# SPDX-License-Identifier: MIT
# Emitter expects events obeying the following grammar:
# stream ::= STREAM-START document* STREAM-END
# document ::= DOCUMENT-START node DOCUMENT-END
# node ::= SCALAR | sequence | mapping
# sequence ::= SEQUENCE-START node* SEQUENCE-END
# mapping ::= MAPPING-START (node node)* MAPPING-END
__all__ = ['Emitter', 'EmitterError']
from .error import YAMLError
from .events import *
class EmitterError(YAMLError):
pass
class ScalarAnalysis:
def __init__(self, scalar, empty, multiline,
allow_flow_plain, allow_block_plain,
allow_single_quoted, allow_double_quoted,
allow_block):
self.scalar = scalar
self.empty = empty
self.multiline = multiline
self.allow_flow_plain = allow_flow_plain
self.allow_block_plain = allow_block_plain
self.allow_single_quoted = allow_single_quoted
self.allow_double_quoted = allow_double_quoted
self.allow_block = allow_block
class Emitter:
DEFAULT_TAG_PREFIXES = {
'!' : '!',
'tag:yaml.org,2002:' : '!!',
}
def __init__(self, stream, canonical=None, indent=None, width=None,
allow_unicode=None, line_break=None):
# The stream should have the methods `write` and possibly `flush`.
self.stream = stream
# Encoding can be overriden by STREAM-START.
self.encoding = None
# Emitter is a state machine with a stack of states to handle nested
# structures.
self.states = []
self.state = self.expect_stream_start
# Current event and the event queue.
self.events = []
self.event = None
# The current indentation level and the stack of previous indents.
self.indents = []
self.indent = None
# Flow level.
self.flow_level = 0
# Contexts.
self.root_context = False
self.sequence_context = False
self.mapping_context = False
self.simple_key_context = False
# Characteristics of the last emitted character:
# - current position.
# - is it a whitespace?
# - is it an indention character
# (indentation space, '-', '?', or ':')?
self.line = 0
self.column = 0
self.whitespace = True
self.indention = True
# Whether the document requires an explicit document indicator
self.open_ended = False
# Formatting details.
self.canonical = canonical
self.allow_unicode = allow_unicode
self.best_indent = 2
if indent and 1 < indent < 10:
self.best_indent = indent
self.best_width = 80
if width and width > self.best_indent*2:
self.best_width = width
self.best_line_break = '\n'
if line_break in ['\r', '\n', '\r\n']:
self.best_line_break = line_break
# Tag prefixes.
self.tag_prefixes = None
# Prepared anchor and tag.
self.prepared_anchor = None
self.prepared_tag = None
# Scalar analysis and style.
self.analysis = None
self.style = None
def dispose(self):
# Reset the state attributes (to clear self-references)
self.states = []
self.state = None
def emit(self, event):
self.events.append(event)
while not self.need_more_events():
self.event = self.events.pop(0)
self.state()
self.event = None
# In some cases, we wait for a few next events before emitting.
def need_more_events(self):
if not self.events:
return True
event = self.events[0]
if isinstance(event, DocumentStartEvent):
return self.need_events(1)
elif isinstance(event, SequenceStartEvent):
return self.need_events(2)
elif isinstance(event, MappingStartEvent):
return self.need_events(3)
else:
return False
def need_events(self, count):
level = 0
for event in self.events[1:]:
if isinstance(event, (DocumentStartEvent, CollectionStartEvent)):
level += 1
elif isinstance(event, (DocumentEndEvent, CollectionEndEvent)):
level -= 1
elif isinstance(event, StreamEndEvent):
level = -1
if level < 0:
return False
return (len(self.events) < count+1)
def increase_indent(self, flow=False, indentless=False):
self.indents.append(self.indent)
if self.indent is None:
if flow:
self.indent = self.best_indent
else:
self.indent = 0
elif not indentless:
