-#!/usr/bin/python
+#!/usr/bin/env python3
+
+"""usage: ./gen-indic-table.py IndicSyllabicCategory.txt IndicPositionalCategory.txt Blocks.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/Blocks.txt
+"""
import sys
if len (sys.argv) != 4:
- print >>sys.stderr, "usage: ./gen-indic-table.py IndicSyllabicCategory.txt IndicMatraCategory.txt Blocks.txt"
- sys.exit (1)
+ sys.exit (__doc__)
-BLACKLISTED_BLOCKS = ["Thai", "Lao", "Tibetan"]
+ALLOWED_SINGLES = [0x00A0, 0x25CC]
+ALLOWED_BLOCKS = [
+ 'Basic Latin',
+ 'Latin-1 Supplement',
+ 'Devanagari',
+ 'Bengali',
+ 'Gurmukhi',
+ 'Gujarati',
+ 'Oriya',
+ 'Tamil',
+ 'Telugu',
+ 'Kannada',
+ 'Malayalam',
+ 'Sinhala',
+ 'Myanmar',
+ 'Khmer',
+ 'Vedic Extensions',
+ 'General Punctuation',
+ 'Superscripts and Subscripts',
+ 'Devanagari Extended',
+ 'Myanmar Extended-B',
+ 'Myanmar Extended-A',
+]
-files = [file (x) 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 f in files]
if not u in combined:
combined[u] = list (defaults)
combined[u][i] = v
-combined = {k:v for k,v in combined.items() if v[2] not in BLACKLISTED_BLOCKS}
+combined = {k:v for k,v in combined.items() if k in ALLOWED_SINGLES or v[2] in ALLOWED_BLOCKS}
data = combined
del combined
num = len (data)
-for u in [0x17CD, 0x17CE, 0x17CF, 0x17D0, 0x17D3]:
- if data[u][0] == 'Other':
- data[u][0] = "Vowel_Dependent"
-
# Move the outliers NO-BREAK SPACE and DOTTED CIRCLE out
singles = {}
-for u in [0x00A0, 0x25CC]:
+for u in ALLOWED_SINGLES:
singles[u] = data[u]
del data[u]
-print "/* == Start of generated table == */"
-print "/*"
-print " * The following table is generated by running:"
-print " *"
-print " * ./gen-indic-table.py IndicSyllabicCategory.txt IndicMatraCategory.txt Blocks.txt"
-print " *"
-print " * on files with these headers:"
-print " *"
+print ("/* == Start of generated table == */")
+print ("/*")
+print (" * The following table is generated by running:")
+print (" *")
+print (" * ./gen-indic-table.py IndicSyllabicCategory.txt IndicPositionalCategory.txt Blocks.txt")
+print (" *")
+print (" * on files with these headers:")
+print (" *")
for h in headers:
for l in h:
- print " * %s" % (l.strip())
-print " */"
-print
-print '#include "hb-ot-shape-complex-indic-private.hh"'
-print
+ print (" * %s" % (l.strip()))
+print (" */")
+print ()
+print ('#include "hb.hh"')
+print ()
+print ('#ifndef HB_NO_OT_SHAPE')
+print ()
+print ('#include "hb-ot-shape-complex-indic.hh"')
+print ()
# Shorten values
short = [{
what = ["INDIC_SYLLABIC_CATEGORY", "INDIC_MATRA_CATEGORY"]
what_short = ["ISC", "IMC"]
+print ('#pragma GCC diagnostic push')
+print ('#pragma GCC diagnostic ignored "-Wunused-macros"')
+cat_defs = []
for i in range (2):
- print
- vv = values[i].keys ()
- vv.sort ()
+ vv = sorted (values[i].keys ())
for v in vv:
v_no_and = v.replace ('_And_', '_')
if v in short[i]:
raise Exception ("Duplicate short value alias", v, all_shorts[i][s])
all_shorts[i][s] = v
short[i][v] = s
- print "#define %s_%s %s_%s %s/* %3d chars; %s */" % \
- (what_short[i], s, what[i], v.upper (), \
- ' '* ((48-1 - len (what[i]) - 1 - len (v)) / 8), \
- values[i][v], v)
-print
-print "#define _(S,M) INDIC_COMBINE_CATEGORIES (ISC_##S, IMC_##M)"
-print
-print
+ cat_defs.append ((what_short[i] + '_' + s, what[i] + '_' + v.upper (), str (values[i][v]), v))
+
+maxlen_s = max ([len (c[0]) for c in cat_defs])
+maxlen_l = max ([len (c[1]) for c in cat_defs])
+maxlen_n = max ([len (c[2]) for c in cat_defs])
+for s in what_short:
+ print ()
+ for c in [c for c in cat_defs if s in c[0]]:
+ print ("#define %s %s /* %s chars; %s */" %
+ (c[0].ljust (maxlen_s), c[1].ljust (maxlen_l), c[2].rjust (maxlen_n), c[3]))
+print ()
+print ('#pragma GCC diagnostic pop')
