Add style information to events generated by Parser.

git-svn-id: http://svn.pyyaml.org/pyyaml/trunk@130 18f92427-320e-0410-9341-c67f048884a3

19 files changed, 415 insertions, 87 deletions

diff --git a/examples/yaml-hl/yaml_hl_ascii.cfg b/examples/yaml-hl/yaml_hl_ascii.cfg
index 851d0f8..122f696 100644
--- a/examples/yaml-hl/yaml_hl_ascii.cfg
+++ b/examples/yaml-hl/yaml_hl_ascii.cfg
@@ -27,9 +27,10 @@ events:
     stream-end:     { end: "\e[0m" }
     document-start:
     document-end:
-    sequence:
-    mapping:
-    collection-end:
+    sequence-start:
+    sequence-end:
+    mapping-start:
+    mapping-end:
     scalar:
 
 replaces: !!pairs
diff --git a/examples/yaml-hl/yaml_hl_html.cfg b/examples/yaml-hl/yaml_hl_html.cfg
index 903810e..5eab7af 100644
--- a/examples/yaml-hl/yaml_hl_html.cfg
+++ b/examples/yaml-hl/yaml_hl_html.cfg
@@ -26,9 +26,10 @@ events:
     stream-end:     { end: </pre> }
     document-start: { start: <span class="document"> }
     document-end:   { end: </span> }
-    sequence:       { start: <span class="sequence"> }
-    mapping:        { start: <span class="mapping"> }
-    collection-end: { end: </span> }
+    sequence-start: { start: <span class="sequence"> }
+    sequence-end:   { end: </span> }
+    mapping-start:  { start: <span class="mapping"> }
+    mapping-end:    { end: </span> }
     scalar:         { start: <span class="scalar">, end: </span> }
 
 replaces: !!pairs
diff --git a/lib/yaml/composer.py b/lib/yaml/composer.py
index 9b8c1e2..7152cbd 100644
--- a/lib/yaml/composer.py
+++ b/lib/yaml/composer.py
@@ -71,9 +71,9 @@ class Composer:
             self.all_anchors[anchor] = event
         if self.parser.check(ScalarEvent):
             node = self.compose_scalar_node()
-        elif self.parser.check(SequenceEvent):
+        elif self.parser.check(SequenceStartEvent):
             node = self.compose_sequence_node()
-        elif self.parser.check(MappingEvent):
+        elif self.parser.check(MappingStartEvent):
             node = self.compose_mapping_node()
         if anchor is not None:
             self.complete_anchors[anchor] = node
@@ -87,7 +87,7 @@ class Composer:
     def compose_sequence_node(self):
         start_event = self.parser.get()
         value = []
-        while not self.parser.check(CollectionEndEvent):
+        while not self.parser.check(SequenceEndEvent):
             value.append(self.compose_node())
         end_event = self.parser.get()
         return SequenceNode(start_event.tag, value,
@@ -96,7 +96,7 @@ class Composer:
     def compose_mapping_node(self):
         start_event = self.parser.get()
         value = {}
-        while not self.parser.check(CollectionEndEvent):
+        while not self.parser.check(MappingEndEvent):
             key_event = self.parser.peek()
             item_key = self.compose_node()
             item_value = self.compose_node()
diff --git a/lib/yaml/events.py b/lib/yaml/events.py
index 8837633..f9a695f 100644
--- a/lib/yaml/events.py
+++ b/lib/yaml/events.py
@@ -1,58 +1,89 @@
 
+# Abstract classes.
+
 class Event:
-    def __init__(self, start_mark, end_mark):
+    def __init__(self, start_mark=None, end_mark=None):
         self.start_mark = start_mark
         self.end_mark = end_mark
     def __repr__(self):
-        attributes = [key for key in self.__dict__
-                if not key.endswith('_mark')]
-        attributes.sort()
+        attributes = [key for key in ['anchor', 'tag', 'value']
+                if hasattr(self, key)]
         arguments = ', '.join(['%s=%r' % (key, getattr(self, key))
                 for key in attributes])
         return '%s(%s)' % (self.__class__.__name__, arguments)
 
 class NodeEvent(Event):
-    def __init__(self, anchor, start_mark, end_mark):
+    def __init__(self, anchor, start_mark=None, end_mark=None):
         self.anchor = anchor
         self.start_mark = start_mark
         self.end_mark = end_mark
 
-class AliasEvent(NodeEvent):
-    pass
-
-class ScalarEvent(NodeEvent):
-    def __init__(self, anchor, tag, value, start_mark, end_mark):
+class CollectionStartEvent(NodeEvent):
+    def __init__(self, anchor, tag, start_mark=None, end_mark=None,
+            flow_style=None):
         self.anchor = anchor
         self.tag = tag
-        self.value = value
         self.start_mark = start_mark
         self.end_mark = end_mark
+        self.flow_style = flow_style
 
