Keep in cache chunks that may be needed in later parts of the download stream

[coyotte508]

Eg sometimes the terms look like this:

    {
      "hash": "1c7d639dd1c16ad28b8acac35ad54bbc4000222f40e40a189913390eac09041b",
      "unpacked_length": 22778309,
      "range": { "start": 0, "end": 346 }
    },
    {
      "hash": "023f73d1ee9ab8c30f58f00e2ab701b4757dd1095720835923584fa27a6f9d20",
      "unpacked_length": 8128,
      "range": { "start": 629, "end": 630 }
    },
    {
      "hash": "1c7d639dd1c16ad28b8acac35ad54bbc4000222f40e40a189913390eac09041b",
      "unpacked_length": 14036784,
      "range": { "start": 346, "end": 571 }
    },
    {
      "hash": "023f73d1ee9ab8c30f58f00e2ab701b4757dd1095720835923584fa27a6f9d20",
      "unpacked_length": 8128,
      "range": { "start": 858, "end": 859 }
    },
    {
      "hash": "1c7d639dd1c16ad28b8acac35ad54bbc4000222f40e40a189913390eac09041b",
      "unpacked_length": 30196753,
      "range": { "start": 571, "end": 1087 }
    },

and in fetch_info, we have this:

    "1c7d639dd1c16ad28b8acac35ad54bbc4000222f40e40a189913390eac09041b": [
      {
        "range": { "start": 0, "end": 1087 },
        "url": "https://s3.us-east-1.amazonaws.com/xethub-xorb-bucket/xorbs/default/1c7d639dd1c16ad28b8acac35ad54bbc4000222f40e40a189913390eac09041b?x-id=GetObject&X-Amz-Algorithm=AWS4-HMAC-SHA256&X-Amz-Credential=AKIA2JU7TKAQLRJCO34O%2F20250312%2Fus-east-1%2Fs3%2Faws4_request&X-Amz-Date=20250312T123613Z&X-Amz-Expires=3600&X-Amz-SignedHeaders=host%3Brange&X-Amz-Signature=8aae8823a8a19578faf34bf92879d8bce4b0cf7459ec17152017fda36d600448",
        "url_range": { "start": 0, "end": 67023668 }
      }
    ],

With current code, the whole section is fetched three times instead of only once.

Instead, we should do only one http call and save the data for when it's needed and clean it up when it's not needed afterwards