The CData content must be split due to encoding issues, which isn't supported (due to bad CData handling on Sun JRE and differences with Xalan.
For a given content it could not be serialized in the document.
Comments don't support & recognition, which means you can't escape the encoding, either they encode or they do not
Result of trying to serialize a 1.1 ncname into a 1.0 document
The serialization encoding cannot support the characters in a given markup (attribute name or element name.
The serialization encoding cannot support the characters in a given markup (attribute name or element name. ächtung in UTF-8 works but not in US-ASCII.
Default serializer, correctness first.
Default serializer, correctness first. Uses the LSSerializer present in each DOM L3 impl. XmlItems are always serialized with the LS, elements and attributes are however for speed reasons, verified for encoding once per QName; the attribute values themselves are written via Text nodes and LS.
Base implementation for a correct serializer using LSSerializer to provide escape character references.
Base implementation for a correct serializer using LSSerializer to provide escape character references.
Developers can override this directly implementing encF, or choose to override createSerializer to further change serialization behaviour
This variety does not use a thread safe global cache, use when your data is progressivly radically different for each run.
This variety does not use a thread safe global cache, use when your data is progressivly radically different for each run.
If you don't like this scheme either simply implement LSSerializerFactoryBase with your own caching needs.
A NamespaceContext represents the prefix->namespace mappings for a given elements children.
A NamespaceContext represents the prefix->namespace mappings for a given elements children. It is only valid for the time of serialization (or comparison).
prefix -> namespace
prefix -> namespace (these should be declared for a given element)
should a new xmlns="" default be added, if so Some(String)
The stream of PullTypes was empty.
PIs also suffer a bad specification
Base exception marker for serialization
Type class for choosing a serializing algo
Interface used for serializing the XML events, same for both stream and tree.
Interface used for serializing the XML events, same for both stream and tree.
The default implementation validates the output via LSSerializer against a given encoding.
An alternative FastSerializer simply outputs strings and is in no way correct but at least faster then going to StreamWriter for incorrectness.
Serializers should use qName.ncName for serialization purposes of both attributes and elements, the QName is provided to allow for validation of ouput.
If a function returns Some it should signal the termination of the serializing. Implementations are expected to respect this approach, its for the benefit of all developers.
List[QName] is provided to help if path relevant information is needed for serialization. For example if a use case requires that a particular path is filtered out. Purely for performance reasons the list is in reverse order with the current QName at the top and root at the bottom.
Serializer factories are responsible for the life cycle of serializers and their resources.
SerializeableXml instances for the core types
Provide push based serializing Iteratee
Performs no validation, sorting of attributes, extra spaces on the end of empty element declarations and uses the short form for empty elements.
Performs no validation, sorting of attributes, extra spaces on the end of empty element declarations and uses the short form for empty elements.
Basically its good for non-pretty printing debugging only
Provides a base class for steam handling
Status of stream processing
Wraps the use of writeTo allowing: xml writeTo output
Adds an extra space after an empty element
This class represents state during a serialization
Default implmementation of serialization
XHTML serialization extra touches
Provides basis for serialisation of streams.
The CData content must be split due to encoding issues, which isn't supported (due to bad CData handling on Sun JRE and differences with Xalan.
NOTE The Xalan behaviour follows the DOMConfiguration behaviour but will split on encoding issues, which breaks the point of using CData, so if you need that, don't use it, let your sax / pull parser handle proper serilization and escaping for you in a normal text field.