jgexml

jgeXml - The Just-Good-Enough XML Parser

jgeXml provides event-driven routines to parse XML 1.0 (both pull and push modes are supported), to write XML and to convert between XML and JSON.

The code has no dependencies on other modules or native libraries.

Events (stateCodes)

sDeclaration
sElement
sAttribute
sValue
sEndElement
sContent
sComment
sProcessingInstruction
sEndDocument

No event is generated for ignoreable whitespace, unlike SAX. Empty elements are normalised into sElement/sEndElement pairs.

Notes

jgeXml is a non-validating parser.

Both when reading and writing, attributes follow after the element event, and in the order they are given in the source.

When converting to JSON, the attributePrefix (to avoid name clashes with child elements) is configurable per parse.

Child elements can be represented as properties or objects in JSON.

The parser by default treats all content as strings when converting to JSON, optionally data can be coerced to primitive numbers or null values.

An experimental limited JSONPath utility is under development.

Limitations

Probably not thread safe.

jgeXml assumes the XML is well-formed.

jgeXml is currently schema agnostic and staunchly atheist when it comes to DTDs. It can parse XML documents with schema information, but it is up to the consumer to interpret the namespace portions of element names. It cannot parse internal DTDs. DOCTYPEs are handled as comments. xmlWrite minimally supports DTDs but you must build them and the DOCTYPE yourself.

It can parse and transform XSD files as XML, conversion to JSON schema is planned.

The parser currently does not support CDATA segments, though the xmlWrite module does.

The parser is string-based; to process streams, read the data into a string first. It may be memory intensive on large documents.

Examples

See testx2x for parsing XML to XML, testx2j for parsing XML to JSON, testj2x for converting JSON to XML, pullparser and pushparser for how to set up and run the parser.