class Serializer(object):
# 'id' and 3+ numbers, but not 000
ANCHOR_TEMPLATE = u'id%03d'
ANCHOR_RE = RegExp(u'id(?!000$)\\d{3,}')
def __init__(
self,
encoding=None,
explicit_start=None,
explicit_end=None,
version=None,
tags=None,
dumper=None,
):
# type: (Any, Optional[bool], Optional[bool], Optional[VersionType], Any, Any) -> None # NOQA
self.dumper = dumper
if self.dumper is not None:
self.dumper._serializer = self
self.use_encoding = encoding
self.use_explicit_start = explicit_start
self.use_explicit_end = explicit_end
if isinstance(version, string_types):
self.use_version = tuple(map(int, version.split('.')))
else:
self.use_version = version # type: ignore
self.use_tags = tags
self.serialized_nodes = {} # type: Dict[Any, Any]
self.anchors = {} # type: Dict[Any, Any]
self.last_anchor_id = 0
self.closed = None # type: Optional[bool]
self._templated_id = None
@property
def emitter(self):
# type: () -> Any
if hasattr(self.dumper, 'typ'):
return self.dumper.emitter
return self.dumper._emitter
@property
def resolver(self):
# type: () -> Any
if hasattr(self.dumper, 'typ'):
self.dumper.resolver
return self.dumper._resolver
def open(self):
# type: () -> None
if self.closed is None:
self.emitter.emit(StreamStartEvent(encoding=self.use_encoding))
self.closed = False
elif self.closed:
raise SerializerError('serializer is closed')
else:
raise SerializerError('serializer is already opened')
def close(self):
# type: () -> None
if self.closed is None:
raise SerializerError('serializer is not opened')
elif not self.closed:
self.emitter.emit(StreamEndEvent())
self.closed = True
# def __del__(self):
# self.close()
def serialize(self, node):
# type: (Any) -> None
if dbg(DBG_NODE):
nprint('Serializing nodes')
node.dump()
if self.closed is None:
raise SerializerError('serializer is not opened')
elif self.closed:
raise SerializerError('serializer is closed')
self.emitter.emit(
DocumentStartEvent(
explicit=self.use_explicit_start, version=self.use_version, tags=self.use_tags
)
)
self.anchor_node(node)
self.serialize_node(node, None, None)
self.emitter.emit(DocumentEndEvent(explicit=self.use_explicit_end))
self.serialized_nodes = {}
self.anchors = {}
self.last_anchor_id = 0
def anchor_node(self, node):
# type: (Any) -> None
if node in self.anchors:
if self.anchors[node] is None:
self.anchors[node] = self.generate_anchor(node)
else:
anchor = None
try:
if node.anchor.always_dump:
anchor = node.anchor.value
except: # NOQA
pass
self.anchors[node] = anchor
if isinstance(node, SequenceNode):
for item in node.value:
self.anchor_node(item)
elif isinstance(node, MappingNode):
for key, value in node.value:
self.anchor_node(key)
self.anchor_node(value)
def generate_anchor(self, node):
# type: (Any) -> Any
try:
anchor = node.anchor.value
except: # NOQA
anchor = None
if anchor is None:
self.last_anchor_id += 1
return self.ANCHOR_TEMPLATE % self.last_anchor_id
return anchor
def serialize_node(self, node, parent, index):
# type: (Any, Any, Any) -> None
alias = self.anchors[node]
if node in self.serialized_nodes:
self.emitter.emit(AliasEvent(alias))
else:
self.serialized_nodes[node] = True
self.resolver.descend_resolver(parent, index)
if isinstance(node, ScalarNode):
# here check if the node.tag equals the one that would result from parsing
# if not equal quoting is necessary for strings
detected_tag = self.resolver.resolve(ScalarNode, node.value, (True, False))
default_tag = self.resolver.resolve(ScalarNode, node.value, (False, True))
implicit = (
(node.tag == detected_tag),
(node.tag == default_tag),
node.tag.startswith('tag:yaml.org,2002:'),
)
self.emitter.emit(
ScalarEvent(
alias,
node.tag,
implicit,
node.value,
style=node.style,
comment=node.comment,
)
)
elif isinstance(node, SequenceNode):
implicit = node.tag == self.resolver.resolve(SequenceNode, node.value, True)
comment = node.comment
end_comment = None
