diff --git a/cmd/w6c/cgen.c b/cmd/w6c/cgen.c index 1c2f3ab8..997c4267 100644 --- a/cmd/w6c/cgen.c +++ b/cmd/w6c/cgen.c @@ -790,7 +790,7 @@ cgexpr(Cg *c, Node *n, Local *locals) /* Use DIV (unsigned) when either operand is an unsigned * integer type — IDIV would sign-extend a u64 with high * bit set into a negative i64 and produce wrong results - * (see strconv.u64toa with v = 1 << 63). */ + * (see strconv.u64tos with v = 1 << 63). */ int unsignd = (n->lhs && type_isunsigned(n->lhs->type)) || (n->rhs && type_isunsigned(n->rhs->type)); ins2(c, A_MOVQ, aimm(0), areg(D_DX)); @@ -819,7 +819,7 @@ cgexpr(Cg *c, Node *n, Local *locals) /* For ordered comparisons on unsigned operands we must * use the JA/JAE/JB/JBE family — signed Jcc would treat * a u64 with the high bit set as negative (e.g. the - * loop guard `n > 0` in strconv.u64toa with n=1<<63). */ + * loop guard `n > 0` in strconv.u64tos with n=1<<63). */ int unsignd = (n->lhs && type_isunsigned(n->lhs->type)) || (n->rhs && type_isunsigned(n->rhs->type)); ins2(c, A_CMPQ, areg(D_BX), areg(D_AX)); diff --git a/lib/ascii/ascii.ww b/lib/ascii/ascii.ww index b99cbf4d..265abf71 100644 --- a/lib/ascii/ascii.ww +++ b/lib/ascii/ascii.ww @@ -1,92 +1,92 @@ -// ascii — byte-class predicates and case folding for the ASCII range. -// Matches Hare's ascii::isdigit family. Bytes outside 0..127 always -// answer `false`. The lexer hot path uses these inline; they are -// expected to inline to a couple of compares. +// ascii — rune-class predicates and case folding for the ASCII range. +// Matches Hare's ascii::isdigit family (rune-taking signature). Runes +// outside 0..127 always answer `false`. The lexer hot path uses these +// inline; they are expected to inline to a couple of compares. -export fn isdigit(c: u8) bool = { - if (c < 48u8) { return false; }; - if (c > 57u8) { return false; }; +export fn isdigit(c: rune) bool = { + if (c < 48) { return false; }; + if (c > 57) { return false; }; return true; }; -export fn isupper(c: u8) bool = { - if (c < 65u8) { return false; }; - if (c > 90u8) { return false; }; +export fn isupper(c: rune) bool = { + if (c < 65) { return false; }; + if (c > 90) { return false; }; return true; }; -export fn islower(c: u8) bool = { - if (c < 97u8) { return false; }; - if (c > 122u8) { return false; }; +export fn islower(c: rune) bool = { + if (c < 97) { return false; }; + if (c > 122) { return false; }; return true; }; -export fn isalpha(c: u8) bool = { +export fn isalpha(c: rune) bool = { if (isupper(c)) { return true; }; return islower(c); }; -export fn isalnum(c: u8) bool = { +export fn isalnum(c: rune) bool = { if (isalpha(c)) { return true; }; return isdigit(c); }; // isspace — the C/Hare set: space, tab, NL, VT, FF, CR. -export fn isspace(c: u8) bool = { - if (c == 32u8) { return true; }; // ' ' - if (c == 9u8) { return true; }; // '\t' - if (c == 10u8) { return true; }; // '\n' - if (c == 11u8) { return true; }; // '\v' - if (c == 12u8) { return true; }; // '\f' - if (c == 13u8) { return true; }; // '\r' +export fn isspace(c: rune) bool = { + if (c == 32) { return true; }; // ' ' + if (c == 9) { return true; }; // '\t' + if (c == 10) { return true; }; // '\n' + if (c == 11) { return true; }; // '\v' + if (c == 12) { return true; }; // '\f' + if (c == 13) { return true; }; // '\r' return false; }; -export fn ishex(c: u8) bool = { +export fn isxdigit(c: rune) bool = { if (isdigit(c)) { return true; }; - if (c >= 65u8) { - if (c <= 70u8) { return true; }; // 'A'..'F' + if (c >= 65) { + if (c <= 70) { return true; }; // 'A'..'F' }; - if (c >= 97u8) { - if (c <= 102u8) { return true; }; // 'a'..'f' + if (c >= 97) { + if (c <= 102) { return true; }; // 'a'..'f' }; return false; }; // digitval — value of `c` as a hex/decimal digit, or -1 if not one. // Useful when scanning numeric literals. -export fn digitval(c: u8) i32 = { - if (isdigit(c)) { return (c - 48u8): i32; }; - if (c >= 65u8) { - if (c <= 70u8) { return ((c - 65u8) + 10u8): i32; }; +export fn digitval(c: rune) i32 = { + if (isdigit(c)) { return (c - 48): i32; }; + if (c >= 65) { + if (c <= 70) { return ((c - 65) + 10): i32; }; }; - if (c >= 97u8) { - if (c <= 102u8) { return ((c - 97u8) + 10u8): i32; }; + if (c >= 97) { + if (c <= 102) { return ((c - 97) + 10): i32; }; }; return -1; }; // isidstart / isidpart — identifier classes used by the lexer. // Alpha or '_' starts; alnum or '_' continues. -export fn isidstart(c: u8) bool = { +export fn isidstart(c: rune) bool = { if (isalpha(c)) { return true; }; - if (c == 95u8) { return true; }; // '_' + if (c == 95) { return true; }; // '_' return false; }; -export fn isidpart(c: u8) bool = { +export fn isidpart(c: rune) bool = { if (isalnum(c)) { return true; }; - if (c == 95u8) { return true; }; + if (c == 95) { return true; }; return false; }; // tolower / toupper — fold ASCII case. Non-letters pass through. -export fn tolower(c: u8) u8 = { - if (isupper(c)) { return c + 32u8; }; +export fn tolower(c: rune) rune = { + if (isupper(c)) { return c + 32; }; return c; }; -export fn toupper(c: u8) u8 = { - if (islower(c)) { return c - 32u8; }; +export fn toupper(c: rune) rune = { + if (islower(c)) { return c - 32; }; return c; }; diff --git a/lib/bufio/bufio.ww b/lib/bufio/bufio.ww index 3042926b..e44482c8 100644 --- a/lib/bufio/bufio.ww +++ b/lib/bufio/bufio.ww @@ -17,7 +17,7 @@ type buf = struct { w: i32, // write cursor (for writers) }; -export fn rinit(b: *buf, s: streamp, data: *u8, cap: i32) void = { +export fn init(b: *buf, s: streamp, data: *u8, cap: i32) void = { b.s = s; b.data = data; b.cap = cap; @@ -25,18 +25,10 @@ export fn rinit(b: *buf, s: streamp, data: *u8, cap: i32) void = { b.w = 0; }; -// peek1 — look at the next byte without consuming. Returns -1 on -// empty buffer; the caller is responsible for refilling via the -// stream when this happens. -export fn peek1(b: *buf) i32 = { - if (b.r < b.w) { - return b.data[b.r]: i32; - }; - return -1; -}; - -// take1 — pop one byte. -1 if empty. -export fn take1(b: *buf) i32 = { +// readbyte — pop one byte. -1 if empty. Hare name (transliterated +// from `read_byte`); the `-1`-for-EOF return is the sanctioned subset +// of Hare's `(u8 | EOF | error)`. +export fn readbyte(b: *buf) i32 = { if (b.r < b.w) { let c: u8 = b.data[b.r]; b.r += 1; @@ -45,27 +37,23 @@ export fn take1(b: *buf) i32 = { return -1; }; -// avail — bytes left to read out of the buffer. -export fn avail(b: *buf) i32 = { - return b.w - b.r; -}; - // Distinct alias so `(str | linerr)` has two variant types the // tagged-union machinery can keep apart at the tag level. The error // variant carries a short description; callers compare with errors.equal // or just inspect by length. type linerr = str; -// takeline — Hare-style fallible line read. Drains the buffer up to -// (but not including) the next '\n' and advances the cursor past the -// newline. Returns the line as a borrowed str on success, or a linerr -// describing why no line was available: +// readline — Hare-style fallible line read (name transliterated from +// `read_line`). Drains the buffer up to (but not including) the next +// '\n' and advances the cursor past the newline. Returns the line as +// a borrowed str on success, or a linerr describing why no line was +// available: // - "eof" when the buffer is empty // - "no newline" when the buffer contains data but no '\n' // // The returned str borrows from the underlying buffer; callers must // consume it (or copy) before refilling. -export fn takeline(b: *buf) (str | linerr) = { +export fn readline(b: *buf) (str | linerr) = { if (b.r >= b.w) { return "eof": linerr; }; let i: i32 = b.r; for (i < b.w) { diff --git a/lib/bytes/bytes.ww b/lib/bytes/bytes.ww index 91167652..a511b331 100644 --- a/lib/bytes/bytes.ww +++ b/lib/bytes/bytes.ww @@ -30,9 +30,10 @@ export fn copy(dst: []u8, src: []u8) i32 = { return n; }; -// indexsub — first index of `sub` in `s`, or -1. Mirrors -// strings.index but on []u8. Empty `sub` matches at 0. -export fn indexsub(s: []u8, sub: []u8) i32 = { +// index — first index of `sub` in `s`, or -1. Mirrors Hare's +// bytes::index (the []u8 needle variant; the u8 needle stays as +// indexbyte until we have union-arg dispatch). Empty `sub` matches at 0. +export fn index(s: []u8, sub: []u8) i32 = { if (sub.len == 0) { return 0; }; if (sub.len > s.len) { return -1; }; let last: i32 = s.len - sub.len; diff --git a/lib/encoding/utf8/utf8.ww b/lib/encoding/utf8/utf8.ww index deb7eeb5..dc1f651e 100644 --- a/lib/encoding/utf8/utf8.ww +++ b/lib/encoding/utf8/utf8.ww @@ -4,7 +4,7 @@ def MAX: rune = 1114111; // 0x10FFFF def BAD: rune = -1; -export fn runelen(r: rune) i32 = { +export fn runesz(r: rune) i32 = { if (r < 0) { return -1; }; if (r < 128) { return 1; }; if (r < 2048) { return 2; }; diff --git a/lib/endian/endian.ww b/lib/endian/endian.ww new file mode 100644 index 00000000..bd304326 --- /dev/null +++ b/lib/endian/endian.ww @@ -0,0 +1,9 @@ +// endian — byte-order conversions. Subset of Hare's endian::; only +// the network-order helpers we need for net::. Host order on amd64 is +// little-endian, so hton* / ntoh* are byte swaps. + +export fn htonu16(in: u16) u16 = { + return ((in << 8) | (in >> 8)) & 0xffff; +}; + +export fn ntohu16(in: u16) u16 = htonu16(in); diff --git a/lib/errors/errors.ww b/lib/errors/errors.ww index d736b040..9c42a576 100644 --- a/lib/errors/errors.ww +++ b/lib/errors/errors.ww @@ -3,14 +3,14 @@ type error = str; -def eEOF: error = "eof"; -def eShortRead: error = "short read"; -def eShortWrite: error = "short write"; -def eClosed: error = "closed"; -def eInvalid: error = "invalid argument"; -def ePerm: error = "permission denied"; -def eNotFound: error = "not found"; -def eExists: error = "already exists"; +def eof: error = "eof"; +def underread: error = "short read"; +def underwrite: error = "short write"; +def closed: error = "closed"; +def invalid: error = "invalid argument"; +def noaccess: error = "permission denied"; +def noentry: error = "not found"; +def exists: error = "already exists"; export fn isnil(e: error) bool = { return e.len == 0; diff --git a/lib/fmt/fmt.ww b/lib/fmt/fmt.ww index fd1df5f5..2bb19697 100644 --- a/lib/fmt/fmt.ww +++ b/lib/fmt/fmt.ww @@ -10,14 +10,17 @@ export fn print(s: str) i64 = { return os.write(1, s.ptr, s.len: u64); }; -export fn println(s: str) void = { - os.write(1, s.ptr, s.len: u64); - os.write(1, "\n".ptr, 1u64); +export fn println(s: str) i64 = { + let n: i64 = os.write(1, s.ptr, s.len: u64); + if (n < 0) { return n; }; + let m: i64 = os.write(1, "\n".ptr, 1u64); + if (m < 0) { return m; }; + return n + m; }; export fn printint(v: i64) void = { let buf: [32]u8; - let n: i32 = strconv.i64toa(buf[0:32], v); + let n: i32 = strconv.i64tos(buf[0:32], v); os.write(1, buf.ptr, n: u64); }; @@ -26,10 +29,13 @@ export fn printlnint(v: i64) void = { os.write(1, "\n".ptr, 1u64); }; -// errln — write a message to stderr with a trailing newline. -export fn errln(s: str) void = { - os.write(2, s.ptr, s.len: u64); - os.write(2, "\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); + if (n < 0) { return n; }; + let m: i64 = os.write(2, "\n".ptr, 1u64); + if (m < 0) { return m; }; + return n + m; }; // fprint / fprintln — same as print/println but on an arbitrary fd. @@ -38,14 +44,17 @@ export fn fprint(fd: i32, s: str) i64 = { return os.write(fd, s.ptr, s.len: u64); }; -export fn fprintln(fd: i32, s: str) void = { - os.write(fd, s.ptr, s.len: u64); - os.write(fd, "\n".ptr, 1u64); +export fn fprintln(fd: i32, s: str) i64 = { + let n: i64 = os.write(fd, s.ptr, s.len: u64); + if (n < 0) { return n; }; + let m: i64 = os.write(fd, "\n".ptr, 1u64); + if (m < 0) { return m; }; + return n + m; }; export fn fprintint(fd: i32, v: i64) void = { let buf: [32]u8; - let n: i32 = strconv.i64toa(buf[0:32], v); + let n: i32 = strconv.i64tos(buf[0:32], v); os.write(fd, buf.ptr, n: u64); }; @@ -56,10 +65,10 @@ export fn errpos(file: str, line: i32, col: i32, msg: str) void = { os.write(2, file.ptr, file.len: u64); os.write(2, ":".ptr, 1u64); let buf: [32]u8; - let n: i32 = strconv.i64toa(buf[0:32], line: i64); + let n: i32 = strconv.i64tos(buf[0:32], line: i64); os.write(2, buf.ptr, n: u64); os.write(2, ":".ptr, 1u64); - n = strconv.i64toa(buf[0:32], col: i64); + n = strconv.i64tos(buf[0:32], col: i64); os.write(2, buf.ptr, n: u64); os.write(2, ": ".ptr, 2u64); os.write(2, msg.ptr, msg.len: u64); diff --git a/lib/math/math.ww b/lib/math/math.ww new file mode 100644 index 00000000..afaddc77 --- /dev/null +++ b/lib/math/math.ww @@ -0,0 +1,13 @@ +// math — numeric helpers. Subset of Hare's math::; only the absolute- +// value pair for the signed integer types we currently care about. The +// return type is unsigned so that abs(I32_MIN) doesn't overflow. + +export fn absi32(n: i32) u32 = { + if (n < 0) { return (-n): u32; }; + return n: u32; +}; + +export fn absi64(n: i64) u64 = { + if (n < 0) { return (-n): u64; }; + return n: u64; +}; diff --git a/lib/net/net.ww b/lib/net/net.ww index fd4ae060..a0b2844e 100644 --- a/lib/net/net.ww +++ b/lib/net/net.ww @@ -42,7 +42,6 @@ export fn listen(fd: i32, backlog: i32) i32 = { return syscall3(SYS_LISTEN, fd: i64, backlog: i64, 0): i32; }; -// htons-equivalent: byte-swap a 16-bit port into network order. -export fn htons(p: u16) u16 = { - return ((p << 8) | (p >> 8)) & 0xffff; -}; +// Byte-order conversions live in lib/endian (Hare's endian::htonu16 +// is the canonical name). Use `endian.htonu16` to put a port into +// network order. diff --git a/lib/os/os.ww b/lib/os/os.ww index fc10c4c6..400c78c1 100644 --- a/lib/os/os.ww +++ b/lib/os/os.ww @@ -119,9 +119,11 @@ export fn filesize(fd: i32) i64 = { return end; }; -// readfull — keep reading until `n` bytes have arrived or the fd +// readall — keep reading until `n` bytes have arrived or the fd // closes early. Returns bytes read (0..=n) or -1 on read error. -export fn readfull(fd: i32, buf: *u8, n: u64) i64 = { +// Hare name (io::readall); the buffer is caller-supplied, matching +// the Plan 9 subset convention. +export fn readall(fd: i32, buf: *u8, n: u64) i64 = { let got: u64 = 0u64; for (got < n) { let r: i64 = read(fd, buf + got, n - got); @@ -132,9 +134,10 @@ export fn readfull(fd: i32, buf: *u8, n: u64) i64 = { return got: i64; }; -// writefull — keep writing until `n` bytes have been accepted or the -// fd refuses progress. Returns bytes written or -1. -export fn writefull(fd: i32, buf: *u8, n: u64) i64 = { +// writeall — keep writing until `n` bytes have been accepted or the +// fd refuses progress. Returns bytes written or -1. Hare name +// (io::writeall). +export fn writeall(fd: i32, buf: *u8, n: u64) i64 = { let sent: u64 = 0u64; for (sent < n) { let r: i64 = write(fd, buf + sent, n - sent); @@ -153,8 +156,8 @@ export fn access(path: *u8, mode: i32) i32 = { return syscall2(SYS_ACCESS, path: i64, mode: i64): i32; }; -// unlink(2). -export fn unlink(path: *u8) i32 = { +// remove — unlink(2). Hare name; the underlying syscall is unlink(2). +export fn remove(path: *u8) i32 = { return syscall1(SYS_UNLINK, path: i64): i32; }; diff --git a/lib/path/path.ww b/lib/path/path.ww index c9f02a98..12f60664 100644 --- a/lib/path/path.ww +++ b/lib/path/path.ww @@ -1,15 +1,8 @@ // path — filesystem path manipulation. UTF-8 paths, '/' separator. +// Mirrors Hare's path:: surface. Reverse byte search lives in strings +// (strings::rbyteindex), not here. -export fn isabs(p: str) bool = { +export fn abs(p: str) bool = { if (p.len == 0) { return false; }; return p[0] == ('/': u8); }; - -export fn lastindex(p: str, c: u8) i32 = { - let i: i32 = p.len - 1; - for (i >= 0) { - if (p[i] == c) { return i; }; - i -= 1; - }; - return -1; -}; diff --git a/lib/strconv/strconv.ww b/lib/strconv/strconv.ww index 44a9c966..a5e5eee4 100644 --- a/lib/strconv/strconv.ww +++ b/lib/strconv/strconv.ww @@ -2,14 +2,16 @@ // buffer. Two error idioms ship side by side: // - Plan 9 style (atoi64): tuple `(value, ok)`. Pre-dates the // tagged-union work; kept for callers that already use it. -// - Hare style (parse64/parseu64): `(value | str)`. The error +// - Hare style (stoi64/stou64): `(value | str)`. The error // variant carries a short, allocation-free message describing // why the parse failed. Prefer this for new code. -// u64toa — write `v` in decimal into `buf` and return the byte count. -// Unsigned-only so callers don't have to think about wraparound when -// printing a u64 that happens to have the high bit set. -export fn u64toa(buf: []u8, v: u64) i32 = { +// 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; let i: i32 = 0; let n: u64 = v; @@ -31,7 +33,7 @@ export fn u64toa(buf: []u8, v: u64) i32 = { return out; }; -export fn i64toa(buf: []u8, v: i64) i32 = { +export fn i64tos(buf: []u8, v: i64) i32 = { let neg: bool = false; let n: i64 = v; if (n < 0) { @@ -81,11 +83,12 @@ export fn atoi64(s: str) (i64, bool) = { return v, true; }; -// parse64 — Hare-style fallible signed decimal parser. The value -// variant is i64; the error variant is a short str describing the -// reason. No locale, no whitespace, no underscores: a leading '-' is -// the only non-digit accepted, and only at position 0. -export fn parse64(s: str) (i64 | str) = { +// stoi64 — Hare-style fallible signed decimal parser. The value +// variant is i64; the error variant is a short str (subset of Hare's +// (invalid | overflow) tagged-union). 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 | str) = { if (s.len == 0) { return "parse: empty"; }; let i: i32 = 0; let neg: bool = false; @@ -103,8 +106,9 @@ export fn parse64(s: str) (i64 | str) = { return v; }; -// parseu64 — fallible unsigned decimal parser. No leading sign. -export fn parseu64(s: str) (u64 | str) = { +// stou64 — fallible unsigned decimal parser. No leading sign. Mirrors +// Hare's strconv::stou64. +export fn stou64(s: str) (u64 | str) = { if (s.len == 0) { return "parse: empty"; }; let v: u64 = 0u64; let i: i32 = 0; diff --git a/lib/strings/strings.ww b/lib/strings/strings.ww index 55ed9d70..1f047f7c 100644 --- a/lib/strings/strings.ww +++ b/lib/strings/strings.ww @@ -1,23 +1,21 @@ // strings — operations over the immutable str type ({ *u8, len }). +// Mirrors Hare's strings::; `len` and `is-empty` aren't functions +// (callers use `s.len` and `s.len == 0` directly). use os; -export fn len(s: str) i32 = { - return s.len; -}; - -export fn isempty(s: str) bool = { - return s.len == 0; -}; - -export fn equal(a: str, b: str) bool = { - if (a.len != b.len) { return false; }; +// compare — bytewise three-way comparison: negative if ab. Matches Hare's strings::compare. ASCII-order, not +// locale-aware. Callers that just need equality use `compare(a, b) == 0`. +export fn compare(a: str, b: str) i32 = { + let n: i32 = a.len; + if (b.len < n) { n = b.len; }; let i: i32 = 0; - for (i < a.len) { - if (a[i] != b[i]) { return false; }; + for (i < n) { + if (a[i] != b[i]) { return (a[i]: i32) - (b[i]: i32); }; i += 1; }; - return true; + return a.len - b.len; }; export fn hasprefix(s: str, p: str) bool = { @@ -41,10 +39,11 @@ export fn hassuffix(s: str, suf: str) bool = { return true; }; -// indexbyte — first index of `c` in `s`, or -1 if absent. Plan 9- -// style sentinel return; callers that prefer a fallible shape can -// wrap this in their own (i32 | str). No allocation. -export fn indexbyte(s: str, c: u8) i32 = { +// byteindex — first index of byte `c` in `s`, or -1 if absent. Hare +// name (strings::byteindex). Plan 9-style sentinel return; callers that +// prefer a fallible shape can wrap this in their own (i32 | str). No +// allocation. +export fn byteindex(s: str, c: u8) i32 = { let i: i32 = 0; for (i < s.len) { if (s[i] == c) { return i; }; @@ -53,6 +52,17 @@ export fn indexbyte(s: str, c: u8) i32 = { return -1; }; +// rbyteindex — last index of byte `c` in `s`, or -1 if absent. Mirrors +// Hare's strings::rbyteindex. +export fn rbyteindex(s: str, c: u8) i32 = { + let i: i32 = s.len - 1; + for (i >= 0) { + if (s[i] == c) { return i; }; + i -= 1; + }; + return -1; +}; + // index — first index of `sub` in `s`, or -1. Naive scan; fine for // short patterns and small strings, which dominate config and CLI // parsing. Empty `sub` matches at 0. diff --git a/lib/types/types.ww b/lib/types/types.ww index 2ae5c120..6aba574d 100644 --- a/lib/types/types.ww +++ b/lib/types/types.ww @@ -1,6 +1,6 @@ -// types — integer limits and helpers, the