self.indent += self.best_indent
# States.
# Stream handlers.
def expect_stream_start(self):
if isinstance(self.event, StreamStartEvent):
if self.event.encoding and not hasattr(self.stream, 'encoding'):
self.encoding = self.event.encoding
self.write_stream_start()
self.state = self.expect_first_document_start
else:
raise EmitterError("expected StreamStartEvent, but got %s"
% self.event)
def expect_nothing(self):
raise EmitterError("expected nothing, but got %s" % self.event)
# Document handlers.
def expect_first_document_start(self):
return self.expect_document_start(first=True)
def expect_document_start(self, first=False):
if isinstance(self.event, DocumentStartEvent):
if (self.event.version or self.event.tags) and self.open_ended:
self.write_indicator('...', True)
self.write_indent()
if self.event.version:
version_text = self.prepare_version(self.event.version)
self.write_version_directive(version_text)
self.tag_prefixes = self.DEFAULT_TAG_PREFIXES.copy()
if self.event.tags:
handles = sorted(self.event.tags.keys())
for handle in handles:
prefix = self.event.tags[handle]
self.tag_prefixes[prefix] = handle
handle_text = self.prepare_tag_handle(handle)
prefix_text = self.prepare_tag_prefix(prefix)
self.write_tag_directive(handle_text, prefix_text)
implicit = (first and not self.event.explicit and not self.canonical
and not self.event.version and not self.event.tags
and not self.check_empty_document())
if not implicit:
self.write_indent()
self.write_indicator('---', True)
if self.canonical:
self.write_indent()
self.state = self.expect_document_root
elif isinstance(self.event, StreamEndEvent):
if self.open_ended:
self.write_indicator('...', True)
self.write_indent()
self.write_stream_end()
self.state = self.expect_nothing
else:
raise EmitterError("expected DocumentStartEvent, but got %s"
% self.event)
def expect_document_end(self):
if isinstance(self.event, DocumentEndEvent):
self.write_indent()
if self.event.explicit:
self.write_indicator('...', True)
self.write_indent()
self.flush_stream()
self.state = self.expect_document_start
else:
raise EmitterError("expected DocumentEndEvent, but got %s"
% self.event)
def expect_document_root(self):
self.states.append(self.expect_document_end)
self.expect_node(root=True)
# Node handlers.
def expect_node(self, root=False, sequence=False, mapping=False,
simple_key=False):
self.root_context = root
self.sequence_context = sequence
self.mapping_context = mapping
self.simple_key_context = simple_key
if isinstance(self.event, AliasEvent):
self.expect_alias()
elif isinstance(self.event, (ScalarEvent, CollectionStartEvent)):
self.process_anchor('&')
self.process_tag()
if isinstance(self.event, ScalarEvent):
self.expect_scalar()
elif isinstance(self.event, SequenceStartEvent):
if self.flow_level or self.canonical or self.event.flow_style \
or self.check_empty_sequence():
self.expect_flow_sequence()
else:
self.expect_block_sequence()
elif isinstance(self.event, MappingStartEvent):
if self.flow_level or self.canonical or self.event.flow_style \
or self.check_empty_mapping():
self.expect_flow_mapping()
else:
self.expect_block_mapping()
else:
raise EmitterError("expected NodeEvent, but got %s" % self.event)
def expect_alias(self):
if self.event.anchor is None:
raise EmitterError("anchor is not specified for alias")
self.process_anchor('*')
self.state = self.states.pop()
def expect_scalar(self):
self.increase_indent(flow=True)
self.process_scalar()
self.indent = self.indents.pop()
self.state = self.states.pop()
# Flow sequence handlers.
def expect_flow_sequence(self):
self.write_indicator('[', True, whitespace=True)
self.flow_level += 1
self.increase_indent(flow=True)
self.state = self.expect_first_flow_sequence_item
def expect_first_flow_sequence_item(self):
if isinstance(self.event, SequenceEndEvent):
self.indent = self.indents.pop()
self.flow_level -= 1
self.write_indicator(']', False)
self.state = self.states.pop()
else:
if self.canonical or self.column > self.best_width:
self.write_indent()
self.states.append(self.expect_flow_sequence_item)
self.expect_node(sequence=True)
def expect_flow_sequence_item(self):
if isinstance(self.event, SequenceEndEvent):
self.indent = self.indents.pop()
self.flow_level -= 1
if self.canonical:
self.write_indicator(',', False)
self.write_indent()
self.write_indicator(']', False)
self.state = self.states.pop()
else:
self.write_indicator(',', False)
if self.canonical or self.column > self.best_width:
self.write_indent()
self.states.append(self.expect_flow_sequence_item)
self.expect_node(sequence=True)
