SGML

Basic omission of the start- and end-element tags for element a, as allowed by the O O tag omission indicators; as shown in the result, the missing tags are generated by SGML

Variant of test6a where an addtional element f may be placed before b in e; this doesn't affect the inference of b and its containing elements

Examples showing that parsed character data can also trigger tag inference; moreover, these examples show special processing of whitespace, followed by non-whitespace character data in content: handling of whitespace (including newline and tab characters) depends on whether PCDATA is accepted in the current context position without tag inference; if it is, then leading whitespace is considered part of parsed character data; otherwise, state transitions (including actions such as closing finished or opening required elements) are driven only with the portion starting PCDATA, any leading whitespace belonging to a separately reported characters event is ignored for validation

Examples for rank-based tag inference where generation of Sn elements around sequence of Hn/Pn is desired