Allow all XML 1.1 characters escapes even if input document is described as XML 1.0 (in addition to ones allowed in 1.l0).

Since this is non-standard option (that is, deviates from XML specification), it is disabled by default.

Since

5.2

Property of type java.net.URL, that will allow specifying context URL to use when resolving relative references, for the main-level entities (external DTD subset, references from the internal DTD subset).

Whether readers will try to cache parsed external DTD subsets or not.

Whether reader is to cache DTDs (when caching enabled) based on public id or not: if not, system id will be primarily used. Although theoretically public IDs should be unique, and should be good caching keys, sometimes broken documents use 'wrong' public IDs, and such by default caching keys are based on system id only.

Property of type java.util.Map, that defines explicit set of internal (generic) entities that will define of override any entities defined in internal or external subsets; except for the 5 pre-defined entities (lt, gt, amp, apos, quot). Can be used to explicitly define entities that would normally come from a DTD.

Property of type XMLResolver, that will allow overriding of default DTD and external parameter entity resolution.

Property of type XMLResolver, that will allow overriding of default external general entity resolution. Note that using this property overrides settings done using javax.xml.stream.XMLInputFactory#RESOLVER (and vice versa).

Size of input buffer (in chars), to use for reading XML content from input stream/reader.

Three-valued property (one of PARSING_MODE_DOCUMENT, PARSING_MODE_FRAGMENT or PARSING_MODE_DOCUMENTS; default being the document mode) that can be used to handle "non-standard" XML content. The default mode (PARSING_MODE_DOCUMENT) allows parsing of only well-formed XML documents, but the other two modes allow more lenient parsing. Fragment mode allows parsing of XML content that does not have a single root element (can have zero or more), nor can have XML or DOCTYPE declarations: this may be useful if parsing a subset of a full XML document. Multi-document (PARSING_MODE_DOCUMENTS) mode on the other hand allows parsing of a stream that contains multiple consequtive well-formed documents, with possibly multiple XML and DOCTYPE declarations.

The main difference from the API perspective is that in first two modes, START_DOCUMENT and END_DOCUMENT are used as usual (as the first and last events returned), whereas the multi-document mode can return multiple pairs of these events: although it is still true that the first event (one cursor points to when reader is instantiated or returned by the event reader), there may be intervening pairs that signal boundary between two adjacent enclosed documents.

Whether stream readers are allowed to do lazy parsing, meaning to parse minimal part of the event when javax.xml.stream.XMLStreamReader#next is called, and only parse the rest as needed (or skip remainder of no extra information is needed). Alternative to lazy parsing is called "eager parsing", and is what most xml parsers use by default.

Enabling lazy parsing can improve performance for tasks where number of textual events are skipped. The downside is that not all well-formedness problems are reported when javax.xml.stream.XMLStreamReader#next is called, but only when the rest of event are read or skipped.

Default value for Woodstox is such that lazy parsing is enabled.

Maximum length of of individual attribute values (in characters)

Since

4.2

Maximum number of attributes allowed for single XML element.

Since

4.2

Maximum length of input document, in characters.

Since

4.2

Maximum number of child elements for any given element.

Since

4.2

Maximum level of nesting of XML elements, starting with root element.

Since

5.4 / 6.4

Maximum number of all elements in a single document.

Since

4.2

Maximum level of nesting of XML elements, starting with root element.

Since

4.2

Maximum number of total (general parsed) entity expansions within input.

Since

4.3

Maximum depth of nested (general parsed) entity expansions.

Since

4.3

Maximum length of individual text (cdata) segments in input, in characters.

Since

4.2

Property to specify shortest non-complete text segment (part of CDATA section or text content) that parser is allowed to return, if not required to coalesce text.

Feature that controls whether linefeeds are normalized into canonical linefeed as mandated by xml specification.

Note that disabling this property (from its default enabled state) will result in non-conforming XML processing. It may be useful for use cases where changes to input content should be minimized.

This read-only property indicates whether null is returned for default name space prefix; Boolean.TRUE indicates it does, Boolean.FALSE that it does not.

Default value for 4.1 is 'false'; this will most likely change for 5.0 since Stax API actually specifies null to be used.

Since

4.1.2

Whether the Reader will recognized DTD++ extensions when parsing DTD subsets.

Whether the Reader will treat character references as entities while parsing XML documents.

Property of type XMLResolver, that will allow graceful handling of references to undeclared (general) entities.

Whether readers will verify that characters in text content are fully valid XML characters (not just Unicode). If true, will check that they are valid (including white space); if false, will not check.

Note that this property will NOT have effect on all encoding problems, specifically:

• UTF-8 decoder will still report invalid UTF-8 byte sequences (and same for other character encodings).
• XML Name character rules follow separate validation which will not be affected

Turning this option off may improve parsing performance; leaving it on guarantees compatibility with XML 1.0 specification regarding character validity rules.

Constants used when no DTD handling is done, and we do not know the 'real' type of an attribute. Seems like CDATA is the safe choice.