in reply to Re^3: memory-efficient hash kind for incremental sort
in thread memory-efficient hash kind for incremental sort

Now citation databases are something I know a bit about. Is there a reason you are not using some of the existing art (e.g. EndNote, Reference Manager, ProCite ...)? What level of metadata are you trying to maintain? One thing you should certainly be aware of is the problem of author ambiguity - multiple individuals can be cited with the same string and multiple strings can represent the same individual - and that problem is not generally tractable without brute force effort and significant knowledge of the field. Are you essentially trying to implement an automated keyword generation?
  • Comment on Re^4: memory-efficient hash kind for incremental sort