1
|
# XPath parsing
|
2
|
|
3
|
import copy
|
4
|
|
5
|
from Parser import Parser
|
6
|
|
7
|
class XpathElem:
|
8
|
def __init__(self, name, value=None, attrs=None, is_attr=False,
|
9
|
is_ptr=False):
|
10
|
if attrs == None: attrs = []
|
11
|
self.name = name
|
12
|
self.value = value
|
13
|
self.attrs = attrs
|
14
|
self.is_attr = is_attr
|
15
|
self.is_ptr = is_ptr
|
16
|
|
17
|
def __repr__(self):
|
18
|
str_ = ''
|
19
|
if self.is_attr: str_ += '@'
|
20
|
str_ += self.name
|
21
|
if self.attrs != []: str_ += repr(self.attrs)
|
22
|
if self.value != None: str_ += '='+repr(self.value)
|
23
|
if self.is_ptr: str_ += '->'
|
24
|
return str_
|
25
|
|
26
|
def __eq__(self, other): return self.__dict__ == other.__dict__
|
27
|
|
28
|
def value(path): return path[-1].value
|
29
|
|
30
|
def set_value(path, value): path[-1].value = value
|
31
|
|
32
|
def backward_id(elem):
|
33
|
if len(elem.attrs) >= 1 and value(elem.attrs[0]) == None and\
|
34
|
len(elem.attrs[0]) == 1: return elem.attrs[0]
|
35
|
else: return None
|
36
|
|
37
|
class XpathParser(Parser):
|
38
|
def _main(self):
|
39
|
self._match_str('/') # optional leading /
|
40
|
return self._path()
|
41
|
|
42
|
def _path(self):
|
43
|
tree = []
|
44
|
trailing_slash = False
|
45
|
while True:
|
46
|
# Split path
|
47
|
if self._match_str('{'):
|
48
|
paths = []
|
49
|
while True:
|
50
|
paths.append(tree + self._path())
|
51
|
if not self._match_str(','): break
|
52
|
self._match_str('}', required=True)
|
53
|
tree = paths[0] # just use first subpath for now
|
54
|
break # nothing allowed after split path
|
55
|
|
56
|
elem = XpathElem(is_attr=self._match_str('@'),
|
57
|
name=self._match_re(r'[\w.*]+', required=True))
|
58
|
|
59
|
# Attrs
|
60
|
if self._match_str('['):
|
61
|
elem.attrs = []
|
62
|
while True:
|
63
|
path = self._path()
|
64
|
if self._match_str('='):
|
65
|
set_value(path, self._match_re(r'[\w.|]*'))
|
66
|
elem.attrs.append(path)
|
67
|
if not self._match_str(','): break
|
68
|
self._match_str(']', required=True)
|
69
|
|
70
|
elem.is_ptr = self._match_str('->')
|
71
|
tree.append(elem)
|
72
|
|
73
|
# Lookahead assertion
|
74
|
if self._match_str('('):
|
75
|
self._match_str('/', required=True) # next / is inside ()
|
76
|
path = self._path()
|
77
|
self._match_str(')', required=True)
|
78
|
elem.attrs.append(path)
|
79
|
tree += path
|
80
|
|
81
|
if not self._match_str('/'): break
|
82
|
|
83
|
# Expand * abbrs
|
84
|
for i in reversed(xrange(len(tree))):
|
85
|
elem = tree[i]
|
86
|
id_ = backward_id(elem)
|
87
|
if id_ != None: elem = id_[0]; offset = -2
|
88
|
elif elem.is_ptr: offset = 2
|
89
|
else: offset = 1
|
90
|
before, abbr, after = elem.name.partition('*')
|
91
|
if abbr != '':
|
92
|
try: elem.name = before+tree[i+offset].name+after
|
93
|
except IndexError: pass # no replacement elem
|
94
|
|
95
|
return tree
|
96
|
|
97
|
def parse(string): return XpathParser(string).parse()
|
98
|
|
99
|
instance_level = 1
|
100
|
|
101
|
def obj(path):
|
102
|
obj_path = copy.deepcopy(path[:instance_level+1])
|
103
|
obj_path[-1].is_ptr = False # prevent pointer w/o target
|
104
|
return obj_path
|
105
|
|
106
|
def set_id(path, id_, has_types=True):
|
107
|
if has_types: id_level = instance_level
|
108
|
else: id_level = 0
|
109
|
path[id_level].attrs.append([XpathElem('id', id_, is_attr=True)])
|
110
|
|
111
|
def is_id(path): return path[0].is_attr and path[0].name == 'id'
|
112
|
|
113
|
def is_instance(elem): return elem.attrs != [] and is_id(elem.attrs[0])
|