-#!/usr/bin/env python
-# flake8: noqa
+#!/usr/bin/env python3
+# flake8: noqa: F821
+
+"""usage: ./gen-use-table.py IndicSyllabicCategory.txt IndicPositionalCategory.txt ArabicShaping.txt DerivedCoreProperties.txt UnicodeData.txt Blocks.txt Scripts.txt IndicSyllabicCategory-Additional.txt IndicPositionalCategory-Additional.txt
+
+Input files:
+* https://unicode.org/Public/UCD/latest/ucd/IndicSyllabicCategory.txt
+* https://unicode.org/Public/UCD/latest/ucd/IndicPositionalCategory.txt
+* https://unicode.org/Public/UCD/latest/ucd/ArabicShaping.txt
+* https://unicode.org/Public/UCD/latest/ucd/DerivedCoreProperties.txt
+* https://unicode.org/Public/UCD/latest/ucd/UnicodeData.txt
+* https://unicode.org/Public/UCD/latest/ucd/Blocks.txt
+* https://unicode.org/Public/UCD/latest/ucd/Scripts.txt
+* ms-use/IndicSyllabicCategory-Additional.txt
+* ms-use/IndicPositionalCategory-Additional.txt
+"""
-from __future__ import print_function, division, absolute_import
-
-import io
import sys
-if len (sys.argv) != 5:
- print ("usage: ./gen-use-table.py IndicSyllabicCategory.txt IndicPositionalCategory.txt UnicodeData.txt Blocks.txt", file=sys.stderr)
- sys.exit (1)
+if len (sys.argv) != 10:
+ sys.exit (__doc__)
-BLACKLISTED_BLOCKS = ["Thai", "Lao"]
+DISABLED_SCRIPTS = {
+ 'Arabic',
+ 'Lao',
+ 'Samaritan',
+ 'Syriac',
+ 'Thai',
+}
-files = [io.open (x, encoding='utf-8') for x in sys.argv[1:]]
+files = [open (x, encoding='utf-8') for x in sys.argv[1:]]
-headers = [[f.readline () for i in range (2)] for j,f in enumerate(files) if j != 2]
+headers = [[f.readline () for i in range (2)] for j,f in enumerate(files) if j != 4]
+for j in range(7, 9):
+ for line in files[j]:
+ line = line.rstrip()
+ if not line:
+ break
+ headers[j - 1].append(line)
headers.append (["UnicodeData.txt does not have a header."])
-data = [{} for f in files]
-values = [{} for f in files]
+data = [{} for _ in files]
+values = [{} for _ in files]
for i, f in enumerate (files):
for line in f:
else:
end = int (uu[1], 16)
- t = fields[1 if i != 2 else 2]
+ t = fields[1 if i not in [2, 4] else 2]
+
+ if i == 2:
+ t = 'jt_' + t
+ elif i == 3 and t != 'Default_Ignorable_Code_Point':
+ continue
+ elif i == 7 and t == 'Consonant_Final_Modifier':
+ # TODO: https://github.com/MicrosoftDocs/typography-issues/issues/336
+ t = 'Syllable_Modifier'
+ elif i == 8 and t == 'NA':
+ t = 'Not_Applicable'
+ i0 = i if i < 7 else i - 7
for u in range (start, end + 1):
- data[i][u] = t
- values[i][t] = values[i].get (t, 0) + end - start + 1
+ data[i0][u] = t
+ values[i0][t] = values[i0].get (t, 0) + end - start + 1
-defaults = ('Other', 'Not_Applicable', 'Cn', 'No_Block')
+defaults = ('Other', 'Not_Applicable', 'jt_X', '', 'Cn', 'No_Block', 'Unknown')
# TODO Characters that are not in Unicode Indic files, but used in USE
-data[0][0x034F] = defaults[0]
data[0][0x1B61] = defaults[0]
data[0][0x1B63] = defaults[0]
data[0][0x1B64] = defaults[0]
data[0][0x11C45] = 'Consonant_Placeholder'
# TODO https://github.com/harfbuzz/harfbuzz/pull/1399
data[0][0x111C8] = 'Consonant_Placeholder'
-for u in range (0xFE00, 0xFE0F + 1):
- data[0][u] = defaults[0]
# Merge data into one dict:
for i,v in enumerate (defaults):
combined = {}
for i,d in enumerate (data):
for u,v in d.items ():
