in many (most ?? ALL???) of the regex searches in wavgen.py, the strings used to contain the regex are not raw strings (i.e. not written as r"").
Therefore, the first time any of these gets compiled you get a LOT of spurious warnings, because for example python tries to interpret "\s" as an escape character, which it isnt.
This is likely trivial to fix by making all those regex definitions raw strings.
But it must be pointed out , at th emoment, at least on recent (i.e. within the last decade) python, these regexes are likely not doing what we think they're doing, because of this issue.... So we may need to tread a little carefully