2000-05-08 14:31:04 -03:00
|
|
|
#
|
|
|
|
# Secret Labs' Regular Expression Engine
|
|
|
|
#
|
|
|
|
# various symbols used by the regular expression engine.
|
|
|
|
# run this script to update the _sre include files!
|
|
|
|
#
|
|
|
|
# Copyright (c) 1998-2000 by Secret Labs AB. All rights reserved.
|
|
|
|
#
|
|
|
|
# Portions of this engine have been developed in cooperation with
|
|
|
|
# CNRI. Hewlett-Packard provided funding for 1.6 integration and
|
|
|
|
# other compatibility work.
|
|
|
|
#
|
|
|
|
|
2000-06-29 16:35:29 -03:00
|
|
|
# should this really be here?
|
|
|
|
|
|
|
|
class error(Exception):
|
|
|
|
pass
|
|
|
|
|
2000-05-08 14:31:04 -03:00
|
|
|
# operators
|
|
|
|
|
|
|
|
FAILURE = "failure"
|
|
|
|
SUCCESS = "success"
|
|
|
|
|
|
|
|
ANY = "any"
|
|
|
|
ASSERT = "assert"
|
2000-06-30 13:13:37 -03:00
|
|
|
ASSERT_NOT = "assert_not"
|
2000-05-08 14:31:04 -03:00
|
|
|
AT = "at"
|
|
|
|
BRANCH = "branch"
|
|
|
|
CALL = "call"
|
|
|
|
CATEGORY = "category"
|
|
|
|
GROUP = "group"
|
|
|
|
GROUP_IGNORE = "group_ignore"
|
|
|
|
IN = "in"
|
|
|
|
IN_IGNORE = "in_ignore"
|
2000-06-29 16:35:29 -03:00
|
|
|
INFO = "info"
|
2000-05-08 14:31:04 -03:00
|
|
|
JUMP = "jump"
|
|
|
|
LITERAL = "literal"
|
|
|
|
LITERAL_IGNORE = "literal_ignore"
|
|
|
|
MARK = "mark"
|
|
|
|
MAX_REPEAT = "max_repeat"
|
|
|
|
MAX_REPEAT_ONE = "max_repeat_one"
|
|
|
|
MIN_REPEAT = "min_repeat"
|
|
|
|
NEGATE = "negate"
|
|
|
|
NOT_LITERAL = "not_literal"
|
|
|
|
NOT_LITERAL_IGNORE = "not_literal_ignore"
|
|
|
|
RANGE = "range"
|
|
|
|
REPEAT = "repeat"
|
2000-06-29 16:35:29 -03:00
|
|
|
REPEAT_ONE = "repeat_one"
|
2000-05-08 14:31:04 -03:00
|
|
|
SUBPATTERN = "subpattern"
|
|
|
|
|
|
|
|
# positions
|
|
|
|
AT_BEGINNING = "at_beginning"
|
2000-06-29 16:35:29 -03:00
|
|
|
AT_BEGINNING_LINE = "at_beginning_line"
|
2000-05-08 14:31:04 -03:00
|
|
|
AT_BOUNDARY = "at_boundary"
|
|
|
|
AT_NON_BOUNDARY = "at_non_boundary"
|
|
|
|
AT_END = "at_end"
|
2000-06-29 16:35:29 -03:00
|
|
|
AT_END_LINE = "at_end_line"
|
2000-05-08 14:31:04 -03:00
|
|
|
|
|
|
|
# categories
|
|
|
|
CATEGORY_DIGIT = "category_digit"
|
|
|
|
CATEGORY_NOT_DIGIT = "category_not_digit"
|
|
|
|
CATEGORY_SPACE = "category_space"
|
|
|
|
CATEGORY_NOT_SPACE = "category_not_space"
|
|
|
|
CATEGORY_WORD = "category_word"
|
|
|
|
CATEGORY_NOT_WORD = "category_not_word"
|
2000-06-29 16:35:29 -03:00
|
|
|
CATEGORY_LINEBREAK = "category_linebreak"
|
|
|
|
CATEGORY_NOT_LINEBREAK = "category_not_linebreak"
|
|
|
|
CATEGORY_LOC_WORD = "category_loc_word"
|
|
|
|
CATEGORY_LOC_NOT_WORD = "category_loc_not_word"
|
|
|
|
CATEGORY_UNI_DIGIT = "category_uni_digit"
|
|
|
|
CATEGORY_UNI_NOT_DIGIT = "category_uni_not_digit"
