Hojun-Cho/ww
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

lib/strconv: graduate to owned-str returns with Hare-shape base param

Browse Source 
i64tos / u64tos / f64tos return a fresh owned str (caller frees via
os.free) instead of writing into a caller-supplied [N]u8. Adds typed
variants (i32tos / i16tos / i8tos and u32 / u16 / u8) and the missing
base parameter on stoi64 / stou64 + typed parse wrappers.

Base values are exported as plain-i32 `def`s (strconv.DEC,
strconv.HEX_UPPER, ...) rather than a `base` enum: cross-module
`strconv.base.DEC` chains miscompile in the cstage cgen — it emits a
memory load through `base(SB)` rather than inlining the constant.
The Sdef path resolves correctly, so callers say `strconv.DEC` and
both cgens lower to an immediate.

Also renames strings.byteindex / rbyteindex to strings.indexbyte /
rindexbyte, matching bytes.indexbyte and reserving the Hare name
`byteindex` for the future `(str | rune)`-needle shape.

fmt drops printint / printlnint / fprintint — those were stand-ins
for variadic `fmt::println(42)`; with the owned-str graduation the
substitute is one call: `fmt.println(strconv.i64tos(42, strconv.DEC))`.

strerror is sketched in a comment but not shipped — match arms over
the wider `error = !(invalid | overflow)` union still expose a
cstage-vs-wwstage spill divergence.
This commit is contained in:
Hojun-Cho
2026-05-13 03:55:16 +09:00
parent 9bc973dc96
commit be8a662f15
14 changed files with 1192 additions and 520 deletions
Download Patch File Download Diff File Expand all files Collapse all files

27
lib/fmt/fmt.ww
View File

@@ -1,10 +1,11 @@
// fmt — minimal formatting writers. All output goes through os.write
// to fd 1 (stdout). No printf-family yet — we don't have varargs in
// the language proper — but the typed entry points cover the common
// cases.
// to a file descriptor. No printf-family yet — we don't have varargs
// in the language proper — so callers compose with strconv.i64tos /
// strings.concat to build the message and then call print / println.
// Hare's `fmt::println(42)` becomes `fmt.println(strconv.i64tos(42,
// strconv.DEC))`.
use os;
use strconv;
export fn print(s: str) i64 = {
return os.write(1, s.ptr, s.len: u64);
@@ -18,17 +19,6 @@ export fn println(s: str) i64 = {
return n + m;
};
export fn printint(v: i64) void = {
let buf: [32]u8;
let n: i32 = strconv.i64tos(buf[0:32], v);
os.write(1, buf.ptr, n: u64);
};
export fn printlnint(v: i64) void = {
printint(v);
os.write(1, "\n".ptr, 1u64);
};
// errorln — write a message to stderr with a trailing newline.
export fn errorln(s: str) i64 = {
let n: i64 = os.write(2, s.ptr, s.len: u64);
@@ -51,10 +41,3 @@ export fn fprintln(fd: i32, s: str) i64 = {
if (m < 0) { return m; };
return n + m;
};
export fn fprintint(fd: i32, v: i64) void = {
let buf: [32]u8;
let n: i32 = strconv.i64tos(buf[0:32], v);
os.write(fd, buf.ptr, n: u64);
};

