6 Commits

Author SHA1 Message Date
Oleksandr Kozachuk 15f8005b6d Merge pull request #3 from ok2/fix/release-strip-breaks-proc-macros
fix(build): exempt build scripts and proc-macros from release strip
2026-08-06 20:46:08 +02:00
Oleksandr Kozachuk d55a27873e fix(build): exempt build scripts and proc-macros from release strip
`wasm-pack build --release` died with "can't find crate" for rustversion,
then thiserror_impl, then every other proc-macro. Cargo strips debuginfo
from release artifacts by default and on macOS that takes the metadata
proc-macro dylibs need to be loadable with it, so rustc could no longer
open them.

Debug builds are unstripped, which is why the whole test suite stayed
green while the browser REPL could not be built for production at all.

Stripping buys nothing for build scripts and proc-macros, so
[profile.release.build-override] exempts them; release binaries stay
stripped.

Also pins wafer-core to 0.2.3 in wafer-web and wafer-cli — both still
asked for 0.2.1. The caret requirement resolved, so nothing broke.
2026-08-06 20:16:41 +02:00
Oleksandr Kozachuk 645b00d6e8 Merge pull request #2 from ok2/feature/swiftforth-number-conversion
feat(core): SwiftForth input number conversion, DPL and NH
2026-08-06 19:57:04 +02:00
Oleksandr Kozachuk 9efb92ddc8 docs(core): correct sign handling note in number conversion docs
The doc comments still claimed a leading + binds as a sign. It does not:
sf64 converts +7 as the double 7 with DPL 1, and the code follows that.
Only a leading - is a sign.
2026-08-06 19:25:48 +02:00
Oleksandr Kozachuk 706c73ce2a feat(core): SwiftForth input number conversion, DPL and NH
Punctuation (`,` `.` `+` `/` `:` and an embedded `-`) after the leftmost
digit now forces double-cell conversion, so `12.34`, `1,234`, `12:30:45`
and `2026-08-06` convert as doubles. Only a trailing `.` worked before,
and `1.5` was an "unknown word" error.

The punctuation is a double-cell marker, not a fractional point, so the
scale has to travel separately: DPL carries the digit count right of the
rightmost punctuation character (negative when there was none), which is
what lets `<# #>` place the point back on output. NH carries the high
cell a single-cell conversion drops, so a token that overflows a cell is
still recoverable as a double.

parse_number and parse_double_number duplicated the prefix and sign
handling and could not share a DPL counter, so they collapse into one
parse_numeric_literal that reports which kind it converted.

Verified token-for-token against sf64. One deliberate divergence: WAFER
keeps accepting a sign before a base prefix (`-$FF`), which sf64 rejects.
2026-08-06 19:19:41 +02:00
Oleksandr Kozachuk a89d7ca704 chore(ci): dprint-format changelog; dedupe sha crates; deny skips for wasmtime 47
CI / check (push) Has been cancelled
2026-08-06 16:15:02 +02:00
10 changed files with 395 additions and 184 deletions
+44
View File
@@ -5,6 +5,50 @@ All notable changes to WAFER are documented in this file.
The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/), The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/),
and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html). and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
## [0.2.3] - 2026-08-06
### Fixed
- **Release builds of `wafer-web` no longer fail on proc-macro loading.**
Cargo strips debuginfo from release artifacts by default, and on macOS
that also strips the metadata proc-macro dylibs need to be loadable, so
`wasm-pack build --release` died with `can't find crate` for
`rustversion`, `thiserror_impl` and every other proc-macro. Build
scripts and proc-macros gain nothing from stripping, so
`[profile.release.build-override]` now exempts them; release binaries
stay stripped. Debug builds were never affected, which is why the test
suite stayed green while the browser REPL could not be built for
production.
- `wafer-web` and `wafer-cli` requested `wafer-core` version `0.2.1`
while the workspace had moved to `0.2.2`. The caret requirement still
resolved, so nothing broke, but the pin is now kept in step.
## [0.2.2] - 2026-08-06
### Added
- **SwiftForth-style input number conversion.** Punctuation (`,` `.` `+`
`/` `:` and an embedded `-`) anywhere after the leftmost digit now forces
double-cell conversion, so `12.34`, `1,234`, `12:30:45` and `2026-08-06`
all convert as doubles without a custom parser. Previously only a
trailing `.` worked and `1.5` was an "unknown word" error. The
punctuation is a double-cell marker, not a fractional point: every
spelling of `1234` (`1234.`, `123.4`, `.1234`) yields the same value.
- **`DPL`** ( -- addr ): digits to the right of the rightmost punctuation
character in the last converted number, negative when the token carried
none. Seeded at -1024 and bumped once per digit, matching `sf64`.
Together with `<# #>` this is how fixed-point input is scaled.
- **`NH`** ( -- addr ): the high-order cell dropped by a single-cell
conversion, so a token that overflows a cell can be recovered as a
double (`4000000000 NH @ D.`).
