Bug#1735: apropos(1) segfaults

[Ian Jackson <ian@chiark.chu.cam.ac.uk>]

Andrew Howell writes ("Bug#1735: apropos(1) segfaults"):
> Bill Mitchell writes:
> > root:/root# apropos crap
> > Segmentation fault (core dumped)
> > root:/root#
> > Script done on Sun Oct 22 18:55:58 1995
>
> I've had this problem before as well, but I discovered that the man database
> was screwed up and forcing it to rebuild it with mandb fixed it. Still it
> shouldn't really core dump even if the database is corrupt.

Then there are at least two or three bugs:

1a. The database got corrupted.
1b. The database didn't get fixed because the corruption wasn't
    detected, or because no appropriate action was taken.  If
    corruption is made sufficiently rare then having it automatically
    un-corrupt it on demand is less necessary.  This is obviously
    better than having it get corrupted and then be fixed.

2. `apropos' dumps core when fed a corrupted database.

Ian.