fix(dot/types): *types.Body to be of type []types.Extrinsic

[kishansagathiya]

Changes

Changing Body from []byte to []Extrinsic, changes its meaning (encoded to not encoded), so the right way to do this should be as follows

• Create a new Struct BodyExtrinsics with its own methods
• Use BodyExtrinsics instead of Body everywhere and changes/use methods accordingly
• Now that old Body struct is not being used anywhere, remove it.
• Rename BodyExtrinsics to Body and accordingly rename its methods

Tests

Issues

• Closes update *types.Body to be of type []types.Extrinsic #1775

Primary Reviewer

[CLAassistant]

All committers have signed the CLA.

[codecov]

Codecov Report

Merging #1807 (760b2af) into development (5c11fb2) will increase coverage by 0.07%.
The diff coverage is 80.00%.

@@               Coverage Diff               @@
##           development    #1807      +/-   ##
===============================================
+ Coverage        59.62%   59.69%   +0.07%     
===============================================
  Files              188      188              
  Lines            19686    19669      -17     
===============================================
+ Hits             11738    11742       +4     
+ Misses            5966     5954      -12     
+ Partials          1982     1973       -9     

Flag	Coverage Δ
unit-tests	59.69% <80.00%> (+0.07%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
dot/rpc/modules/author.go	72.58% <ø> (ø)
dot/types/block_data.go	6.25% <ø> (ø)
dot/state/block.go	41.55% <57.14%> (-0.09%)	⬇️
dot/types/block.go	68.00% <66.66%> (-7.00%)	⬇️
dot/types/body.go	54.54% <70.83%> (+3.75%)	⬆️
dot/core/service.go	53.63% <100.00%> (+0.85%)	⬆️
dot/network/test_helpers.go	83.00% <100.00%> (+0.17%)	⬆️
dot/rpc/modules/childstate.go	91.48% <100.00%> (+3.61%)	⬆️
dot/state/test_helpers.go	63.77% <100.00%> (+0.57%)	⬆️
dot/sync/syncer.go	46.83% <100.00%> (+0.77%)	⬆️
... and 11 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update a04deb6...760b2af. Read the comment docs.

[arijitAD]

Any reason for the body to be a pointer? Since it is already a slice.

[kishansagathiya]

I have kept what was being used previously.
Not using pointer would be safer in general. Changing this from pointer, would mean edit this at a lot places. But, I think doing so would be the right way. Will change it.

[kishansagathiya]

At most place where body is returned it is returned as a pointer.

So if passed without pointer to the slice (Body and not *Body), you can modify it, but to an extent. From what I read, you can modify, values but not length of the slice. If you try to append to it, slice won't have values that you expected. So far, I have not seen anywhere Body getting modified.

I also see that at places a pointer to Body and a pointer is stored for even other slices.

type BlockData struct {
	Hash          common.Hash
	Header        *Header
	Body          *Body
	Receipt       *[]byte
	MessageQueue  *[]byte
	Justification *[]byte
}

@noot is returning or storing *Body (and also other slices, like MessageQueue *[]byte) done deliberately? Should I care returning Body instead of *Body?

[noot]

all the fields in the BlockData should be a pointer as they are optional types. please leave this as is

[noot]

actually, for this functions specifically, I'm not sure if it really matters. I think it could be either

[noot]

looks good! just a few comments

[noot]

@kishansagathiya could you resolve the conflicts when you get a chance?

[noot]

remove

[timwu20]

@kishansagathiya can you remove this?

[noot]

nice work :D

[timwu20]

I think this whole function can be return scale.Marshal(*b)

[timwu20]

if []Extrinsic is not supposed to be encoded, why are we marshalling here? I see that the old code did this, but should this not return [][]byte given that we are using the Extrinsic type to represent the decoded extrinsic? @noot @kishansagathiya

[kishansagathiya]

The marshaling part is correct, but I think this function should be better named as AsEncodedBytesArray. That would avoid the confusion that it is causing. And the value returned by this function is being converted into [][]byte and []string anyway, so returning [][]byte is better

[jimjbrettj]

Are we sure that the marshalling here is correct? On my branch jimmy/integrateScaleIntoTypes I had my own tests

gossamer/dot/types/block_test.go

Line 111 in 0cb69f5

func TestMustEncodeBlock(t *testing.T) {

for block encoding written which I compare the encoding to the expected byte string (hard coded so I don't test against itself) and when I merge these changes that test fails because the encoding doesn't equal what is expected. So there appears to be difference between the encodings, but I am not certain what the correct format is. For example, the body.AsEncodedExtrinsics() function specifies that it is scale encoded but that is not actually the case because there is not length prefix for the encoded slice. I am not positive if it should be scale encoded or if its just incorrect documentation but it does appear that the body encoding now is different than it was before. Perhaps @noot could help provide some clarification on this?

[github-actions]

🎉 This PR is included in version 0.6.0 🎉

The release is available on GitHub release

Your semantic-release bot 📦🚀