- if i >= 2 and not u in combined:
- continue
if not u in combined:
+ if i >= 4:
+ continue
combined[u] = list (defaults)
combined[u][i] = v
-combined = {k:v for k,v in combined.items() if v[3] not in BLACKLISTED_BLOCKS}
+combined = {k: v for k, v in combined.items() if v[6] not in DISABLED_SCRIPTS}
data = combined
del combined
-num = len (data)
property_names = [
'Number_Joiner',
'Number',
'Brahmi_Joining_Number',
+ 'Hieroglyph',
+ 'Hieroglyph_Joiner',
+ 'Hieroglyph_Segment_Begin',
+ 'Hieroglyph_Segment_End',
# Indic_Positional_Category
'Not_Applicable',
'Right',
'Top',
'Bottom',
'Top_And_Bottom',
+ 'Top_And_Bottom_And_Left',
'Top_And_Right',
'Top_And_Left',
'Top_And_Left_And_Right',
'Bottom_And_Right',
'Top_And_Bottom_And_Right',
'Overstruck',
+ # Joining_Type
+ 'jt_C',
+ 'jt_D',
+ 'jt_L',
+ 'jt_R',
+ 'jt_T',
+ 'jt_U',
+ 'jt_X',
]
-try:
- basestring
-except NameError:
- basestring = str
-
class PropertyValue(object):
def __init__(self, name_):
self.name = name_
def __str__(self):
return self.name
def __eq__(self, other):
- return self.name == (other if isinstance(other, basestring) else other.name)
+ return self.name == (other if isinstance(other, str) else other.name)
def __ne__(self, other):
return not (self == other)
def __hash__(self):
globals().update(property_values)
-def is_BASE(U, UISC, UGC):
+def is_BASE(U, UISC, UDI, UGC, AJT):
return (UISC in [Number, Consonant, Consonant_Head_Letter,
- #SPEC-DRAFT Consonant_Placeholder,
Tone_Letter,
- Vowel_Independent #SPEC-DRAFT
+ Vowel_Independent,
] or
+ # TODO: https://github.com/MicrosoftDocs/typography-issues/issues/484
+ AJT in [jt_C, jt_D, jt_L, jt_R] and UISC != Joiner or
(UGC == Lo and UISC in [Avagraha, Bindu, Consonant_Final, Consonant_Medial,
Consonant_Subjoined, Vowel, Vowel_Dependent]))
-def is_BASE_IND(U, UISC, UGC):
- #SPEC-DRAFT return (UISC in [Consonant_Dead, Modifying_Letter] or UGC == Po)
- return (UISC in [Consonant_Dead, Modifying_Letter] or
- (UGC == Po and not U in [0x104B, 0x104E, 0x1B5B, 0x1B5C, 0x1B5F, 0x2022, 0x111C8, 0x11A3F, 0x11A45, 0x11C44, 0x11C45]) or
- False # SPEC-DRAFT-OUTDATED! U == 0x002D
- )
-def is_BASE_NUM(U, UISC, UGC):
+def is_BASE_NUM(U, UISC, UDI, UGC, AJT):
return UISC == Brahmi_Joining_Number
-def is_BASE_OTHER(U, UISC, UGC):
- if UISC == Consonant_Placeholder: return True #SPEC-DRAFT
- #SPEC-DRAFT return U in [0x00A0, 0x00D7, 0x2015, 0x2022, 0x25CC, 0x25FB, 0x25FC, 0x25FD, 0x25FE]
+def is_BASE_OTHER(U, UISC, UDI, UGC, AJT):
+ if UISC == Consonant_Placeholder: return True
return U in [0x2015, 0x2022, 0x25FB, 0x25FC, 0x25FD, 0x25FE]
-def is_CGJ(U, UISC, UGC):
- return U == 0x034F
-def is_CONS_FINAL(U, UISC, UGC):
+def is_CGJ(U, UISC, UDI, UGC, AJT):
+ # Also includes VARIATION_SELECTOR, WJ, and ZWJ
+ return U == 0x200D or UDI and UGC in [Mc, Me, Mn]
+def is_CONS_FINAL(U, UISC, UDI, UGC, AJT):
return ((UISC == Consonant_Final and UGC != Lo) or
UISC == Consonant_Succeeding_Repha)
-def is_CONS_FINAL_MOD(U, UISC, UGC):
- #SPEC-DRAFT return UISC in [Consonant_Final_Modifier, Syllable_Modifier]
- return UISC == Syllable_Modifier
-def is_CONS_MED(U, UISC, UGC):
+def is_CONS_FINAL_MOD(U, UISC, UDI, UGC, AJT):
+ return UISC == Syllable_Modifier
+def is_CONS_MED(U, UISC, UDI, UGC, AJT):
