The current version of BibTEX is not built for 8-bit character processing, even though TEX 3 does support it. This forces users who want access to the high bit IBM character set to either keep their bibliographies in 7-bit form, or forgoe compatibility with most implementations of BibTEX.
In order to deal with this problem, BibDB provides automatic and user transparent conversion between 7-bit text and 8-bit text. Using the configuration file it is possible to tell BibDB to keep the database file in 7-bit (TEX) or 8-bit format, and independently to display and edit the file in BibDB in 7-bit or 8-bit mode. BibDB can also be used to convert files to and from 8-bit form, thus restoring to the avid 8-bit user (mostly Germans, I have found!) compatibility with the rest of the world.
When the program is in 8-bit mode, entries read from the file are converted to 8-bit form before being displayed. In addition, text entered by the user into entries and patterns is automatically converted. This ensures that you do not need a special keyboard in order to conveniently use 8-bit mode, as e.g. the string `` \"u" will be automatically converted to ``ü". A point to watch though is entering accented ``i" characters — the string `` \" \i a" will be converted to ``ïa", but the string `` \" \ia" will remain unchanged.
When using BibTEX styles such as alpha.bst, BibTEX 0.99c requires all TEX form accented characters in fields such as ``author" and ``editor" to be enclosed in braces, in order for its character counting algorithm to operate properly. BibDB supports this requirement through designating the desired fields as ``authorlike" in the configuration file. Braces will then be added or removed as needed during conversion to and from 8-bit mode. Note that BibDB will not add appropriate braces unless a mode conversion is performed — that is up to the user.