Suggestion for potential addition for better storage management.

[namanshergill]

Idea 1: Maybe an option to specify the max amount of items (or size) the cache stores at a time, in the interceptor, Currently it seems like the package caches everything ~~and never deletes them (even in the case mentioned below), unless a new response overwrites it~~, which could lead to uncomfortable cache sizes for some apps.

~~Idea 2: Delete the data from cache if it comes up as expired when checked.~~

Whoops sorry, wrote this while half asleep, confused maxAge and maxStale.

[namanshergill]

I'm still going through the code so I apologise if I missed and mentioned something the package already does!

[llfbandit]

You're right there's nothing to implictly delete old cached values against a given limit. There's no access time set to the values to handle this.

As I use this package for one my project, I'm not too worried by the data size. I store many quite big JSON entities for a year period and the database is sized around 3-4MB when fully populated.

FYI, this package is designed for small / medium data, not to store images and so on like we can do when navigating on web browsers.

Despite this, each store calls clean(staleOnly: true) to remove values with maxStale when instanciated which not cover your proposal.

This is a good suggestion. I'll think about this.

[namanshergill]

You're right there's nothing to implictly delete old cached values against a given limit. There's no access time set to the values to handle this.

As I use this package for one my project, I'm not too worried by the data size. I store many quite big JSON entities for a year period and the database is sized around 3-4MB when fully populated.

FYI, this package is designed for small / medium data, not to store images and so on like we can do when navigating on web browsers.

Ah, I did not know it actually took only that much space!

Despite this, each store calls clean(staleOnly: true) to remove values with maxStale when instanciated which not cover your proposal.

This is a good suggestion. I'll think about this.

Oh, so basically all stale values are deleted? I thought it was only done for a specific entry when it was checked and it turned out to be stale. I did feel I must've missed something in the code.

There's no access time set to the values to handle this.

Makes sense. Plus if the package purges all stale data by itself anyway I guess it will always stay a manageable size in most cases, as you mentioned.

[llfbandit]

Re-opening, this is still a valid request.