+print ()
+print ("#define _(S,M) INDIC_COMBINE_CATEGORIES (ISC_##S, IMC_##M)")
+print ()
+print ()
total = 0
used = 0
def print_block (block, start, end, data):
global total, used, last_block
if block and block != last_block:
- print
- print
- print " /* %s */" % block
+ print ()
+ print ()
+ print (" /* %s */" % block)
num = 0
assert start % 8 == 0
assert (end+1) % 8 == 0
for u in range (start, end+1):
if u % 8 == 0:
- print
- print " /* %04X */" % u,
+ print ()
+ print (" /* %04X */" % u, end="")
if u in data:
num += 1
d = data.get (u, defaults)
- sys.stdout.write ("%9s" % ("_(%s,%s)," % (short[0][d[0]], short[1][d[1]])))
+ print ("%9s" % ("_(%s,%s)," % (short[0][d[0]], short[1][d[1]])), end="")
total += end - start + 1
used += num
if block:
last_block = block
-uu = data.keys ()
-uu.sort ()
+uu = sorted (data.keys ())
last = -100000
num = 0
offset = 0
starts = []
ends = []
-print "static const INDIC_TABLE_ELEMENT_TYPE indic_table[] = {"
+print ("static const INDIC_TABLE_ELEMENT_TYPE indic_table[] = {")
for u in uu:
if u <= last:
continue
if last >= 0:
ends.append (last + 1)
offset += ends[-1] - starts[-1]
- print
- print
- print "#define indic_offset_0x%04xu %d" % (start, offset)
+ print ()
+ print ()
+ print ("#define indic_offset_0x%04xu %d" % (start, offset))
starts.append (start)
print_block (block, start, end, data)
last = end
ends.append (last + 1)
offset += ends[-1] - starts[-1]
-print
-print
+print ()
+print ()
occupancy = used * 100. / total
page_bits = 12
-print "}; /* Table items: %d; occupancy: %d%% */" % (offset, occupancy)
-print
-print "INDIC_TABLE_ELEMENT_TYPE"
-print "hb_indic_get_categories (hb_codepoint_t u)"
-print "{"
-print " switch (u >> %d)" % page_bits
-print " {"
-pages = set([u>>page_bits for u in starts+ends+singles.keys()])
+print ("}; /* Table items: %d; occupancy: %d%% */" % (offset, occupancy))
+print ()
+print ("INDIC_TABLE_ELEMENT_TYPE")
+print ("hb_indic_get_categories (hb_codepoint_t u)")
+print ("{")
+print (" switch (u >> %d)" % page_bits)
+print (" {")
+pages = set ([u>>page_bits for u in starts+ends+list (singles.keys ())])
for p in sorted(pages):
- print " case 0x%0Xu:" % p
+ print (" case 0x%0Xu:" % p)
+ for u,d in singles.items ():
+ if p != u>>page_bits: continue
+ print (" if (unlikely (u == 0x%04Xu)) return _(%s,%s);" % (u, short[0][d[0]], short[1][d[1]]))
for (start,end) in zip (starts, ends):
if p not in [start>>page_bits, end>>page_bits]: continue
offset = "indic_offset_0x%04xu" % start
- print " if (hb_in_range (u, 0x%04Xu, 0x%04Xu)) return indic_table[u - 0x%04Xu + %s];" % (start, end-1, start, offset)
- for u,d in singles.items ():
- if p != u>>page_bits: continue
- print " if (unlikely (u == 0x%04Xu)) return _(%s,%s);" % (u, short[0][d[0]], short[1][d[1]])
- print " break;"
- print ""
-print " default:"
-print " break;"
-print " }"
-print " return _(x,x);"
-print "}"
-print
-print "#undef _"
+ print (" if (hb_in_range<hb_codepoint_t> (u, 0x%04Xu, 0x%04Xu)) return indic_table[u - 0x%04Xu + %s];" % (start, end-1, start, offset))
+ print (" break;")
+ print ("")
+print (" default:")
+print (" break;")
+print (" }")
+print (" return _(x,x);")
+print ("}")
+print ()
+print ("#undef _")
for i in range (2):
- print
- vv = values[i].keys ()
- vv.sort ()
+ print ()
+ vv = sorted (values[i].keys ())
for v in vv:
- print "#undef %s_%s" % \
- (what_short[i], short[i][v])
-print
-print "/* == End of generated table == */"
+ print ("#undef %s_%s" %
+ (what_short[i], short[i][v]))
+print ()
+print ('#endif')
+print ()
+print ("/* == End of generated table == */")
# Maintain at least 30% occupancy in the table */
if occupancy < 30:
projects / platform / upstream / harfbuzz.git / blobdiff