tor-browser

The Tor Browser
git clone https://git.dasho.dev/tor-browser.git
Log | Files | Refs | README | LICENSE

gen-arabic-joining-list.py (2402B)

      1 #!/usr/bin/env python3
      2 
      3 """usage: ./gen-arabic-joining-table.py ArabicShaping.txt Scripts.txt
      4 
      5 Input files:
      6 * https://unicode.org/Public/UCD/latest/ucd/ArabicShaping.txt
      7 * https://unicode.org/Public/UCD/latest/ucd/Scripts.txt
      8 """
      9 
     10 import os.path, sys
     11 
     12 if len (sys.argv) != 3:
     13 sys.exit (__doc__)
     14 
     15 files = [open (x, encoding='utf-8') for x in sys.argv[1:]]
     16 
     17 headers = [[f.readline (), f.readline ()] for f in files]
     18 while files[0].readline ().find ('##################') < 0:
     19 pass
     20 
     21 def read (f):
     22 mapping = {}
     23 for line in f:
     24 
     25 	j = line.find ('#')
     26 	if j >= 0:
     27 		line = line[:j]
     28 
     29 	fields = [x.strip () for x in line.split (';')]
     30 	if len (fields) == 1:
     31 		continue
     32 
     33 	uu = fields[0].split ('..')
     34 	start = int (uu[0], 16)
     35 	if len (uu) == 1:
     36 		end = start
     37 	else:
     38 		end = int (uu[1], 16)
     39 
     40 	t = fields[1]
     41 
     42 	for u in range (start, end + 1):
     43 		mapping[u] = t
     44 
     45 return mapping
     46 
     47 def read_joining_uu (f):
     48 values = set ()
     49 for line in f:
     50 
     51 	if line[0] == '#':
     52 		continue
     53 
     54 	fields = [x.strip () for x in line.split (';')]
     55 	if len (fields) == 1:
     56 		continue
     57 	if fields[2] in {'T', 'U'}:
     58 		continue
     59 
     60 	values.add (int (fields[0], 16))
     61 
     62 return sorted (values)
     63 
     64 def print_has_arabic_joining (scripts, joining_uu):
     65 
     66 print ("static bool")
     67 print ("has_arabic_joining (hb_script_t script)")
     68 print ("{")
     69 print ("  /* List of scripts that have data in arabic-table. */")
     70 print ("  switch ((int) script)")
     71 print ("  {")
     72 
     73 for script in sorted ({scripts[u] for u in joining_uu if scripts[u] not in {'Common', 'Inherited'}}):
     74 	print ("    case HB_SCRIPT_{}:".format (script.upper ()))
     75 
     76 print ("      return true;")
     77 print ()
     78 print ("    default:")
     79 print ("      return false;")
     80 print ("  }")
     81 print ("}")
     82 print ()
     83 
     84 print ("/* == Start of generated function == */")
     85 print ("/*")
     86 print (" * The following function is generated by running:")
     87 print (" *")
     88 print (" *   ./gen-arabic-joining-list.py ArabicShaping.txt Scripts.txt")
     89 print (" *")
     90 print (" * on files with these headers:")
     91 print (" *")
     92 for h in headers:
     93 for l in h:
     94 	print (" * %s" % (l.strip ()))
     95 print (" */")
     96 print ()
     97 print ("#ifndef HB_OT_SHAPER_ARABIC_JOINING_LIST_HH")
     98 print ("#define HB_OT_SHAPER_ARABIC_JOINING_LIST_HH")
     99 print ()
    100 
    101 print_has_arabic_joining (read (files[1]), read_joining_uu (files[0]))
    102 
    103 print ()
    104 print ("#endif /* HB_OT_SHAPER_ARABIC_JOINING_LIST_HH */")
    105 print ()
    106 print ("/* == End of generated function == */")