trunk/maint/GenerateUtt.py

#! /usr/bin/python

# Generate utt tables.

# The source file pcre_tables.c contains (amongst other things), a table that
# is indexed by script name. In order to reduce the number of relocations when
# loading the library, the names are held as a single large string, with
# offsets in the table. This is tedious to maintain by hand. Therefore, this
# script is used to generate the table. The output is sent to stdout.

# Modified by PH 17-March-2009 to generate the more verbose form that works
# for UTF-support in EBCDIC as well as ASCII environments.
# Modified by PH 01-March-2010 to add new scripts from Unicode 5.2.0.

script_names = ['Arabic', 'Armenian', 'Bengali', 'Bopomofo', 'Braille', 'Buginese', 'Buhid', 'Canadian_Aboriginal', \
 'Cherokee', 'Common', 'Coptic', 'Cypriot', 'Cyrillic', 'Deseret', 'Devanagari', 'Ethiopic', 'Georgian', \
 'Glagolitic', 'Gothic', 'Greek', 'Gujarati', 'Gurmukhi', 'Han', 'Hangul', 'Hanunoo', 'Hebrew', 'Hiragana', \
 'Inherited', 'Kannada', 'Katakana', 'Kharoshthi', 'Khmer', 'Lao', 'Latin', 'Limbu', 'Linear_B', 'Malayalam', \
 'Mongolian', 'Myanmar', 'New_Tai_Lue', 'Ogham', 'Old_Italic', 'Old_Persian', 'Oriya', 'Osmanya', 'Runic', \
 'Shavian', 'Sinhala', 'Syloti_Nagri', 'Syriac', 'Tagalog', 'Tagbanwa', 'Tai_Le', 'Tamil', 'Telugu', 'Thaana', \
 'Thai', 'Tibetan', 'Tifinagh', 'Ugaritic', 'Yi', \
 # New for Unicode 5.0
 'Balinese', 'Cuneiform', 'Nko', 'Phags_Pa', 'Phoenician', \
 # New for Unicode 5.1
 'Carian', 'Cham', 'Kayah_Li', 'Lepcha', 'Lycian', 'Lydian', 'Ol_Chiki', 'Rejang', 'Saurashtra', 'Sundanese', 'Vai', \
 # New for Unicode 5.2
 'Avestan', 'Bamum', 'Egyptian_Hieroglyphs', 'Imperial_Aramaic', \
 'Inscriptional_Pahlavi', 'Inscriptional_Parthian', \
 'Javanese', 'Kaithi', 'Lisu', 'Meetei_Mayek', \
 'Old_South_Arabian', 'Old_Turkic', 'Samaritan', 'Tai_Tham', 'Tai_Viet'
 ]

category_names = ['Cc', 'Cf', 'Cn', 'Co', 'Cs', 'Ll', 'Lm', 'Lo', 'Lt', 'Lu',
  'Mc', 'Me', 'Mn', 'Nd', 'Nl', 'No', 'Pc', 'Pd', 'Pe', 'Pf', 'Pi', 'Po', 'Ps',
  'Sc', 'Sk', 'Sm', 'So', 'Zl', 'Zp', 'Zs' ]

general_category_names = ['C', 'L', 'M', 'N', 'P', 'S', 'Z']

utt_table  = zip(script_names, ['PT_SC'] * len(script_names))
utt_table += zip(category_names, ['PT_PC'] * len(category_names))
utt_table += zip(general_category_names, ['PT_GC'] * len(general_category_names))
utt_table.append(('L&', 'PT_LAMP'))
utt_table.append(('Any', 'PT_ANY'))

utt_table.sort()

# We have to use STR_ macros to define the strings so that it all works in
# UTF-8 mode on EBCDIC platforms.

for utt in utt_table:
        print '#define STRING_%s0' % (utt[0].replace('&', '_AMPERSAND')),
        for c in utt[0]:
                if c == '_':
                        print 'STR_UNDERSCORE',
                elif c == '&':
                        print 'STR_AMPERSAND',
                else:
                        print 'STR_%s' % c,;
        print '"\\0"'

# Print the actual table, using the string names

print ''
print 'const char _pcre_utt_names[] = ';
last = ''
for utt in utt_table:
        if utt == utt_table[-1]:
                last = ';'
        print '  STRING_%s0%s' % (utt[0].replace('&', '_AMPERSAND'), last)
# This was how it was done before the EBCDIC-compatible modification.
#        print '  "%s\\0"%s' % (utt[0], last)