Verified token-for-token against SwiftForth `sf64`: DPL values, double
promotion and sign handling agree on every probed form. One deliberate
divergence — WAFER also accepts a sign before a base prefix (`-$FF`), which
`sf64` rejects; the Forth 2012 spelling `$-FF` works in both. A leading `+`
is punctuation rather than a sign in both engines, so `+7` is the double 7
with `DPL` = 1.
## [0.2.1] - 2026-08-06 ## [0.2.1] - 2026-08-06
### Fixed ### Fixed
+1 -1
View File
@@ -79,7 +79,7 @@ Handle in `interpret_token_immediate()` or `compile_token()` as a special case.
## Testing ## Testing
- Run `cargo test --workspace` before committing (currently 542 unit + 1 benchmark + 11 compliance + 9 comparison + 5 crypto) - Run `cargo test --workspace` before committing (currently 562 unit + 1 benchmark + 11 compliance + 9 comparison + 5 crypto)
- Forth 2012 compliance: `cargo test -p wafer-core --test compliance` - Forth 2012 compliance: `cargo test -p wafer-core --test compliance`
- Cross-engine comparison (vs gforth): `cargo test -p wafer-core --test comparison` - Cross-engine comparison (vs gforth): `cargo test -p wafer-core --test comparison`
- Performance benchmarks (release mode): `cargo test -p wafer-core --test comparison -- --nocapture --ignored` - Performance benchmarks (release mode): `cargo test -p wafer-core --test comparison -- --nocapture --ignored`
Generated
+23 -87
View File
@@ -132,15 +132,6 @@ dependencies = [
"generic-array", "generic-array",
] ]
[[package]]
name = "block-buffer"
version = "0.12.1"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "d2f6c7dbe95a6ed67ad9f18e57daf93a2f034c524b99fd2b76d18fdfeb6660aa"
dependencies = [
"hybrid-array",
]
[[package]] [[package]]
name = "bumpalo" name = "bumpalo"
version = "3.20.3" version = "3.20.3"
@@ -182,9 +173,9 @@ checksum = "f079e83a288787bcd14a6aea84cee5c87a67c5a3e660c30f557a3d24761b3527"
[[package]] [[package]]
name = "clap" name = "clap"
version = "4.6.5" version = "4.6.6"
source = "registry+https://github.com/rust-lang/crates.io-index" source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "301b56658598e48f3648647ac6fc887be7e7108eddfa4e9b63fcf3ec58c0cadf" checksum = "473c7e07f409a8d772161724aa8db6a765a2532a70f9667eeb7b49d3d02fbdca"
dependencies = [ dependencies = [
"clap_builder", "clap_builder",
"clap_derive", "clap_derive",
@@ -192,9 +183,9 @@ dependencies = [
[[package]] [[package]]
name = "clap_builder" name = "clap_builder"
version = "4.6.5" version = "4.6.6"
source = "registry+https://github.com/rust-lang/crates.io-index" source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "94a65403d1a1bd28f7dc68eb8506e8874808ee5eecb59298de588e2e1407a078" checksum = "7b48fea5a88e9ae728a2dcbedbfc0e730f7d60da42e1cb049a83c9fb8b789889"
dependencies = [ dependencies = [
"anstream", "anstream",
"anstyle", "anstyle",
@@ -255,12 +246,6 @@ dependencies = [
"windows-sys", "windows-sys",
] ]
[[package]]
name = "const-oid"
version = "0.10.2"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "a6ef517f0926dd24a1582492c791b6a4818a4d94e789a334894aa15b0d12f55c"
[[package]] [[package]]
name = "cpp_demangle" name = "cpp_demangle"
version = "0.5.1" version = "0.5.1"
@@ -279,15 +264,6 @@ dependencies = [
"libc", "libc",
] ]
[[package]]
name = "cpufeatures"
version = "0.3.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "8b2a41393f66f16b0823bb79094d54ac5fbd34ab292ddafb9a0456ac9f87d201"
dependencies = [
"libc",
]
[[package]] [[package]]
name = "cranelift-assembler-x64" name = "cranelift-assembler-x64"
version = "0.134.3" version = "0.134.3"
@@ -352,7 +328,7 @@ dependencies = [
"rustc-hash", "rustc-hash",
"serde", "serde",
"serde_derive", "serde_derive",
"sha2 0.10.9", "sha2",
"smallvec", "smallvec",
"target-lexicon", "target-lexicon",
"wasmtime-internal-core", "wasmtime-internal-core",
@@ -478,15 +454,6 @@ dependencies = [
"typenum", "typenum",
] ]
[[package]]
name = "crypto-common"
version = "0.2.2"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "ce6e4c961d6cd6c9a86db418387425e8bdeaf05b3c8bc1411e6dca4c252f1453"
dependencies = [
"hybrid-array",
]
[[package]] [[package]]
name = "debugid" name = "debugid"
version = "0.8.0" version = "0.8.0"
@@ -502,19 +469,8 @@ version = "0.10.7"
source = "registry+https://github.com/rust-lang/crates.io-index" source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "9ed9a281f7bc9b7576e61468ba615a66a5c8cfdff42420a70aa82701a3b1e292" checksum = "9ed9a281f7bc9b7576e61468ba615a66a5c8cfdff42420a70aa82701a3b1e292"
dependencies = [ dependencies = [
"block-buffer 0.10.4", "block-buffer",
"crypto-common 0.1.7", "crypto-common",
]
[[package]]
