Suppose you have designated the following criterion for split points:
[HTMLParaStyles]
Heading2 = Split
Now suppose your DITA document has content organized as follows:
Heading1
Heading2
Body text
Heading2
Body text
Heading1
. . .
And you want split points as follows:
Heading1
Heading2
Body text
------------------- split here
Heading2
Body text
------------------- split here
Heading1
. . .
That is, to avoid a title-only topic, you want to suppress a split between the first Heading1 and the first Heading2.
Instead of the first Heading2 you could use a different paragraph format that is not assigned the Split property, such as Heading2First, and add that format to the paragraph catalog. Or, you could use [HelpContentsLevels] to assign level numbers to your heading formats, and set SmartSplit=Yes:
[HelpContentsLevels]
; FM paragraph format name = TOC level (MS or Java)
Heading1 = 1
Heading2 = 2
[HTMLParaStyles]
Heading1 = Split
Heading2 = Split
[HTMLOptions]
; SmartSplit = No (default) or Yes (suppress splits in sequences of
; heads which lack text bodies, as long as the [HelpContentsLevels]
; values continue to increase)
SmartSplit = Yes
The effect of these settings is to suppress splits as long as the heading level number increases for each paragraph in an unbroken sequence of paragraphs. For example, these settings would keep all of the following together:
Heading1
Heading2
Heading3
Body text
In other words, SmartSplit prevents splits when both of the following are true:
Note: It is not a good idea to use SmartSplit in conjunction with trails of links; see §29.2.1 Understanding trails of links.