seed module that the -// rest of the stdlib depends on. Plan 9-flavoured: the names are -// short and the constants are platform-fixed (we are amd64 only). +// types — integer limits. Mirrors Hare's types::limits (I8_MAX, …) +// platform-fixed for amd64. Numeric helpers live in lib/math, matching +// Hare's split between types::limits and math::. def I8_MAX: i8 = 127; def I16_MAX: i16 = 32767; @@ -16,33 +16,3 @@ def U8_MAX: u8 = 255; def U16_MAX: u16 = 65535; def U32_MAX: u32 = 4294967295; def U64_MAX: u64 = 18446744073709551615; - -export fn mini32(a: i32, b: i32) i32 = { - if (a < b) { return a; }; - return b; -}; - -export fn maxi32(a: i32, b: i32) i32 = { - if (a > b) { return a; }; - return b; -}; - -export fn mini64(a: i64, b: i64) i64 = { - if (a < b) { return a; }; - return b; -}; - -export fn maxi64(a: i64, b: i64) i64 = { - if (a > b) { return a; }; - return b; -}; - -export fn absi32(x: i32) i32 = { - if (x < 0) { return -x; }; - return x; -}; - -export fn absi64(x: i64) i64 = { - if (x < 0) { return -x; }; - return x; -}; diff --git a/lib/ww/ast.ww b/lib/ww/ast.ww index d8772200..a733e966 100644 --- a/lib/ww/ast.ww +++ b/lib/ww/ast.ww @@ -256,12 +256,12 @@ fn pr(fd: i32, n: *node, d: i32) void = { if (n.kind == N_INTLIT) { putc1(fd, 32u8); let buf: [32]u8; - let m: i32 = strconv.u64toa(buf[0:32], n.uval); + let m: i32 = strconv.u64tos(buf[0:32], n.uval); os.write(fd, buf.ptr, m: u64); } else { if (n.kind == N_RUNELIT) { putc1(fd, 32u8); let buf: [32]u8; - let m: i32 = strconv.u64toa(buf[0:32], n.uval); + let m: i32 = strconv.u64tos(buf[0:32], n.uval); os.write(fd, buf.ptr, m: u64); } else { if ( n.kind == N_STRLIT || diff --git a/lib/ww/lex/lex.ww b/lib/ww/lex/lex.ww index 55cc919b..56ca2fc2 100644 --- a/lib/ww/lex/lex.ww +++ b/lib/ww/lex/lex.ww @@ -230,18 +230,18 @@ fn escape(l: *lex, out: *i32) bool = { let lo: i32 = lget(l); if (hi < 0) { return false; }; if (lo < 0) { return false; }; - if (!ascii.ishex(hi: u8)) { + if (!ascii.isxdigit(hi: rune)) { let cp: pos; curpos(l, &cp); errat(l, &cp, "bad \\x escape"); return false; }; - if (!ascii.ishex(lo: u8)) { + if (!ascii.isxdigit(lo: rune)) { let cp: pos; curpos(l, &cp); errat(l, &cp, "bad \\x escape"); return false; }; - let h: i32 = ascii.digitval(hi: u8); - let lv: i32 = ascii.digitval(lo: u8); + let h: i32 = ascii.digitval(hi: rune); + let lv: i32 = ascii.digitval(lo: rune); *out = (h << 4) | lv; return true; }; @@ -255,7 +255,7 @@ fn scandecimalrun(l: *lex) void = { for (true) { let c: i32 = lpeek(l, 0u64); if (c < 0) { break; }; - if (!ascii.isdigit(c: u8)) { + if (!ascii.isdigit(c: rune)) { if (c != 95) { break; }; }; lget(l); @@ -266,7 +266,7 @@ fn scanhexrun(l: *lex) void = { for (true) { let c: i32 = lpeek(l, 0u64); if (c < 0) { break; }; - if (!ascii.ishex(c: u8)) { + if (!ascii.isxdigit(c: rune)) { if (c != 95) { break; }; }; lget(l); @@ -305,7 +305,7 @@ fn scanexp(l: *lex) void = { for (true) { let c: i32 = lpeek(l, 0u64); if (c < 0) { break; }; - if (!ascii.isdigit(c: u8)) { break; }; + if (!ascii.isdigit(c: rune)) { break; }; lget(l); }; }; @@ -390,12 +390,12 @@ fn lexnum(l: *lex, start: *pos, out: *tok) void = { let pc: i32 = lpeek(l, 0u64); if (pc >= 0) { - if (ascii.isidstart(pc: u8)) { + if (ascii.isidstart(pc: rune)) { let sb: u64 = l.lpos; for (true) { let cc: i32 = lpeek(l, 0u64); if (cc < 0) { break; }; - if (!ascii.isidpart(cc: u8)) { break; }; + if (!ascii.isidpart(cc: rune)) { break; }; lget(l); }; let sl: u64 = l.lpos - sb; @@ -439,7 +439,7 @@ fn lexident(l: *lex, start: *pos, out: *tok) void = { for (true) { let c: i32 = lpeek(l, 0u64); if (c < 0) { break; }; - if (!ascii.isidpart(c: u8)) { break; }; + if (!ascii.isidpart(c: rune)) { break; }; lget(l); }; let n: u64 = l.lpos - begin; @@ -588,8 +588,8 @@ export fn lexnext(l: *lex, out: *tok) void = { let c: i32 = lpeek(l, 0u64); if (c >= 0) { - if (ascii.isidstart(c: u8)) { lexident(l, &start, out); return; }; - if (ascii.isdigit(c: u8)) { lexnum(l, &start, out); return; }; + if (ascii.isidstart(c: rune)) { lexident(l, &start, out); return; }; + if (ascii.isdigit(c: rune)) { lexnum(l, &start, out); return; }; }; if (c == 34) { lget(l); lexstr(l, &start, out); return; }; diff --git a/lib/ww/lex/tok.ww b/lib/ww/lex/tok.ww index 18eea126..da7ac879 100644 --- a/lib/ww/lex/tok.ww +++ b/lib/ww/lex/tok.ww @@ -372,10 +372,10 @@ export fn tokprint(fd: i32, t: *tok) void = { }; fputcbyte(fd, 58u8); // ':' let buf: [32]u8; - let n: i32 = strconv.i64toa(buf[0:32], t.line: i64); + let n: i32 = strconv.i64tos(buf[0:32], t.line: i64); os.write(fd, buf.ptr, n: u64); fputcbyte(fd, 58u8); - n = strconv.i64toa(buf[0:32], t.col: i64); + n = strconv.i64tos(buf[0:32], t.col: i64); os.write(fd, buf.ptr, n: u64); fputcbyte(fd, 32u8); // ' ' fputsstr(fd, tokname(t.kind)); @@ -391,11 +391,11 @@ export fn tokprint(fd: i32, t: *tok) void = { fputq(fd, ttext.ptr, ttext.len); } else { if (t.kind == TK_INT) { fputcbyte(fd, 32u8); - n = strconv.u64toa(buf[0:32], t.uval); + n = strconv.u64tos(buf[0:32], t.uval); os.write(fd, buf.ptr, n: u64); } else { if (t.kind == TK_RUNE) { fputcbyte(fd, 32u8); - n = strconv.u64toa(buf[0:32], t.uval); + n = strconv.u64tos(buf[0:32], t.uval); os.write(fd, buf.ptr, n: u64); };};};};}; // TK_FLOAT is intentionally not handled here — %g formatting diff --git a/selfhost/cmd/w6a/main.combined.ww b/selfhost/cmd/w6a/main.combined.ww index 4f9dd1f3..69d2dbeb 100644 --- a/selfhost/cmd/w6a/main.combined.ww +++ b/selfhost/cmd/w6a/main.combined.ww @@ -120,9 +120,11 @@ export fn filesize(fd: i32) i64 = { return end; }; -// readfull — keep reading until `n` bytes have arrived or the fd +// readall — keep reading until `n` bytes have arrived or the fd // closes early. Returns bytes read (0..=n) or -1 on read error. -export fn readfull(fd: i32, buf: *u8, n: u64) i64 = { +// Hare name (io::readall); the buffer is caller-supplied, matching +// the Plan 9 subset convention. +export fn readall(fd: i32, buf: *u8, n: u64) i64 = { let got: u64 = 0u64; for (got < n) { let r: i64 = read(fd, buf + got, n - got); @@ -133,9 +135,10 @@ export fn readfull(fd: i32, buf: *u8, n: u64) i64 = { return got: i64; }; -// writefull — keep writing until `n` bytes have been accepted or the -// fd refuses progress. Returns bytes written or -1. -export fn writefull(fd: i32, buf: *u8, n: u64) i64 = { +// writeall — keep writing until `n` bytes have been accepted or the +// fd refuses progress. Returns bytes written or -1. Hare name +// (io::writeall). +export fn writeall(fd: i32, buf: *u8, n: u64) i64 = { let sent: u64 = 0u64; for (sent < n) { let r: i64 = write(fd, buf + sent, n - sent); @@ -154,8 +157,8 @@ export fn access(path: *u8, mode: i32) i32 = { return syscall2(SYS_ACCESS, path: i64, mode: i64): i32; }; -// unlink(2). -export fn unlink(path: *u8) i32 = { +// remove — unlink(2). Hare name; the underlying syscall is unlink(2). +export fn remove(path: *u8) i32 = { return syscall1(SYS_UNLINK, path: i64): i32; }; @@ -2040,27 +2043,27 @@ export fn emitelf(a: *asm_, fd: i32) i32 = { wru16(eh.ptr, 60u64, NSECT); // e_shnum wru16(eh.ptr, 62u64, 5u16); // e_shstrndx - if (os.writefull(fd, eh.ptr, 64u64) != 64i64) { return -1; }; + if (os.writeall(fd, eh.ptr, 64u64) != 64i64) { return -1; }; if (a.textlen > 0u64) { - if (os.writefull(fd, a.text, a.textlen) != a.textlen: i64) { return -1; }; + if (os.writeall(fd, a.text, a.textlen) != a.textlen: i64) { return -1; }; }; if (rela.n > 0u64) { - if (os.writefull(fd, rela.p, rela.n) != rela.n: i64) { return -1; }; + if (os.writeall(fd, rela.p, rela.n) != rela.n: i64) { return -1; }; }; if (sym.n > 0u64) { - if (os.writefull(fd, sym.p, sym.n) != sym.n: i64) { return -1; }; + if (os.writeall(fd, sym.p, sym.n) != sym.n: i64) { return -1; }; }; if (str_.n > 0u64) { - if (os.writefull(fd, str_.p, str_.n) != str_.n: i64) { return -1; }; + if (os.writeall(fd, str_.p, str_.n) != str_.n: i64) { return -1; }; }; if (shstr.n > 0u64) { - if (os.writefull(fd, shstr.p, shstr.n) != shstr.n: i64) { return -1; }; + if (os.writeall(fd, shstr.p, shstr.n) != shstr.n: i64) { return -1; }; }; // Pad to 8 before shdrs. let written: u64 = EHDR_SZ + a.textlen + rela.n + sym.n + str_.n + shstr.n; for ((written & 7u64) != 0u64) { - os.writefull(fd, &zero, 1u64); + os.writeall(fd, &zero, 1u64); written += 1u64; }; @@ -2069,7 +2072,7 @@ export fn emitelf(a: *asm_, fd: i32) i32 = { // SHT_NULL let sn: i32 = 0; for (sn < 64) { shbuf[sn] = 0u8; sn += 1; }; - os.writefull(fd, shbuf.ptr, 64u64); + os.writeall(fd, shbuf.ptr, 64u64); // .text sn = 0; for (sn < 64) { shbuf[sn] = 0u8; sn += 1; }; @@ -2079,7 +2082,7 @@ export fn emitelf(a: *asm_, fd: i32) i32 = { wru64(shbuf.ptr, 24u64, offtext); wru64(shbuf.ptr, 32u64, a.textlen); wru64(shbuf.ptr, 48u64, 1u64); // sh_addralign - os.writefull(fd, shbuf.ptr, 64u64); + os.writeall(fd, shbuf.ptr, 64u64); // .rela.text sn = 0; for (sn < 64) { shbuf[sn] = 0u8; sn += 1; }; @@ -2092,7 +2095,7 @@ export fn emitelf(a: *asm_, fd: i32) i32 = { wru32(shbuf.ptr, 44u64, 1u32); // sh_info = .text idx wru64(shbuf.ptr, 48u64, 8u64); wru64(shbuf.ptr, 56u64, RELA_SZ); - os.writefull(fd, shbuf.ptr, 64u64); + os.writeall(fd, shbuf.ptr, 64u64); // .symtab sn = 0; for (sn < 64) { shbuf[sn] = 0u8; sn += 1; }; @@ -2104,7 +2107,7 @@ export fn emitelf(a: *asm_, fd: i32) i32 = { wru32(shbuf.ptr, 44u64, 1u32); // sh_info = one local (STN_UNDEF) wru64(shbuf.ptr, 48u64, 8u64); wru64(shbuf.ptr, 56u64, SYM_SZ); - os.writefull(fd, shbuf.ptr, 64u64); + os.writeall(fd, shbuf.ptr, 64u64); // .strtab sn = 0; for (sn < 64) { shbuf[sn] = 0u8; sn += 1; }; @@ -2113,7 +2116,7 @@ export fn emitelf(a: *asm_, fd: i32) i32 = { wru64(shbuf.ptr, 24u64, offstr); wru64(shbuf.ptr, 32u64, str_.n); wru64(shbuf.ptr, 48u64, 1u64); - os.writefull(fd, shbuf.ptr, 64u64); + os.writeall(fd, shbuf.ptr, 64u64); // .shstrtab sn = 0; for (sn < 64) { shbuf[sn] = 0u8; sn += 1; }; @@ -2122,7 +2125,7 @@ export fn emitelf(a: *asm_, fd: i32) i32 = { wru64(shbuf.ptr, 24u64, offshstr); wru64(shbuf.ptr, 32u64, shstr.n); wru64(shbuf.ptr, 48u64, 1u64); - os.writefull(fd, shbuf.ptr, 64u64); + os.writeall(fd, shbuf.ptr, 64u64); return 0; }; @@ -2168,7 +2171,7 @@ fn slurp(path: *u8) (*u8, u64) = { if (n < 0i64) { os.close(fd); return