-class CollectionEvent(NodeEvent):
-    def __init__(self, anchor, tag, start_mark, end_mark):
-        self.anchor = anchor
-        self.tag = tag
+class CollectionEndEvent(Event):
+    pass
+
+# Implementations.
+
+class StreamStartEvent(Event):
+    def __init__(self, start_mark=None, end_mark=None,
+            encoding=None, canonical=None, indent=None, width=None):
         self.start_mark = start_mark
         self.end_mark = end_mark
+        self.encoding = encoding
+        self.canonical = canonical
+        self.indent = indent
+        self.width = width
 
-class SequenceEvent(CollectionEvent):
+class StreamEndEvent(Event):
     pass
 
-class MappingEvent(CollectionEvent):
-    pass
+class DocumentStartEvent(Event):
+    def __init__(self, start_mark=None, end_mark=None,
+            implicit=None, version=None, tags=None):
+        self.start_mark = start_mark
+        self.end_mark = end_mark
+        self.implicit = implicit
+        self.version = version
+        self.tags = tags
 
-class CollectionEndEvent(Event):
+class DocumentEndEvent(Event):
+    def __init__(self, start_mark=None, end_mark=None,
+            implicit=None):
+        self.start_mark = start_mark
+        self.end_mark = end_mark
+        self.implicit = implicit
+
+class AliasEvent(NodeEvent):
     pass
 
-class DocumentStartEvent(Event):
+class ScalarEvent(NodeEvent):
+    def __init__(self, anchor, tag, value, start_mark=None, end_mark=None,
+            implicit=None, style=None):
+        self.anchor = anchor
+        self.tag = tag
+        self.value = value
+        self.start_mark = start_mark
+        self.end_mark = end_mark
+        self.implicit = implicit
+        self.style = style
+
+class SequenceStartEvent(CollectionStartEvent):
     pass
 
-class DocumentEndEvent(Event):
+class SequenceEndEvent(CollectionEndEvent):
     pass
 
-class StreamStartEvent(Event):
+class MappingStartEvent(CollectionStartEvent):
     pass
 
-class StreamEndEvent(Event):
+class MappingEndEvent(CollectionEndEvent):
     pass
 
diff --git a/lib/yaml/parser.py b/lib/yaml/parser.py
index 9870699..4e0db9f 100644
--- a/lib/yaml/parser.py
+++ b/lib/yaml/parser.py
@@ -130,7 +130,8 @@ class Parser:
 
         # Parse start of stream.
         token = self.scanner.get()
-        yield StreamStartEvent(token.start_mark, token.end_mark)
+        yield StreamStartEvent(token.start_mark, token.end_mark,
+                encoding=token.encoding)
 
         # Parse implicit document.
         if not self.scanner.check(DirectiveToken, DocumentStartToken,
@@ -138,21 +139,25 @@ class Parser:
             self.tag_handles = self.DEFAULT_TAGS
             token = self.scanner.peek()
             start_mark = end_mark = token.start_mark
-            yield DocumentStartEvent(start_mark, end_mark)
+            yield DocumentStartEvent(start_mark, end_mark,
+                    implicit=True)
             for event in self.parse_block_node():
                 yield event
             token = self.scanner.peek()
             start_mark = end_mark = token.start_mark
+            implicit = True
             while self.scanner.check(DocumentEndToken):
                 token = self.scanner.get()
                 end_mark = token.end_mark
-            yield DocumentEndEvent(start_mark, end_mark)
+                implicit = True
+            yield DocumentEndEvent(start_mark, end_mark,
+                    implicit=implicit)
 
         # Parse explicit documents.
         while not self.scanner.check(StreamEndToken):
             token = self.scanner.peek()
             start_mark = token.start_mark
-            self.process_directives()
+            version, tags = self.process_directives()
             if not self.scanner.check(DocumentStartToken):
                 raise ParserError(None, None,
                         "expected '<document start>', but found %r"
@@ -160,7 +165,8 @@ class Parser:
                         self.scanner.peek().start_mark)
             token = self.scanner.get()
             end_mark = token.end_mark
-            yield DocumentStartEvent(start_mark, end_mark)
+            yield DocumentStartEvent(start_mark, end_mark,
+                    implicit=False, version=version, tags=tags)
             if self.scanner.check(DirectiveToken,
                     DocumentStartToken, DocumentEndToken, StreamEndToken):
                 yield self.process_empty_scalar(token.end_mark)
@@ -169,10 +175,13 @@ class Parser:
                     yield event
             token = self.scanner.peek()
             start_mark = end_mark = token.start_mark
+            implicit=True
             while self.scanner.check(DocumentEndToken):
                 token = self.scanner.get()
                 end_mark = token.end_mark
-            yield DocumentEndEvent(start_mark, end_mark)
+                implicit=False
+            yield DocumentEndEvent(start_mark, end_mark,
+                    implicit=implicit)
 