print '\nconst ucp_type_table _pcre_utt[] = { '
offset = 0
last = ','
for utt in utt_table:
        if utt[1] in ('PT_ANY', 'PT_LAMP'):
                value = '0'
        else:
                value = 'ucp_' + utt[0]
        if utt == utt_table[-1]:
                last = ''
        print '  { %3d, %s, %s }%s ' % (offset, utt[1], value, last)
        offset += len(utt[0]) + 1
print '};'
1	#! /usr/bin/python
2
3	# Generate utt tables.
4
5	# The source file pcre_tables.c contains (amongst other things), a table that
6	# is indexed by script name. In order to reduce the number of relocations when
7	# loading the library, the names are held as a single large string, with
8	# offsets in the table. This is tedious to maintain by hand. Therefore, this
9	# script is used to generate the table. The output is sent to stdout.
10
11	# Modified by PH 17-March-2009 to generate the more verbose form that works
12	# for UTF-support in EBCDIC as well as ASCII environments.
13	# Modified by PH 01-March-2010 to add new scripts from Unicode 5.2.0.
14
15	script_names = ['Arabic', 'Armenian', 'Bengali', 'Bopomofo', 'Braille', 'Buginese', 'Buhid', 'Canadian_Aboriginal', \
16	'Cherokee', 'Common', 'Coptic', 'Cypriot', 'Cyrillic', 'Deseret', 'Devanagari', 'Ethiopic', 'Georgian', \
17	'Glagolitic', 'Gothic', 'Greek', 'Gujarati', 'Gurmukhi', 'Han', 'Hangul', 'Hanunoo', 'Hebrew', 'Hiragana', \
18	'Inherited', 'Kannada', 'Katakana', 'Kharoshthi', 'Khmer', 'Lao', 'Latin', 'Limbu', 'Linear_B', 'Malayalam', \
19	'Mongolian', 'Myanmar', 'New_Tai_Lue', 'Ogham', 'Old_Italic', 'Old_Persian', 'Oriya', 'Osmanya', 'Runic', \
20	'Shavian', 'Sinhala', 'Syloti_Nagri', 'Syriac', 'Tagalog', 'Tagbanwa', 'Tai_Le', 'Tamil', 'Telugu', 'Thaana', \
21	'Thai', 'Tibetan', 'Tifinagh', 'Ugaritic', 'Yi', \
22	# New for Unicode 5.0
23	'Balinese', 'Cuneiform', 'Nko', 'Phags_Pa', 'Phoenician', \
24	# New for Unicode 5.1
25	'Carian', 'Cham', 'Kayah_Li', 'Lepcha', 'Lycian', 'Lydian', 'Ol_Chiki', 'Rejang', 'Saurashtra', 'Sundanese', 'Vai', \
26	# New for Unicode 5.2
27	'Avestan', 'Bamum', 'Egyptian_Hieroglyphs', 'Imperial_Aramaic', \
28	'Inscriptional_Pahlavi', 'Inscriptional_Parthian', \
29	'Javanese', 'Kaithi', 'Lisu', 'Meetei_Mayek', \
30	'Old_South_Arabian', 'Old_Turkic', 'Samaritan', 'Tai_Tham', 'Tai_Viet'
31	]
32
33	category_names = ['Cc', 'Cf', 'Cn', 'Co', 'Cs', 'Ll', 'Lm', 'Lo', 'Lt', 'Lu',
34	'Mc', 'Me', 'Mn', 'Nd', 'Nl', 'No', 'Pc', 'Pd', 'Pe', 'Pf', 'Pi', 'Po', 'Ps',
35	'Sc', 'Sk', 'Sm', 'So', 'Zl', 'Zp', 'Zs' ]
36
37	general_category_names = ['C', 'L', 'M', 'N', 'P', 'S', 'Z']
38
39	utt_table = zip(script_names, ['PT_SC'] * len(script_names))
40	utt_table += zip(category_names, ['PT_PC'] * len(category_names))
41	utt_table += zip(general_category_names, ['PT_GC'] * len(general_category_names))
42	utt_table.append(('L&', 'PT_LAMP'))
43	utt_table.append(('Any', 'PT_ANY'))
44
45	utt_table.sort()
46
47	# We have to use STR_ macros to define the strings so that it all works in
48	# UTF-8 mode on EBCDIC platforms.
49
50	for utt in utt_table:
51	print '#define STRING_%s0' % (utt[0].replace('&', '_AMPERSAND')),
52	for c in utt[0]:
53	if c == '_':
54	print 'STR_UNDERSCORE',
55	elif c == '&':
56	print 'STR_AMPERSAND',
57	else:
58	print 'STR_%s' % c,;
59	print '"\\0"'
60
61	# Print the actual table, using the string names
62
63	print ''
64	print 'const char _pcre_utt_names[] = ';
65	last = ''
66	for utt in utt_table:
67	if utt == utt_table[-1]:
68	last = ';'
69	print ' STRING_%s0%s' % (utt[0].replace('&', '_AMPERSAND'), last)
70	# This was how it was done before the EBCDIC-compatible modification.
71	# print ' "%s\\0"%s' % (utt[0], last)
72
73	print '\nconst ucp_type_table _pcre_utt[] = { '
74	offset = 0
75	last = ','
76	for utt in utt_table:
77	if utt[1] in ('PT_ANY', 'PT_LAMP'):
78	value = '0'
79	else:
80	value = 'ucp_' + utt[0]
81	if utt == utt_table[-1]:
82	last = ''
83	print ' { %3d, %s, %s }%s ' % (offset, utt[1], value, last)
84	offset += len(utt[0]) + 1
85	print '};'