summaryrefslogtreecommitdiffstats
path: root/Lib/sre_compile.py
blob: 53da005e763a727057281ff4ba9f7cc0079b4dec (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
#
# Secret Labs' Regular Expression Engine
# $Id$
#
# convert template to internal format
#
# Copyright (c) 1997-2000 by Secret Labs AB.  All rights reserved.
#
# This code can only be used for 1.6 alpha testing.  All other use
# require explicit permission from Secret Labs AB.
#
# Portions of this engine have been developed in cooperation with
# CNRI.  Hewlett-Packard provided funding for 1.6 integration and
# other compatibility work.
#

import array, string, sys

import _sre

from sre_constants import *

# find an array type code that matches the engine's code size
for WORDSIZE in "BHil":
    if len(array.array(WORDSIZE, [0]).tostring()) == _sre.getcodesize():
	break
else:
    raise RuntimeError, "cannot find a useable array type"

# FIXME: <fl> should move some optimizations from the parser to here!

class Code:
    def __init__(self):
	self.data = []
    def __len__(self):
	return len(self.data)
    def __getitem__(self, index):
	return self.data[index]
    def __setitem__(self, index, code):
	self.data[index] = code
    def append(self, code):
	self.data.append(code)
    def todata(self):
	# print self.data
	try:
	    return array.array(WORDSIZE, self.data).tostring()
	except OverflowError:
	    print self.data
	    raise

def _compile(code, pattern, flags, level=0):
    append = code.append
    for op, av in pattern:
	if op is ANY:
	    if flags & SRE_FLAG_DOTALL:
		append(OPCODES[op]) # any character at all!
	    else:
		append(OPCODES[CATEGORY])
		append(CHCODES[CATEGORY_NOT_LINEBREAK])
	elif op in (SUCCESS, FAILURE):
	    append(OPCODES[op])
	elif op is AT:
	    append(OPCODES[op])
	    if flags & SRE_FLAG_MULTILINE:
		append(ATCODES[AT_MULTILINE[av]])
	    else:
		append(ATCODES[av])
	elif op is BRANCH:
	    append(OPCODES[op])
	    tail = []
	    for av in av[1]:
		skip = len(code); append(0)
		_compile(code, av, flags, level)
		append(OPCODES[JUMP])
		tail.append(len(code)); append(0)
		code[skip] = len(code) - skip
	    append(0) # end of branch
	    for tail in tail:
		code[tail] = len(code) - tail
	elif op is CALL:
	    append(OPCODES[op])
	    skip = len(code); append(0)
	    _compile(code, av, flags, level+1)
	    append(OPCODES[SUCCESS])
	    code[skip] = len(code) - skip
	elif op is CATEGORY: # not used by current parser
	    append(OPCODES[op])
	    if flags & SRE_FLAG_LOCALE:
		append(CH_LOCALE[CHCODES[av]])
	    else:
		append(CHCODES[av])
	elif op is GROUP:
	    if flags & SRE_FLAG_IGNORECASE:
		append(OPCODES[OP_IGNORE[op]])
	    else:
		append(OPCODES[op])
	    append(av-1)
	elif op is IN:
	    if flags & SRE_FLAG_IGNORECASE:
		append(OPCODES[OP_IGNORE[op]])
		def fixup(literal):
		    return ord(literal.lower())
	    else:
		append(OPCODES[op])
		fixup = ord
	    skip = len(code); append(0)
	    for op, av in av:
		append(OPCODES[op])
		if op is NEGATE:
		    pass
		elif op is LITERAL:
		    append(fixup(av))
		elif op is RANGE:
		    append(fixup(av[0]))
		    append(fixup(av[1]))
		elif op is CATEGORY:
		    if flags & SRE_FLAG_LOCALE:
			append(CH_LOCALE[CHCODES[av]])
		    else:
			append(CHCODES[av])
		else:
		    raise ValueError, "unsupported set operator"
	    append(OPCODES[FAILURE])
	    code[skip] = len(code) - skip
	elif op in (LITERAL, NOT_LITERAL):
	    if flags & SRE_FLAG_IGNORECASE:
		append(OPCODES[OP_IGNORE[op]])
		append(ord(av.lower()))
	    else:
		append(OPCODES[op])
		append(ord(av))
	elif op is MARK:
	    append(OPCODES[op])
	    append(av)
 	elif op in (REPEAT, MIN_REPEAT, MAX_REPEAT):
	    lo, hi = av[2].getwidth()
 	    if lo == 0:
 		raise SyntaxError, "cannot repeat zero-width items"
	    if lo == hi == 1 and op is MAX_REPEAT:
		append(OPCODES[MAX_REPEAT_ONE])
		skip = len(code); append(0)
		append(av[0])
		append(av[1])
		_compile(code, av[2], flags, level+1)
		append(OPCODES[SUCCESS])
		code[skip] = len(code) - skip
	    else:
		append(OPCODES[op])
		skip = len(code); append(0)
		append(av[0])
		append(av[1])
		_compile(code, av[2], flags, level+1)
		if op is MIN_REPEAT:
		    append(OPCODES[MIN_UNTIL])
		else:
		    append(OPCODES[MAX_UNTIL])
		code[skip] = len(code) - skip
	elif op is SUBPATTERN:
 	    group = av[0]
 	    if group:
 		append(OPCODES[MARK])
 		append((group-1)*2)
	    _compile(code, av[1], flags, level+1)
 	    if group:
 		append(OPCODES[MARK])
 		append((group-1)*2+1)
	else:
	    raise ValueError, ("unsupported operand type", op)

def compile(p, flags=0):
    # convert pattern list to internal format
    if type(p) in (type(""), type(u"")):
	import sre_parse
	pattern = p
	p = sre_parse.parse(p)
    else:
	pattern = None
    flags = p.pattern.flags | flags
    code = Code()
    _compile(code, p.data, flags)
    code.append(OPCODES[SUCCESS])
    data = code.todata()
    if 0: # debugging
	print
	print "-" * 68
	import sre_disasm
	sre_disasm.disasm(data)
	print "-" * 68
    return _sre.compile(
	pattern, flags,
	data,
	p.pattern.groups-1, p.pattern.groupdict
	)