         # Parse end of stream.
         token = self.scanner.get()
@@ -201,9 +210,14 @@ class Parser:
                             "duplicate tag handle %r" % handle.encode('utf-8'),
                             token.start_mark)
                 self.tag_handles[handle] = prefix
+        if self.tag_handles:
+            value = self.yaml_version, self.tag_handles.copy()
+        else:
+            value = self.yaml_version, None
         for key in self.DEFAULT_TAGS:
             if key not in self.tag_handles:
                 self.tag_handles[key] = self.DEFAULT_TAGS[key]
+        return value
 
     def parse_block_node(self):
         return self.parse_node(block=True)
@@ -232,19 +246,22 @@ class Parser:
             start_mark = end_mark = tag_mark = None
             if self.scanner.check(AnchorToken):
                 token = self.scanner.get()
-                start_mark = end_mark = token.start_mark
+                start_mark = token.start_mark
+                end_mark = token.end_mark
                 anchor = token.value
                 if self.scanner.check(TagToken):
                     token = self.scanner.get()
-                    end_mark = tag_mark = token.start_mark
+                    tag_mark = token.start_mark
+                    end_mark = token.end_mark
                     tag = token.value
             elif self.scanner.check(TagToken):
                 token = self.scanner.get()
-                start_mark = end_mark = tag_mark = token.start_mark
+                start_mark = tag_mark = token.start_mark
+                end_mark = token.end_mark
                 tag = token.value
                 if self.scanner.check(AnchorToken):
                     token = self.scanner.get()
-                    end_mark = token.start_mark
+                    end_mark = token.end_mark
                     anchor = token.value
             if tag is not None:
                 handle, suffix = tag
@@ -258,38 +275,48 @@ class Parser:
                     tag = suffix
             if tag is None:
                 if not (self.scanner.check(ScalarToken) and
-                        self.scanner.peek().plain):
+                        self.scanner.peek().implicit):
                     tag = u'!'
             if start_mark is None:
-                start_mark = self.scanner.peek().start_mark
+                start_mark = end_mark = self.scanner.peek().start_mark
             event = None
             collection_events = None
             if indentless_sequence and self.scanner.check(BlockEntryToken):
                 end_mark = self.scanner.peek().end_mark
-                event = SequenceEvent(anchor, tag, start_mark, end_mark)
+                event = SequenceStartEvent(anchor, tag, start_mark, end_mark)
                 collection_events = self.parse_indentless_sequence()
             else:
                 if self.scanner.check(ScalarToken):
                     token = self.scanner.get()
                     end_mark = token.end_mark
                     event = ScalarEvent(anchor, tag, token.value,
-                            start_mark, end_mark)
+                            start_mark, end_mark,
+                            implicit=token.implicit, style=token.style)
                 elif self.scanner.check(FlowSequenceStartToken):
                     end_mark = self.scanner.peek().end_mark
-                    event = SequenceEvent(anchor, tag, start_mark, end_mark)
+                    event = SequenceStartEvent(anchor, tag, start_mark, end_mark,
+                            flow_style=True)
                     collection_events = self.parse_flow_sequence()
                 elif self.scanner.check(FlowMappingStartToken):
                     end_mark = self.scanner.peek().end_mark
-                    event = MappingEvent(anchor, tag, start_mark, end_mark)
+                    event = MappingStartEvent(anchor, tag, start_mark, end_mark,
+                            flow_style=True)
                     collection_events = self.parse_flow_mapping()
                 elif block and self.scanner.check(BlockSequenceStartToken):
                     end_mark = self.scanner.peek().start_mark
-                    event = SequenceEvent(anchor, tag, start_mark, end_mark)
+                    event = SequenceStartEvent(anchor, tag, start_mark, end_mark,
+                            flow_style=False)
                     collection_events = self.parse_block_sequence()
                 elif block and self.scanner.check(BlockMappingStartToken):
                     end_mark = self.scanner.peek().start_mark
-                    event = MappingEvent(anchor, tag, start_mark, end_mark)
+                    event = MappingStartEvent(anchor, tag, start_mark, end_mark,
+                            flow_style=False)
                     collection_events = self.parse_block_mapping()
+                elif anchor is not None or tag is not None:
+                    # Empty scalars are allowed even if a tag or an anchor is
+                    # specified.
+                    event = ScalarEvent(anchor, tag, u'', start_mark, end_mark,
+                            implicit=True)
                 else:
                     if block:
                         node = 'block'
@@ -320,7 +347,7 @@ class Parser:
             raise ParserError("while scanning a block collection", start_mark,
                     "expected <block end>, but found %r" % token.id, token.start_mark)
         token = self.scanner.get()
-        yield CollectionEndEvent(token.start_mark, token.end_mark)
+        yield SequenceEndEvent(token.start_mark, token.end_mark)
 
     def parse_indentless_sequence(self):
         # (BLOCK-ENTRY block_node?)+
@@ -333,7 +360,7 @@ class Parser:
             else:
                 yield self.process_empty_scalar(token.end_mark)
         token = self.scanner.peek()
-        yield CollectionEndEvent(token.start_mark, token.start_mark)
+        yield SequenceEndEvent(token.start_mark, token.start_mark)
 
