21.6.3 Understanding Mif2Go support for FrameMaker 8+ Unicode
FrameMaker (version 8 and later versions) supports only one part of Unicode, the BMP (Basic Multilingual Plane), which is from U+0000 to U+FFFF. FrameMaker does not support the “astral planes”, above 0xFFFF. FrameMaker does store the characters in those planes, using 4-byte UTF-8 encoding, but will not display them, showing just a placeholder (the question mark or a space) instead. However, if you export to HTML using Mif2Go, the characters will appear in the output, and anyone with the appropriate font installed can see them in their browser.
Mif2Go provides improved Unicode processing for FrameMaker version 8 and later versions, supporting characters in the “astral planes” above Plane 0 (the Basic Multilingual Plane or BMP), such as the characters in Plane 1 (the Supplementary Multilingual Plane or SMP) and Plane 2 (the Supplementary Ideographic Plane or SIP). UTF-8 can encode characters through Plane 31, if any are ever defined; only Planes 0, 1, 2, 14, 15, and 16 are used in the current specification (Version 5.0). Each plane can include 64K characters. Mif2Go produces CJK output using UTF-8, including characters above the BMP, such as Chinese Extension B. Viewing such characters requires a font that shows them; but even if such a font is not present on your system, Mif2Go produces the correct characters.
Mif2Go does not support non-Unicode double-byte character sets, except for Asian and Cyrillic code pages used by HTML Help; see §9.13 Generating HTML Help in non-Western languages and §13.4.3 Specifying character encoding for HTML.
§22.7.6 Assigning CSS classes based on Unicode character ranges
§38.2.4 Saving FrameMaker 8 files as FrameMaker 8 MIF
> 21 Mapping text formats to HTML/XML > 21.6 Mapping special characters > 21.6.3 Understanding Mif2Go support for FrameMaker 8+ Unicode