# Flow mapping handlers.
def expect_flow_mapping(self):
self.write_indicator('{', True, whitespace=True)
self.flow_level += 1
self.increase_indent(flow=True)
self.state = self.expect_first_flow_mapping_key
def expect_first_flow_mapping_key(self):
if isinstance(self.event, MappingEndEvent):
self.indent = self.indents.pop()
self.flow_level -= 1
self.write_indicator('}', False)
self.state = self.states.pop()
else:
if self.canonical or self.column > self.best_width:
self.write_indent()
if not self.canonical and self.check_simple_key():
self.states.append(self.expect_flow_mapping_simple_value)
self.expect_node(mapping=True, simple_key=True)
else:
self.write_indicator('?', True)
self.states.append(self.expect_flow_mapping_value)
self.expect_node(mapping=True)
def expect_flow_mapping_key(self):
if isinstance(self.event, MappingEndEvent):
self.indent = self.indents.pop()
self.flow_level -= 1
if self.canonical:
self.write_indicator(',', False)
self.write_indent()
self.write_indicator('}', False)
self.state = self.states.pop()
else:
self.write_indicator(',', False)
if self.canonical or self.column > self.best_width:
self.write_indent()
if not self.canonical and self.check_simple_key():
self.states.append(self.expect_flow_mapping_simple_value)
self.expect_node(mapping=True, simple_key=True)
else:
self.write_indicator('?', True)
self.states.append(self.expect_flow_mapping_value)
self.expect_node(mapping=True)
def expect_flow_mapping_simple_value(self):
self.write_indicator(':', False)
self.states.append(self.expect_flow_mapping_key)
self.expect_node(mapping=True)
def expect_flow_mapping_value(self):
if self.canonical or self.column > self.best_width:
self.write_indent()
self.write_indicator(':', True)
self.states.append(self.expect_flow_mapping_key)
self.expect_node(mapping=True)
# Block sequence handlers.
def expect_block_sequence(self):
indentless = (self.mapping_context and not self.indention)
self.increase_indent(flow=False, indentless=indentless)
self.state = self.expect_first_block_sequence_item
def expect_first_block_sequence_item(self):
return self.expect_block_sequence_item(first=True)
def expect_block_sequence_item(self, first=False):
if not first and isinstance(self.event, SequenceEndEvent):
self.indent = self.indents.pop()
self.state = self.states.pop()
else:
self.write_indent()
self.write_indicator('-', True, indention=True)
self.states.append(self.expect_block_sequence_item)
self.expect_node(sequence=True)
# Block mapping handlers.
def expect_block_mapping(self):
self.increase_indent(flow=False)
self.state = self.expect_first_block_mapping_key
def expect_first_block_mapping_key(self):
return self.expect_block_mapping_key(first=True)
def expect_block_mapping_key(self, first=False):
if not first and isinstance(self.event, MappingEndEvent):
self.indent = self.indents.pop()
self.state = self.states.pop()
else:
self.write_indent()
if self.check_simple_key():
self.states.append(self.expect_block_mapping_simple_value)
self.expect_node(mapping=True, simple_key=True)
else:
self.write_indicator('?', True, indention=True)
self.states.append(self.expect_block_mapping_value)
self.expect_node(mapping=True)
def expect_block_mapping_simple_value(self):
self.write_indicator(':', False)
self.states.append(self.expect_block_mapping_key)
self.expect_node(mapping=True)
def expect_block_mapping_value(self):
self.write_indent()
self.write_indicator(':', True, indention=True)
self.states.append(self.expect_block_mapping_key)
self.expect_node(mapping=True)