|
|
|
|
CATEGORY_UNI_SPACE = "category_uni_space"
|
|
|
|
CATEGORY_UNI_NOT_SPACE = "category_uni_not_space"
|
|
|
|
CATEGORY_UNI_WORD = "category_uni_word"
|
|
|
|
CATEGORY_UNI_NOT_WORD = "category_uni_not_word"
|
|
|
|
CATEGORY_UNI_LINEBREAK = "category_uni_linebreak"
|
|
|
|
CATEGORY_UNI_NOT_LINEBREAK = "category_uni_not_linebreak"
|
2000-05-08 14:31:04 -03:00
|
|
|
|
2000-06-29 16:35:29 -03:00
|
|
|
OPCODES = [
|
2000-05-08 14:31:04 -03:00
|
|
|
|
|
|
|
# failure=0 success=1 (just because it looks better that way :-)
|
|
|
|
FAILURE, SUCCESS,
|
|
|
|
|
|
|
|
ANY,
|
2000-06-30 13:13:37 -03:00
|
|
|
ASSERT, ASSERT_NOT,
|
2000-05-08 14:31:04 -03:00
|
|
|
AT,
|
|
|
|
BRANCH,
|
|
|
|
CALL,
|
|
|
|
CATEGORY,
|
|
|
|
GROUP, GROUP_IGNORE,
|
|
|
|
IN, IN_IGNORE,
|
2000-06-29 16:35:29 -03:00
|
|
|
INFO,
|
2000-05-08 14:31:04 -03:00
|
|
|
JUMP,
|
|
|
|
LITERAL, LITERAL_IGNORE,
|
|
|
|
MARK,
|
2000-06-29 16:35:29 -03:00
|
|
|
MAX_REPEAT,
|
2000-05-08 14:31:04 -03:00
|
|
|
MAX_REPEAT_ONE,
|
2000-06-29 16:35:29 -03:00
|
|
|
MIN_REPEAT,
|
2000-05-08 14:31:04 -03:00
|
|
|
NOT_LITERAL, NOT_LITERAL_IGNORE,
|
|
|
|
NEGATE,
|
|
|
|
RANGE,
|
|
|
|
REPEAT
|
|
|
|
|
|
|
|
]
|
|
|
|
|
2000-06-29 16:35:29 -03:00
|
|
|
ATCODES = [
|
|
|
|
AT_BEGINNING, AT_BEGINNING_LINE, AT_BOUNDARY,
|
|
|
|
AT_NON_BOUNDARY, AT_END, AT_END_LINE
|
|
|
|
]
|
|
|
|
|
|
|
|
CHCODES = [
|
|
|
|
CATEGORY_DIGIT, CATEGORY_NOT_DIGIT, CATEGORY_SPACE,
|
|
|
|
CATEGORY_NOT_SPACE, CATEGORY_WORD, CATEGORY_NOT_WORD,
|
|
|
|
CATEGORY_LINEBREAK, CATEGORY_NOT_LINEBREAK, CATEGORY_LOC_WORD,
|
|
|
|
CATEGORY_LOC_NOT_WORD, CATEGORY_UNI_DIGIT, CATEGORY_UNI_NOT_DIGIT,
|
|
|
|
CATEGORY_UNI_SPACE, CATEGORY_UNI_NOT_SPACE, CATEGORY_UNI_WORD,
|
|
|
|
CATEGORY_UNI_NOT_WORD, CATEGORY_UNI_LINEBREAK,
|
|
|
|
CATEGORY_UNI_NOT_LINEBREAK
|
|
|
|
]
|
|
|
|
|
|
|
|
def makedict(list):
|
|
|
|
d = {}
|
|
|
|
i = 0
|
|
|
|
for item in list:
|
2000-06-30 13:13:37 -03:00
|
|
|
d[item] = i
|
|
|
|
i = i + 1
|
2000-06-29 16:35:29 -03:00
|
|
|
return d
|
|
|
|
|
|
|
|
OPCODES = makedict(OPCODES)
|
|
|
|
ATCODES = makedict(ATCODES)
|
|
|
|
CHCODES = makedict(CHCODES)
|
2000-05-08 14:31:04 -03:00
|
|
|
|
|
|
|
# replacement operations for "ignore case" mode
|
2000-06-29 16:35:29 -03:00
|
|
|
OP_IGNORE = {
|
2000-05-08 14:31:04 -03:00
|
|
|
GROUP: GROUP_IGNORE,
|
|
|
|
IN: IN_IGNORE,
|
|
|
|
LITERAL: LITERAL_IGNORE,
|
|
|
|
NOT_LITERAL: NOT_LITERAL_IGNORE
|
|
|
|
}
|
|
|
|
|
2000-06-29 16:35:29 -03:00
|
|
|
AT_MULTILINE = {
|
|
|
|
AT_BEGINNING: AT_BEGINNING_LINE,
|
|
|
|
AT_END: AT_END_LINE
|
2000-05-08 14:31:04 -03:00
|
|
|
}
|
|
|
|
|
2000-06-29 16:35:29 -03:00
|
|
|
CH_LOCALE = {
|
|
|
|
CATEGORY_DIGIT: CATEGORY_DIGIT,