seq_comment = None
if node.flow_style is True:
if comment: # eol comment on flow style sequence
seq_comment = comment[0]
# comment[0] = None
if comment and len(comment) > 2:
end_comment = comment[2]
else:
end_comment = None
self.emitter.emit(
SequenceStartEvent(
alias,
node.tag,
implicit,
flow_style=node.flow_style,
comment=node.comment,
)
)
index = 0
for item in node.value:
self.serialize_node(item, node, index)
index += 1
self.emitter.emit(SequenceEndEvent(comment=[seq_comment, end_comment]))
elif isinstance(node, MappingNode):
implicit = node.tag == self.resolver.resolve(MappingNode, node.value, True)
comment = node.comment
end_comment = None
map_comment = None
if node.flow_style is True:
if comment: # eol comment on flow style sequence
map_comment = comment[0]
# comment[0] = None
if comment and len(comment) > 2:
end_comment = comment[2]
self.emitter.emit(
MappingStartEvent(
alias,
node.tag,
implicit,
flow_style=node.flow_style,
comment=node.comment,
nr_items=len(node.value),
)
)
for key, value in node.value:
self.serialize_node(key, node, None)
self.serialize_node(value, node, key)
self.emitter.emit(MappingEndEvent(comment=[map_comment, end_comment]))
self.resolver.ascend_resolver()
from google.appengine._internal.ruamel.yaml.nodes import MappingNode, ScalarNode, SequenceNode # NOQA
from google.appengine._internal.ruamel.yaml.util import RegExp # NOQA
__all__ = ['BaseResolver', 'Resolver', 'VersionedResolver']
# fmt: off
# resolvers consist of
# - a list of applicable version
# - a tag
# - a regexp
# - a list of first characters to match
implicit_resolvers = [
([(1, 2)],
u'tag:yaml.org,2002:bool',
RegExp(u'''^(?:true|True|TRUE|false|False|FALSE)$''', re.X),
list(u'tTfF')),
([(1, 1)],
u'tag:yaml.org,2002:bool',
RegExp(u'''^(?:yes|Yes|YES|no|No|NO
|true|True|TRUE|false|False|FALSE
|on|On|ON|off|Off|OFF)$''', re.X),
list(u'yYnNtTfFoO')),
([(1, 2)],
u'tag:yaml.org,2002:float',
RegExp(u'''^(?:
[-+]?(?:[0-9][0-9_]*)\\.[0-9_]*(?:[eE][-+]?[0-9]+)?
|[-+]?(?:[0-9][0-9_]*)(?:[eE][-+]?[0-9]+)
|[-+]?\\.[0-9_]+(?:[eE][-+][0-9]+)?
|[-+]?\\.(?:inf|Inf|INF)
|\\.(?:nan|NaN|NAN))$''', re.X),
class Reader(object):
# Reader:
# - determines the data encoding and converts it to a unicode string,
# - checks if characters are in allowed range,
# - adds '\0' to the end.
# Reader accepts
# - a `str` object (PY2) / a `bytes` object (PY3),
# - a `unicode` object (PY2) / a `str` object (PY3),
# - 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, stream, loader=None):
# type: (Any, Any) -> None
self.loader = loader
if self.loader is not None and getattr(self.loader, '_reader',
None) is None:
self.loader._reader = self
self.reset_reader()
self.stream = stream # type: Any # as .read is called
def reset_reader(self):
# type: () -> None
self.name = None # type: Any
self.stream_pointer = 0
self.eof = True
self.buffer = ""
self.pointer = 0
self.raw_buffer = None # type: Any
self.raw_decode = None
self.encoding = None # type: Optional[Text]
self.index = 0
self.line = 0
self.column = 0
@property
def stream(self):
# type: () -> Any
try:
return self._stream
except AttributeError:
raise YAMLStreamError('input stream needs to specified')
@stream.setter
def stream(self, val):
# type: (Any) -> None
if val is None:
return
self._stream = None
if isinstance(val, text_type):
self.name = '<unicode string>'
self.check_printable(val)
self.buffer = val + u'\0' # type: ignore
elif isinstance(val, binary_type):
self.name = '<byte string>'
self.raw_buffer = val
self.determine_encoding()
else:
if not hasattr(val, 'read'):
raise YAMLStreamError(
'stream argument needs to have a read() method')
self._stream = val
self.name = getattr(self.stream, 'name', '<file>')
self.eof = False
self.raw_buffer = None
self.determine_encoding()
def peek(self, index=0):
# type: (int) -> Text
try:
return self.buffer[self.pointer + index]
except IndexError:
self.update(index + 1)
return self.buffer[self.pointer + index]
def prefix(self, length=1):
# type: (int) -> Any
if self.pointer + length >= len(self.buffer):
self.update(length)
return self.buffer[self.pointer:self.pointer + length]
def forward_1_1(self, length=1):
# type: (int) -> None
if self.pointer + length + 1 >= len(self.buffer):
self.update(length + 1)
while length != 0:
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] != u'\n'):
self.line += 1
self.column = 0
elif ch != u'\uFEFF':
self.column += 1
length -= 1
def forward(self, length=1):
# type: (int) -> None
if self.pointer + length + 1 >= len(self.buffer):
self.update(length + 1)
while length != 0:
ch = self.buffer[self.pointer]
self.pointer += 1
self.index += 1
if ch == u'\n' or (ch == u'\r'
and self.buffer[self.pointer] != u'\n'):
self.line += 1
self.column = 0
elif ch != u'\uFEFF':
self.column += 1
length -= 1
def get_mark(self):
# type: () -> Any
if self.stream is None:
return StringMark(self.name, self.index, self.line, self.column,
self.buffer, self.pointer)
else:
return FileMark(self.name, self.index, self.line, self.column)
def determine_encoding(self):
# type: () -> None
while not self.eof and (self.raw_buffer is None
or len(self.raw_buffer) < 2):
self.update_raw()
if isinstance(self.raw_buffer, binary_type):
if self.raw_buffer.startswith(codecs.BOM_UTF16_LE):
self.raw_decode = codecs.utf_16_le_decode # type: ignore
self.encoding = 'utf-16-le'
elif self.raw_buffer.startswith(codecs.BOM_UTF16_BE):
self.raw_decode = codecs.utf_16_be_decode # type: ignore
self.encoding = 'utf-16-be'
else:
self.raw_decode = codecs.utf_8_decode # type: ignore
self.encoding = 'utf-8'
self.update(1)
if UNICODE_SIZE == 2:
NON_PRINTABLE = RegExp(u'[^\x09\x0A\x0D\x20-\x7E\x85'
u'\xA0-\uD7FF'
u'\uE000-\uFFFD'
u']')
else:
NON_PRINTABLE = RegExp(u'[^\x09\x0A\x0D\x20-\x7E\x85'
u'\xA0-\uD7FF'
u'\uE000-\uFFFD'
u'\U00010000-\U0010FFFF'
u']')
_printable_ascii = ('\x09\x0A\x0D' +
"".join(map(chr, range(0x20, 0x7F)))).encode('ascii')
@classmethod
def _get_non_printable_ascii(cls, data): # type: ignore
# type: (Text, bytes) -> Optional[Tuple[int, Text]]
ascii_bytes = data.encode('ascii')
non_printables = ascii_bytes.translate(
None, cls._printable_ascii) # type: ignore
if not non_printables:
return None
non_printable = non_printables[:1]
return ascii_bytes.index(non_printable), non_printable.decode('ascii')
@classmethod
def _get_non_printable_regex(cls, data):
# type: (Text) -> Optional[Tuple[int, Text]]
match = cls.NON_PRINTABLE.search(data)
if not bool(match):
return None
return match.start(), match.group()
@classmethod
def _get_non_printable(cls, data):
# type: (Text) -> Optional[Tuple[int, Text]]
try:
return cls._get_non_printable_ascii(data) # type: ignore
except UnicodeEncodeError:
return cls._get_non_printable_regex(data)
def check_printable(self, data):
# type: (Any) -> None
non_printable_match = self._get_non_printable(data)
if non_printable_match is not None:
start, character = non_printable_match
position = self.index + (len(self.buffer) - self.pointer) + start
raise ReaderError(
self.name,
position,
ord(character),
'unicode',
'special characters are not allowed',
)
def update(self, length):
# type: (int) -> None
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 as exc:
if PY3:
character = self.raw_buffer[exc.start]
else:
character = exc.object[exc.start]
if self.stream is not None:
position = self.stream_pointer - len(
self.raw_buffer) + exc.start
elif 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 += '\0'
self.raw_buffer = None
break
def update_raw(self, size=None):
# type: (Optional[int]) -> None
if size is None:
size = 4096 if PY3 else 1024
data = self.stream.read(size)
if self.raw_buffer is None:
self.raw_buffer = data
else:
self.raw_buffer += data
self.stream_pointer += len(data)
if not data:
self.eof = True