# Consonant_Initial_Postfixed is new in Unicode 11; not in the spec.
return (UISC == Consonant_Medial and UGC != Lo or
UISC == Consonant_Initial_Postfixed)
-def is_CONS_MOD(U, UISC, UGC):
- return UISC in [Nukta, Gemination_Mark, Consonant_Killer]
-def is_CONS_SUB(U, UISC, UGC):
- #SPEC-DRAFT return UISC == Consonant_Subjoined
+def is_CONS_MOD(U, UISC, UDI, UGC, AJT):
+ return (UISC in [Nukta, Gemination_Mark, Consonant_Killer] and
+ not is_SYM_MOD(U, UISC, UDI, UGC, AJT))
+def is_CONS_SUB(U, UISC, UDI, UGC, AJT):
return UISC == Consonant_Subjoined and UGC != Lo
-def is_CONS_WITH_STACKER(U, UISC, UGC):
+def is_CONS_WITH_STACKER(U, UISC, UDI, UGC, AJT):
return UISC == Consonant_With_Stacker
-def is_HALANT(U, UISC, UGC):
+def is_HALANT(U, UISC, UDI, UGC, AJT):
return (UISC in [Virama, Invisible_Stacker]
- and not is_HALANT_OR_VOWEL_MODIFIER(U, UISC, UGC)
- and not is_SAKOT(U, UISC, UGC))
-def is_HALANT_OR_VOWEL_MODIFIER(U, UISC, UGC):
- # https://github.com/harfbuzz/harfbuzz/issues/1102
+ and not is_HALANT_OR_VOWEL_MODIFIER(U, UISC, UDI, UGC, AJT)
+ and not is_SAKOT(U, UISC, UDI, UGC, AJT))
+def is_HALANT_OR_VOWEL_MODIFIER(U, UISC, UDI, UGC, AJT):
+ # Split off of HALANT
# https://github.com/harfbuzz/harfbuzz/issues/1379
- return U in [0x11046, 0x1134D]
-def is_HALANT_NUM(U, UISC, UGC):
+ return U == 0x1134D
+def is_HALANT_NUM(U, UISC, UDI, UGC, AJT):
return UISC == Number_Joiner
-def is_ZWNJ(U, UISC, UGC):
+def is_HIEROGLYPH(U, UISC, UDI, UGC, AJT):
+ return UISC == Hieroglyph
+def is_HIEROGLYPH_JOINER(U, UISC, UDI, UGC, AJT):
+ return UISC == Hieroglyph_Joiner
+def is_HIEROGLYPH_SEGMENT_BEGIN(U, UISC, UDI, UGC, AJT):
+ return UISC == Hieroglyph_Segment_Begin
+def is_HIEROGLYPH_SEGMENT_END(U, UISC, UDI, UGC, AJT):
+ return UISC == Hieroglyph_Segment_End
+def is_ZWNJ(U, UISC, UDI, UGC, AJT):
return UISC == Non_Joiner
-def is_ZWJ(U, UISC, UGC):
- return UISC == Joiner
-def is_Word_Joiner(U, UISC, UGC):
- return U == 0x2060
-def is_OTHER(U, UISC, UGC):
- #SPEC-OUTDATED return UGC == Zs # or any other SCRIPT_COMMON characters
- return (UISC == Other
- and not is_SYM(U, UISC, UGC)
- and not is_SYM_MOD(U, UISC, UGC)
- and not is_CGJ(U, UISC, UGC)
- and not is_Word_Joiner(U, UISC, UGC)
- and not is_VARIATION_SELECTOR(U, UISC, UGC)
+def is_OTHER(U, UISC, UDI, UGC, AJT):
+ # Also includes BASE_IND, Rsv, and SYM
+ return ((UGC in [Cn, Po] or UISC in [Consonant_Dead, Joiner, Modifying_Letter, Other])
+ and not is_BASE(U, UISC, UDI, UGC, AJT)
+ and not is_BASE_OTHER(U, UISC, UDI, UGC, AJT)
+ and not is_CGJ(U, UISC, UDI, UGC, AJT)
+ and not is_SYM_MOD(U, UISC, UDI, UGC, AJT)
)
-def is_Reserved(U, UISC, UGC):
- return UGC == 'Cn'
-def is_REPHA(U, UISC, UGC):
+def is_REPHA(U, UISC, UDI, UGC, AJT):