name = "digest"
version = "0.11.3"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "f1dd6dbb5841937940781866fa1281a1ff7bd3bf827091440879f9994983d5c2"
dependencies = [
"block-buffer 0.12.1",
"const-oid",
"crypto-common 0.2.2",
] ]
[[package]] [[package]]
@@ -795,15 +751,6 @@ dependencies = [
"windows-sys", "windows-sys",
] ]
[[package]]
name = "hybrid-array"
version = "0.4.14"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "707114b52a152fa7bdb290cd7cd5912d9467273b6d74e21b8d81aca1f8533f6b"
dependencies = [
"typenum",
]
[[package]] [[package]]
name = "id-arena" name = "id-arena"
version = "2.3.0" version = "2.3.0"
@@ -1400,13 +1347,13 @@ dependencies = [
[[package]] [[package]]
name = "sha1" name = "sha1"
version = "0.11.0" version = "0.10.7"
source = "registry+https://github.com/rust-lang/crates.io-index" source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "aacc4cc499359472b4abe1bf11d0b12e688af9a805fa5e3016f9a386dc2d0214" checksum = "a978451301f4db1d02937a4ab3ccce137717b81826e79b7d49ffe3244a13c3b8"
dependencies = [ dependencies = [
"cfg-if", "cfg-if",
"cpufeatures 0.3.0", "cpufeatures",
"digest 0.11.3", "digest",
] ]
[[package]] [[package]]
@@ -1416,19 +1363,8 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "a7507d819769d01a365ab707794a4084392c824f54a7a6a7862f8c3d0892b283" checksum = "a7507d819769d01a365ab707794a4084392c824f54a7a6a7862f8c3d0892b283"
dependencies = [ dependencies = [
"cfg-if", "cfg-if",
"cpufeatures 0.2.17", "cpufeatures",
"digest 0.10.7", "digest",
]
[[package]]
name = "sha2"
version = "0.11.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "446ba717509524cb3f22f17ecc096f10f4822d76ab5c0b9822c5f9c284e825f4"
dependencies = [
"cfg-if",
"cpufeatures 0.3.0",
"digest 0.11.3",
] ]
[[package]] [[package]]
@@ -1653,7 +1589,7 @@ checksum = "0b928f33d975fc6ad9f86c8f283853ad26bdd5b10b7f1542aa2fa15e2289105a"
[[package]] [[package]]
name = "wafer" name = "wafer"
version = "0.2.1" version = "0.2.3"
dependencies = [ dependencies = [
"anyhow", "anyhow",
"clap", "clap",
@@ -1664,13 +1600,13 @@ dependencies = [
[[package]] [[package]]
name = "wafer-core" name = "wafer-core"
version = "0.2.1" version = "0.2.3"
dependencies = [ dependencies = [
"anyhow", "anyhow",
"insta", "insta",
"proptest", "proptest",
"sha1", "sha1",
"sha2 0.11.0", "sha2",
"thiserror 2.0.19", "thiserror 2.0.19",
"wasm-encoder 0.255.0", "wasm-encoder 0.255.0",
"wasmparser 0.255.0", "wasmparser 0.255.0",
@@ -1679,7 +1615,7 @@ dependencies = [
[[package]] [[package]]
name = "wafer-web" name = "wafer-web"
version = "0.2.1" version = "0.2.3"
dependencies = [ dependencies = [
"anyhow", "anyhow",
"js-sys", "js-sys",
@@ -1966,7 +1902,7 @@ dependencies = [
"semver", "semver",
"serde", "serde",
"serde_derive", "serde_derive",
"sha2 0.10.9", "sha2",
"smallvec", "smallvec",
"target-lexicon", "target-lexicon",
"wasm-encoder 0.252.0", "wasm-encoder 0.252.0",
@@ -1989,7 +1925,7 @@ dependencies = [
"rustix", "rustix",
"serde", "serde",
"serde_derive", "serde_derive",
"sha2 0.10.9", "sha2",
"toml", "toml",
"wasmtime-environ", "wasmtime-environ",
"windows-sys", "windows-sys",
@@ -2239,18 +2175,18 @@ dependencies = [
[[package]] [[package]]
name = "zerocopy" name = "zerocopy"
version = "0.8.55" version = "0.8.56"
source = "registry+https://github.com/rust-lang/crates.io-index" source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "b5a105cd7b140f6eeec8acff2ea38135d3cab283ada58540f629fe51e46696eb" checksum = "556764e583adb45a9f8d413c2a147fa7e8d821e48e12b14fd560b607998b75eb"
dependencies = [ dependencies = [
"zerocopy-derive", "zerocopy-derive",
] ]
[[package]] [[package]]
name = "zerocopy-derive" name = "zerocopy-derive"
version = "0.8.55" version = "0.8.56"
source = "registry+https://github.com/rust-lang/crates.io-index" source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "0fe976fb70c78cd64cccfe3a6fc142244e8a77b70959b30faf9d0ac37ee228eb" checksum = "f2ab42fc20575779bd240faa45f94a74256f755c0fa9e89f0ede20d91d0cdfc1"
dependencies = [ dependencies = [
"proc-macro2", "proc-macro2",
"quote", "quote",
+11 -3
View File
@@ -3,7 +3,7 @@ members = ["crates/*"]
resolver = "2" resolver = "2"
[workspace.package] [workspace.package]
version = "0.2.1" version = "0.2.3"
edition = "2024" edition = "2024"
license = "MIT OR Apache-2.0" license = "MIT OR Apache-2.0"
repository = "https://github.com/ok2/wafer" repository = "https://github.com/ok2/wafer"
@@ -48,6 +48,14 @@ anyhow = "1"
thiserror = "2" thiserror = "2"
proptest = "1" proptest = "1"
insta = "1" insta = "1"
sha1 = "0.11" sha1 = "0.10"
sha2 = "0.11" sha2 = "0.10"
send_wrapper = "0.6" send_wrapper = "0.6"
# Cargo strips debuginfo from release artifacts by default, and on macOS that
# also strips the metadata proc-macro dylibs need to be loadable — release
# builds then fail with "can't find crate" for every proc-macro (rustversion,
# thiserror_impl, ...). Build scripts and proc-macros gain nothing from
# stripping, so exempt them; the release binaries stay stripped.
[profile.release.build-override]
strip = false
+1 -1
View File
@@ -9,7 +9,7 @@ license.workspace = true
workspace = true workspace = true
[dependencies] [dependencies]
wafer-core = { path = "../core", version = "0.2.1" } wafer-core = { path = "../core", version = "0.2.3" }
wasmtime = { workspace = true } wasmtime = { workspace = true }
anyhow = { workspace = true } anyhow = { workspace = true }
clap = { version = "4", features = ["derive"] } clap = { version = "4", features = ["derive"] }
+20
View File
@@ -108,6 +108,23 @@ pub const SYSVAR_HLD: u32 = SYSVAR_BASE + 28;
pub const SYSVAR_LEAVE_FLAG: u32 = SYSVAR_BASE + 32; pub const SYSVAR_LEAVE_FLAG: u32 = SYSVAR_BASE + 32;
/// Throw code left by a compiled stack-guard fault for `_STACK_FAULT_`. /// Throw code left by a compiled stack-guard fault for `_STACK_FAULT_`.
pub const SYSVAR_FAULT_CODE: u32 = SYSVAR_BASE + 36; pub const SYSVAR_FAULT_CODE: u32 = SYSVAR_BASE + 36;
/// DPL: digits right of the rightmost punctuation in the last converted
/// number; negative when the token carried no punctuation.
pub const SYSVAR_DPL: u32 = SYSVAR_BASE + 40;
/// NH: high-order cell of the last single-cell conversion, so an
/// out-of-range token can be recovered as a double.
pub const SYSVAR_NH: u32 = SYSVAR_BASE + 44;
/// Seed for [`SYSVAR_DPL`] before conversion starts.
///
/// `SwiftForth` seeds DPL with a negative value and bumps it once per digit,
/// so an unpunctuated token still ends up negative. Punctuation resets the
/// counter to zero, which makes the final value the digit count right of the
/// rightmost punctuation character.
///
/// The exact seed is observable: `sf64` reports DPL as -1020 after `1234`
/// and -1023 after `-1`, both of which pin it to -1024.
pub const DPL_INIT: i32 = -1024;
#[cfg(test)] #[cfg(test)]
mod tests { mod tests {
@@ -149,6 +166,9 @@ mod tests {
SYSVAR_NUM_TIB, SYSVAR_NUM_TIB,
SYSVAR_HLD, SYSVAR_HLD,
SYSVAR_LEAVE_FLAG, SYSVAR_LEAVE_FLAG,
SYSVAR_FAULT_CODE,
SYSVAR_DPL,
SYSVAR_NH,
]; ];
for offset in all_offsets { for offset in all_offsets {
assert!(offset + CELL_SIZE <= SYSVAR_BASE + SYSVAR_SIZE); assert!(offset + CELL_SIZE <= SYSVAR_BASE + SYSVAR_SIZE);
+273 -80
View File
@@ -23,12 +23,64 @@ use crate::ir::IrOp;
#[cfg(feature = "crypto")] #[cfg(feature = "crypto")]
use crate::memory::HASH_SCRATCH_BASE; use crate::memory::HASH_SCRATCH_BASE;
use crate::memory::{ use crate::memory::{
CELL_SIZE, DATA_STACK_TOP, FLOAT_SIZE, FLOAT_STACK_BASE, FLOAT_STACK_TOP, INPUT_BUFFER_BASE, CELL_SIZE, DATA_STACK_TOP, DPL_INIT, FLOAT_SIZE, FLOAT_STACK_BASE, FLOAT_STACK_TOP,
INPUT_BUFFER_SIZE, RETURN_STACK_TOP, SYSVAR_BASE_VAR, SYSVAR_FAULT_CODE, SYSVAR_HERE, INPUT_BUFFER_BASE, INPUT_BUFFER_SIZE, RETURN_STACK_TOP, SYSVAR_BASE_VAR, SYSVAR_DPL,
SYSVAR_LEAVE_FLAG, SYSVAR_NUM_TIB, SYSVAR_STATE, SYSVAR_TO_IN, SYSVAR_FAULT_CODE, SYSVAR_HERE, SYSVAR_LEAVE_FLAG, SYSVAR_NH, SYSVAR_NUM_TIB, SYSVAR_STATE,
SYSVAR_TO_IN,
}; };
use crate::optimizer::optimize; use crate::optimizer::optimize;
// ---------------------------------------------------------------------------
// Number conversion
// ---------------------------------------------------------------------------
/// Characters that force double-cell conversion, following `SwiftForth`'s
/// input number conversion rules.
///
/// A leading `-` is the one exception: it binds as a sign, which keeps `-1`
/// a single-cell number while `1-2` converts as a double.
const DOUBLE_PUNCTUATION: [u8; 6] = *b",.+-/:";
/// Split a leading minus off a token, returning whether it was negative.
///
/// Only `-` is a sign. A leading `+` stays punctuation, matching `sf64`,
/// where `+7` converts as the double 7 with `DPL` = 1.
///
/// The sign may sit between a base-override prefix and the digits (`$-FF`,
/// the Forth 2012 spelling) or, as a WAFER extension, before the prefix
/// (`-$FF`), so this runs at both positions.
fn strip_sign(s: &str) -> (bool, &str) {
match s.as_bytes().first() {
Some(b'-') => (true, &s[1..]),
_ => (false, s),
}
}
/// A numeric token that converted successfully.
#[derive(Debug, Clone, Copy, PartialEq, Eq)]
struct NumberLiteral {
/// The accumulated 64-bit value, sign applied.
value: i64,
/// Digits right of the rightmost punctuation character. Negative when the
/// token carried no punctuation, which is how `DPL` reports "single-cell".
dpl: i32,
/// Whether punctuation forced double-cell conversion.
is_double: bool,
}
impl NumberLiteral {
/// Low-order cell — the value a single-cell conversion leaves on the stack.
fn lo(self) -> i32 {
self.value as i32
}
/// High-order cell. For a single-cell conversion this is what `NH` holds,
/// letting an out-of-range token be recovered as a double.
fn hi(self) -> i32 {
(self.value >> 32) as i32
}
}
// --------------------------------------------------------------------------- // ---------------------------------------------------------------------------
// Control-flow compilation state // Control-flow compilation state
// --------------------------------------------------------------------------- // ---------------------------------------------------------------------------
@@ -1194,22 +1246,18 @@ impl<R: Runtime> ForthVM<R> {
return Ok(()); return Ok(());
} }
// Try to parse as double-number (trailing dot) // Try to convert as a number; punctuation makes it double-cell
if let Some((lo, hi)) = self.parse_double_number(token) { if let Some(lit) = self.parse_numeric_literal(token) {
self.push_data_stack(lo)?; self.record_number_conversion(lit);
self.push_data_stack(hi)?; self.push_data_stack(lit.lo())?;
if self.recording_toplevel && self.state == 0 { if self.recording_toplevel && self.state == 0 {
self.toplevel_ir.push(IrOp::PushI32(lo)); self.toplevel_ir.push(IrOp::PushI32(lit.lo()));
self.toplevel_ir.push(IrOp::PushI32(hi));
} }
return Ok(()); if lit.is_double {
} self.push_data_stack(lit.hi())?;
// Try to parse as number
if let Some(n) = self.parse_number(token) {
self.push_data_stack(n)?;
if self.recording_toplevel && self.state == 0 { if self.recording_toplevel && self.state == 0 {
self.toplevel_ir.push(IrOp::PushI32(n)); self.toplevel_ir.push(IrOp::PushI32(lit.hi()));
}
} }
return Ok(()); return Ok(());
} }
@@ -1571,16 +1619,13 @@ impl<R: Runtime> ForthVM<R> {
return Ok(()); return Ok(());
} }
// Try to parse as double-number (trailing dot) // Try to convert as a number; punctuation makes it double-cell
if let Some((lo, hi)) = self.parse_double_number(token) { if let Some(lit) = self.parse_numeric_literal(token) {
self.push_ir(IrOp::PushI32(lo)); self.record_number_conversion(lit);
self.push_ir(IrOp::PushI32(hi)); self.push_ir(IrOp::PushI32(lit.lo()));
return Ok(()); if lit.is_double {
self.push_ir(IrOp::PushI32(lit.hi()));
} }
// Try to parse as number
if let Some(n) = self.parse_number(token) {
self.push_ir(IrOp::PushI32(n));
return Ok(()); return Ok(());
} }
@@ -2731,85 +2776,111 @@ impl<R: Runtime> ForthVM<R> {
// Number parsing // Number parsing
// ----------------------------------------------------------------------- // -----------------------------------------------------------------------
/// Try to parse a token as a number. /// Try to convert a token to a number, following `SwiftForth`'s input
fn parse_number(&self, token: &str) -> Option<i32> { /// number conversion rules.
///
/// Punctuation (`,` `.` `+` `-` `/` `:`) forces double-cell conversion, so
/// `12.34`, `1,234`, `12:30:45` and `2026-08-06` all convert as doubles.
/// Only a leading `-` escapes this and binds as a sign, which keeps `-1`
/// single-cell; a leading `+` stays punctuation, so `+7` is the double 7.
///
/// `DPL` counts up once per digit from [`DPL_INIT`] and resets to zero at
/// every punctuation character, so it ends up holding the digit count right
/// of the rightmost punctuation, and stays negative for unpunctuated tokens.
fn parse_numeric_literal(&self, token: &str) -> Option<NumberLiteral> {
let token = token.trim(); let token = token.trim();
if token.is_empty() { if token.is_empty() {
return None; return None;
} }
// Check for negative prefix // A leading sign binds to the number; it is not double punctuation.
let (negative, rest) = if let Some(stripped) = token.strip_prefix('-') { let (neg_outer, rest) = strip_sign(token);
(true, stripped)
} else {
(false, token)
};
if rest.is_empty() { if rest.is_empty() {
return None; return None;
} }
// Parse based on prefix // Character literal: 'x' → ASCII value of x. No digits, so DPL stays
let result = if let Some(hex) = rest.strip_prefix('$') { // at its seed and the result is always single-cell.
i64::from_str_radix(hex, 16).ok() if rest.len() == 3 && rest.as_bytes()[0] == b'\'' && rest.as_bytes()[2] == b'\'' {
} else if let Some(dec) = rest.strip_prefix('#') { let value = i64::from(rest.as_bytes()[1]);
dec.parse::<i64>().ok() return Some(NumberLiteral {
} else if let Some(bin) = rest.strip_prefix('%') { value: if neg_outer { -value } else { value },
i64::from_str_radix(bin, 2).ok() dpl: DPL_INIT,
} else if rest.len() == 3 && rest.as_bytes()[0] == b'\'' && rest.as_bytes()[2] == b'\'' { is_double: false,
// Character literal: 'x' → ASCII value of x });
Some(rest.as_bytes()[1] as i64)
} else {
i64::from_str_radix(rest, self.base).ok()
};
result.map(|n| if negative { -(n as i32) } else { n as i32 })
} }
/// Try to parse a token as a double-number (token ends with `.`). // A base-override prefix sits before the leftmost digit.
/// Returns (lo, hi) where the double-cell value is (hi << 32) | lo. let (radix, after_prefix) = match rest.as_bytes()[0] {
fn parse_double_number(&self, token: &str) -> Option<(i32, i32)> { b'$' => (16, &rest[1..]),
let token = token.trim(); b'#' => (10, &rest[1..]),
if token.is_empty() { b'%' => (2, &rest[1..]),
_ => (self.base, rest),
};
// Forth 2012 spells a signed based number `#-1289`, so the sign can
// also follow the prefix. Either way it precedes the leftmost digit
// and so is a sign rather than double punctuation.
let (neg_inner, digits) = strip_sign(after_prefix);
let negative = neg_outer ^ neg_inner;
if digits.is_empty() {
return None; return None;
} }
// Check for trailing dot (double-number indicator) // Walk the digit string, stripping punctuation and tracking DPL.
let without_dot = token.strip_suffix('.')?; let mut buf = String::with_capacity(digits.len());
if without_dot.is_empty() { let mut dpl = DPL_INIT;
let mut is_double = false;
for &b in digits.as_bytes() {
if DOUBLE_PUNCTUATION.contains(&b) {
is_double = true;
dpl = 0;
} else {
buf.push(char::from(b));
dpl += 1;
}
}
if buf.is_empty() {
return None; return None;
} }
// Check for negative prefix // i128 accumulation so the full u64 range survives conversion.
let (negative, rest) = if let Some(stripped) = without_dot.strip_prefix('-') { let magnitude = i128::from_str_radix(&buf, radix).ok()?;
(true, stripped) let value = if negative {
-(magnitude as i64)
} else { } else {
(false, without_dot) magnitude as i64
}; };
if rest.is_empty() { Some(NumberLiteral {
return None; value,
} dpl,
is_double,
// Parse based on prefix -- use i128 to handle the full u64 range
let result: Option<i128> = if let Some(hex) = rest.strip_prefix('$') {
i128::from_str_radix(hex, 16).ok()
} else if let Some(dec) = rest.strip_prefix('#') {
dec.parse::<i128>().ok()
} else if let Some(bin) = rest.strip_prefix('%') {
i128::from_str_radix(bin, 2).ok()
} else {
i128::from_str_radix(rest, self.base).ok()
};
result.map(|n| {
let val: i64 = if negative { -(n as i64) } else { n as i64 };
let lo = val as i32;
let hi = (val >> 32) as i32;
(lo, hi)
}) })
} }
/// Publish the outcome of a conversion in `DPL` and `NH`.
///
/// `NH` only carries meaning after a single-cell conversion, where it holds
/// the high-order cell that the stack result dropped.
fn record_number_conversion(&mut self, lit: NumberLiteral) {
self.rt.mem_write_i32(SYSVAR_DPL, lit.dpl);
if !lit.is_double {
self.rt.mem_write_i32(SYSVAR_NH, lit.hi());
}
}
/// Try to parse a token as a single-cell number, ignoring `DPL`/`NH`.
/// Used where only a plain cell value is meaningful.
fn parse_number(&self, token: &str) -> Option<i32> {
self.parse_numeric_literal(token)
.filter(|lit| !lit.is_double)
.map(NumberLiteral::lo)
}
// ----------------------------------------------------------------------- // -----------------------------------------------------------------------
// Float literal parsing // Float literal parsing
// ----------------------------------------------------------------------- // -----------------------------------------------------------------------
@@ -3092,6 +3163,7 @@ impl<R: Runtime> ForthVM<R> {
self.register_to_in()?; self.register_to_in()?;
self.register_state_var()?; self.register_state_var()?;
self.register_base_var()?; self.register_base_var()?;
self.register_number_conversion_vars()?;
// Double-cell arithmetic // Double-cell arithmetic
self.register_m_star()?; self.register_m_star()?;
@@ -5151,6 +5223,22 @@ impl<R: Runtime> ForthVM<R> {
Ok(()) Ok(())
} }
/// DPL ( -- addr ) and NH ( -- addr ): input number conversion results.
///
/// `DPL` holds the digit count right of the rightmost punctuation
/// character in the last converted number, or a negative value when the
/// token carried none. `NH` holds the high-order cell dropped by a
/// single-cell conversion, so an out-of-range token can be recovered as a
/// double.
fn register_number_conversion_vars(&mut self) -> anyhow::Result<()> {
self.rt.mem_write_i32(SYSVAR_DPL, DPL_INIT);
self.rt.mem_write_i32(SYSVAR_NH, 0);
self.register_primitive("DPL", false, vec![IrOp::PushI32(SYSVAR_DPL as i32)])?;
self.register_primitive("NH", false, vec![IrOp::PushI32(SYSVAR_NH as i32)])?;
Ok(())
}
/// M* ( n1 n2 -- d ) signed multiply producing double-cell result. /// M* ( n1 n2 -- d ) signed multiply producing double-cell result.
fn register_m_star(&mut self) -> anyhow::Result<()> { fn register_m_star(&mut self) -> anyhow::Result<()> {
let func: HostFn = Box::new(move |ctx: &mut dyn HostAccess| { let func: HostFn = Box::new(move |ctx: &mut dyn HostAccess| {
@@ -10828,6 +10916,111 @@ mod tests {
assert_eq!(eval_stack("1E 2.5E 1E F~"), vec![0]); // |1-2.5| = 1.5 >= 1 assert_eq!(eval_stack("1E 2.5E 1E F~"), vec![0]); // |1-2.5| = 1.5 >= 1
} }
#[test]
fn punctuation_anywhere_converts_as_double() {
// The punctuation is a double-cell marker, not a fractional point:
// every form below carries the same digits, so the value is the same.
// eval_stack reports top-first, so a double reads as [hi, lo].
for token in ["1234.", "123.4", "12.34", "1.234", ".1234"] {
assert_eq!(eval_stack(token), vec![0, 1234], "token {token}");
}
// SwiftForth accepts comma, colon, slash, plus and dash too, which is
// what makes dates and times convert without a custom parser.
assert_eq!(eval_stack("1,234"), vec![0, 1234]);
assert_eq!(eval_stack("12:30:45"), vec![0, 123045]);
assert_eq!(eval_stack("2026-08-06"), vec![0, 20260806]);
assert_eq!(eval_stack("12/34"), vec![0, 1234]);
assert_eq!(eval_stack("1+234"), vec![0, 1234]);
}
#[test]
fn dpl_counts_digits_right_of_last_punctuation() {
for (token, dpl) in [("1234.", 0), ("123.4", 1), ("12.34", 2), (".1234", 4)] {
assert_eq!(eval_stack(&format!("{token} 2DROP DPL @")), vec![dpl]);
}
// Only the rightmost punctuation counts.
assert_eq!(eval_stack("12:30:45 2DROP DPL @"), vec![2]);
}
#[test]
fn dpl_stays_negative_for_unpunctuated_numbers() {
// A leading minus is a sign, not punctuation, so these stay single-cell.
for token in ["1234", "-1", "$FF"] {
let dpl = eval_stack(&format!("{token} DROP DPL @"))[0];
assert!(dpl < 0, "token {token} left DPL = {dpl}");
}
assert_eq!(eval_stack("-1"), vec![-1]);
// DPL counts up from its seed once per digit.
assert_eq!(eval_stack("1234 DROP DPL @"), vec![DPL_INIT + 4]);
assert_eq!(eval_stack("-1 DROP DPL @"), vec![DPL_INIT + 1]);
}
#[test]
fn leading_plus_is_punctuation_not_a_sign() {
// sf64 converts `+7` as the double 7 with DPL = 1: unlike `-`, a
// leading `+` does not bind to the number.
assert_eq!(eval_stack("+7"), vec![0, 7]);
assert_eq!(eval_stack("+7 2DROP DPL @"), vec![1]);
// Same after a base prefix.
assert_eq!(eval_stack("#+7"), vec![0, 7]);
assert_eq!(eval_stack("$+F"), vec![0, 15]);
}
#[test]
fn repeated_punctuation_only_counts_from_the_last_one() {
// sf64: `12..34` is 1234 with DPL 2, `1-2-3` is 123 with DPL 1.
assert_eq!(eval_stack("12..34"), vec![0, 1234]);
assert_eq!(eval_stack("12..34 2DROP DPL @"), vec![2]);
assert_eq!(eval_stack("1-2-3"), vec![0, 123]);
assert_eq!(eval_stack("1-2-3 2DROP DPL @"), vec![1]);
}
#[test]
fn nh_recovers_an_out_of_range_single_number() {
// 4000000000 overflows a signed cell, so the stack value is truncated.
assert_eq!(eval_stack("4000000000"), vec![-294967296]);
// NH carries the high cell, making the true value recoverable.
assert_eq!(eval_stack("4000000000 NH @"), vec![0, -294967296]);
assert_eq!(eval_output("4000000000 NH @ D."), "4000000000 ");
}
#[test]
fn float_literals_still_win_over_double_punctuation() {
// `1.5E0` has an embedded dot, but "15E0" is not a decimal number,
// so conversion falls through to the float parser.
assert_eq!(eval_output("1.5E0 F."), "1.500000 ");
assert_eq!(eval_output("-3.25E0 F."), "-3.250000 ");
assert_eq!(eval_output("1E-3 F."), "0.001000 ");
}
#[test]
fn double_punctuation_respects_base_prefixes() {
assert_eq!(eval_stack("$FF."), vec![0, 255]);
assert_eq!(eval_stack("$F.F"), vec![0, 255]);
assert_eq!(eval_stack("%1010."), vec![0, 10]);
assert_eq!(eval_stack("#12.34"), vec![0, 1234]);
assert_eq!(eval_stack("-$FF."), vec![-1, -255]);
}
#[test]
fn sign_after_a_base_prefix_is_a_sign_not_punctuation() {
// Forth 2012 spells signed based numbers with the sign after the
// prefix. The dash precedes the leftmost digit, so it must not
// trigger double-cell conversion.
assert_eq!(eval_stack("#-1289"), vec![-1289]);
assert_eq!(eval_stack("$-12eF"), vec![-4847]);
assert_eq!(eval_stack("%-10010110"), vec![-150]);
// The sign may also precede the prefix, and both spellings cancel.
assert_eq!(eval_stack("-$FF"), vec![-255]);
assert_eq!(eval_stack("-$-FF"), vec![255]);
}
#[test]
fn punctuated_numbers_compile_into_definitions() {
assert_eq!(eval_stack(": D1 12.34 ; D1"), vec![0, 1234]);
assert_eq!(eval_output(": STAMP 2026-08-06 D. ; STAMP"), "20260806 ");
}
#[test] #[test]
fn optimizer_doesnt_break_basic_arithmetic() { fn optimizer_doesnt_break_basic_arithmetic() {
assert_eq!(eval_stack("5 3 +"), vec![8]); assert_eq!(eval_stack("5 3 +"), vec![8]);
+10
View File
@@ -350,6 +350,16 @@ pub const WORD_DOCS: &[(&str, &str, &str)] = &[
"Convert digits, accumulating into ud.", "Convert digits, accumulating into ud.",
), ),
("BASE", "( -- addr )", "Variable holding the number base."), ("BASE", "( -- addr )", "Variable holding the number base."),
(
"DPL",
"( -- addr )",
"Variable: digits right of the last punctuation; negative if none.",
),
(
"NH",
"( -- addr )",
"Variable: high cell dropped by the last single-cell conversion.",
),
("HEX", "( -- )", "Set BASE to sixteen."), ("HEX", "( -- )", "Set BASE to sixteen."),
("DECIMAL", "( -- )", "Set BASE to ten."), ("DECIMAL", "( -- )", "Set BASE to ten."),
// -- Core: strings -- // -- Core: strings --
+1 -1
View File
@@ -12,7 +12,7 @@ workspace = true
crate-type = ["cdylib", "rlib"] crate-type = ["cdylib", "rlib"]
[dependencies] [dependencies]
wafer-core = { path = "../core", version = "0.2.1", default-features = false, features = ["crypto"] } wafer-core = { path = "../core", version = "0.2.3", default-features = false, features = ["crypto"] }
wasm-bindgen = "0.2" wasm-bindgen = "0.2"
js-sys = "0.3" js-sys = "0.3"
send_wrapper = { workspace = true } send_wrapper = { workspace = true }
+2 -2
View File
@@ -18,11 +18,11 @@ confidence-threshold = 0.8
[bans] [bans]
multiple-versions = "deny" multiple-versions = "deny"
wildcards = "deny" wildcards = "deny"
# Transitive duplicates from wasmtime v31 -- will resolve when upgrading # Transitive duplicates from wasmtime v47 dependencies
skip = [ skip = [
"getrandom", "getrandom",
"syn",
"hashbrown", "hashbrown",
"r-efi",
"thiserror", "thiserror",
"thiserror-impl", "thiserror-impl",
"wasm-encoder", "wasm-encoder",