The XSD specifying the structure of the Text Block XMLs is given in the Appendix. Also, an example can be found there. If needed, the XSD can also be written to the file TBimport.xsd with the following command line invocation:
tbmigrate /xsd
An ITP Text Block consists of exactly one <tbk> node at top level with the attribute "xsv", to indicate the Text Block version. This <tbk> node can contain multiple paragraphs and (un)ordered lists of paragraphs, represented by the nodes <par> and <lst>. The attribute "ordered" on a <lst> node indicates whether a list is ordered or unordered. Paragraph nodes can either contain normal text or header text, and fields indicated by the attribute "font". A paragraph can also have an indentation, described by the attribute "indentation".
Within paragraphs, you can have multiple texts, Fields, and special characters, stored in the nodes <txt>, <fld> and <chr>. All nodes have the attributes "bold", "italic", and "underline" to indicate their layout.
The <chr> nodes represent a non-breaking space, a non-breaking hyphen, or a line break. This is indicated by the attribute "type". Each <chr> can represent only one character.
The attribute "set" in a <fld> node indicates from which Field Set a Field comes. The value of this node is the name of the Field. Each <fld> node can represent only one Field.
The <txt> nodes just contain their text, possibly inside a CDATA section. Note that white space in these nodes is significant and will show up in the resulting Text Block.
The XML should conform to the XSD. Also, field and field set names are limited to characters from the set Latin-1 character and subject to the naming rules of the ITP language.