# Checkers.
def check_empty_sequence(self):
return (isinstance(self.event, SequenceStartEvent) and self.events
and isinstance(self.events[0], SequenceEndEvent))
def check_empty_mapping(self):
return (isinstance(self.event, MappingStartEvent) and self.events
and isinstance(self.events[0], MappingEndEvent))
def check_empty_document(self):
if not isinstance(self.event, DocumentStartEvent) or not self.events:
return False
event = self.events[0]
return (isinstance(event, ScalarEvent) and event.anchor is None
and event.tag is None and event.implicit and event.value == '')
def check_simple_key(self):
length = 0
if isinstance(self.event, NodeEvent) and self.event.anchor is not None:
if self.prepared_anchor is None:
self.prepared_anchor = self.prepare_anchor(self.event.anchor)
length += len(self.prepared_anchor)
if isinstance(self.event, (ScalarEvent, CollectionStartEvent)) \
and self.event.tag is not None:
if self.prepared_tag is None:
self.prepared_tag = self.prepare_tag(self.event.tag)
length += len(self.prepared_tag)
if isinstance(self.event, ScalarEvent):
if self.analysis is None:
self.analysis = self.analyze_scalar(self.event.value)
length += len(self.analysis.scalar)
return (length < 128 and (isinstance(self.event, AliasEvent)
or (isinstance(self.event, ScalarEvent)
and not self.analysis.empty and not self.analysis.multiline)
or self.check_empty_sequence() or self.check_empty_mapping()))
# Anchor, Tag, and Scalar processors.
def process_anchor(self, indicator):
if self.event.anchor is None:
self.prepared_anchor = None
return
if self.prepared_anchor is None:
self.prepared_anchor = self.prepare_anchor(self.event.anchor)
if self.prepared_anchor:
self.write_indicator(indicator+self.prepared_anchor, True)
self.prepared_anchor = None
def process_tag(self):
tag = self.event.tag
if isinstance(self.event, ScalarEvent):
if self.style is None:
self.style = self.choose_scalar_style()
if ((not self.canonical or tag is None) and
((self.style == '' and self.event.implicit[0])
or (self.style != '' and self.event.implicit[1]))):
self.prepared_tag = None
return
if self.event.implicit[0] and tag is None:
tag = '!'
self.prepared_tag = None
else:
if (not self.canonical or tag is None) and self.event.implicit:
self.prepared_tag = None
return
if tag is None:
raise EmitterError("tag is not specified")
if self.prepared_tag is None:
self.prepared_tag = self.prepare_tag(tag)
if self.prepared_tag:
self.write_indicator(self.prepared_tag, True)
self.prepared_tag = None
def choose_scalar_style(self):
if self.analysis is None:
self.analysis = self.analyze_scalar(self.event.value)
if self.event.style == '"' or self.canonical:
return '"'
if not self.event.style and self.event.implicit[0]:
if (not (self.simple_key_context and
(self.analysis.empty or self.analysis.multiline))
and (self.flow_level and self.analysis.allow_flow_plain
or (not self.flow_level and self.analysis.allow_block_plain))):
return ''
if self.event.style and self.event.style in '|>':
if (not self.flow_level and not self.simple_key_context
and self.analysis.allow_block):
return self.event.style
if not self.event.style or self.event.style == '\'':
if (self.analysis.allow_single_quoted and
not (self.simple_key_context and self.analysis.multiline)):
return '\''
return '"'
def process_scalar(self):
if self.analysis is None:
self.analysis = self.analyze_scalar(self.event.value)
if self.style is None:
self.style = self.choose_scalar_style()
split = (not self.simple_key_context)
#if self.analysis.multiline and split \
# and (not self.style or self.style in '\'\"'):
# self.write_indent()
if self.style == '"':
self.write_double_quoted(self.analysis.scalar, split)
elif self.style == '\'':
self.write_single_quoted(self.analysis.scalar, split)
elif self.style == '>':
self.write_folded(self.analysis.scalar)
elif self.style == '|':
self.write_literal(self.analysis.scalar)
else:
self.write_plain(self.analysis.scalar, split)
self.analysis = None
self.style = None
# Analyzers.
def prepare_version(self, version):
major, minor = version
if major != 1:
raise EmitterError("unsupported YAML version: %d.%d" % (major, minor))
return '%d.%d' % (major, minor)
def prepare_tag_handle(self, handle):
if not handle:
raise EmitterError("tag handle must not be empty")
if handle[0] != '!' or handle[-1] != '!':
raise EmitterError("tag handle must start and end with '!': %r" % handle)
for ch in handle[1:-1]:
if not ('0'