Lexichem TK 2.4.2

Important note

  • Custom dictionary entry points added in version 2.4.0 (but not documented) have been removed pending redesign/reimplementation.

New features

  • New translate examples have been added for all languages (C++, C#, Java, Python).

Minor bug fixes

  • A memory leak that occurred when naming certain structures with Von Baeyer rings has been fixed.

  • String length overflows in language translation code have been fixed. These overflows were functionally benign but did show up with certain sanitizing compilers.

  • The internal name dictionary has been cleaned up:

    • The dictionary lookup for names for which one entry is a strict substring of another(eg: AZT and AZTREONAM) has been fixed.

    • All nitro groups are now charge-separated (4-valent).

    • Dash vs. space separators are now handled gracefully. Dictionary entries with spaces now also match input names containing dashes.

  • Warnings for name conversion failures have been modified. When a name cannot be converted, the warning is now “Failed to parse name: BADNAME”. Additional developer-oriented messages about “unexpected tokens” are now debug-level messages and are not seen by default.

Documentation fixes

  • PDF documentation will no longer throw a warning about “Arial-Bold bad BBox” in some PDF readers.