GitLab

This commit:
- reduce the API to 2 simpler functions
- add consistent and clear control over if the root should be visited
- add control over the concurrency factor

This fixes the inconsistent EnumerateChildrenAsync behavior by renaming
everything.

1. EnumerateChildrenAsync has been renamed to WalkParallel because (a) it
visited the root so EnumerateChildren was incorrect and (b) it isn't async.
2. EnumerateChildren has been renamed to Walk and now _also_ visits the root for
consistency.

Anyone needing to skip the root can do so manually (`if c.Equals(root)` ...).

Given all the bitswap session improvements, I think it's time to give our users
some candy. This should allow us to actually take advantage of the increased
wantlist sizes.

fixes https://github.com/ipfs/go-ipfs/issues/5793

This adds also EnumerateChildren*Depth which call a visit function which
receives the Cid and the depth in the dag at which it has been seen.

alternative to #5243 that updates go-cid and all packages that depend on it

License: MIT
Signed-off-by: Steven Allen <steven@stebalien.com>

* go-log
* sys
* go-crypto

License: MIT
Signed-off-by: Steven Allen <steven@stebalien.com>

License: MIT
Signed-off-by: Łukasz Magiera <magik6k@gmail.com>

Otherwise, GetMany on the children of a node with duplicate links may fail

License: MIT
Signed-off-by: Steven Allen <steven@stebalien.com>

No idea why this was changed this was introduced in:

08f342e8bada4f4eb0c5462a3623f0c2b828240f (part of #3598)

License: MIT
Signed-off-by: Steven Allen <steven@stebalien.com>

License: MIT
Signed-off-by: Łukasz Magiera <magik6k@gmail.com>

License: MIT
Signed-off-by: Hector Sanjuan <hector@protocol.ai>

License: MIT
Signed-off-by: Hector Sanjuan <hector@protocol.ai>

License: MIT
Signed-off-by: Jeromy <jeromyj@gmail.com>

License: MIT
Signed-off-by: Jeromy <jeromyj@gmail.com>

License: MIT
Signed-off-by: Steven Allen <steven@stebalien.com>

Also:

* Update the blockstore/blockservice methods to match.
* Construct a new temporary offline dag instead of having a
  GetOfflineLinkService method.

License: MIT
Signed-off-by: Steven Allen <steven@stebalien.com>

License: MIT
Signed-off-by: Steven Allen <steven@stebalien.com>

License: MIT
Signed-off-by: ForrestWeston <forrest@protocol.ai>

License: MIT
Signed-off-by: Steven Allen <steven@stebalien.com>

License: MIT
Signed-off-by: Łukasz Magiera <magik6k@gmail.com>

1. Modern storage devices (i.e., SSDs) tend to be highly parallel.
2. Allows us to read and write at the same time (avoids pausing while flushing).

fixes https://github.com/ipfs/go-ipfs/issues/898#issuecomment-331849064

License: MIT
Signed-off-by: Steven Allen <steven@stebalien.com>

License: MIT
Signed-off-by: Łukasz Magiera <magik6k@gmail.com>

License: MIT
Signed-off-by: Steven Allen <steven@stebalien.com>

License: MIT
Signed-off-by: Jeromy <jeromyj@gmail.com>

License: MIT
Signed-off-by: Jeromy <jeromyj@gmail.com>

License: MIT
Signed-off-by: Łukasz Magiera <magik6k@gmail.com>

License: MIT
Signed-off-by: Arthur Elliott <clownpriest@gmail.com>

License: MIT
Signed-off-by: Arthur Elliott <clownpriest@gmail.com>

I previously optimized the IPLD cbor decoder to *not* encode and then re-decode
objects when constructing them with `WrapObject`. Unfortunately, we rely on this
to canonicalize the object before computing the tree/links.

License: MIT
Signed-off-by: Steven Allen <steven@stebalien.com>

Later, we should:

1. Pull the other node formats out of IPFS (at least the raw one).
2. Pull out the decoder registration/management into a `go-ipld` library.

License: MIT
Signed-off-by: Steven Allen <steven@stebalien.com>

License: MIT
Signed-off-by: Jeromy <jeromyj@gmail.com>

License: MIT
Signed-off-by: Jeromy <jeromyj@gmail.com>

License: MIT
Signed-off-by: Jeromy <jeromyj@gmail.com>

License: MIT
Signed-off-by: Jeromy <jeromyj@gmail.com>

License: MIT
Signed-off-by: Łukasz Magiera <magik6k@gmail.com>