     def parse_block_mapping(self):
         # BLOCK-MAPPING_START
@@ -365,7 +392,7 @@ class Parser:
             raise ParserError("while scanning a block mapping", start_mark,
                     "expected <block end>, but found %r" % token.id, token.start_mark)
         token = self.scanner.get()
-        yield CollectionEndEvent(token.start_mark, token.end_mark)
+        yield MappingEndEvent(token.start_mark, token.end_mark)
 
     def parse_flow_sequence(self):
         # flow_sequence     ::= FLOW-SEQUENCE-START
@@ -383,8 +410,9 @@ class Parser:
         while not self.scanner.check(FlowSequenceEndToken):
             if self.scanner.check(KeyToken):
                 token = self.scanner.get()
-                yield MappingEvent(None, u'!',
-                        token.start_mark, token.end_mark)
+                yield MappingStartEvent(None, u'!',
+                        token.start_mark, token.end_mark,
+                        flow_style=True)
                 if not self.scanner.check(ValueToken,
                         FlowEntryToken, FlowSequenceEndToken):
                     for event in self.parse_flow_node():
@@ -402,7 +430,7 @@ class Parser:
                     token = self.scanner.peek()
                     yield self.process_empty_scalar(token.start_mark)
                 token = self.scanner.peek()
-                yield CollectionEndEvent(token.start_mark, token.start_mark)
+                yield MappingEndEvent(token.start_mark, token.start_mark)
             else:
                 for event in self.parse_flow_node():
                     yield event
@@ -413,7 +441,7 @@ class Parser:
             if self.scanner.check(FlowEntryToken):
                 self.scanner.get()
         token = self.scanner.get()
-        yield CollectionEndEvent(token.start_mark, token.end_mark)
+        yield SequenceEndEvent(token.start_mark, token.end_mark)
 
     def parse_flow_mapping(self):
         # flow_mapping      ::= FLOW-MAPPING-START
@@ -457,8 +485,9 @@ class Parser:
             raise ParserError("while scanning a flow mapping", start_mark,
                     "expected '}', but found %r" % token.id, token.start_mark)
         token = self.scanner.get()
-        yield CollectionEndEvent(token.start_mark, token.end_mark)
+        yield MappingEndEvent(token.start_mark, token.end_mark)
 
     def process_empty_scalar(self, mark):
-        return ScalarEvent(None, None, u'', mark, mark)
+        return ScalarEvent(None, None, u'', mark, mark,
+                implicit=True)
 
diff --git a/lib/yaml/reader.py b/lib/yaml/reader.py
index 9778943..ab16a13 100644
--- a/lib/yaml/reader.py
+++ b/lib/yaml/reader.py
@@ -100,6 +100,7 @@ class Reader:
         self.pointer = 0
         self.raw_buffer = None
         self.raw_decode = None
+        self.encoding = None
         self.index = 0
         self.line = 0
         self.column = 0
@@ -156,10 +157,13 @@ class Reader:
         if not isinstance(self.raw_buffer, unicode):
             if self.raw_buffer.startswith(codecs.BOM_UTF16_LE):
                 self.raw_decode = utf_16_le_decode
+                self.encoding = 'utf-16-le'
             elif self.raw_buffer.startswith(codecs.BOM_UTF16_BE):
                 self.raw_decode = utf_16_be_decode
+                self.encoding = 'utf-16-be'
             else:
                 self.raw_decode = utf_8_decode
+                self.encoding = 'utf-8'
         self.update(1)
 
     NON_PRINTABLE = re.compile(u'[^\x09\x0A\x0D\x20-\x7E\x85\xA0-\uD7FF\uE000-\uFFFD]')
diff --git a/lib/yaml/scanner.py b/lib/yaml/scanner.py
index 9c536b4..ff4a8cc 100644
--- a/lib/yaml/scanner.py
+++ b/lib/yaml/scanner.py
@@ -379,8 +379,9 @@ class Scanner:
         # Read the token.
         mark = self.reader.get_mark()
         
-        # Add STREAM-END.
-        self.tokens.append(StreamStartToken(mark, mark))
+        # Add STREAM-START.
+        self.tokens.append(StreamStartToken(mark, mark,
+            encoding=self.reader.encoding))
         
 
     def fetch_stream_end(self):
@@ -638,12 +639,12 @@ class Scanner:
         self.tokens.append(self.scan_tag())
 
     def fetch_literal(self):
-        self.fetch_block_scalar(folded=False)
+        self.fetch_block_scalar(style='|')
 
     def fetch_folded(self):
-        self.fetch_block_scalar(folded=True)
+        self.fetch_block_scalar(style='>')
 
-    def fetch_block_scalar(self, folded):
+    def fetch_block_scalar(self, style):
 