nil, 0u64; }; let nz: u64 = n: u64; let buf: *u8 = os.alloc(nz + 1u64): *u8; - let got: i64 = os.readfull(fd, buf, nz); + let got: i64 = os.readall(fd, buf, nz); os.close(fd); if (got != n) { return nil, 0u64; }; buf[nz] = 0u8; diff --git a/selfhost/cmd/w6a/main.ww b/selfhost/cmd/w6a/main.ww index cdad9244..82ef34f6 100644 --- a/selfhost/cmd/w6a/main.ww +++ b/selfhost/cmd/w6a/main.ww @@ -38,7 +38,7 @@ fn slurp(path: *u8) (*u8, u64) = { if (n < 0i64) { os.close(fd); return nil, 0u64; }; let nz: u64 = n: u64; let buf: *u8 = os.alloc(nz + 1u64): *u8; - let got: i64 = os.readfull(fd, buf, nz); + let got: i64 = os.readall(fd, buf, nz); os.close(fd); if (got != n) { return nil, 0u64; }; buf[nz] = 0u8; diff --git a/selfhost/cmd/w6a/obj.ww b/selfhost/cmd/w6a/obj.ww index 5cdadc16..acf6b1fc 100644 --- a/selfhost/cmd/w6a/obj.ww +++ b/selfhost/cmd/w6a/obj.ww @@ -205,27 +205,27 @@ export fn emitelf(a: *asm_, fd: i32) i32 = { wru16(eh.ptr, 60u64, NSECT); // e_shnum wru16(eh.ptr, 62u64, 5u16); // e_shstrndx - if (os.writefull(fd, eh.ptr, 64u64) != 64i64) { return -1; }; + if (os.writeall(fd, eh.ptr, 64u64) != 64i64) { return -1; }; if (a.textlen > 0u64) { - if (os.writefull(fd, a.text, a.textlen) != a.textlen: i64) { return -1; }; + if (os.writeall(fd, a.text, a.textlen) != a.textlen: i64) { return -1; }; }; if (rela.n > 0u64) { - if (os.writefull(fd, rela.p, rela.n) != rela.n: i64) { return -1; }; + if (os.writeall(fd, rela.p, rela.n) != rela.n: i64) { return -1; }; }; if (sym.n > 0u64) { - if (os.writefull(fd, sym.p, sym.n) != sym.n: i64) { return -1; }; + if (os.writeall(fd, sym.p, sym.n) != sym.n: i64) { return -1; }; }; if (str_.n > 0u64) { - if (os.writefull(fd, str_.p, str_.n) != str_.n: i64) { return -1; }; + if (os.writeall(fd, str_.p, str_.n) != str_.n: i64) { return -1; }; }; if (shstr.n > 0u64) { - if (os.writefull(fd, shstr.p, shstr.n) != shstr.n: i64) { return -1; }; + if (os.writeall(fd, shstr.p, shstr.n) != shstr.n: i64) { return -1; }; }; // Pad to 8 before shdrs. let written: u64 = EHDR_SZ + a.textlen + rela.n + sym.n + str_.n + shstr.n; for ((written & 7u64) != 0u64) { - os.writefull(fd, &zero, 1u64); + os.writeall(fd, &zero, 1u64); written += 1u64; }; @@ -234,7 +234,7 @@ export fn emitelf(a: *asm_, fd: i32) i32 = { // SHT_NULL let sn: i32 = 0; for (sn < 64) { shbuf[sn] = 0u8; sn += 1; }; - os.writefull(fd, shbuf.ptr, 64u64); + os.writeall(fd, shbuf.ptr, 64u64); // .text sn = 0; for (sn < 64) { shbuf[sn] = 0u8; sn += 1; }; @@ -244,7 +244,7 @@ export fn emitelf(a: *asm_, fd: i32) i32 = { wru64(shbuf.ptr, 24u64, offtext); wru64(shbuf.ptr, 32u64, a.textlen); wru64(shbuf.ptr, 48u64, 1u64); // sh_addralign - os.writefull(fd, shbuf.ptr, 64u64); + os.writeall(fd, shbuf.ptr, 64u64); // .rela.text sn = 0; for (sn < 64) { shbuf[sn] = 0u8; sn += 1; }; @@ -257,7 +257,7 @@ export fn emitelf(a: *asm_, fd: i32) i32 = { wru32(shbuf.ptr, 44u64, 1u32); // sh_info = .text idx wru64(shbuf.ptr, 48u64, 8u64); wru64(shbuf.ptr, 56u64, RELA_SZ); - os.writefull(fd, shbuf.ptr, 64u64); + os.writeall(fd, shbuf.ptr, 64u64); // .symtab sn = 0; for (sn < 64) { shbuf[sn] = 0u8; sn += 1; }; @@ -269,7 +269,7 @@ export fn emitelf(a: *asm_, fd: i32) i32 = { wru32(shbuf.ptr, 44u64, 1u32); // sh_info = one local (STN_UNDEF) wru64(shbuf.ptr, 48u64, 8u64); wru64(shbuf.ptr, 56u64, SYM_SZ); - os.writefull(fd, shbuf.ptr, 64u64); + os.writeall(fd, shbuf.ptr, 64u64); // .strtab sn = 0; for (sn < 64) { shbuf[sn] = 0u8; sn += 1; }; @@ -278,7 +278,7 @@ export fn emitelf(a: *asm_, fd: i32) i32 = { wru64(shbuf.ptr, 24u64, offstr); wru64(shbuf.ptr, 32u64, str_.n); wru64(shbuf.ptr, 48u64, 1u64); - os.writefull(fd, shbuf.ptr, 64u64); + os.writeall(fd, shbuf.ptr, 64u64); // .shstrtab sn = 0; for (sn < 64) { shbuf[sn] = 0u8; sn += 1; }; @@ -287,7 +287,7 @@ export fn emitelf(a: *asm_, fd: i32) i32 = { wru64(shbuf.ptr, 24u64, offshstr); wru64(shbuf.ptr, 32u64, shstr.n); wru64(shbuf.ptr, 48u64, 1u64); - os.writefull(fd, shbuf.ptr, 64u64); + os.writeall(fd, shbuf.ptr, 64u64); return 0; }; diff --git a/selfhost/cmd/w6c/main.combined.ww b/selfhost/cmd/w6c/main.combined.ww index ad8004b4..bc943174 100644 --- a/selfhost/cmd/w6c/main.combined.ww +++ b/selfhost/cmd/w6c/main.combined.ww @@ -120,9 +120,11 @@ export fn filesize(fd: i32) i64 = { return end; }; -// readfull — keep reading until `n` bytes have arrived or the fd +// readall — keep reading until `n` bytes have arrived or the fd // closes early. Returns bytes read (0..=n) or -1 on read error. -export fn readfull(fd: i32, buf: *u8, n: u64) i64 = { +// Hare name (io::readall); the buffer is caller-supplied, matching +// the Plan 9 subset convention. +export fn readall(fd: i32, buf: *u8, n: u64) i64 = { let got: u64 = 0u64; for (got < n) { let r: i64 = read(fd, buf + got, n - got); @@ -133,9 +135,10 @@ export fn readfull(fd: i32, buf: *u8, n: u64) i64 = { return got: i64; }; -// writefull — keep writing until `n` bytes have been accepted or the -// fd refuses progress. Returns bytes written or -1. -export fn writefull(fd: i32, buf: *u8, n: u64) i64 = { +// writeall — keep writing until `n` bytes have been accepted or the +// fd refuses progress. Returns bytes written or -1. Hare name +// (io::writeall). +export fn writeall(fd: i32, buf: *u8, n: u64) i64 = { let sent: u64 = 0u64; for (sent < n) { let r: i64 = write(fd, buf + sent, n - sent); @@ -154,8 +157,8 @@ export fn access(path: *u8, mode: i32) i32 = { return syscall2(SYS_ACCESS, path: i64, mode: i64): i32; }; -// unlink(2). -export fn unlink(path: *u8) i32 = { +// remove — unlink(2). Hare name; the underlying syscall is unlink(2). +export fn remove(path: *u8) i32 = { return syscall1(SYS_UNLINK, path: i64): i32; }; @@ -321,14 +324,16 @@ export fn freearena(a: *arena) void = { // buffer. Two error idioms ship side by side: // - Plan 9 style (atoi64): tuple `(value, ok)`. Pre-dates the // tagged-union work; kept for callers that already use it. -// - Hare style (parse64/parseu64): `(value | str)`. The error +// - Hare style (stoi64/stou64): `(value | str)`. The error // variant carries a short, allocation-free message describing // why the parse failed. Prefer this for new code. -// u64toa — write `v` in decimal into `buf` and return the byte count. -// Unsigned-only so callers don't have to think about wraparound when -// printing a u64 that happens to have the high bit set. -export fn u64toa(buf: []u8, v: u64) i32 = { +// 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; let i: i32 = 0; let n: u64 = v; @@ -350,7 +355,7 @@ export fn u64toa(buf: []u8, v: u64) i32 = { return out; }; -export fn i64toa(buf: []u8, v: i64) i32 = { +export fn i64tos(buf: []u8, v: i64) i32 = { let neg: bool = false; let n: i64 = v; if (n < 0) { @@ -400,11 +405,12 @@ export fn atoi64(s: str) (i64, bool) = { return v, true; }; -// parse64 — Hare-style fallible signed decimal parser. The value -// variant is i64; the error variant is a short str describing the -// reason. No locale, no whitespace, no underscores: a leading '-' is -// the only non-digit accepted, and only at position 0. -export fn parse64(s: str) (i64 | str) = { +// stoi64 — Hare-style fallible signed decimal parser. The value +// variant is i64; the error variant is a short str (subset of Hare's +// (invalid | overflow) tagged-union). 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 | str) = { if (s.len == 0) { return "parse: empty"; }; let i: i32 = 0; let neg: bool = false; @@ -422,8 +428,9 @@ export fn parse64(s: str) (i64 | str) = { return v; }; -// parseu64 — fallible unsigned decimal parser. No leading sign. -export fn parseu64(s: str) (u64 | str) = { +// stou64 — fallible unsigned decimal parser. No leading sign. Mirrors +// Hare's strconv::stou64. +export fn stou64(s: str) (u64 | str) = { if (s.len == 0) { return "parse: empty"; }; let v: u64 = 0u64; let i: i32 = 0; @@ -812,10 +819,10 @@ export fn tokprint(fd: i32, t: *tok) void = { }; fputcbyte(fd, 58u8); // ':' let buf: [32]u8; - let n: i32 = strconv.i64toa(buf[0:32], t.line: i64); + let n: i32 = strconv.i64tos(buf[0:32], t.line: i64); os.write(fd, buf.ptr, n: u64); fputcbyte(fd, 58u8); - n = strconv.i64toa(buf[0:32], t.col: i64); + n = strconv.i64tos(buf[0:32], t.col: i64); os.write(fd, buf.ptr, n: u64); fputcbyte(fd, 32u8); // ' ' fputsstr(fd, tokname(t.kind)); @@ -831,11 +838,11 @@ export fn tokprint(fd: i32, t: *tok) void = { fputq(fd, ttext.ptr, ttext.len); } else { if (t.kind == TK_INT) { fputcbyte(fd, 32u8); - n = strconv.u64toa(buf[0:32], t.uval); + n = strconv.u64tos(buf[0:32], t.uval); os.write(fd, buf.ptr, n: u64); } else { if (t.kind == TK_RUNE) { fputcbyte(fd, 32u8); - n = strconv.u64toa(buf[0:32], t.uval); + n = strconv.u64tos(buf[0:32], t.uval); os.write(fd, buf.ptr, n: u64); };};};};}; // TK_FLOAT is intentionally not handled here — %g formatting @@ -846,96 +853,96 @@ export fn tokprint(fd: i32, t: *tok) void = { }; // MODULE: ascii -// ascii — byte-class predicates and case folding for the ASCII range. -// Matches Hare's ascii::isdigit family. Bytes outside 0..127 always -// answer `false`. The lexer hot path uses these inline; they are -// expected to inline to a couple of compares. +// ascii — rune-class predicates and case folding for the ASCII range. +// Matches Hare's ascii::isdigit family (rune-taking signature). Runes +// outside 0..127 always answer `false`. The lexer hot path uses these +// inline; they are expected to inline to a couple of compares. -export fn isdigit(c: u8) bool = { - if (c < 48u8) { return false; }; - if (c > 57u8) { return false; }; +export fn isdigit(c: rune) bool = { + if (c < 48) { return false; }; + if (c > 57) { return false; }; return true; }; -export fn isupper(c: u8) bool = { - if (c < 65u8) { return false; }; - if (c > 90u8) { return false; }; +export fn isupper(c: rune) bool = { + if (c < 65) { return false; }; + if (c > 90) { return false; }; return true; }; -export fn islower(c: u8) bool = { - if (c < 97u8) { return false; }; - if (c > 122u8) { return false; }; +export fn islower(c: rune) bool = { + if (c < 97) { return false; }; + if (c > 122) { return false; }; return true; }; -export fn isalpha(c: u8) bool = { +export fn isalpha(c: rune) bool = { if (isupper(c)) { return