Counting index hunks by iteration is slightly inefficient

[sourcefrog]

From #133 (comment)

Current code (

conserve/src/index.rs

Lines 237 to 256 in bd6759f

	pub fn count_hunks(&self) -> Result<u32> {
	// TODO: Might be faster to list the directory than to probe for all of them.
	// TODO: Perhaps, list the directories and cope cleanly with
	// one hunk being missing.
	for i in 0.. {
	let path = hunk_relpath(i);
	if !self
	.transport
	.exists(&path)
	.map_err(|source| Error::ReadIndex {
	source,
	path: path.into(),
	})?
	{
	// If hunk 1 is missing, 1 hunks exists.
	return Ok(i);
	}
	}
	unreachable!();
	}

) to determine the number of index hunks probes for each of them, which is a bit inefficient. However this is only done once per restore, so it's not really a hot path.

Better ways to do this:

• look in the band tail for a count, if it's there
• otherwise, list the index directories