To parse a whole document, xmlParse file content takes a filename (for generating error reports) and the string content of that file. A parse error causes program failure, with message to stderr.

To parse just a DTD, dtdParse file content takes a filename (for generating error reports) and the string content of that file. If no DTD was found, you get Nothing rather than an error. However, if a DTD is found but contains errors, the program crashes.

To parse a partial document, e.g. from an XML-based stream protocol, where you may later want to get more document elements from the same stream. Arguments are: a parser for the item you want, and the already-lexed input to parse from. Returns the item you wanted (or an error message), plus the remainder of the input.

Return an entire XML document including prolog and trailing junk.

Return a complete element including all its inner content.

Return a content particle, e.g. text, element, reference, etc

Return an XML comment.

Return parsed freetext (i.e. until the next markup)

Return either a general entity reference, or a character reference.

Return a DOCTYPE decl, indicating a DTD.

Parse a processing instruction.

Parse only the parts between angle brackets in an element tag.

Return a qualified name (although the namespace qualification is not processed here; this is merely to get the correct type).

Return just a name, e.g. element name, attribute name.

Return the next token from the input only if it matches the given token.

For use with stream parsers - returns the complete opening element tag.

For use with stream parsers - accepts a closing tag, provided it matches the given element name.

Some empty symbol tables for GE and PE references.

XParser is just a specialisation of the PolyStateLazy parser.