true; }; return islower(c); }; -export fn isalnum(c: u8) bool = { +export fn isalnum(c: rune) bool = { if (isalpha(c)) { return true; }; return isdigit(c); }; // isspace — the C/Hare set: space, tab, NL, VT, FF, CR. -export fn isspace(c: u8) bool = { - if (c == 32u8) { return true; }; // ' ' - if (c == 9u8) { return true; }; // '\t' - if (c == 10u8) { return true; }; // '\n' - if (c == 11u8) { return true; }; // '\v' - if (c == 12u8) { return true; }; // '\f' - if (c == 13u8) { return true; }; // '\r' +export fn isspace(c: rune) bool = { + if (c == 32) { return true; }; // ' ' + if (c == 9) { return true; }; // '\t' + if (c == 10) { return true; }; // '\n' + if (c == 11) { return true; }; // '\v' + if (c == 12) { return true; }; // '\f' + if (c == 13) { return true; }; // '\r' return false; }; -export fn ishex(c: u8) bool = { +export fn isxdigit(c: rune) bool = { if (isdigit(c)) { return true; }; - if (c >= 65u8) { - if (c <= 70u8) { return true; }; // 'A'..'F' + if (c >= 65) { + if (c <= 70) { return true; }; // 'A'..'F' }; - if (c >= 97u8) { - if (c <= 102u8) { return true; }; // 'a'..'f' + if (c >= 97) { + if (c <= 102) { return true; }; // 'a'..'f' }; return false; }; // digitval — value of `c` as a hex/decimal digit, or -1 if not one. // Useful when scanning numeric literals. -export fn digitval(c: u8) i32 = { - if (isdigit(c)) { return (c - 48u8): i32; }; - if (c >= 65u8) { - if (c <= 70u8) { return ((c - 65u8) + 10u8): i32; }; +export fn digitval(c: rune) i32 = { + if (isdigit(c)) { return (c - 48): i32; }; + if (c >= 65) { + if (c <= 70) { return ((c - 65) + 10): i32; }; }; - if (c >= 97u8) { - if (c <= 102u8) { return ((c - 97u8) + 10u8): i32; }; + if (c >= 97) { + if (c <= 102) { return ((c - 97) + 10): i32; }; }; return -1; }; // isidstart / isidpart — identifier classes used by the lexer. // Alpha or '_' starts; alnum or '_' continues. -export fn isidstart(c: u8) bool = { +export fn isidstart(c: rune) bool = { if (isalpha(c)) { return true; }; - if (c == 95u8) { return true; }; // '_' + if (c == 95) { return true; }; // '_' return false; }; -export fn isidpart(c: u8) bool = { +export fn isidpart(c: rune) bool = { if (isalnum(c)) { return true; }; - if (c == 95u8) { return true; }; + if (c == 95) { return true; }; return false; }; // tolower / toupper — fold ASCII case. Non-letters pass through. -export fn tolower(c: u8) u8 = { - if (isupper(c)) { return c + 32u8; }; +export fn tolower(c: rune) rune = { + if (isupper(c)) { return c + 32; }; return c; }; -export fn toupper(c: u8) u8 = { - if (islower(c)) { return c - 32u8; }; +export fn toupper(c: rune) rune = { + if (islower(c)) { return c - 32; }; return c; }; @@ -1172,18 +1179,18 @@ fn escape(l: *lex, out: *i32) bool = { let lo: i32 = lget(l); if (hi < 0) { return false; }; if (lo < 0) { return false; }; - if (!ascii.ishex(hi: u8)) { + if (!ascii.isxdigit(hi: rune)) { let cp: pos; curpos(l, &cp); errat(l, &cp, "bad \\x escape"); return false; }; - if (!ascii.ishex(lo: u8)) { + if (!ascii.isxdigit(lo: rune)) { let cp: pos; curpos(l, &cp); errat(l, &cp, "bad \\x escape"); return false; }; - let h: i32 = ascii.digitval(hi: u8); - let lv: i32 = ascii.digitval(lo: u8); + let h: i32 = ascii.digitval(hi: rune); + let lv: i32 = ascii.digitval(lo: rune); *out = (h << 4) | lv; return true; }; @@ -1197,7 +1204,7 @@ fn scandecimalrun(l: *lex) void = { for (true) { let c: i32 = lpeek(l, 0u64); if (c < 0) { break; }; - if (!ascii.isdigit(c: u8)) { + if (!ascii.isdigit(c: rune)) { if (c != 95) { break; }; }; lget(l); @@ -1208,7 +1215,7 @@ fn scanhexrun(l: *lex) void = { for (true) { let c: i32 = lpeek(l, 0u64); if (c < 0) { break; }; - if (!ascii.ishex(c: u8)) { + if (!ascii.isxdigit(c: rune)) { if (c != 95) { break; }; }; lget(l); @@ -1247,7 +1254,7 @@ fn scanexp(l: *lex) void = { for (true) { let c: i32 = lpeek(l, 0u64); if (c < 0) { break; }; - if (!ascii.isdigit(c: u8)) { break; }; + if (!ascii.isdigit(c: rune)) { break; }; lget(l); }; }; @@ -1332,12 +1339,12 @@ fn lexnum(l: *lex, start: *pos, out: *tok) void = { let pc: i32 = lpeek(l, 0u64); if (pc >= 0) { - if (ascii.isidstart(pc: u8)) { + if (ascii.isidstart(pc: rune)) { let sb: u64 = l.lpos; for (true) { let cc: i32 = lpeek(l, 0u64); if (cc < 0) { break; }; - if (!ascii.isidpart(cc: u8)) { break; }; + if (!ascii.isidpart(cc: rune)) { break; }; lget(l); }; let sl: u64 = l.lpos - sb; @@ -1381,7 +1388,7 @@ fn lexident(l: *lex, start: *pos, out: *tok) void = { for (true) { let c: i32 = lpeek(l, 0u64); if (c < 0) { break; }; - if (!ascii.isidpart(c: u8)) { break; }; + if (!ascii.isidpart(c: rune)) { break; }; lget(l); }; let n: u64 = l.lpos - begin; @@ -1530,8 +1537,8 @@ export fn lexnext(l: *lex, out: *tok) void = { let c: i32 = lpeek(l, 0u64); if (c >= 0) { - if (ascii.isidstart(c: u8)) { lexident(l, &start, out); return; }; - if (ascii.isdigit(c: u8)) { lexnum(l, &start, out); return; }; + if (ascii.isidstart(c: rune)) { lexident(l, &start, out); return; }; + if (ascii.isdigit(c: rune)) { lexnum(l, &start, out); return; }; }; if (c == 34) { lget(l); lexstr(l, &start, out); return; }; @@ -1895,12 +1902,12 @@ fn pr(fd: i32, n: *node, d: i32) void = { if (n.kind == N_INTLIT) { putc1(fd, 32u8); let buf: [32]u8; - let m: i32 = strconv.u64toa(buf[0:32], n.uval); + let m: i32 = strconv.u64tos(buf[0:32], n.uval); os.write(fd, buf.ptr, m: u64); } else { if (n.kind == N_RUNELIT) { putc1(fd, 32u8); let buf: [32]u8; - let m: i32 = strconv.u64toa(buf[0:32], n.uval); + let m: i32 = strconv.u64tos(buf[0:32], n.uval); os.write(fd, buf.ptr, m: u64); } else { if ( n.kind == N_STRLIT || @@ -7408,13 +7415,13 @@ fn emitline(s: str) void = { os.write(1, s.ptr, s.len: u64); }; fn emitint(v: i64) void = { let buf: [32]u8; - let n: i32 = strconv.i64toa(buf[0:32], v); + let n: i32 = strconv.i64tos(buf[0:32], v); os.write(1, buf.ptr, n: u64); }; fn emituint(v: u64) void = { let buf: [32]u8; - let n: i32 = strconv.u64toa(buf[0:32], v); + let n: i32 = strconv.u64tos(buf[0:32], v); os.write(1, buf.ptr, n: u64); }; @@ -7452,7 +7459,7 @@ fn mklabel(c: *cgen, base: str) str = { i += 1; j += 1; }; buf[i] = 95u8; i += 1; // '_' - let n: i32 = strconv.i64toa(buf[i:128], c.labelseq: i64); + let n: i32 = strconv.i64tos(buf[i:128], c.labelseq: i64); c.labelseq += 1; let total: i32 = i + n; let p: *u8 = amalloc(c.a, (total: u64) + 1u64): *u8; @@ -7491,7 +7498,7 @@ fn internstrlit(c: *cgen, bytes: str) str = { // New label "_S_". let buf: [32]u8; buf[0] = 95u8; buf[1] = 83u8; buf[2] = 95u8; // "_S_" - let n: i32 = strconv.i64toa(buf[3:32], c.strlitseq: i64); + let n: i32 = strconv.i64tos(buf[3:32], c.strlitseq: i64); c.strlitseq += 1; let total: i32 = 3 + n; let p: *u8 = amalloc(c.a, (total: u64) + 1u64): *u8; @@ -7930,7 +7937,7 @@ fn slurp(path: *u8) (*u8, u64) = { if (n < 0i64) { os.close(fd); return nil, 0u64; }; let nz: u64 = n: u64; let buf: *u8 = os.alloc(nz + 1u64): *u8; - let got: i64 = os.readfull(fd, buf, nz); + let got: i64 = os.readall(fd, buf, nz); os.close(fd); if (got != n) { return nil, 0u64; }; buf[nz] = 0u8; diff --git a/selfhost/cmd/w6c/main.ww b/selfhost/cmd/w6c/main.ww index 905eaf99..df03c843 100644 --- a/selfhost/cmd/w6c/main.ww +++ b/selfhost/cmd/w6c/main.ww @@ -46,7 +46,7 @@ fn slurp(path: *u8) (*u8, u64) = { if (n < 0i64) { os.close(fd); return nil, 0u64; }; let nz: u64 = n: u64; let buf: *u8 = os.alloc(nz + 1u64): *u8; - let got: i64 = os.readfull(fd, buf, nz); + let got: i64 = os.readall(fd, buf, nz); os.close(fd); if (got != n) { return nil, 0u64; }; buf[nz] = 0u8; diff --git a/selfhost/cmd/w6l/dyn.ww b/selfhost/cmd/w6l/dyn.ww index 6353b041..0abbdf44 100644 --- a/selfhost/cmd/w6l/dyn.ww +++ b/selfhost/cmd/w6l/dyn.ww @@ -123,13 +123,13 @@ fn dbasename(p: *u8) *u8 = { // ---- file slurp -------------------------------------------------------- -fn readallso(path: *u8) (*u8, u64) = { +fn slurpso(path: *u8) (*u8, u64) = { let fd: i32 = os.open(path, os.O_RDONLY, 0i32); if (fd < 0) { return nil, 0u64; }; let n: i64 = os.filesize(fd); if (n < 0i64) { os.close(fd); return nil, 0u64; }; let buf: *u8 = os.alloc(n: u64): *u8; - let got: i64 = os.readfull(fd, buf, n: u64); + let got: i64 = os.readall(fd, buf, n: u64); os.close(fd); if (got != n) { return nil, 0u64; }; return buf, n: u64; @@ -165,7 +165,7 @@ fn vdnameat(buf: *u8, verdefoff: u64, verdefsize: u64, export fn loadso(l: *lnk, path: *u8) i32 = { let buf: *u8; let blen: u64; - buf, blen = readallso(path); + buf, blen = slurpso(path); if (buf == nil) { os.write(2, "w6l: cannot read .so\n".ptr, 20u64); return -1; diff --git a/selfhost/cmd/w6l/dynout.ww b/selfhost/cmd/w6l/dynout.ww index 52f0401d..d0b5fb93 100644 --- a/selfhost/cmd/w6l/dynout.ww +++ b/selfhost/cmd/w6l/dynout.ww @@ -720,7 +720,7 @@ export fn emitdynelf(l: *lnk, fd: i32, base: u64, entry: u64) i32 = { dbcopy(filebuf, gotpltoff, gotpltbuf, gotpltsz); dbcopy(filebuf, dynamicoff, dynamicbuf, dynamicsz); - let wrote: i64 = os.writefull(fd, filebuf, fileend); + let wrote: i64 = os.writeall(fd, filebuf, fileend); if (wrote != fileend: i64) { return 1; }; diff --git a/selfhost/cmd/w6l/main.combined.ww b/selfhost/cmd/w6l/main.combined.ww index b7ec7674..476e1b0d 100644 --- a/selfhost/cmd/w6l/main.combined.ww +++ b/selfhost/cmd/w6l/main.combined.ww @@ -120,9 +120,11 @@ export fn filesize(fd: i32) i64 = { return end; }; -// readfull — keep reading until `n` bytes have arrived or the fd +// readall — keep reading until `n` bytes have arrived or the fd // closes early. Returns bytes read (0..