A “fat” Teddy implementation that is generic over both the vector type
and the minimum length of the patterns being searched for.
Represents the low and high nybble masks that will be used during “fat”
Teddy search.
A vector generic mask for the low and high nybbles in a set of patterns.
Each 8-bit lane j in a vector corresponds to a bitset where the ith bit
is set if and only if the nybble j is in the bucket i at a particular
position.
A match type specialized to the Teddy implementations below.
A “slim” Teddy implementation that is generic over both the vector type
and the minimum length of the patterns being searched for.
Represents the low and high nybble masks that will be used during
search. Each mask is 32 bytes wide, although only the first 16 bytes are
used for 128-bit vectors.
The common elements of all “slim” and “fat” Teddy search implementations.