         # A simple key may follow a block scalar.
         self.allow_simple_key = True
@@ -652,15 +653,15 @@ class Scanner:
         self.remove_possible_simple_key()
 
         # Scan and add SCALAR.
-        self.tokens.append(self.scan_block_scalar(folded))
+        self.tokens.append(self.scan_block_scalar(style))
 
     def fetch_single(self):
-        self.fetch_flow_scalar(double=False)
+        self.fetch_flow_scalar(style='\'')
 
     def fetch_double(self):
-        self.fetch_flow_scalar(double=True)
+        self.fetch_flow_scalar(style='"')
 
-    def fetch_flow_scalar(self, double):
+    def fetch_flow_scalar(self, style):
 
         # A flow scalar could be a simple key.
         self.save_possible_simple_key()
@@ -669,7 +670,7 @@ class Scanner:
         self.allow_simple_key = False
 
         # Scan and add SCALAR.
-        self.tokens.append(self.scan_flow_scalar(double))
+        self.tokens.append(self.scan_flow_scalar(style))
 
     def fetch_plain(self):
 
@@ -986,9 +987,14 @@ class Scanner:
         end_mark = self.reader.get_mark()
         return TagToken(value, start_mark, end_mark)
 
-    def scan_block_scalar(self, folded):
+    def scan_block_scalar(self, style):
         # See the specification for details.
 
+        if style == '>':
+            folded = True
+        else:
+            folded = False
+
         chunks = []
         start_mark = self.reader.get_mark()
 
@@ -1021,6 +1027,7 @@ class Scanner:
             line_break = self.scan_line_break()
             breaks, end_mark = self.scan_block_scalar_breaks(indent)
             if self.reader.column == indent and self.reader.peek() != u'\0':
+
                 # Unfortunately, folding rules are ambiguous.
                 #
                 # This is the folding according to the specification:
@@ -1053,7 +1060,8 @@ class Scanner:
             chunks.extend(breaks)
 
         # We are done.
-        return ScalarToken(u''.join(chunks), False, start_mark, end_mark)
+        return ScalarToken(u''.join(chunks), False, start_mark, end_mark,
+                style)
 
     def scan_block_scalar_indicators(self, start_mark):
         # See the specification for details.
@@ -1137,13 +1145,17 @@ class Scanner:
                 self.reader.forward()
         return chunks, end_mark
 
-    def scan_flow_scalar(self, double):
+    def scan_flow_scalar(self, style):
         # See the specification for details.
         # Note that we loose indentation rules for quoted scalars. Quoted
         # scalars don't need to adhere indentation because " and ' clearly
         # mark the beginning and the end of them. Therefore we are less
         # restrictive then the specification requires. We only need to check
         # that document separators are not included in scalars.
+        if style == '"':
+            double = True
+        else:
+            double = False
         chunks = []
         start_mark = self.reader.get_mark()
         quote = self.reader.peek()
@@ -1154,7 +1166,8 @@ class Scanner:
             chunks.extend(self.scan_flow_scalar_non_spaces(double, start_mark))
         self.reader.forward()
         end_mark = self.reader.get_mark()
-        return ScalarToken(u''.join(chunks), False, start_mark, end_mark)
+        return ScalarToken(u''.join(chunks), False, start_mark, end_mark,
+                style)
 
     ESCAPE_REPLACEMENTS = {
         u'0':   u'\0',
diff --git a/lib/yaml/tokens.py b/lib/yaml/tokens.py
index 59b36af..933e56f 100644
--- a/lib/yaml/tokens.py
+++ b/lib/yaml/tokens.py
@@ -30,6 +30,11 @@ class DocumentEndToken(Token):
 
 class StreamStartToken(Token):
     id = '<stream start>'
+    def __init__(self, start_mark=None, end_mark=None,
+            encoding=None):
+        self.start_mark = start_mark
+        self.end_mark = end_mark
+        self.encoding = encoding
 
 class StreamEndToken(Token):
     id = '<stream end>'
@@ -90,9 +95,10 @@ class TagToken(Token):
 
 class ScalarToken(Token):
     id = '<scalar>'
-    def __init__(self, value, plain, start_mark, end_mark):
+    def __init__(self, value, implicit, start_mark, end_mark, style=None):
         self.value = value
-        self.plain = plain
+        self.implicit = implicit
         self.start_mark = start_mark
         self.end_mark = end_mark
+        self.style = style
 