=n) or -1 on read error. -export fn readfull(fd: i32, buf: *u8, n: u64) i64 = { +// Hare name (io::readall); the buffer is caller-supplied, matching +// the Plan 9 subset convention. +export fn readall(fd: i32, buf: *u8, n: u64) i64 = { let got: u64 = 0u64; for (got < n) { let r: i64 = read(fd, buf + got, n - got); @@ -133,9 +135,10 @@ export fn readfull(fd: i32, buf: *u8, n: u64) i64 = { return got: i64; }; -// writefull — keep writing until `n` bytes have been accepted or the -// fd refuses progress. Returns bytes written or -1. -export fn writefull(fd: i32, buf: *u8, n: u64) i64 = { +// writeall — keep writing until `n` bytes have been accepted or the +// fd refuses progress. Returns bytes written or -1. Hare name +// (io::writeall). +export fn writeall(fd: i32, buf: *u8, n: u64) i64 = { let sent: u64 = 0u64; for (sent < n) { let r: i64 = write(fd, buf + sent, n - sent); @@ -154,8 +157,8 @@ export fn access(path: *u8, mode: i32) i32 = { return syscall2(SYS_ACCESS, path: i64, mode: i64): i32; }; -// unlink(2). -export fn unlink(path: *u8) i32 = { +// remove — unlink(2). Hare name; the underlying syscall is unlink(2). +export fn remove(path: *u8) i32 = { return syscall1(SYS_UNLINK, path: i64): i32; }; @@ -497,13 +500,13 @@ def RELA_ADDEND: u64 = 16u64; // ---- file slurp -------------------------------------------------------- -fn readall(path: *u8) (*u8, u64) = { +fn slurp(path: *u8) (*u8, u64) = { let fd: i32 = os.open(path, os.O_RDONLY, 0i32); if (fd < 0) { return nil, 0u64; }; let n: i64 = os.filesize(fd); if (n < 0i64) { os.close(fd); return nil, 0u64; }; let buf: *u8 = os.alloc(n: u64): *u8; - let got: i64 = os.readfull(fd, buf, n: u64); + let got: i64 = os.readall(fd, buf, n: u64); os.close(fd); if (got != n) { return nil, 0u64; }; return buf, n: u64; @@ -756,7 +759,7 @@ fn loadarchive(l: *lnk, path: *u8, buf: *u8, len: u64) i32 = { export fn load(l: *lnk, path: *u8) i32 = { let bufp: *u8; let buflen: u64; - bufp, buflen = readall(path); + bufp, buflen = slurp(path); if (bufp == nil) { os.write(2, "w6l: cannot read object\n".ptr, 23u64); return -1; @@ -1033,13 +1036,13 @@ fn dbasename(p: *u8) *u8 = { // ---- file slurp -------------------------------------------------------- -fn readallso(path: *u8) (*u8, u64) = { +fn slurpso(path: *u8) (*u8, u64) = { let fd: i32 = os.open(path, os.O_RDONLY, 0i32); if (fd < 0) { return nil, 0u64; }; let n: i64 = os.filesize(fd); if (n < 0i64) { os.close(fd); return nil, 0u64; }; let buf: *u8 = os.alloc(n: u64): *u8; - let got: i64 = os.readfull(fd, buf, n: u64); + let got: i64 = os.readall(fd, buf, n: u64); os.close(fd); if (got != n) { return nil, 0u64; }; return buf, n: u64; @@ -1075,7 +1078,7 @@ fn vdnameat(buf: *u8, verdefoff: u64, verdefsize: u64, export fn loadso(l: *lnk, path: *u8) i32 = { let buf: *u8; let blen: u64; - buf, blen = readallso(path); + buf, blen = slurpso(path); if (buf == nil) { os.write(2, "w6l: cannot read .so\n".ptr, 20u64); return -1; @@ -2146,7 +2149,7 @@ export fn emitdynelf(l: *lnk, fd: i32, base: u64, entry: u64) i32 = { dbcopy(filebuf, gotpltoff, gotpltbuf, gotpltsz); dbcopy(filebuf, dynamicoff, dynamicbuf, dynamicsz); - let wrote: i64 = os.writefull(fd, filebuf, fileend); + let wrote: i64 = os.writeall(fd, filebuf, fileend); if (wrote != fileend: i64) { return 1; }; @@ -2246,10 +2249,10 @@ export fn emitelf(l: *lnk, fd: i32, base: u64, entry: u64) i32 = { wru64(hdr, 112u64, TEXT_OFF); // p_align // Write [0..0x1000) then .text. - let n1: i64 = os.writefull(fd, hdr, TEXT_OFF); + let n1: i64 = os.writeall(fd, hdr, TEXT_OFF); if (n1 != TEXT_OFF: i64) { return -1; }; if (l.textlen > 0u64) { - let n2: i64 = os.writefull(fd, l.text, l.textlen); + let n2: i64 = os.writeall(fd, l.text, l.textlen); if (n2 != l.textlen: i64) { return -1; }; }; return 0; diff --git a/selfhost/cmd/w6l/obj.ww b/selfhost/cmd/w6l/obj.ww index 197d4d28..b60007ad 100644 --- a/selfhost/cmd/w6l/obj.ww +++ b/selfhost/cmd/w6l/obj.ww @@ -75,13 +75,13 @@ def RELA_ADDEND: u64 = 16u64; // ---- file slurp -------------------------------------------------------- -fn readall(path: *u8) (*u8, u64) = { +fn slurp(path: *u8) (*u8, u64) = { let fd: i32 = os.open(path, os.O_RDONLY, 0i32); if (fd < 0) { return nil, 0u64; }; let n: i64 = os.filesize(fd); if (n < 0i64) { os.close(fd); return nil, 0u64; }; let buf: *u8 = os.alloc(n: u64): *u8; - let got: i64 = os.readfull(fd, buf, n: u64); + let got: i64 = os.readall(fd, buf, n: u64); os.close(fd); if (got != n) { return nil, 0u64; }; return buf, n: u64; @@ -334,7 +334,7 @@ fn loadarchive(l: *lnk, path: *u8, buf: *u8, len: u64) i32 = { export fn load(l: *lnk, path: *u8) i32 = { let bufp: *u8; let buflen: u64; - bufp, buflen = readall(path); + bufp, buflen = slurp(path); if (bufp == nil) { os.write(2, "w6l: cannot read object\n".ptr, 23u64); return -1; diff --git a/selfhost/cmd/w6l/out.ww b/selfhost/cmd/w6l/out.ww index 16c61750..cf19d5f8 100644 --- a/selfhost/cmd/w6l/out.ww +++ b/selfhost/cmd/w6l/out.ww @@ -90,10 +90,10 @@ export fn emitelf(l: *lnk, fd: i32, base: u64, entry: u64) i32 = { wru64(hdr, 112u64, TEXT_OFF); // p_align // Write [0..0x1000) then .text. - let n1: i64 = os.writefull(fd, hdr, TEXT_OFF); + let n1: i64 = os.writeall(fd, hdr, TEXT_OFF); if (n1 != TEXT_OFF: i64) { return -1; }; if (l.textlen > 0u64) { - let n2: i64 = os.writefull(fd, l.text, l.textlen); + let n2: i64 = os.writeall(fd, l.text, l.textlen); if (n2 != l.textlen: i64) { return -1; }; }; return 0; diff --git a/selfhost/cmd/wcc/cgen.ww b/selfhost/cmd/wcc/cgen.ww index 03ea7ecf..1e527d5e 100644 --- a/selfhost/cmd/wcc/cgen.ww +++ b/selfhost/cmd/wcc/cgen.ww @@ -279,13 +279,13 @@ fn emitline(s: str) void = { os.write(1, s.ptr, s.len: u64); }; fn emitint(v: i64) void = { let buf: [32]u8; - let n: i32 = strconv.i64toa(buf[0:32], v); + let n: i32 = strconv.i64tos(buf[0:32], v); os.write(1, buf.ptr, n: u64); }; fn emituint(v: u64) void = { let buf: [32]u8; - let n: i32 = strconv.u64toa(buf[0:32], v); + let n: i32 = strconv.u64tos(buf[0:32], v); os.write(1, buf.ptr, n: u64); }; @@ -323,7 +323,7 @@ fn mklabel(c: *cgen, base: str) str = { i += 1; j += 1; }; buf[i] = 95u8; i += 1; // '_' - let n: i32 = strconv.i64toa(buf[i:128], c.labelseq: i64); + let n: i32 = strconv.i64tos(buf[i:128], c.labelseq: i64); c.labelseq += 1; let total: i32 = i + n; let p: *u8 = amalloc(c.a, (total: u64) + 1u64): *u8; @@ -362,7 +362,7 @@ fn internstrlit(c: *cgen, bytes: str) str = { // New label "_S_". let buf: [32]u8; buf[0] = 95u8; buf[1] = 83u8; buf[2] = 95u8; // "_S_" - let n: i32 = strconv.i64toa(buf[3:32], c.strlitseq: i64); + let n: i32 = strconv.i64tos(buf[3:32], c.strlitseq: i64); c.strlitseq += 1; let total: i32 = 3 + n; let p: *u8 = amalloc(c.a, (total: u64) + 1u64): *u8; diff --git a/selfhost/cmd/wcc/err.ww b/selfhost/cmd/wcc/err.ww index 3e14642b..5205d8c7 100644 --- a/selfhost/cmd/wcc/err.ww +++ b/selfhost/cmd/wcc/err.ww @@ -16,7 +16,7 @@ let nerrors: i32 = 0; let nwarnings: i32 = 0; export fn fatal(msg: str) void = { - fmt.errln(msg); + fmt.errorln(msg); os.exit(1); }; diff --git a/selfhost/cmd/ww/main.combined.ww b/selfhost/cmd/ww/main.combined.ww index 1a7f1e4c..9d482fc6 100644 --- a/selfhost/cmd/ww/main.combined.ww +++ b/selfhost/cmd/ww/main.combined.ww @@ -120,9 +120,11 @@ export fn filesize(fd: i32) i64 = { return end; }; -// readfull — keep reading until `n` bytes have arrived or the fd +// readall — keep reading until `n` bytes have arrived or the fd // closes early. Returns bytes read (0..=n) or -1 on read error. -export fn readfull(fd: i32, buf: *u8, n: u64) i64 = { +// Hare name (io::readall); the buffer is caller-supplied, matching +// the Plan 9 subset convention. +export fn readall(fd: i32, buf: *u8, n: u64) i64 = { let got: u64 = 0u64; for (got < n) { let r: i64 = read(fd, buf + got, n - got); @@ -133,9 +135,10 @@ export fn readfull(fd: i32, buf: *u8, n: u64) i64 = { return got: i64; }; -// writefull — keep writing until `n` bytes have been accepted or the -// fd refuses progress. Returns bytes written or -1. -export fn writefull(fd: i32, buf: *u8, n: u64) i64 = { +// writeall — keep writing until `n` bytes have been accepted or the +// fd refuses progress. Returns bytes written or -1. Hare name +// (io::writeall). +export fn writeall(fd: i32, buf: *u8, n: u64) i64 = { let sent: u64 = 0u64; for (sent < n) { let r: i64 = write(fd, buf + sent, n - sent); @@ -154,8 +157,8 @@ export fn access(path: *u8, mode: i32) i32 = { return syscall2(SYS_ACCESS, path: i64, mode: i64): i32; }; -// unlink(2). -export fn unlink(path: *u8) i32 = { +// remove — unlink(2). Hare name; the underlying syscall is unlink(2). +export fn remove(path: *u8) i32 = { return syscall1(SYS_UNLINK, path: i64): i32; }; @@ -599,14 +602,14 @@ fn locateimport(a: *arena, dirs: *u8, name: *u8, namelen: u64) *u8 = { // ---- file slurp ------------------------------------------------------- -fn readall(pathcs: *u8) (*u8, u64) = { +fn slurp(pathcs: *u8) (*u8, u64) = { let fd: i32 = os.open(pathcs, os.O_RDONLY, 0i32); if (fd < 0) { return nil, 0u64; }; let n: i64 = os.filesize(fd); if (n < 0i64) { os.close(fd); return nil, 0u64; }; let nu: u64 = n: u64; let buf: *u8 = os.alloc(nu + 1u64): *u8; - let got: i64 = os.readfull(fd, buf, nu); + let got: i64 = os.readall(fd, buf, nu); os.close(fd); if (got != n) { return nil, 0u64; }; buf[nu] = 0u8; @@ -707,7 +710,7 @@ fn expand(c: *expctx, pathcs: *u8) void = { let bufp: *u8; let blen: u64; - bufp, blen = readall(pathcs); + bufp, blen = slurp(pathcs); if (bufp == nil) { os.write(2, "ww: cannot read source\n".ptr, 23u64); return; @@ -743,12 +746,12 @@ fn expand(c: *expctx, pathcs: *u8) void = { let mn: u64; mp, mn = modulename(pathcs, plen); if (mn > 0u64) { - os.writefull(c.out, "// MODULE: ".ptr, 11u64); - os.writefull(c.out, mp, mn); - os.writefull(c.out, "\n".ptr, 1u64); + os.writeall(c.out, "// MODULE: ".ptr, 11u64); + os.writeall(c.out, mp, mn); + os.writeall(c.out, "\n".ptr, 1u64); }; - os.writefull(c.out, bufp, blen); - os.writefull(c.out, "\n".ptr, 1u64); + os.writeall(c.out, bufp, blen); + os.writeall(c.out, "\n".ptr, 1u64); }; // ---- Build pipeline --------------------------------------------------- @@ -1342,7 +1345,7 @@ fn dorun(selfdir: *u8, argv: **u8, argc: i32, start: i32) i32 = { lf.libs = libs; lf.nlibs = nlibs; if (buildone(selfdir, resolved, tmp, incs, &lf) != 0) { - os.unlink(tmp); + os.remove(tmp); return 1; }; @@ -1359,7 +1362,7 @@ fn dorun(selfdir: *u8, argv: **u8, argc: i32, start: i32) i32 = { }; execargv[nextra + 1] = nil; let rc: i32 = procrun(tmp, execargv); - os.unlink(tmp); + os.remove(tmp); return rc; }; @@ -1374,14 +1377,14 @@ fn runsingletest(selfdir: *u8, src: *u8) i32 = { let tmp: *u8 = os.alloc(PATH_MAX): *u8; makeruntmp(tmp); if (buildone(selfdir, src, tmp, "\0".ptr, nil) != 0) { - os.unlink(tmp); + os.remove(tmp); return 1; }; let execargv: **u8 = os.alloc(16u64): **u8; execargv[0] = tmp; execargv[1] = nil; let rc: i32 = procrun(tmp, execargv); - os.unlink(tmp); + os.remove(tmp); return rc; }; @@ -1446,7 +1449,7 @@ fn rundirtests(selfdir: *u8, dir: *u8) i32 = { os.write(2, "\n".ptr, 1u64); }; }; - os.unlink(tmp); + os.remove(tmp); }; off += reclen; }; diff --git a/selfhost/cmd/ww/main.ww b/selfhost/cmd/ww/main.ww index 50834982..60088bf3 100644 --- a/selfhost/cmd/ww/main.ww +++ b/selfhost/cmd/ww/main.ww @@ -280,14 +280,14 @@ fn locateimport(a: *arena, dirs: *u8, name: *u8, namelen: u64) *u8 = { // ---- file slurp ------------------------------------------------------- -fn readall(pathcs: *u8) (*u8, u64) = { +fn slurp(pathcs: *u8) (*u8, u64) = { let fd: i32 = os.open(pathcs, os.O_RDONLY, 0i32); if (fd < 0) { return nil, 0u64; }; let n: i64 = os.filesize(fd); if (n < 0i64) { os.close(fd); return nil, 0u64; }; let nu: u64 = n: u64; let buf: *u8 = os.alloc(nu + 1u64): *u8; - let got: i64 = os.readfull(fd, buf, nu); + let got: i64 = os.readall(fd, buf, nu); os.close(fd); if (got != n) { return nil, 0u64; }; buf[nu] = 0u8; @@ -388,7 +388,7 @@ fn expand(c: *expctx, pathcs: *u8) void = { let bufp: *u8; let blen: u64; - bufp, blen = readall(pathcs); + bufp, blen = slurp(pathcs); if (bufp == nil) { os.write(2, "ww: cannot read source\n".ptr, 23u64); return; @@ -424,12 +424,12 @@ fn expand(c: *expctx, pathcs: *u8) void = { let mn: u64; mp, mn = modulename(pathcs, plen); if (mn > 0u64) { - os.writefull(c.out, "// MODULE: ".ptr, 11u64); - os.writefull(c.out, mp, mn); - os.writefull(c.out, "\n".ptr, 1u64); + os.writeall(c.out, "// MODULE: ".ptr, 11u64); + os.writeall(c.out, mp, mn); + os.writeall(c.out, "\n".ptr, 1u64); }; - os.writefull(c.out, bufp, blen); - os.writefull(c.out, "\n".ptr, 1u64); + os.writeall(c.out, bufp, blen); + os.writeall(c.out, "\n".ptr, 1u64); }; // ---- Build pipeline --------------------------------------------------- @@ -1023,7 +1023,7 @@ fn dorun(selfdir: *u8, argv: **u8, argc: i32, start: i32) i32 = { lf.libs = libs; lf.nlibs = nlibs; if (buildone(selfdir, resolved, tmp, incs, &lf) != 0) { - os.unlink(tmp); + os.remove(tmp); return 1; }; @@ -1040,7 +1040,7 @@ fn dorun(selfdir: *u8, argv: **u8, argc: i32, start: i32) i32 = { }; execargv[nextra + 1] = nil; let rc: i32 = procrun(tmp, execargv); - os.unlink(tmp); + os.remove(tmp); return rc; }; @@ -1055,14 +1055,14 @@ fn runsingletest(selfdir: *u8, src: *u8) i32 = { let tmp: *u8 = os.alloc(PATH_MAX): *u8; makeruntmp(tmp); if (buildone(selfdir, src, tmp, "\0".ptr, nil) != 0) { - os.unlink(tmp); + os.remove(tmp); return 1; }; let execargv: **u8 = os.alloc(16u64): **u8; execargv[0] = tmp; execargv[1] = nil; let rc: i32 = procrun(tmp, execargv); - os.unlink(tmp); + os.remove(tmp); return rc; }; @@ -1127,7 +1127,7 @@ fn rundirtests(selfdir: *u8, dir: *u8) i32 = { os.write(2, "\n".ptr, 1u64); }; }; - os.unlink(tmp); + os.remove(tmp); }; off += reclen; }; diff --git a/selfhost/cmd/wwdump/main.combined.ww b/selfhost/cmd/wwdump/main.combined.ww index 85f30b10..9e6fe859 100644 --- a/selfhost/cmd/wwdump/main.combined.ww +++ b/selfhost/cmd/wwdump/main.combined.ww @@ -120,9 +120,11 @@ export fn filesize(fd: i32) i64 = { return end; }; -// readfull — keep reading until `n` bytes have arrived or the fd +// readall — keep reading until `n` bytes have arrived or the fd // closes early. Returns bytes read (0..=n) or -1 on read error. -export fn readfull(fd: i32, buf: *u8, n: u64) i64 = { +// Hare name (io::readall); the buffer is caller-supplied, matching +// the Plan 9 subset convention. +export fn readall(fd: i32, buf: *u8, n: u64) i64 = { let got: u64 = 0u64; for (got < n) { let r: i64 = read(fd, buf + got, n - got); @@ -133,9 +135,10 @@ export fn readfull(fd: i32, buf: *u8, n: u64) i64 = { return got: i64; }; -// writefull — keep writing until `n` bytes have been accepted or the -// fd refuses progress. Returns bytes written or -1. -export fn writefull(fd: i32, buf: *u8, n: u64) i64 = { +// writeall — keep writing until `n` bytes have been accepted or the +// fd refuses progress. Returns bytes written or -1. Hare name +// (io::writeall). +export fn writeall(fd: i32, buf: *u8, n: u64) i64 = { let sent: u64 = 0u64; for (sent < n) { let r: i64 = write(fd, buf + sent, n - sent); @@ -154,8 +157,8 @@ export fn access(path: *u8, mode: i32) i32 = { return syscall2(SYS_ACCESS, path: i64, mode: i64): i32; }; -// unlink(2). -export fn unlink(path: *u8) i32 = { +// remove — unlink(2). Hare name; the underlying syscall is unlink(2). +export fn remove(path: *u8) i32 = { return syscall1(SYS_UNLINK, path: i64): i32; }; @@ -321,14 +324,16 @@ export fn freearena(a: *arena) void = { // buffer. Two error idioms ship side by side: // - Plan 9 style (atoi64): tuple `(value, ok)`. Pre-dates the // tagged-union work; kept for callers that already use it. -// - Hare style (parse64/parseu64): `(value | str)`. The error +// - Hare style (stoi64/stou64): `(value | str)`. The error // variant carries a short, allocation-free message describing // why the parse failed. Prefer this for new code. -// u64toa — write `v` in decimal into `buf` and return the byte count. -// Unsigned-only so callers don't have to think about wraparound when -// printing a u64 that happens to have the high bit set. -export fn u64toa(buf: []u8, v: u64) i32 = { +// 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; let i: i32 = 0; let n: u64 = v; @@ -350,7 +355,7 @@ export fn u64toa(buf: []u8, v: u64) i32 = { return out; }; -export fn i64toa(buf: []u8, v: i64) i32 = { +export fn i64tos(buf: []u8, v: i64) i32 = { let neg: bool = false; let n: i64 = v; if (n < 0) { @@ -400,11 +405,12 @@ export fn atoi64(s: str) (i64, bool) = { return v, true; }; -// parse64 — Hare-style fallible signed decimal parser. The value -// variant is i64; the error variant is a short str describing the -// reason. No locale, no whitespace, no underscores: a leading '-' is -// the only non-digit accepted, and only at position 0. -export fn parse64(s: str) (i64 | str) = { +// stoi64 — Hare-style fallible signed decimal parser. The value +// variant is i64; the error variant is a short str (subset of Hare's +// (invalid | overflow) tagged-union). 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 | str) = { if (s.len == 0) { return "parse: empty"; }; let i: i32 = 0; let neg: bool = false; @@ -422,8 +428,9 @@ export fn parse64(s: str) (i64 | str) = { return v; }; -// parseu64 — fallible unsigned decimal parser. No leading sign. -export fn parseu64(s: str) (u64 | str) = { +// stou64 — fallible unsigned decimal parser. No leading sign. Mirrors +// Hare's strconv::stou64. +export fn stou64(s: str) (u64 | str) = { if (s.len == 0) { return "parse: empty"; }; let v: u64 = 0u64; let i: i32 = 0; @@ -812,10 +819,10 @@ export fn tokprint(fd: i32, t: *tok) void = { }; fputcbyte(fd, 58u8); // ':' let buf: [32]u8; - let n: i32 = strconv.i64toa(buf[0:32], t.line: i64); + let n: i32 = strconv.i64tos(buf[0:32], t.line: i64); os.write(fd, buf.ptr, n: u64); fputcbyte(fd, 58u8); - n = strconv.i64toa(buf[0:32], t.col: i64); + n = strconv.i64tos(buf[0:32], t.col: i64); os.write(fd, buf.ptr, n: u64); fputcbyte(fd, 32u8); // ' ' fputsstr(fd, tokname(t.kind)); @@ -831,11 +838,11 @@ export fn tokprint(fd: i32, t: *tok) void = { fputq(fd, ttext.ptr, ttext.len); } else { if (t.kind == TK_INT) { fputcbyte(fd, 32u8); - n = strconv.u64toa(buf[0:32], t.uval); + n = strconv.u64tos(buf[0:32], t.uval); os.write(fd, buf.ptr, n: u64); } else { if (t.kind == TK_RUNE) { fputcbyte(fd, 32u8); - n = strconv.u64toa(buf[0:32], t.uval); + n = strconv.u64tos(buf[0:32], t.uval); os.write(fd, buf.ptr, n: u64); };};};};}; // TK_FLOAT is intentionally not handled here — %g formatting @@ -846,96 +853,96 @@ export fn tokprint(fd: i32, t: *tok) void = { }; // MODULE: ascii -// ascii — byte-class predicates and case folding for the ASCII range. -// Matches Hare's ascii::isdigit family. Bytes outside 0..127 always -// answer `false`. The lexer hot path uses these inline; they are -// expected to inline to a couple of compares. +// ascii — rune-class predicates and case folding for the ASCII range. +// Matches Hare's ascii::isdigit family (rune-taking signature). Runes +// outside 0..127 always answer `false`. The lexer hot path uses these +// inline; they are expected to inline to a couple of compares. -export fn isdigit(c: u8) bool = { - if (c < 48u8) { return false; }; - if (c > 57u8) { return false; }; +export fn isdigit(c: rune) bool = { + if (c < 48) { return false; }; + if (c > 57) { return false; }; return true; }; -export fn isupper(c: u8) bool = { - if (c < 65u8) { return false; }; - if (c > 90u8) { return false; }; +export fn isupper(c: rune) bool = { + if (c < 65) { return false; }; + if (c > 90) { return false; }; return true; }; -export fn islower(c: u8) bool = { - if (c < 97u8) { return false; }; - if (c > 122u8) { return false; }; +export fn islower(c: rune) bool = { + if (c < 97) { return false; }; + if (c > 122) { return false; }; return true; }; -export fn isalpha(c: u8) bool = { +export fn isalpha(c: rune) bool = { if (isupper(c)) { return true; }; return islower(c); }; -export fn isalnum(c: u8) bool = { +export fn isalnum(c: rune) bool = { if (isalpha(c)) { return true; }; return isdigit(c); }; // isspace — the C/Hare set: space, tab, NL, VT, FF, CR. -export fn isspace(c: u8) bool = { - if (c == 32u8) { return true; }; // ' ' - if (c == 9u8) { return true; }; // '\t' - if (c == 10u8) { return true; }; // '\n' - if (c == 11u8) { return true; }; // '\v' - if (c == 12u8) { return true; }; // '\f' - if (c == 13u8) { return true; }; // '\r' +export fn isspace(c: rune) bool = { + if (c == 32) { return true; }; // ' ' + if (c == 9) { return true; }; // '\t' + if (c == 10) { return true; }; // '\n' + if (c == 11) { return true; }; // '\v' + if (c == 12) { return true; }; // '\f' + if (c == 13) { return true; }; // '\r' return false; }; -export fn ishex(c: u8) bool = { +export fn isxdigit(c: rune) bool = { if (isdigit(c)) { return true; }; - if (c >= 65u8) { - if (c <= 70u8) { return true; }; // 'A'..'F' + if (c >= 65) { + if (c <= 70) { return true; }; // 'A'..'F' }; - if (c >= 97u8) { - if (c <= 102u8) { return true; }; // 'a'..'f' + if (c >= 97) { + if (c <= 102) { return true; }; // 