In Word, it takes the form of a normal Word Table of Contents. After generating the Word output, the page numbers are not populated. You need to refresh the table of contents by right-clicking on a field and choosing Update Field.
In HTML, the table of contents takes the form of a side panel.
In PDF, it takes the form of an introductory table of contents.

Document attributes

By default, the table of contents includes two levels of heading.

More levels of heading can be set by using the document attribute :toclevels:.

e.g. :toclevels: 3 for three levels of heading included.

The number of levels of heading to include can be set separately for HTML/PDF and for DOC, by using the document attributes :htmltoclevels: and :doctoclevels:.

Using the `toc` command to generate a ToC

A manual table of contents command has been added to Metanorma [added in https://github.com/metanorma/metanorma-standoc/releases/tag/v1.10.5].

This is intended for a local table of contents, e.g., at the start of an appendix.

The table of contents toc command is utilized as follows.

toc:["xpath1:depth1","xpath2:depth2","xpath3:depth3"...]

Where,

depth is a number indicating the depth of indentation of the ToC entry (default value 1)
xpath is the XPath of the title being indexed in the Metanorma XML.

For instance, in the following command:

toc:["//clause[@id = 'clause1'\]/clause/title","//clause[@id = 'clause1'\]/clause/clause/title:2"]

all titles of subclauses of the clause with ID clause1 are at the first depth level of the manual table of contents (defaults to depth 1)
all titles of sub-subclauses of the clause with ID clause1 are at the second depth level of the manual table of contents

This means that the table of contents will have of the first-level clause with ID clause1, running three levels deep.

Manually specifying a ToC

A manual table of contents can also be provided as a clause of type toc [added in https://github.com/metanorma/metanorma-standoc/releases/tag/v1.10.7] without using the toc command.

Any table of contents entries within that manual table of contents are expected to be provided as cross-references within unordered lists.

The table of contents can contain subclauses, each with its own list of cross-references, and other text.

Example 1. Example of manually specifying a ToC

[type=toc]
=== Table of contents

This is a table of contents.

==== First subsection

* <<xref1>>
* <<xref2,This is a modified title>>

==== Second subsection

* <<xref3>>
* <<xref4>>

Variant titles

Clauses in Metanorma normally have only one title, and that title is used to identify the clause in the table of contents.

However, a manually created table of contents can make use of variant titles instead.

Variant titles [added in https://github.com/metanorma/metanorma-standoc/releases/tag/v1.10.5] are entered as paragraphs with a variant-title role attribute within a clause, as follows:

=== Proper title

[.variant-title,type=toc]
This is the variant title

Text of section.

Variant titles are not rendered in the body of the text. However, any variant titles of type toc are used instead of the title as the entry text for a manual table of contents.

Note	Of course, if an explicit cross-reference text is given in a `toc` clause, that takes priority.

Variant titles are also used to generate the automated all-of-document table of contents for HTML and PDF documents [added in https://github.com/metanorma/isodoc/releases/tag/v1.8.5]. They are not used in Word DOC output, because tables of contents for Word documents are generated dynamically from the current heading.

Variant titles are not currently used to generate the tables of other entities.

Author's documentation

Table of contents