How can I copy one *bit* stream to another most efficiently?

I need a function to copy N bits from src bit stream to the offset O in the destination bit stream.

bit_copy(unsigned char *dest_bit_stream, unsigned char offset,

unsigned char *src_bit_stream, unsigned char length);

if I call it as bit_copy(dest, 13, src, 17), it should copy as follows:

src bits dest bits

[0-2] --> [13-15]

[3-10] --> [16-23]

[11-16] --> [24-29]

I have ideas in mind (such as shifting the dest 13 bits left and use bit operations), but I'm not experienced with assembly and I think there should be better C-solutions.

Any comment is greatly appreciated