Incremental file unpacking (was "reading a small amount of data actually reads a whole file from the zip")

[Former user]

Originally reported on Google Code with ID 32 ``` We're looking into using fuse-zip in conjunction with httpfs to read headers from files in remote zips, *without* downloading the whole zip, or even whole files within the zip.

To that end I did some strace analysis of fuse-zip operations. The results are here: https://gist.github.com/8249054df8e03889bdc3

In summary, reading the first 10 bytes of a 10MB file within a zip causes fuse-zip to read the whole 10MB, rather than stopping after the uncompressed data becomes >= 10 bytes.

```

Reported by `craig.ds` on 2012-01-10 09:22:29

[Former user]

``` Currently fuse-zip does not supports incremental file reading. The whole file are loaded in fopen() call.

May be fopen behaviour will be improved in the future releases, but you should not expect this in the nearest months. ```

Reported by `alexander.galanin` on 2012-04-20 16:59:04 - Labels added: Type-Enhancement

[Alexander Galanin]

Issue #53 was marked as a duplicate of this issue.