Adds support for eexps with 12- and 20-bit addresses #825
Conversation
| EExpressionWith4BitAddress, // 0x00-0x3F - | ||
| EExpressionWith12BitAddress, // 0x40-0x4F - | ||
| EExpressionWith20BitAddress, // 0x50-0x5F - |
There was a problem hiding this comment.
🗺️ Of this block, these three lines (EExpressionWith*BitAddress) are the new/interesting ones. The others are highlighted because the comments were automatically realigned.
| EExpressionWith4BitAddress | ||
| | EExpressionWith12BitAddress | ||
| | EExpressionWith20BitAddress | ||
| | EExpressionWithLengthPrefix => { | ||
| ParseValueExprResult::EExp(self.read_e_expression(opcode)) |
There was a problem hiding this comment.
🗺️ I've been considering adding a less granular OpcodeKind or similar alongside OpcodeType to capture high-level groups like EExpression or Value for places like this where we don't actually care about the encoding just yet.
| &[ | ||
| // Opcode with low nibble setting bias | ||
| 0x50 | low_opcode_nibble as u8, | ||
| ], |
There was a problem hiding this comment.
| &[ | |
| // Opcode with low nibble setting bias | |
| 0x50 | low_opcode_nibble as u8, | |
| ], | |
| // Opcode with low nibble setting bias | |
| 0x50 | low_opcode_nibble as u8, |
There was a problem hiding this comment.
The extend_from_slices_copy method takes a slice-of-slices (&[&[u8]]). For potentially interesting information about the motivation for this method, check out the PR that introduced it.
Now that I'm re-reading this, though, I could rewrite it to call (biased as u16).to_le_bytes() in advance and pack the two bytes into this slice manually. I think I'll do that.
There was a problem hiding this comment.
I see that now. I missed that the call on L701 was extend_from_slice_copy whereas the call on L713 was extend_from_slices_copy.
Prior to this PR, e-expressions with an address between
0..64could be encoded as a single-byte opcode and larger e-expressions could only be encoded as a length-prefixed e-expression.This patch adds read and write support for binary e-expressions that have a trailing 1- or 2-byte
FixedUIntto complete the address.By submitting this pull request, I confirm that my contribution is made under the terms of the Apache 2.0 license.