return UISC in [Consonant_Preceding_Repha, Consonant_Prefixed]
-def is_SAKOT(U, UISC, UGC):
+def is_SAKOT(U, UISC, UDI, UGC, AJT):
+ # Split off of HALANT
return U == 0x1A60
-def is_SYM(U, UISC, UGC):
- if U == 0x25CC: return False #SPEC-DRAFT
- #SPEC-DRAFT return UGC in [So, Sc] or UISC == Symbol_Letter
- return UGC in [So, Sc] and U not in [0x1B62, 0x1B68]
-def is_SYM_MOD(U, UISC, UGC):
+def is_SYM_MOD(U, UISC, UDI, UGC, AJT):
return U in [0x1B6B, 0x1B6C, 0x1B6D, 0x1B6E, 0x1B6F, 0x1B70, 0x1B71, 0x1B72, 0x1B73]
-def is_VARIATION_SELECTOR(U, UISC, UGC):
- return 0xFE00 <= U <= 0xFE0F
-def is_VOWEL(U, UISC, UGC):
+def is_VOWEL(U, UISC, UDI, UGC, AJT):
# https://github.com/harfbuzz/harfbuzz/issues/376
return (UISC == Pure_Killer or
(UGC != Lo and UISC in [Vowel, Vowel_Dependent] and U not in [0xAA29]))
-def is_VOWEL_MOD(U, UISC, UGC):
+def is_VOWEL_MOD(U, UISC, UDI, UGC, AJT):
# https://github.com/harfbuzz/harfbuzz/issues/376
return (UISC in [Tone_Mark, Cantillation_Mark, Register_Shifter, Visarga] or
(UGC != Lo and (UISC == Bindu or U in [0xAA29])))
use_mapping = {
'B': is_BASE,
- 'IND': is_BASE_IND,
'N': is_BASE_NUM,
'GB': is_BASE_OTHER,
'CGJ': is_CGJ,
'H': is_HALANT,
'HVM': is_HALANT_OR_VOWEL_MODIFIER,
'HN': is_HALANT_NUM,
+ 'G': is_HIEROGLYPH,
+ 'J': is_HIEROGLYPH_JOINER,
+ 'SB': is_HIEROGLYPH_SEGMENT_BEGIN,
+ 'SE': is_HIEROGLYPH_SEGMENT_END,
'ZWNJ': is_ZWNJ,
- 'ZWJ': is_ZWJ,
- 'WJ': is_Word_Joiner,
'O': is_OTHER,
- 'Rsv': is_Reserved,
'R': is_REPHA,
- 'S': is_SYM,
'Sk': is_SAKOT,
'SM': is_SYM_MOD,
- 'VS': is_VARIATION_SELECTOR,
'V': is_VOWEL,
'VM': is_VOWEL_MOD,
}
},
'M': {
'Abv': [Top],
- 'Blw': [Bottom, Bottom_And_Left],
+ 'Blw': [Bottom, Bottom_And_Left, Bottom_And_Right],
'Pst': [Right],
- 'Pre': [Left],
+ 'Pre': [Left, Top_And_Bottom_And_Left],
},
'CM': {
'Abv': [Top],
- 'Blw': [Bottom],
+ 'Blw': [Bottom, Overstruck],
},
'V': {
'Abv': [Top, Top_And_Bottom, Top_And_Bottom_And_Right, Top_And_Right],
'Blw': [Bottom, Overstruck, Bottom_And_Right],
- 'Pst': [Right, Top_And_Left, Top_And_Left_And_Right, Left_And_Right],
- 'Pre': [Left],
+ 'Pst': [Right],
+ 'Pre': [Left, Top_And_Left, Top_And_Left_And_Right, Left_And_Right],
},
'VM': {
'Abv': [Top],
'Blw': [Bottom],
'Pst': [Not_Applicable],
},
+ 'R': None,
'SUB': None,
}
def map_to_use(data):
out = {}
items = use_mapping.items()
- for U,(UISC,UIPC,UGC,UBlock) in data.items():
+ for U, (UISC, UIPC, AJT, UDI, UGC, UBlock, _) in data.items():
# Resolve Indic_Syllabic_Category
- # TODO: These don't have UISC assigned in Unicode 12.0, but have UIPC
+ # TODO: These don't have UISC assigned in Unicode 13.0.0, but have UIPC
if 0x1CE2 <= U <= 0x1CE8: UISC = Cantillation_Mark
# Tibetan:
- # TODO: These don't have UISC assigned in Unicode 12.0, but have UIPC
