Problems in FrameMaker 7 with CE fonts?
This paper is based on the platform Windows 2000/Windows XP, which supports the standard MS virtual font substitution mechanism for Non-Unicode applications. That means, the operating system supports automatically the display of a certain part of a big font, like "Arial CE", via enabling a specific country setting.
Using languages within FrameMaker
Adobe does not support FrameMaker for easter- and southern-european languages. Nevertheless it is possible to use for example a Greek font, but with restrictions:
- All dialog boxes remain in Latin-1 font, i.e. it’s not possible to type in Greek indices, to search/replace for Greek characters.
- No dictionary exist for specific languages.
The following characters are not usable within FrameMaker or are displayed the wrong way (font-depending): CE - Ź and ť
The following characters are not usable within FrameMaker (language-depending):
Using languages for creating PDF documents
- Create the postscript file using "Acrobat Distiller" driver and define the required PDF settings.
- Create the PDF file with Acrobat Distiller, using "UnicodeBookmarks" (enable prolog/epilogue, pay attention to the installation instructions, switch on language specific mapping table)
Result:
- The text is converted correctly, i.e. all characters, which are displayed in FM.
- Bookmark characters, which are non-western standard, are not converted correctly and may lead to corrupted bookmarks.
Problems coming up in bookmarks
With the default settings of FrameMaker 2 problems arise in the PDF file:
- Wrong font in bookmarks
The bookmark texts are converted 1:1 to the target font without transforming it to the language specific subset.
In other words: Acrobat uses as standard a Latin-1 font. It may also use a unicode font with subsets of language-specific character sets. This must be indicated via a special code (\376\377) at the beginning of a text.
FrameMaker does not pay attention to the source language font and converts all character according to its ANSI code instead of transforming it into the language-specific subset. - Wrong coding of characters
Two serious bugs occur in FrameMaker:- FrameMaker writes some characters the wrong way into the postscript file. This leads to wrong characters or corrupt bookmarks.
- In some cases 2 different source characters are mapped on the same target character in the postscript file! That means: The entries in the postscript file are wrong, the distiller is not able to distinguish that identical characters.
- Changes in Shlomos' configuration file 'Unicode.ini'
The PlugIn has to be configured, i.e. in the file 'Unicode.ini' a specific mapping table has to be created for mapping each source character to the correct target character.
TS_charset 182 <0389> put % capital letter ETA with Tonos
Comments
