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
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
|
# This module contains abstractions for the input stream. You don't have to
# looks further, there are no pretty code.
#
# We define two classes here.
#
# Marker(source, line, column)
# It's just a record and its only use is producing nice error messages.
# Parser does not use it for any other purposes.
#
# Reader(source, data)
# Reader determines the encoding of `data` and converts it to unicode.
# Reader provides the following methods and attributes:
# reader.peek(length=1) - return the next `length` characters
# reader.forward(length=1) - move the current position to `length` characters.
# reader.index - the number of the current character.
# reader.line, stream.column - the line and the column of the current character.
__all__ = ['Marker', 'Reader', 'ReaderError']
from error import YAMLError
import codecs, re
# Unfortunately, codec functions in Python 2.3 does not support the `finish`
# arguments, so we have to write our own wrappers.
try:
codecs.utf_8_decode('', 'strict', False)
from codecs import utf_8_decode, utf_16_le_decode, utf_16_be_decode
except TypeError:
def utf_16_le_decode(data, errors, finish=False):
if not finish and len(data) % 2 == 1:
data = data[:-1]
return codecs.utf_16_le_decode(data, errors)
def utf_16_be_decode(data, errors, finish=False):
if not finish and len(data) % 2 == 1:
data = data[:-1]
return codecs.utf_16_be_decode(data, errors)
def utf_8_decode(data, errors, finish=False):
if not finish:
# We are trying to remove a possible incomplete multibyte character
# from the suffix of the data.
# The first byte of a multi-byte sequence is in the range 0xc0 to 0xfd.
# All further bytes are in the range 0x80 to 0xbf.
# UTF-8 encoded UCS characters may be up to six bytes long.
count = 0
while count < 5 and count < len(data) \
and '\x80' <= data[-count-1] <= '\xBF':
count -= 1
if count < 5 and count < len(data) \
and '\xC0' <= data[-count-1] <= '\xFD':
data = data[:-count-1]
return codecs.utf_8_decode(data, errors)
class Marker:
def __init__(self, name, line, column, buffer, pointer):
self.name = name
self.line = line
self.column = column
self.buffer = buffer
self.pointer = pointer
def get_snippet(self, indent=4, max_length=75):
if self.buffer is None:
return None
head = ''
start = self.pointer
while start > 0 and self.buffer[start-1] not in u'\0\r\n\x85\u2028\u2029':
start -= 1
if self.pointer-start > max_length/2-1:
head = ' ... '
start += 5
break
tail = ''
end = self.pointer
while end < len(self.buffer) and self.buffer[end] not in u'\0\r\n\x85\u2028\u2029':
end += 1
if end-self.pointer > max_length/2-1:
tail = ' ... '
end -= 5
break
snippet = self.buffer[start:end].encode('utf-8')
return ' '*indent + head + snippet + tail + '\n' \
+ ' '*(indent+self.pointer-start+len(head)) + '^'
def __str__(self):
snippet = self.get_snippet()
where = " in \"%s\", line %d, column %d" \
% (self.name, self.line+1, self.column+1)
if snippet is not None:
where += ":\n"+snippet
return where
class ReaderError(YAMLError):
def __init__(self, name, position, character, encoding, reason):
self.name = name
self.character = character
self.position = position
self.encoding = encoding
self.reason = reason
def __str__(self):
if isinstance(self.character, str):
return "'%s' codec can't decode byte #x%02x: %s\n" \
" in \"%s\", position %d" \
% (self.encoding, ord(self.character), self.reason,
self.name, self.position)
else:
return "unacceptable character #x%04x: %s\n" \
" in \"%s\", position %d" \
% (ord(self.character), self.reason,
self.name, self.position)
class Reader:
# Reader:
# - determines the data encoding and converts it to unicode,
# - checks if characters are in allowed range,
# - adds '\0' to the end.
# Reader accepts
# - a `str` object,
# - a `unicode` object,
# - a file-like object with its `read` method returning `str`,
# - a file-like object with its `read` method returning `unicode`.
# Yeah, it's ugly and slow.
def __init__(self, data):
self.name = None
self.stream = None
self.stream_pointer = 0
self.eof = True
self.buffer = u''
self.pointer = 0
self.raw_buffer = None
self.raw_decode = None
self.index = 0
self.line = 0
self.column = 0
if isinstance(data, unicode):
self.name = "<unicode string>"
self.check_printable(data)
self.buffer = data+u'\0'
elif isinstance(data, str):
self.name = "<string>"
self.raw_buffer = data
self.determine_encoding()
else:
self.stream = data
self.name = getattr(data, 'name', "<file>")
self.eof = False
self.raw_buffer = ''
self.determine_encoding()
def peek(self, length=1):
if self.pointer+length >= len(self.buffer):
self.update(length)
return self.buffer[self.pointer:self.pointer+length]
def forward(self, length=1):
if self.pointer+length+1 >= len(self.buffer):
self.update(length+1)
for k in range(length):
ch = self.buffer[self.pointer]
self.pointer += 1
self.index += 1
if ch in u'\n\x85\u2028\u2029' \
or (ch == u'\r' and self.buffer[self.pointer+1] != u'\n'):
self.line += 1
self.column = 0
elif ch != u'\uFEFF':
self.column += 1
def get_marker(self):
if self.stream is None:
return Marker(self.name, self.line, self.column,
self.buffer, self.pointer)
else:
return Marker(self.name, self.line, self.column, None, None)
def determine_encoding(self):
while not self.eof and len(self.raw_buffer) < 2:
self.update_raw()
if not isinstance(self.raw_buffer, unicode):
if self.raw_buffer.startswith(codecs.BOM_UTF16_LE):
self.raw_decode = utf_16_le_decode
elif self.raw_buffer.startswith(codecs.BOM_UTF16_BE):
self.raw_decode = utf_16_be_decode
else:
self.raw_decode = utf_8_decode
self.update(1)
NON_PRINTABLE = re.compile(u'[^\x09\x0A\x0D\x20-\x7E\x85\xA0-\uD7FF\uE000-\uFFFD]')
def check_printable(self, data):
match = self.NON_PRINTABLE.search(data)
if match:
character = match.group()
position = self.index+(len(self.buffer)-self.pointer)+match.start()
raise ReaderError(self.name, position, character,
'unicode', "special characters are not allowed")
def update(self, length):
if self.raw_buffer is None:
return
self.buffer = self.buffer[self.pointer:]
self.pointer = 0
while len(self.buffer) < length:
if not self.eof:
self.update_raw()
if self.raw_decode is not None:
try:
data, converted = self.raw_decode(self.raw_buffer,
'strict', self.eof)
except UnicodeDecodeError, exc:
character = exc.object[exc.start]
if self.stream is not None:
position = self.stream_pointer-len(self.raw_buffer)+exc.start
else:
position = exc.start
raise ReaderError(self.name, position, character,
exc.encoding, exc.reason)
else:
data = self.raw_buffer
converted = len(data)
self.check_printable(data)
self.buffer += data
self.raw_buffer = self.raw_buffer[converted:]
if self.eof:
self.buffer += u'\0'
self.raw_buffer = None
break
def update_raw(self, size=1024):
data = self.stream.read(size)
if data:
self.raw_buffer += data
self.stream_pointer += len(data)
else:
self.eof = True
#try:
# import psyco
# psyco.bind(Reader)
#except ImportError:
# pass
|