+ # TODO: These don't have UISC assigned in Unicode 13.0.0, but have UIPC
if 0x0F18 <= U <= 0x0F19 or 0x0F3E <= U <= 0x0F3F: UISC = Vowel_Dependent
- if 0x0F86 <= U <= 0x0F87: UISC = Tone_Mark
- # Overrides to allow NFC order matching syllable
- # https://github.com/harfbuzz/harfbuzz/issues/1012
- if UBlock == 'Tibetan' and is_VOWEL (U, UISC, UGC):
- if UIPC == Top:
- UIPC = Bottom
-
- # TODO: https://github.com/harfbuzz/harfbuzz/pull/982
- # also https://github.com/harfbuzz/harfbuzz/issues/1012
- if UBlock == 'Chakma' and is_VOWEL (U, UISC, UGC):
- if UIPC == Top:
- UIPC = Bottom
- elif UIPC == Bottom:
- UIPC = Top
# TODO: https://github.com/harfbuzz/harfbuzz/pull/627
if 0x1BF2 <= U <= 0x1BF3: UISC = Nukta; UIPC = Bottom
# the nasalization marks, maybe only for U+1CE9..U+1CF1.
if U == 0x1CED: UISC = Tone_Mark
- # TODO: https://github.com/harfbuzz/harfbuzz/issues/1105
- if U == 0x11134: UISC = Gemination_Mark
+ # TODO: https://github.com/microsoft/font-tools/issues/1
+ if U == 0xA982: UISC = Consonant_Succeeding_Repha
- values = [k for k,v in items if v(U,UISC,UGC)]
- assert len(values) == 1, "%s %s %s %s" % (hex(U), UISC, UGC, values)
+ values = [k for k,v in items if v(U, UISC, UDI, UGC, AJT)]
+ assert len(values) == 1, "%s %s %s %s %s %s" % (hex(U), UISC, UDI, UGC, AJT, values)
USE = values[0]
# Resolve Indic_Positional_Category
- # TODO: These should die, but have UIPC in Unicode 12.0
+ # TODO: These should die, but have UIPC in Unicode 13.0.0
if U in [0x953, 0x954]: UIPC = Not_Applicable
- # TODO: In USE's override list but not in Unicode 12.0
- if U == 0x103C: UIPC = Left
-
- # TODO: https://github.com/harfbuzz/harfbuzz/pull/2012
- if U == 0x1C29: UIPC = Left
-
- # TODO: These are not in USE's override list that we have, nor are they in Unicode 12.0
+ # TODO: These are not in USE's override list that we have, nor are they in Unicode 13.0.0
if 0xA926 <= U <= 0xA92A: UIPC = Top
# TODO: https://github.com/harfbuzz/harfbuzz/pull/1037
# and https://github.com/harfbuzz/harfbuzz/issues/1631
if U in [0x11302, 0x11303, 0x114C1]: UIPC = Top
- if U == 0x1171E: UIPC = Left
if 0x1CF8 <= U <= 0x1CF9: UIPC = Top
- assert (UIPC in [Not_Applicable, Visual_Order_Left] or
- USE in use_positions), "%s %s %s %s %s" % (hex(U), UIPC, USE, UISC, UGC)
+ # TODO: https://github.com/harfbuzz/harfbuzz/pull/982
+ # also https://github.com/harfbuzz/harfbuzz/issues/1012
+ if 0x1112A <= U <= 0x1112B: UIPC = Top
+ if 0x11131 <= U <= 0x11132: UIPC = Top
+
+ assert (UIPC in [Not_Applicable, Visual_Order_Left] or U == 0x0F7F or
+ USE in use_positions), "%s %s %s %s %s %s %s" % (hex(U), UIPC, USE, UISC, UDI, UGC, AJT)
pos_mapping = use_positions.get(USE, None)
if pos_mapping:
values = [k for k,v in pos_mapping.items() if v and UIPC in v]