337
lib/strconv/strconv.ww
View File

@@ -1,114 +1,270 @@
// strconv — number↔string conversions. Decimal i64 to/from a fixed
// buffer. Error shapes mirror Hare's strconv types: (T | invalid |
// overflow) where each error is a named alias over a payload type
// (Hare uses !size / !void; ww uses i32 / void without the `!` mark).
// strconv — number↔string conversions.
//
// Mirrors Hare's strconv:: surface. The *tos functions return a fresh
// owned `str`; release via os.free(r.ptr, r.len: u64) when done.
// Hare returns `const str` into a static buffer; ww allocates per
// call because the wwstage cgen doesn't currently support mutating a
// module-level `*u8` (so a lazy-init shared buffer isn't expressible
// today). Graduate to the static-buffer shape once that lands.
use os;
// invalid — input wasn't a valid number in the requested format.
// Payload is the byte index of the first offending position. Mirrors
// Hare's strconv::invalid = !size (we use i32 instead of size).
// Payload is the byte index of the first offending position.
// Mirrors Hare's strconv::invalid = !size.
export type invalid = !i32;
// overflow — input was valid but doesn't fit the target type. No
// payload (a single yes/no signal). Mirrors Hare's !void shape.
// overflow — input was valid but doesn't fit the target type.
// Mirrors Hare's strconv::overflow = !void.
export type overflow = !void;
// u64tos — write `v` in decimal into `buf` and return the byte count.
// Hare name; the buffer-in shape is the sanctioned Plan 9 subset of
// Hare's `u64tos(u, base) const str`. Unsigned-only so callers don't
// have to think about wraparound when printing a u64 with the high
// bit set.
export fn u64tos(buf: []u8, v: u64) i32 = {
let tmp: [32]u8;
// error — any error from a strconv call. Mirrors Hare's strconv::error.
export type error = !(invalid | overflow);
// base — numeric base for parsing/formatting. Plain i32 (not a named
// enum) because cross-module `strconv.base.DEC` chains miscompile in
// the cstage cgen — it emits a memory load through `base(SB)` rather
// than inlining the enum value. Hare names them as `strconv::base`
// enum values; we expose them as module-level `def`s so callers say
// `strconv.DEC` and the cgen inlines the immediate.
//
// HEX is HEX_UPPER; HEX_LOWER is a separate pseudo-base that produces
// lowercase a-f digits.
export def DEFAULT: i32 = 0;
export def BIN: i32 = 2;
export def OCT: i32 = 8;
export def DEC: i32 = 10;
export def HEX_UPPER: i32 = 16;
export def HEX: i32 = 16;
export def HEX_LOWER: i32 = 17;
fn basenum(b: i32) i64 = {
if (b == BIN) { return 2; };
if (b == OCT) { return 8; };
if (b == HEX) { return 16; };
if (b == HEX_UPPER) { return 16; };
if (b == HEX_LOWER) { return 16; };
return 10; // DEC and DEFAULT
};
fn basedigit(d: i64, b: i32) u8 = {
if (d < 10) { return (d + 48): u8; };
let off: i64 = d - 10;
if (b == HEX_LOWER) { return (off + 97): u8; };
return (off + 65): u8;
};
// u64tos — convert v to a base-b numeric string. Returns owned str;
// release via os.free(r.ptr, r.len: u64). Mirrors Hare's
// strconv::u64tos (Hare returns const str into a static buffer).
export fn u64tos(v: u64, b: i32) str = {
let nb: u64 = basenum(b): u64;
let tmp: [65]u8;
let i: i32 = 0;
let n: u64 = v;
