Internal improvement: Move selector computation from codec into abi-utils
#5465
Conversation
haltman-at
requested review from
gnidan,
benjamincburns,
cds-amal,
lsqproduction,
eggplantzzz and
sukanyaparashar
haltman-at
added
new-feature: @truffle/abi-utils
and removed
new-feature: @truffle/abi-utils
labels
This makes sense! Good call. |
|
More specifically the reason it came up is because I'm planning a substantial Codec breakage to deal with this internal function pointers problem. :P Btw, you may notice I added the |
packages/abi-utils/lib/signature.ts
Outdated
| const tupleMatch = parameter.type.match(/tuple(.*)/); | ||
| if (tupleMatch === null) { | ||
| //does not start with "tuple" |
There was a problem hiding this comment.
The comment says does not start with tuple but the regex doesn't use ^. Is that correct?
There was a problem hiding this comment.
Oops! Good catch. This code will work anyway, because there's no other place that tuple can occur, but I should change it to not rely on that, thanks.
|
Putting the label back on at @gnidan's insistence. |
This PR moves code for computing signatures and selectors from
codecintoabi-utils. Note that we continue to re-export these functions from Codec, in order to avoid a breaking change. I also added a few tests of these functions inabi-utils-- not property-based, sorry, I couldn't figure out a good way to do that. On top of that, I expanded theabi-utilsREADME to cover this new functionality.Since this PR is mostly just moving code around and not changing it, there's not a lot to say here about the code itself. However, I wanted to record why I'm doing this.
Basically, it comes back to the abi-to-sol problem. I want to be able to make breakiing changes to Codec without a repeat of the problem that caused last time (see #5396 for an explanation). This is the only code from Codec that abi-to-sol uses. Moving it to abi-utils will of course not get rid of the problem, but it reduces it, since it means that the problem will only come up when we break abi-utils rather than when we break codec. Since we break abi-utils hardly ever, and break codec somewhat frequently, this acts as a stopgap to allow us to avoid problems until we figure out a better solution.