diff --git a/tests/data/aliases.events b/tests/data/aliases.events
new file mode 100644
index 0000000..9139b51
--- /dev/null
+++ b/tests/data/aliases.events
@@ -0,0 +1,8 @@
+- !StreamStart
+- !DocumentStart
+- !SequenceStart
+- !Scalar { anchor: 'myanchor', tag: '!mytag', value: 'data' }
+- !Alias { anchor: 'myanchor' }
+- !SequenceEnd
+- !DocumentEnd
+- !StreamEnd
diff --git a/tests/data/documents.events b/tests/data/documents.events
new file mode 100644
index 0000000..223a314
--- /dev/null
+++ b/tests/data/documents.events
@@ -0,0 +1,11 @@
+- !StreamStart
+- !DocumentStart
+- !Scalar { implicit: true }
+- !DocumentEnd
+- !DocumentStart
+- !Scalar { implicit: true }
+- !DocumentEnd
+- !DocumentStart
+- !Scalar { implicit: true }
+- !DocumentEnd
+- !StreamEnd
diff --git a/tests/data/mappings.events b/tests/data/mappings.events
new file mode 100644
index 0000000..e66ec09
--- /dev/null
+++ b/tests/data/mappings.events
@@ -0,0 +1,44 @@
+- !StreamStart
+
+- !DocumentStart
+- !MappingStart
+- !Scalar { implicit: true, value: 'key' }
+- !Scalar { implicit: true, value: 'value' }
+- !Scalar { implicit: true, value: 'empty mapping' }
+- !MappingStart
+- !MappingEnd
+- !Scalar { implicit: true, value: 'empty mapping with tag' }
+- !MappingStart { tag: '!mytag' }
+- !MappingEnd
+- !Scalar { implicit: true, value: 'block mapping' }
+- !MappingStart
+- !MappingStart
+- !Scalar { implicit: true, value: 'complex' }
+- !Scalar { implicit: true, value: 'key' }
+- !Scalar { implicit: true, value: 'complex' }
+- !Scalar { implicit: true, value: 'key' }
+- !MappingEnd
+- !MappingStart
+- !Scalar { implicit: true, value: 'complex' }
+- !Scalar { implicit: true, value: 'key' }
+- !MappingEnd
+- !MappingEnd
+- !Scalar { implicit: true, value: 'flow mapping' }
+- !MappingStart { flow: true }
+- !Scalar { implicit: true, value: 'key' }
+- !Scalar { implicit: true, value: 'value' }
+- !MappingStart
+- !Scalar { implicit: true, value: 'complex' }
+- !Scalar { implicit: true, value: 'key' }
+- !Scalar { implicit: true, value: 'complex' }
+- !Scalar { implicit: true, value: 'key' }
+- !MappingEnd
+- !MappingStart
+- !Scalar { implicit: true, value: 'complex' }
+- !Scalar { implicit: true, value: 'key' }
+- !MappingEnd
+- !MappingEnd
+- !MappingEnd
+- !DocumentEnd
+
+- !StreamEnd
diff --git a/tests/data/scalars.events b/tests/data/scalars.events
new file mode 100644
index 0000000..0d3f024
--- /dev/null
+++ b/tests/data/scalars.events
@@ -0,0 +1,24 @@
+- !StreamStart
+
+- !DocumentStart
+- !MappingStart
+- !Scalar { implicit: true, value: 'empty scalar' }
+- !Scalar { implicit: true, value: '' }
+- !Scalar { implicit: true, value: 'implicit scalar' }
+- !Scalar { implicit: true, value: 'data' }
+- !Scalar { implicit: true, value: 'quoted scalar' }
+- !Scalar { value: 'data', style: '"' }
+- !Scalar { implicit: true, value: 'block scalar' }
+- !Scalar { value: 'data', style: '|' }
+- !Scalar { implicit: true, value: 'empty scalar with tag' }
+- !Scalar { implicit: true, tag: '!mytag', value: '' }
+- !Scalar { implicit: true, value: 'implicit scalar with tag' }
+- !Scalar { implicit: true, tag: '!mytag', value: 'data' }
+- !Scalar { implicit: true, value: 'quoted scalar with tag' }
+- !Scalar { value: 'data', style: '"', tag: '!mytag' }
+- !Scalar { implicit: true, value: 'block scalar with tag' }
+- !Scalar { value: 'data', style: '|', tag: '!mytag' }
+- !MappingEnd
+- !DocumentEnd
+
+- !StreamEnd
diff --git a/tests/data/sequences.events b/tests/data/sequences.events
new file mode 100644
index 0000000..0fb514e
--- /dev/null