if (n == 0u64) { tmp[0] = 48u8; i = 1; };
for (n > 0u64) {
tmp[i] = ((n % 10u64) + 48u64): u8;
n = n / 10u64;
let d: i64 = (n % nb): i64;
tmp[i] = basedigit(d, b);
n = n / nb;
i += 1;
};
if (i == 0) {
tmp[0] = 48u8;
i = 1;
};
let buf: *u8 = os.alloc(i: u64): *u8;
let out: i32 = 0;
for (i > 0) {
i -= 1;
buf[out] = tmp[i];
out += 1;
};
return out;
let r: str;
r.ptr = buf;
r.len = out;
return r;
};
export fn i64tos(buf: []u8, v: i64) i32 = {
// i64tos — convert v to a base-b numeric string. Returns owned str;
// release via os.free. Mirrors Hare's strconv::i64tos.
export fn i64tos(v: i64, b: i32) str = {
let neg: bool = false;
let n: i64 = v;
if (n < 0) {
neg = true;
n = -n;
};
let tmp: [32]u8;
if (n < 0) { neg = true; n = -n; };
let nb: i64 = basenum(b);
let tmp: [65]u8;
let i: i32 = 0;
if (n == 0) { tmp[0] = 48u8; i = 1; };
for (n > 0) {
tmp[i] = ((n % 10) + 48): u8;
n = n / 10;
let d: i64 = n % nb;
tmp[i] = basedigit(d, b);
n = n / nb;
i += 1;
};
if (i == 0) {
tmp[0] = 48u8;
i = 1;
};
let extra: i32 = 0;
if (neg) { extra = 1; };
let total: i32 = i + extra;
let buf: *u8 = os.alloc(total: u64): *u8;
let out: i32 = 0;
if (neg) {
buf[out] = 45u8; // '-'
out += 1;
};
if (neg) { buf[out] = 45u8; out += 1; }; // '-'
for (i > 0) {
i -= 1;
buf[out] = tmp[i];
out += 1;
};
return out;
let r: str;
r.ptr = buf;
r.len = out;
return r;
};
// stoi64 — Hare-style fallible signed decimal parser. No locale, no
// whitespace, no underscores: a leading '-' is the only non-digit
// accepted, and only at position 0.
export fn stoi64(s: str) (i64 | invalid | overflow) = {
export fn i32tos(v: i32, b: i32) str = { return i64tos(v: i64, b); };
export fn i16tos(v: i16, b: i32) str = { return i64tos(v: i64, b); };
export fn i8tos(v: i8, b: i32) str = { return i64tos(v: i64, b); };
export fn u32tos(v: u32, b: i32) str = { return u64tos(v: u64, b); };
export fn u16tos(v: u16, b: i32) str = { return u64tos(v: u64, b); };
export fn u8tos(v: u8, b: i32) str = { return u64tos(v: u64, b); };
// digval — value of digit byte `c` under base `b`, or -1 if not a
// valid digit. Letters are accepted case-insensitively under HEX /
// HEX_UPPER; only lowercase under HEX_LOWER.
fn digval(c: u8, b: i32) i32 = {
if (c >= 48u8) { if (c <= 57u8) { return (c - 48u8): i32; }; };
if (b == HEX_LOWER) {
if (c >= 97u8) { if (c <= 102u8) { return ((c - 97u8) + 10u8): i32; }; };
return -1;
};
if (c >= 65u8) { if (c <= 70u8) { return ((c - 65u8) + 10u8): i32; }; };
if (c >= 97u8) { if (c <= 102u8) { return ((c - 97u8) + 10u8): i32; }; };
return -1;
};
// stoi64 — parse signed base-b number. Mirrors Hare's strconv::stoi64.
// No locale, no whitespace, no underscores: optional leading '-' then
// digits. Returns invalid with the offending index or overflow on
// out-of-range.
export fn stoi64(s: str, b: i32) (i64 | invalid | overflow) = {
if (s.len == 0) { return 0: invalid; };
let i: i32 = 0;
let neg: bool = false;
if (s[0] == 45u8) { neg = true; i = 1; };
if (i >= s.len) { return i: invalid; };
let nb: i32 = basenum(b): i32;
let v: i64 = 0;
for (i < s.len) {
let c: u8 = s[i];
if (c < 48u8) { return i: invalid; };
if (c > 57u8) { return i: invalid; };
v = v * 10 + ((c: i64) - 48);
let d: i32 = digval(c, b);
if (d < 0) { return i: invalid; };
if (d >= nb) { return i: invalid; };
