| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| |
|
|
| |
Restricting encoding is silly.
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
This commit makes the wrapper types more generally applicable.
encoding.TextMarshaler is supported by most codec implementations (e.g.
for yaml).
The tests now ensure that package json actually recognizes the custom
marshaler implementation irrespective of how it is implemented.
The Uint type has new tests, too. These are tricky because uint size
depends on the CPU word size. Turns out that there was one incorrect
case where decoding returned ErrUint64Range instead of ErrUintRange.
|
| |
|
|
|
| |
All other functions return errors from package hexutil, ensure that
Decode does too.
|
| |
|
|
| |
This follows the change to common/math big integer parsing in PR #3699.
|
| |
|
|
|
| |
The code was too clever and failed to include zeros on a big.Word
boundary.
|
| | |
|
|
|
The new package is purpose-built to handle the encoding consumed and
produced by the RPC API.
|
