+_Zcaron_name = font[0x017D].glyphname  # Ž
+_zcaron_name = font[0x017E].glyphname  # ž
+_dz_gap = 20
+for _cp, _left, _right in [
+    (0x01F1, 'D', 'Z'),             # DZ capital DZ
+    (0x01F2, 'D', 'z'),             # Dz titlecase Dz
+    (0x01F3, 'd', 'z'),             # dz lowercase dz
+    (0x01C4, 'D', _Zcaron_name),    # DŽ capital DŽ
+    (0x01C5, 'D', _zcaron_name),    # Dž titlecase Dž
+    (0x01C6, 'd', _zcaron_name),    # dž lowercase dž
+]:
+    _g = font.createMappedChar(_cp)
+    _g.clear()
+    _g.addReference(_left)
+    _dx = int(round(font[_left].boundingBox()[2] + _dz_gap - font[_right].boundingBox()[0]))
+    _g.addReference(_right, psMat.translate(_dx, 0))
+    _g.width = _dx + font[_right].width
+
+
 # ---------------------------------------------------------------------------
 # Save
 # ---------------------------------------------------------------------------

xkcd-script/samples/gen_charmap.py

@@ -43,6 +43,7 @@
     (0x00A0, 0x0100, "Latin-1 Supplement"),
     (0x0100, 0x0180, "Latin Extended-A"),
     (0x0180, 0x0250, "Latin Extended-B"),
+    (0x02B0, 0x0300, "Spacing Modifier Letters"),
     (0x0300, 0x0370, "Combining Diacritical Marks"),
     (0x0370, 0x0400, "Greek and Coptic"),
     (0x1E00, 0x1F00, "Latin Extended Additional"),