v = v * (nb: i64) + (d: i64);
i += 1;
};
if (neg) { v = -v; };
return v;
};
// stou64 — fallible unsigned decimal parser. No leading sign.
export fn stou64(s: str) (u64 | invalid | overflow) = {
// stou64 — parse unsigned base-b number. Mirrors Hare's strconv::stou64.
export fn stou64(s: str, b: i32) (u64 | invalid | overflow) = {
if (s.len == 0) { return 0: invalid; };
let nb: u64 = basenum(b): u64;
let v: u64 = 0u64;
let i: i32 = 0;
for (i < s.len) {
let c: u8 = s[i];
if (c < 48u8) { return i: invalid; };
if (c > 57u8) { return i: invalid; };
v = v * 10u64 + ((c: u64) - 48u64);
let d: i32 = digval(c, b);
if (d < 0) { return i: invalid; };
if ((d: u64) >= nb) { return i: invalid; };
v = v * nb + (d: u64);
i += 1;
};
return v;
};
// f64tos — write `v` in decimal into `buf` and return the byte count.
// Hare name; this is the buffer-in Plan 9 subset of Hare's
// `f64tos(n) const str`. Today's surface:
export fn stoi32(s: str, b: i32) (i32 | invalid | overflow) = {
let r = stoi64(s, b);
match (r) {
case let v: i64 => {
if (v > 2147483647i64) { return overflow{}; };
if (v < -2147483648i64) { return overflow{}; };
return v: i32;
};
case let e: invalid => return e;
case let e: overflow => return e;
};
return 0: invalid; // unreachable; appeases the path-cov checker
};
export fn stoi16(s: str, b: i32) (i16 | invalid | overflow) = {
let r = stoi64(s, b);
match (r) {
case let v: i64 => {
if (v > 32767i64) { return overflow{}; };
if (v < -32768i64) { return overflow{}; };
return v: i16;
};
case let e: invalid => return e;
case let e: overflow => return e;
};
return 0: invalid;
};
export fn stoi8(s: str, b: i32) (i8 | invalid | overflow) = {
let r = stoi64(s, b);
match (r) {
case let v: i64 => {
if (v > 127i64) { return overflow{}; };
if (v < -128i64) { return overflow{}; };
return v: i8;
};
case let e: invalid => return e;
case let e: overflow => return e;
};
return 0: invalid;
};
export fn stou32(s: str, b: i32) (u32 | invalid | overflow) = {
let r = stou64(s, b);
match (r) {
case let v: u64 => {
if (v > 4294967295u64) { return overflow{}; };
return v: u32;
};
case let e: invalid => return e;
case let e: overflow => return e;
};
return 0: invalid;
};
export fn stou16(s: str, b: i32) (u16 | invalid | overflow) = {
let r = stou64(s, b);
match (r) {
case let v: u64 => {
if (v > 65535u64) { return overflow{}; };
return v: u16;
};
case let e: invalid => return e;
case let e: overflow => return e;
};
return 0: invalid;
};
export fn stou8(s: str, b: i32) (u8 | invalid | overflow) = {
let r = stou64(s, b);
match (r) {
case let v: u64 => {
if (v > 255u64) { return overflow{}; };
return v: u8;
};
case let e: invalid => return e;
case let e: overflow => return e;
};
return 0: invalid;
};
// f64tos — convert v to a decimal string. Returns owned str; release
// via os.free. Mirrors Hare's strconv::f64tos (current ww impl is
// fixed-point only, max 6 fractional digits, no NaN/Inf support —
// see graduate-to-Ryū note below).
//
// Surface:
//
// - finite values only. NaN/±Inf detection needs an f64→u64 bit
// reinterpret cast that the cgen doesn't expose yet.
@@ -119,23 +275,20 @@ export fn stou64(s: str) (u64 | invalid | overflow) = {
// in scientific notation via Ryū; we will graduate when the
// compiler grows the bit-reinterpret cast.
//