+++ b/tests/data/sequences.events
@@ -0,0 +1,81 @@
+- !StreamStart
+
+- !DocumentStart
+- !SequenceStart
+- !SequenceEnd
+- !DocumentEnd
+
+- !DocumentStart
+- !SequenceStart { tag: '!mytag' }
+- !SequenceEnd
+- !DocumentEnd
+
+- !DocumentStart
+- !SequenceStart
+- !SequenceStart
+- !SequenceEnd
+- !SequenceStart { tag: '!mytag' }
+- !SequenceEnd
+- !SequenceStart
+- !Scalar
+- !Scalar { value: 'data' }
+- !Scalar { tag: '!mytag', value: 'data' }
+- !SequenceEnd
+- !SequenceStart
+- !SequenceStart
+- !SequenceStart
+- !Scalar
+- !SequenceEnd
+- !SequenceEnd
+- !SequenceEnd
+- !SequenceStart
+- !SequenceStart { tag: '!mytag' }
+- !SequenceStart
+- !Scalar { value: 'data' }
+- !SequenceEnd
+- !SequenceEnd
+- !SequenceEnd
+- !SequenceEnd
+- !DocumentEnd
+
+- !DocumentStart
+- !SequenceStart
+- !MappingStart
+- !Scalar { value: 'key1' }
+- !SequenceStart
+- !Scalar { value: 'data1' }
+- !Scalar { value: 'data2' }
+- !SequenceEnd
+- !Scalar { value: 'key2' }
+- !SequenceStart { tag: '!mytag1' }
+- !Scalar { value: 'data3' }
+- !SequenceStart
+- !Scalar { value: 'data4' }
+- !Scalar { value: 'data5' }
+- !SequenceEnd
+- !SequenceStart { tag: '!mytag2' }
+- !Scalar { value: 'data6' }
+- !Scalar { value: 'data7' }
+- !SequenceEnd
+- !SequenceEnd
+- !MappingEnd
+- !SequenceEnd
+- !DocumentEnd
+
+- !DocumentStart
+- !SequenceStart
+- !SequenceStart { flow: true }
+- !SequenceStart
+- !SequenceEnd
+- !Scalar
+- !Scalar { value: 'data' }
+- !Scalar { tag: '!mytag', value: 'data' }
+- !SequenceStart { tag: '!mytag' }
+- !Scalar { value: 'data' }
+- !Scalar { value: 'data' }
+- !SequenceEnd
+- !SequenceEnd
+- !SequenceEnd
+- !DocumentEnd
+
+- !StreamEnd
diff --git a/tests/data/spec-08-12.canonical b/tests/data/spec-08-12.canonical
index dc3b81a..93899f4 100644
--- a/tests/data/spec-08-12.canonical
+++ b/tests/data/spec-08-12.canonical
@@ -6,4 +6,5 @@
   !!str "Tagged",
   *A,
   !!str "",
+  !!str "",
 ]
diff --git a/tests/data/spec-08-12.data b/tests/data/spec-08-12.data
index d936ba2..3d4c6b7 100644
--- a/tests/data/spec-08-12.data
+++ b/tests/data/spec-08-12.data
@@ -3,6 +3,6 @@
   &anchor "Anchored",
   !!str 'Tagged',
   *anchor, # Alias node
-#  !!str,   # Empty plain scalar
+  !!str ,  # Empty plain scalar
   '',   # Empty plain scalar
 ]
diff --git a/tests/test_appliance.py b/tests/test_appliance.py
index eb54faa..ed76e47 100644
--- a/tests/test_appliance.py
+++ b/tests/test_appliance.py
@@ -243,10 +243,10 @@ class CanonicalParser:
             if self.test_token(ScalarToken):
                 self.events.append(ScalarEvent(anchor, tag, self.get_value(), None, None))
             elif self.test_token(FlowSequenceStartToken):
-                self.events.append(SequenceEvent(anchor, tag, None, None))
+                self.events.append(SequenceStartEvent(anchor, tag, None, None))
                 self.parse_sequence()
             elif self.test_token(FlowMappingStartToken):
-                self.events.append(MappingEvent(anchor, tag, None, None))
+                self.events.append(MappingStartEvent(anchor, tag, None, None))
                 self.parse_mapping()
             else:
                 raise Error("SCALAR, '[', or '{' is expected, got "+repr(self.tokens[self.index]))
@@ -261,7 +261,7 @@ class CanonicalParser:
                 if not self.test_token(FlowSequenceEndToken):
                     self.parse_node()
         self.consume_token(FlowSequenceEndToken)
-        self.events.append(CollectionEndEvent(None, None))
+        self.events.append(SequenceEndEvent(None, None))
 
