python Cachetools can items have different ttl?

Question

Im using @cachetools.func.ttl_cache(maxsize=3, ttl=3600, timer=time.time, typed=False) to cache different data frames. The function being wrapped doesn't build the DF itself, but given an argument calls the right function.

Depending on the argument the DF may be time consuming or fast to build, given that I want to modify the item ttl (time-to-live). So that item 1 may have ttl=3600 while item 2 ttl=10800.

Is that functionality supported? using a global variable or any other way? docs

Answer 1

Will is most likely correct, but just in case you want to try, I think subclass TTLCache and overwrite one function should work:

from cachetools import Cache, TTLCache

cache = TTLItemCache(maxsize=2, ttl=100)
# ttl=100
cache.__setitem__('key1', 'val1')
# ttl=200
cache.__setitem__('key2', 'val2', ttl=200)

class TTLItemCache(TTLCache):
    def __setitem__(self, key, value, cache_setitem=Cache.__setitem__, ttl=None):
        super(TTLItemCache, self).__setitem__(key, value)
        if ttl:
            link = self._TTLCache__links.get(key, None)
            if link:
                link.expire += ttl - self.ttl