// Round-trip is therefore lossy past 6 fractional digits; callers
// that need bit-exact recovery should not use this until the
// graduate-to-Ryū step lands. `f64tos(buf, 1.0)` writes "1" (no
// decimal point), `f64tos(buf, 1.5)` writes "1.5", `f64tos(buf,
// 0.1)` writes "0.1".
// Round-trip is therefore lossy past 6 fractional digits.
//
// No float literals in the body — 990's wwdump diff requires this
// file's TK_FLOAT count to match between C and ww front-ends, and
// the ww-side wwdump currently skips TK_FLOAT.fval while the C side
// %g-formats it. Same trick lib/ww/lex/lex.ww's parsef64 uses:
// build f64 constants via int-to-f64 casts.
export fn f64tos(buf: []u8, v: f64) i32 = {
export fn f64tos(v: f64) str = {
let tmp: [64]u8;
let out: i32 = 0;
let f: f64 = v;
let zero: f64 = 0: f64;
if (f < zero) {
buf[out] = 45u8; // '-'
tmp[out] = 45u8; // '-'
out += 1;
f = -f;
};
@@ -145,8 +298,14 @@ export fn f64tos(buf: []u8, v: f64) i32 = {
if (f >= cap) {
let s: str = "huge";
let k: i32 = 0;
for (k < s.len) { buf[out] = s[k]; out += 1; k += 1; };
return out;
for (k < s.len) { tmp[out] = s[k]; out += 1; k += 1; };
let buf: *u8 = os.alloc(out: u64): *u8;
let q: i32 = 0;
for (q < out) { buf[q] = tmp[q]; q += 1; };
let r: str;
r.ptr = buf;
r.len = out;
return r;
};
let ip: i64 = f: i64;
// Fractional part scaled to 6 decimal digits, with round-to-
@@ -163,26 +322,38 @@ export fn f64tos(buf: []u8, v: f64) i32 = {
ip += 1;
fp = 0;
};
let itmp: [32]u8;
let in: i32 = i64tos(itmp[0:32], ip);
let intstr: str = i64tos(ip, DEC);
let k: i32 = 0;
for (k < in) { buf[out] = itmp[k]; out += 1; k += 1; };
if (fp == 0) { return out; };
buf[out] = 46u8; // '.'
out += 1;
let ftmp: [16]u8;
let m: i32 = u64tos(ftmp[0:16], fp: u64);
// Pad fractional to 6 digits with leading zeros (e.g. 0.05 →
// fp=50000, m=5, pad one '0' before "50000").
let z: i32 = 6 - m;
for (z > 0) { buf[out] = 48u8; out += 1; z -= 1; };
k = 0;
for (k < m) { buf[out] = ftmp[k]; out += 1; k += 1; };
// Trim trailing zeros in the fractional part (we know fp != 0,
// so the loop stops before erasing the dot).
for (out > 0) {
if (buf[out - 1] != 48u8) { break; };
out -= 1;
for (k < intstr.len) { tmp[out] = intstr.ptr[k]; out += 1; k += 1; };
os.free(intstr.ptr: *void, intstr.len: u64);
if (fp != 0) {
tmp[out] = 46u8; // '.'
out += 1;
let fracstr: str = u64tos(fp: u64, DEC);
// Pad fractional to 6 digits with leading zeros (e.g. 0.05 →
// fp=50000, fracstr="50000", pad one '0' before).
let z: i32 = 6 - fracstr.len;
for (z > 0) { tmp[out] = 48u8; out += 1; z -= 1; };
k = 0;
for (k < fracstr.len) { tmp[out] = fracstr.ptr[k]; out += 1; k += 1; };
os.free(fracstr.ptr: *void, fracstr.len: u64);
// Trim trailing zeros in the fractional part.
for (out > 0) {
if (tmp[out - 1] != 48u8) { break; };
out -= 1;
};
};
return out;
let buf: *u8 = os.alloc(out: u64): *u8;
let q: i32 = 0;
for (q < out) { buf[q] = tmp[q]; q += 1; };
let r: str;
r.ptr = buf;
r.len = out;
return r;
};
// strerror — Hare has strconv::strerror; ww doesn't ship it yet
// because a `match (e) { case invalid => ... }` arm over the wider
// `error = !(invalid | overflow)` union exposes a cstage-vs-wwstage
// cgen divergence (one cgen spills the unused payload slot, the
// other elides it). Restore once the cgens converge.

