[spec] Leading + character when parsing amount
#101
Comments
antonok-edm
changed the title
+ character when parsing amount+ character when parsing amount
|
Yes, forbidding a leading plus would be good. This was intentional in the reference implementation, but we could stand to make the spec more strict. Also, awesome project! I can't view your Gitlab, so I'm guessing it's still private? |
|
Since you're doing an awesome job of reviewing the spec and finding places where it's ambiguous or inconsistent, would you and @chitalian take a look at #77 and the latest spec there? Transfer requests should be mostly the same, but transaction requests are new. Wallets are actively implementing this now, but it's still open for feedback! |
Sounds good 😄
Yep, we're in the never-ending "almost ready to publish" phase 😅 We'll be submitting it as an entry in the Riptide hackathon though, so it'll be up soon! You're welcome to add it to the Edit: it's published now!
I took a quick scan over it, I will add my thoughts there. |
@chitalian and I are working on a Rust port of the
@solana/payJavaScript SDK. Many float implementations, including the one in the Rust standard library and theBigNumberlibrary used by the official SDK, consider a leading+to be valid when parsing numbers. We noticed that the description in the spec does not address the behavior of a leading+character. For reference, it looks like the official SDK happens to forbid a leading+, although it's unclear from the code whether or not it's intentional. Forbidding a leading+feels like the correct choice, but it would be great to have that clarified in the spec.The text was updated successfully, but these errors were encountered: