Add v1util.EstargzReadCloser utility. #870
Conversation
Codecov Report
@@ Coverage Diff @@
## master #870 +/- ##
==========================================
+ Coverage 74.28% 74.34% +0.05%
==========================================
Files 106 107 +1
Lines 4476 4486 +10
==========================================
+ Hits 3325 3335 +10
Misses 656 656
Partials 495 495
Continue to review full report at Codecov.
|
mattmoor
force-pushed
the
estargz-utils
branch
from
bcc3fa1
to
dcf13b4
Compare
|
Now with 100% coverage 😛 |
|
Commented on Slack, but for visibility:
|
pkg/v1/v1util/zip.go
Outdated
| // returns an io.ReadCloser from which compressed data may be read. | ||
| // Refer to estargz for the options: | ||
| // https://pkg.go.dev/github.com/containerd/stargz-snapshotter@v0.2.0/estargz#Option | ||
| func EstargzReadCloser(r io.ReadCloser, opts ...estargz.Option) (io.ReadCloser, v1.Hash, error) { |
There was a problem hiding this comment.
Can we put this under pkg/internal/estargz while we're experimenting and expose it later? I'm not sure that I want to commit to this interface quite yet, and v1util is depended on by basically every package in this module, so it can't get pruned if it's in here.
Also estargz.NewRead[Close?]er reads nicely 😛
There was a problem hiding this comment.
This should be done, and I blew up the rest of v1util in: #872 for consistency.
pkg/v1/v1util/zip.go
Outdated
| func EstargzReadCloser(r io.ReadCloser, opts ...estargz.Option) (io.ReadCloser, v1.Hash, error) { | ||
| defer r.Close() | ||
|
|
||
| bs, err := ioutil.ReadAll(r) |
There was a problem hiding this comment.
Assuming this is just for prototyping and we won't always buffer the whole thing for a final implementation? We should be teeing r into a hasher, right?
There was a problem hiding this comment.
Yeah, I don't love this implementation either, but I think the estargz stuff wants to be able to seek around to build the index and reorder things for the final estargz file. The v1.Hash is not the overall hash, but the hash of the table of contents. I owe a better function comment than this has, so let me take a crack at that now.
Are you me?? |
Why not patch upstream? |
Very finite amounts of time. |
So you want to maintain a fork of something... right... 🤣 |
This adds the first utility function we will need to start producing estargz layers.
mattmoor
force-pushed
the
estargz-utils
branch
from
6beb862
to
3983a88
Compare
mattmoor
force-pushed
the
estargz-utils
branch
from
2648c64
to
ed4e70f
Compare
This adds the first utility function we will need to start producing estargz layers.
Related: #866