feat!: adopt piece cid v2 #21
Conversation
test/piece.spec.js
Outdated
| // assert.deepEqual(piece.paddedSize, data.out.paddedSize) | ||
|
|
||
| const json = piece.toJSON() | ||
| // assert.deepEqual(piece.toInfo(), { | ||
| // link: parseLink(data.out.cid), | ||
| // height: Math.log2(data.out.size / Node.Size), | ||
| // }) | ||
|
|
||
| assert.deepEqual(json, { | ||
| link: { | ||
| '/': data.out.cid, | ||
| }, | ||
| height: Math.log2(data.out.size / Node.Size), | ||
| }) | ||
|
|
||
| const view = Piece.fromJSON(json) | ||
| assert.deepEqual(view.link, piece.link) | ||
| assert.deepEqual(view.size, piece.size) | ||
| assert.deepEqual(view.height, piece.height) | ||
| // const view = Piece.fromJSON(json) | ||
| // assert.deepEqual(view.link, piece.link) | ||
| // assert.deepEqual(view.size, piece.size) | ||
| // assert.deepEqual(view.height, piece.height) |
There was a problem hiding this comment.
looks like something in progress stayed here?
test/aggregate.spec.js
Outdated
| // assert.deepEqual( | ||
| // JSON.stringify(build), | ||
| // JSON.stringify({ | ||
| // link: build.link, | ||
| // height: Math.log2((1 << 20) / Node.Size), | ||
| // }) | ||
| // ) |
There was a problem hiding this comment.
looks like something in progress stayed here?
test/aggregate.spec.js
Outdated
| /Pieces are too large to fit in the index/ | ||
| ) | ||
| }, | ||
|
|
||
| 'pass bad value to estimate': async (assert) => { | ||
| 'skip! pass bad value to estimate': async (assert) => { |
There was a problem hiding this comment.
what is the reason to skip this one?
There was a problem hiding this comment.
It became impossible to pass bad values (see line 172) because we switch to height and therefor size is always pow2. I probably will remove this test, I just had it disabled as I was trying to figure out what to do with it.
|
|
||
| get link() { | ||
| if (this._link == null) { | ||
| this._link = Link.create( |
There was a problem hiding this comment.
@Gozala this creates a LegacyLink for the AggregateBuild.toInfo() which should not be expect right?
There was a problem hiding this comment.
I'm not sure I understand the question here. PieceInfo corresponds to same named go struct which uses older version of CID, which is why we create a legacy link here.
There was a problem hiding this comment.
It is related to getting an Aggregate CID being Piece CIDv2. I think this is important for us for metrics at least, be able to know by Aggregate CID their encoded height/size
There was a problem hiding this comment.
Now, it looks like only CID we can get from here is through AggregateBuild.toInfo(). Maybe we just want a getter for Piece CIDv2 from aggregate
There was a problem hiding this comment.
See what I mean in practise https://github.com/web3-storage/w3up/pull/850/files
test/aggregate.spec.js
Outdated
| // assert.deepEqual( | ||
| // JSON.stringify(build), | ||
| // JSON.stringify({ | ||
| // link: build.link, | ||
| // height: Math.log2((1 << 20) / Node.Size), | ||
| // }) | ||
| // ) |
There was a problem hiding this comment.
| // assert.deepEqual( | |
| // JSON.stringify(build), | |
| // JSON.stringify({ | |
| // link: build.link, | |
| // height: Math.log2((1 << 20) / Node.Size), | |
| // }) | |
| // ) |
Upgrades `data-segment` lib with Piece CID v2 storacha/data-segment#21 Needs: - [x] figure out legacy CID for storacha/data-segment#21 (review)
Aligns implementation around Piece CID V2