|
|
|
|
CATEGORY_NOT_DIGIT: CATEGORY_NOT_DIGIT,
|
|
|
|
CATEGORY_SPACE: CATEGORY_SPACE,
|
|
|
|
CATEGORY_NOT_SPACE: CATEGORY_NOT_SPACE,
|
|
|
|
CATEGORY_WORD: CATEGORY_LOC_WORD,
|
|
|
|
CATEGORY_NOT_WORD: CATEGORY_LOC_NOT_WORD,
|
|
|
|
CATEGORY_LINEBREAK: CATEGORY_LINEBREAK,
|
|
|
|
CATEGORY_NOT_LINEBREAK: CATEGORY_NOT_LINEBREAK
|
2000-05-08 14:31:04 -03:00
|
|
|
}
|
|
|
|
|
2000-06-29 16:35:29 -03:00
|
|
|
CH_UNICODE = {
|
|
|
|
CATEGORY_DIGIT: CATEGORY_UNI_DIGIT,
|
|
|
|
CATEGORY_NOT_DIGIT: CATEGORY_UNI_NOT_DIGIT,
|
|
|
|
CATEGORY_SPACE: CATEGORY_UNI_SPACE,
|
|
|
|
CATEGORY_NOT_SPACE: CATEGORY_UNI_NOT_SPACE,
|
|
|
|
CATEGORY_WORD: CATEGORY_UNI_WORD,
|
|
|
|
CATEGORY_NOT_WORD: CATEGORY_UNI_NOT_WORD,
|
|
|
|
CATEGORY_LINEBREAK: CATEGORY_UNI_LINEBREAK,
|
|
|
|
CATEGORY_NOT_LINEBREAK: CATEGORY_UNI_NOT_LINEBREAK
|
|
|
|
}
|
|
|
|
|
|
|
|
# flags
|
|
|
|
SRE_FLAG_TEMPLATE = 1
|
|
|
|
SRE_FLAG_IGNORECASE = 2
|
|
|
|
SRE_FLAG_LOCALE = 4
|
|
|
|
SRE_FLAG_MULTILINE = 8
|
|
|
|
SRE_FLAG_DOTALL = 16
|
|
|
|
SRE_FLAG_UNICODE = 32
|
|
|
|
SRE_FLAG_VERBOSE = 64
|
|
|
|
|
2000-05-08 14:31:04 -03:00
|
|
|
if __name__ == "__main__":
|
|
|
|
import string
|
2000-06-29 16:35:29 -03:00
|
|
|
def dump(f, d, prefix):
|
2000-06-30 13:13:37 -03:00
|
|
|
items = d.items()
|
|
|
|
items.sort(lambda a, b: cmp(a[1], b[1]))
|
|
|
|
for k, v in items:
|
|
|
|
f.write("#define %s_%s %s\n" % (prefix, string.upper(k), v))
|
2000-05-08 14:31:04 -03:00
|
|
|
f = open("sre_constants.h", "w")
|
2000-06-30 13:13:37 -03:00
|
|
|
f.write("""\
|
|
|
|
/*
|
|
|
|
* Secret Labs' Regular Expression Engine
|
|
|
|
*
|
|
|
|
* regular expression matching engine
|
|
|
|
*
|
|
|
|
* NOTE: This file is generated by sre_constants.py. If you need
|
|
|
|
* to change anything in here, edit sre_constants.py and run it.
|
|
|
|
*
|
|
|
|
* Copyright (c) 1997-2000 by Secret Labs AB. All rights reserved.
|
|
|
|
*
|
|
|
|
* See the _sre.c file for information on usage and redistribution.
|
|
|
|
*/
|
|
|
|
|
|
|
|
""")
|
|
|
|
|
2000-06-29 16:35:29 -03:00
|
|
|
dump(f, OPCODES, "SRE_OP")
|
|
|
|
dump(f, ATCODES, "SRE")
|
|
|
|
dump(f, CHCODES, "SRE")
|
|
|
|
f.write("#define SRE_FLAG_TEMPLATE %d\n" % SRE_FLAG_TEMPLATE)
|
|
|
|
f.write("#define SRE_FLAG_IGNORECASE %d\n" % SRE_FLAG_IGNORECASE)
|
|
|
|
f.write("#define SRE_FLAG_LOCALE %d\n" % SRE_FLAG_LOCALE)
|
|
|
|
f.write("#define SRE_FLAG_MULTILINE %d\n" % SRE_FLAG_MULTILINE)
|
|
|
|
f.write("#define SRE_FLAG_DOTALL %d\n" % SRE_FLAG_DOTALL)
|
|
|
|
f.write("#define SRE_FLAG_UNICODE %d\n" % SRE_FLAG_UNICODE)
|
|
|
|
f.write("#define SRE_FLAG_VERBOSE %d\n" % SRE_FLAG_VERBOSE)
|
2000-05-08 14:31:04 -03:00
|
|
|
f.close()
|
|
|
|
print "done"
|