- assert len(values) == 1, "%s %s %s %s %s %s" % (hex(U), UIPC, USE, UISC, UGC, values)
+ assert len(values) == 1, "%s %s %s %s %s %s %s %s" % (hex(U), UIPC, USE, UISC, UDI, UGC, AJT, values)
USE = USE + values[0]
out[U] = (USE, UBlock)
print ("/*")
print (" * The following table is generated by running:")
print (" *")
-print (" * ./gen-use-table.py IndicSyllabicCategory.txt IndicPositionalCategory.txt UnicodeData.txt Blocks.txt")
+print (" * {} IndicSyllabicCategory.txt IndicPositionalCategory.txt ArabicShaping.txt DerivedCoreProperties.txt UnicodeData.txt Blocks.txt Scripts.txt IndicSyllabicCategory-Additional.txt IndicPositionalCategory-Additional.txt".format (sys.argv[0]))
print (" *")
print (" * on files with these headers:")
print (" *")
print (" * %s" % (l.strip()))
print (" */")
print ()
-print ('#include "hb.hh"')
+print ("#ifndef HB_OT_SHAPE_COMPLEX_USE_TABLE_HH")
+print ("#define HB_OT_SHAPE_COMPLEX_USE_TABLE_HH")
print ()
-print ('#ifndef HB_NO_OT_SHAPE')
+print ('#include "hb.hh"')
print ()
-print ('#include "hb-ot-shape-complex-use.hh"')
+print ('#include "hb-ot-shape-complex-use-machine.hh"')
print ()
total = 0
print ('#pragma GCC diagnostic ignored "-Wunused-macros"')
for k,v in sorted(use_mapping.items()):
if k in use_positions and use_positions[k]: continue
- print ("#define %s USE_%s /* %s */" % (k, k, v.__name__[3:]))
+ print ("#define %s USE(%s) /* %s */" % (k, k, v.__name__[3:]))
for k,v in sorted(use_positions.items()):
if not v: continue
for suf in v.keys():
tag = k + suf
- print ("#define %s USE_%s" % (tag, tag))
+ print ("#define %s USE(%s)" % (tag, tag))
print ('#pragma GCC diagnostic pop')
print ("")
-print ("static const USE_TABLE_ELEMENT_TYPE use_table[] = {")
+print ("static const uint8_t use_table[] = {")
for u in uu:
if u <= last:
continue
+ if data[u][0] == 'O':
+ continue
block = data[u][1]
start = u//8*8
if start != last + 1:
if start - last <= 1+16*3:
print_block (None, last+1, start-1, data)
- last = start-1
else:
if last >= 0:
ends.append (last + 1)
page_bits = 12
print ("}; /* Table items: %d; occupancy: %d%% */" % (offset, occupancy))
print ()
-print ("USE_TABLE_ELEMENT_TYPE")
+print ("static inline uint8_t")
print ("hb_use_get_category (hb_codepoint_t u)")
print ("{")
print (" switch (u >> %d)" % page_bits)
print (" default:")
print (" break;")
print (" }")
-print (" return USE_O;")
+print (" return USE(O);")
print ("}")
print ()
for k in sorted(use_mapping.keys()):
print ("#undef %s" % tag)
print ()
print ()
-print ('#endif')
+print ("#endif /* HB_OT_SHAPE_COMPLEX_USE_TABLE_HH */")
print ("/* == End of generated table == */")
# Maintain at least 50% occupancy in the table */
projects / platform / upstream / harfbuzz.git / blobdiff