A bit vector is just an array of bits; subsets of bits within the bytes have some meaning. That allows more compact storage for certain types of data. For example, if you only needed boolean on-off values, you can store eight values in one byte.
If your values require more than byte sized bits, the bits you require can still be packed more efficiently; if you need 10 bits, you can pack eight of those in an ten byte string (rather than the 16 or 32 bytes you might otherwise use).
The use of vector in this context probably came from jump tables: the bits represent a place for code to jump to, and therefor are at least vaguely related to the physics/engineering definition of a vector quantity (maginitude and direction). I still think of these as bit maps or bit fields, but apparently I’m out of touch.
Perl has the “vec” function for bit fields. Its granularity is a little bit limited: the number of bits you want to examine or set has to be a power of 2, so you can’t (for example) conveniently work with three bit fields. You’d need to use four bits, a small waste, or handle all the nasty details yourself with substr and > operators. Using “vec” is a lot more pleasant. Here’s an example sets some individual bits, and reads them back from the string.
#!/usr/bin/perl
my $bitf;
vec($bitf,0,1)=1;
vec($bitf,1,4)=7;
for($x=0;$x
If you don't see why setting bits 4 to 7 as 7 produces "110", it's just simple banary, starting from bit 4: bit 4 has the value of 1 if set, bit 5 has 2, bit 6 has 4.
*Originally published at APLawrence.com
A.P. Lawrence provides SCO Unix and Linux consulting services http://www.pcunix.com