     # mapping: MAPPING-START (map_entry (ENTRY map_entry)*)? ENTRY? MAPPING-END
     def parse_mapping(self):
@@ -273,7 +273,7 @@ class CanonicalParser:
                 if not self.test_token(FlowMappingEndToken):
                     self.parse_map_entry()
         self.consume_token(FlowMappingEndToken)
-        self.events.append(CollectionEndEvent(None, None))
+        self.events.append(MappingEndEvent(None, None))
 
     # map_entry: KEY node VALUE node
     def parse_map_entry(self):
diff --git a/tests/test_emitter.py b/tests/test_emitter.py
new file mode 100644
index 0000000..2da6696
--- /dev/null
+++ b/tests/test_emitter.py
@@ -0,0 +1,73 @@
+
+import test_appliance, sys, StringIO
+
+from yaml import *
+import yaml
+
+class TestEmitterOnCanonical(test_appliance.TestAppliance):
+
+    def _testEmitterOnCanonical(self, test_name, canonical_filename):
+        events = list(iter(Parser(Scanner(Reader(file(canonical_filename, 'rb'))))))
+        #writer = sys.stdout
+        writer = StringIO.StringIO()
+        emitter = Emitter(writer)
+        #print "-"*30
+        #print "ORIGINAL DATA:"
+        #print file(canonical_filename, 'rb').read()
+        for event in events:
+            emitter.emit(event)
+        data = writer.getvalue()
+        new_events = list(parse(data))
+        self.failUnlessEqual(len(events), len(new_events))
+        for event, new_event in zip(events, new_events):
+            self.failUnlessEqual(event.__class__, new_event.__class__)
+
+TestEmitterOnCanonical.add_tests('testEmitterOnCanonical', '.canonical')
+
+class EventsConstructor(Constructor):
+
+    def construct_event(self, node):
+        if isinstance(node, ScalarNode):
+            mapping = {}
+        else:
+            mapping = self.construct_mapping(node)
+        class_name = str(node.tag[1:])+'Event'
+        if class_name in ['AliasEvent', 'ScalarEvent', 'SequenceStartEvent', 'MappingStartEvent']:
+            mapping.setdefault('anchor', None)
+        if class_name in ['ScalarEvent', 'SequenceStartEvent', 'MappingStartEvent']:
+            mapping.setdefault('tag', None)
+        if class_name == 'ScalarEvent':
+            mapping.setdefault('value', '')
+        value = getattr(yaml, class_name)(**mapping)
+        return value
+
+EventsConstructor.add_constructor(None, EventsConstructor.construct_event)
+
+class TestEmitter(test_appliance.TestAppliance):
+
+    def _testEmitter(self, test_name, events_filename):
+        events = load_document(file(events_filename, 'rb'), Constructor=EventsConstructor)
+        self._dump(events_filename, events)
+        writer = StringIO.StringIO()
+        emitter = Emitter(writer)
+        for event in events:
+            emitter.emit(event)
+        data = writer.getvalue()
+        new_events = list(parse(data))
+        self.failUnlessEqual(len(events), len(new_events))
+        for event, new_event in zip(events, new_events):
+            self.failUnlessEqual(event.__class__, new_event.__class__)
+
+    def _dump(self, events_filename, events):
+        writer = sys.stdout
+        emitter = Emitter(writer)
+        print "="*30
+        print "EVENTS:"
+        print file(events_filename, 'rb').read()
+        print '-'*30
+        print "OUTPUT:"
+        for event in events:
+            emitter.emit(event)
+        
+TestEmitter.add_tests('testEmitter', '.events')
+
diff --git a/tests/test_structure.py b/tests/test_structure.py
index ca3cb60..cb9a017 100644
--- a/tests/test_structure.py
+++ b/tests/test_structure.py
@@ -35,17 +35,17 @@ class TestStructure(test_appliance.TestAppliance):
                 return True
             else:
                 return None
-        elif parser.check(SequenceEvent):
+        elif parser.check(SequenceStartEvent):
             parser.get()
             sequence = []
-            while not parser.check(CollectionEndEvent):
+            while not parser.check(SequenceEndEvent):
                 sequence.append(self._convert(parser))
             parser.get()
             return sequence
-        elif parser.check(MappingEvent):
+        elif parser.check(MappingStartEvent):
             parser.get()
             mapping = []
-            while not parser.check(CollectionEndEvent):
+            while not parser.check(MappingEndEvent):
                 key = self._convert(parser)
                 value = self._convert(parser)
                 mapping.append((key, value))
@@ -92,7 +92,7 @@ class TestParser(test_appliance.TestAppliance):
                 #self.failUnlessEqual(event1.anchor, event2.anchor)
                 #self.failUnlessEqual(event1.tag, event2.tag)
                 self.failUnlessEqual(event1.value, event2.value)
-            if isinstance(event1, CollectionEvent):
+            if isinstance(event1, CollectionStartEvent):
                 #self.failUnlessEqual(event1.anchor, event2.anchor)
                 #self.failUnlessEqual(event1.tag, event2.tag)
                 pass
@@ -213,7 +213,7 @@ class TestParserOnCanonical(test_appliance.TestAppliance):
                 self.failUnlessEqual(event1.anchor, event2.anchor)
                 self.failUnlessEqual(event1.tag, event2.tag)
                 self.failUnlessEqual(event1.value, event2.value)
-            if isinstance(event1, CollectionEvent):
+            if isinstance(event1, CollectionStartEvent):
                 self.failUnlessEqual(event1.anchor, event2.anchor)
                 self.failUnlessEqual(event1.tag, event2.tag)
 
diff --git a/tests/test_yaml.py b/tests/test_yaml.py
index bb5a9f1..a2759c4 100644
--- a/tests/test_yaml.py
+++ b/tests/test_yaml.py
@@ -9,6 +9,7 @@ from test_structure import *
 from test_errors import *
 from test_detector import *
 from test_constructor import *
+#from test_emitter import *
 from test_syck import *
 
 def main(module='__main__'):