14
lib/strings/strings.ww
View File

@@ -39,9 +39,12 @@ export fn hassuffix(s: str, suf: str) bool = {
return true;
};
// byteindex — first index of byte `c` in `s`. Hare-shaped optional:
// (i32 | void). void variant indicates "not found".
export fn byteindex(s: str, c: u8) (i32 | void) = {
// indexbyte — first byte position of byte `c` in `s`. Mirrors
// Hare's strings::byteindex when the needle is a single ASCII rune,
// renamed to match bytes.indexbyte and to disambiguate from Hare's
// `byteindex(haystack, needle: (str | rune))` which we don't have
// the union-arg ABI for yet.
export fn indexbyte(s: str, c: u8) (i32 | void) = {
let i: i32 = 0;
for (i < s.len) {
if (s[i] == c) { return i; };
@@ -50,9 +53,8 @@ export fn byteindex(s: str, c: u8) (i32 | void) = {
return;
};
// rbyteindex — last index of byte `c` in `s`. Mirrors Hare's
// strings::rbyteindex.
export fn rbyteindex(s: str, c: u8) (i32 | void) = {
// rindexbyte — last byte position of byte `c` in `s`.
export fn rindexbyte(s: str, c: u8) (i32 | void) = {
let i: i32 = s.len - 1;
for (i >= 0) {
if (s[i] == c) { return i; };

10
lib/ww/ast.ww
View File

@@ -269,14 +269,12 @@ fn pr(fd: i32, n: *node, d: i32) void = {
if (n.kind == nkind.N_INTLIT) {
putc1(fd, 32u8);
let buf: [32]u8;
let m: i32 = strconv.u64tos(buf[0:32], n.uval);
os.write(fd, buf.ptr, m: u64);
let s: str = strconv.u64tos(n.uval, strconv.DEC);
os.write(fd, s.ptr, s.len: u64);
} else { if (n.kind == nkind.N_RUNELIT) {
putc1(fd, 32u8);
let buf: [32]u8;
let m: i32 = strconv.u64tos(buf[0:32], n.uval);
os.write(fd, buf.ptr, m: u64);
let s: str = strconv.u64tos(n.uval, strconv.DEC);
os.write(fd, s.ptr, s.len: u64);
} else { if (
n.kind == nkind.N_STRLIT ||
n.kind == nkind.N_IDENT ||

17
lib/ww/lex/tok.ww
View File

@@ -382,12 +382,11 @@ export fn tokprint(fd: i32, t: *tok) void = {
fputsstr(fd, "<none>");
};
fputcbyte(fd, 58u8); // ':'
let buf: [32]u8;
let n: i32 = strconv.i64tos(buf[0:32], t.line: i64);
os.write(fd, buf.ptr, n: u64);
let ls: str = strconv.i64tos(t.line: i64, strconv.DEC);
os.write(fd, ls.ptr, ls.len: u64);
fputcbyte(fd, 58u8);
n = strconv.i64tos(buf[0:32], t.col: i64);
os.write(fd, buf.ptr, n: u64);
let cs: str = strconv.i64tos(t.col: i64, strconv.DEC);
os.write(fd, cs.ptr, cs.len: u64);
fputcbyte(fd, 32u8); // ' '
fputsstr(fd, tokname(t.kind));
@@ -402,12 +401,12 @@ export fn tokprint(fd: i32, t: *tok) void = {
fputq(fd, ttext.ptr, ttext.len);
} else { if (t.kind == tkind.TK_INT) {
fputcbyte(fd, 32u8);
n = strconv.u64tos(buf[0:32], t.uval);
os.write(fd, buf.ptr, n: u64);
let us: str = strconv.u64tos(t.uval, strconv.DEC);
os.write(fd, us.ptr, us.len: u64);
} else { if (t.kind == tkind.TK_RUNE) {
fputcbyte(fd, 32u8);
n = strconv.u64tos(buf[0:32], t.uval);
os.write(fd, buf.ptr, n: u64);
let us: str = strconv.u64tos(t.uval, strconv.DEC);
os.write(fd, us.ptr, us.len: u64);
};};};};};
// tkind.TK_FLOAT is intentionally not handled here — %g formatting
// won't byte-match across implementations. Diff fixtures must