This pull request allows either strings or bytes to be supplied as input to the constructor and methods such as lcs().
To avoid errors in operations with conflicting types, the input type is stored and checked when necessary, under field input_type.
One inconvenient of handling raw bytes is that the terminal symbol becomes a multi-byte sequence (e.g. 3 bytes) when converted from the generated unicode char. Some conditionals were adapted to handle this larger terminal symbol length. One function used in those conditionals is _suffix_contains_terminal_symbol(). The validations on this function handle some cases I stumbled upon (e.g. suffixes with 2 bytes from the terminal symbol; suffixes with one original byte plus 2 bytes from the terminal symbol). Feel free to suggest an alternative approach if you want something more robust.
This pull request allows either strings or bytes to be supplied as input to the constructor and methods such as
lcs()
.To avoid errors in operations with conflicting types, the input type is stored and checked when necessary, under field
input_type
.One inconvenient of handling raw bytes is that the terminal symbol becomes a multi-byte sequence (e.g. 3 bytes) when converted from the generated unicode char. Some conditionals were adapted to handle this larger terminal symbol length. One function used in those conditionals is
_suffix_contains_terminal_symbol()
. The validations on this function handle some cases I stumbled upon (e.g. suffixes with 2 bytes from the terminal symbol; suffixes with one original byte plus 2 bytes from the terminal symbol). Feel free to suggest an alternative approach if you want something more robust.