'a'..'f' }; return false; }; // digitval — value of `c` as a hex/decimal digit, or -1 if not one. // Useful when scanning numeric literals. -export fn digitval(c: u8) i32 = { - if (isdigit(c)) { return (c - 48u8): i32; }; - if (c >= 65u8) { - if (c <= 70u8) { return ((c - 65u8) + 10u8): i32; }; +export fn digitval(c: rune) i32 = { + if (isdigit(c)) { return (c - 48): i32; }; + if (c >= 65) { + if (c <= 70) { return ((c - 65) + 10): i32; }; }; - if (c >= 97u8) { - if (c <= 102u8) { return ((c - 97u8) + 10u8): i32; }; + if (c >= 97) { + if (c <= 102) { return ((c - 97) + 10): i32; }; }; return -1; }; // isidstart / isidpart — identifier classes used by the lexer. // Alpha or '_' starts; alnum or '_' continues. -export fn isidstart(c: u8) bool = { +export fn isidstart(c: rune) bool = { if (isalpha(c)) { return true; }; - if (c == 95u8) { return true; }; // '_' + if (c == 95) { return true; }; // '_' return false; }; -export fn isidpart(c: u8) bool = { +export fn isidpart(c: rune) bool = { if (isalnum(c)) { return true; }; - if (c == 95u8) { return true; }; + if (c == 95) { return true; }; return false; }; // tolower / toupper — fold ASCII case. Non-letters pass through. -export fn tolower(c: u8) u8 = { - if (isupper(c)) { return c + 32u8; }; +export fn tolower(c: rune) rune = { + if (isupper(c)) { return c + 32; }; return c; }; -export fn toupper(c: u8) u8 = { - if (islower(c)) { return c - 32u8; }; +export fn toupper(c: rune) rune = { + if (islower(c)) { return c - 32; }; return c; }; @@ -1172,18 +1179,18 @@ fn escape(l: *lex, out: *i32) bool = { let lo: i32 = lget(l); if (hi < 0) { return false; }; if (lo < 0) { return false; }; - if (!ascii.ishex(hi: u8)) { + if (!ascii.isxdigit(hi: rune)) { let cp: pos; curpos(l, &cp); errat(l, &cp, "bad \\x escape"); return false; }; - if (!ascii.ishex(lo: u8)) { + if (!ascii.isxdigit(lo: rune)) { let cp: pos; curpos(l, &cp); errat(l, &cp, "bad \\x escape"); return false; }; - let h: i32 = ascii.digitval(hi: u8); - let lv: i32 = ascii.digitval(lo: u8); + let h: i32 = ascii.digitval(hi: rune); + let lv: i32 = ascii.digitval(lo: rune); *out = (h << 4) | lv; return true; }; @@ -1197,7 +1204,7 @@ fn scandecimalrun(l: *lex) void = { for (true) { let c: i32 = lpeek(l, 0u64); if (c < 0) { break; }; - if (!ascii.isdigit(c: u8)) { + if (!ascii.isdigit(c: rune)) { if (c != 95) { break; }; }; lget(l); @@ -1208,7 +1215,7 @@ fn scanhexrun(l: *lex) void = { for (true) { let c: i32 = lpeek(l, 0u64); if (c < 0) { break; }; - if (!ascii.ishex(c: u8)) { + if (!ascii.isxdigit(c: rune)) { if (c != 95) { break; }; }; lget(l); @@ -1247,7 +1254,7 @@ fn scanexp(l: *lex) void = { for (true) { let c: i32 = lpeek(l, 0u64); if (c < 0) { break; }; - if (!ascii.isdigit(c: u8)) { break; }; + if (!ascii.isdigit(c: rune)) { break; }; lget(l); }; }; @@ -1332,12 +1339,12 @@ fn lexnum(l: *lex, start: *pos, out: *tok) void = { let pc: i32 = lpeek(l, 0u64); if (pc >= 0) { - if (ascii.isidstart(pc: u8)) { + if (ascii.isidstart(pc: rune)) { let sb: u64 = l.lpos; for (true) { let cc: i32 = lpeek(l, 0u64); if (cc < 0) { break; }; - if (!ascii.isidpart(cc: u8)) { break; }; + if (!ascii.isidpart(cc: rune)) { break; }; lget(l); }; let sl: u64 = l.lpos - sb; @@ -1381,7 +1388,7 @@ fn lexident(l: *lex, start: *pos, out: *tok) void = { for (true) { let c: i32 = lpeek(l, 0u64); if (c < 0) { break; }; - if (!ascii.isidpart(c: u8)) { break; }; + if (!ascii.isidpart(c: rune)) { break; }; lget(l); }; let n: u64 = l.lpos - begin; @@ -1530,8 +1537,8 @@ export fn lexnext(l: *lex, out: *tok) void = { let c: i32 = lpeek(l, 0u64); if (c >= 0) { - if (ascii.isidstart(c: u8)) { lexident(l, &start, out); return; }; - if (ascii.isdigit(c: u8)) { lexnum(l, &start, out); return; }; + if (ascii.isidstart(c: rune)) { lexident(l, &start, out); return; }; + if (ascii.isdigit(c: rune)) { lexnum(l, &start, out); return; }; }; if (c == 34) { lget(l); lexstr(l, &start, out); return; }; @@ -1895,12 +1902,12 @@ fn pr(fd: i32, n: *node, d: i32) void = { if (n.kind == N_INTLIT) { putc1(fd, 32u8); let buf: [32]u8; - let m: i32 = strconv.u64toa(buf[0:32], n.uval); + let m: i32 = strconv.u64tos(buf[0:32], n.uval); os.write(fd, buf.ptr, m: u64); } else { if (n.kind == N_RUNELIT) { putc1(fd, 32u8); let buf: [32]u8; - let m: i32 = strconv.u64toa(buf[0:32], n.uval); + let m: i32 = strconv.u64tos(buf[0:32], n.uval); os.write(fd, buf.ptr, m: u64); } else { if ( n.kind == N_STRLIT || @@ -7408,13 +7415,13 @@ fn emitline(s: str) void = { os.write(1, s.ptr, s.len: u64); }; fn emitint(v: i64) void = { let buf: [32]u8; - let n: i32 = strconv.i64toa(buf[0:32], v); + let n: i32 = strconv.i64tos(buf[0:32], v); os.write(1, buf.ptr, n: u64); }; fn emituint(v: u64) void = { let buf: [32]u8; - let n: i32 = strconv.u64toa(buf[0:32], v); + let n: i32 = strconv.u64tos(buf[0:32], v); os.write(1, buf.ptr, n: u64); }; @@ -7452,7 +7459,7 @@ fn mklabel(c: *cgen, base: str) str = { i += 1; j += 1; }; buf[i] = 95u8; i += 1; // '_' - let n: i32 = strconv.i64toa(buf[i:128], c.labelseq: i64); + let n: i32 = strconv.i64tos(buf[i:128], c.labelseq: i64); c.labelseq += 1; let total: i32 = i + n; let p: *u8 = amalloc(c.a, (total: u64) + 1u64): *u8; @@ -7491,7 +7498,7 @@ fn internstrlit(c: *cgen, bytes: str) str = { // New label "_S_". let buf: [32]u8; buf[0] = 95u8; buf[1] = 83u8; buf[2] = 95u8; // "_S_" - let n: i32 = strconv.i64toa(buf[3:32], c.strlitseq: i64); + let n: i32 = strconv.i64tos(buf[3:32], c.strlitseq: i64); c.strlitseq += 1; let total: i32 = 3 + n; let p: *u8 = amalloc(c.a, (total: u64) + 1u64): *u8; @@ -7979,7 +7986,7 @@ export fn main(argc: i32, argv: **u8) i32 = { let a: *arena = newarena(); let buf: *u8 = amalloc(a, sz: u64): *u8; - let r: i64 = os.readfull(fd, buf, sz: u64); + let r: i64 = os.readall(fd, buf, sz: u64); os.close(fd); if (r != sz) { os.write(2, "wwdump: short read\n".ptr, 19u64); @@ -8018,11 +8025,11 @@ export fn main(argc: i32, argv: **u8) i32 = { os.write(1, argstr(path).ptr, argstrlen(path): u64); os.write(1, ": ".ptr, 2u64); let buf: [32]u8; - let n: i32 = strconv.i64toa(buf[0:32], ck.nresolved: i64); + let n: i32 = strconv.i64tos(buf[0:32], ck.nresolved: i64); os.write(1, buf.ptr, n: u64); os.write(1, "/".ptr, 1u64); let total: i32 = ck.nresolved + ck.nunresolved; - n = strconv.i64toa(buf[0:32], total: i64); + n = strconv.i64tos(buf[0:32], total: i64); os.write(1, buf.ptr, n: u64); os.write(1, " resolved\n".ptr, 10u64); if (ck.nunresolved > 0) { return 1; }; diff --git a/selfhost/cmd/wwdump/main.ww b/selfhost/cmd/wwdump/main.ww index 6701b4ba..b9ef1bdf 100644 --- a/selfhost/cmd/wwdump/main.ww +++ b/selfhost/cmd/wwdump/main.ww @@ -95,7 +95,7 @@ export fn main(argc: i32, argv: **u8) i32 = { let a: *arena = newarena(); let buf: *u8 = amalloc(a, sz: u64): *u8; - let r: i64 = os.readfull(fd, buf, sz: u64); + let r: i64 = os.readall(fd, buf, sz: u64); os.close(fd); if (r != sz) { os.write(2, "wwdump: short read\n".ptr, 19u64); @@ -134,11 +134,11 @@ export fn main(argc: i32, argv: **u8) i32 = { os.write(1, argstr(path).ptr, argstrlen(path): u64); os.write(1, ": ".ptr, 2u64); let buf: [32]u8; - let n: i32 = strconv.i64toa(buf[0:32], ck.nresolved: i64); + let n: i32 = strconv.i64tos(buf[0:32], ck.nresolved: i64); os.write(1, buf.ptr, n: u64); os.write(1, "/".ptr, 1u64); let total: i32 = ck.nresolved + ck.nunresolved; - n = strconv.i64toa(buf[0:32], total: i64); + n = strconv.i64tos(buf[0:32], total: i64); os.write(1, buf.ptr, n: u64); os.write(1, " resolved\n".ptr, 10u64); if (ck.nunresolved > 0) { return 1; }; diff --git a/selfhost/test/smoke.combined.ww b/selfhost/test/smoke.combined.ww index 0092a6de..b906a1f3 100644 --- a/selfhost/test/smoke.combined.ww +++ b/selfhost/test/smoke.combined.ww @@ -120,9 +120,11 @@ export fn filesize(fd: i32) i64 = { return end; }; -// readfull — keep reading until `n` bytes have arrived or the fd +// readall — keep reading until `n` bytes have arrived or the fd // closes early. Returns bytes read (0..=n) or -1 on read error. -export fn readfull(fd: i32, buf: *u8, n: u64) i64 = { +// Hare name (io::readall); the buffer is caller-supplied, matching +// the Plan 9 subset convention. +export fn readall(fd: i32, buf: *u8, n: u64) i64 = { let got: u64 = 0u64; for (got < n) { let r: i64 = read(fd, buf + got, n - got); @@ -133,9 +135,10 @@ export fn readfull(fd: i32, buf: *u8, n: u64) i64 = { return got: i64; }; -// writefull — keep writing until `n` bytes have been accepted or the -// fd refuses progress. Returns bytes written or -1. -export fn writefull(fd: i32, buf: *u8, n: u64) i64 = { +// writeall — keep writing until `n` bytes have been accepted or the +// fd refuses progress. Returns bytes written or -1. Hare name +// (io::writeall). +export fn writeall(fd: i32, buf: *u8, n: u64) i64 = { let sent: u64 = 0u64; for (sent < n) { let r: i64 = write(fd, buf + sent, n - sent); @@ -154,8 +157,8 @@ export fn access(path: *u8, mode: i32) i32 = { return syscall2(SYS_ACCESS, path: i64, mode: i64): i32; }; -// unlink(2). -export fn unlink(path: *u8) i32 = { +// remove — unlink(2). Hare name; the underlying syscall is unlink(2). +export fn remove(path: *u8) i32 = { return syscall1(SYS_UNLINK, path: i64): i32; }; @@ -213,14 +216,16 @@ export fn getdents64(fd: i32, buf: *u8, n: u64) i64 = { // buffer. Two error idioms ship side by side: // - Plan 9 style (atoi64): tuple `(value, ok)`. Pre-dates the // tagged-union work; kept for callers that already use it. -// - Hare style (parse64/parseu64): `(value | str)`. The error +// - Hare style (stoi64/stou64): `(value | str)`. The error // variant carries a short, allocation-free message describing // why the parse failed. Prefer this for new code. -// u64toa — write `v` in decimal into `buf` and return the byte count. -// Unsigned-only so callers don't have to think about wraparound when -// printing a u64 that happens to have the high bit set. -export fn u64toa(buf: []u8, v: u64) i32 = { +// 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; let i: i32 = 0; let n: u64 = v; @@ -242,7 +247,7 @@ export fn u64toa(buf: []u8, v: u64) i32 = { return out; }; -export fn i64toa(buf: []u8, v: i64) i32 = { +export fn i64tos(buf: []u8, v: i64) i32 = { let neg: bool = false; let n: i64 = v; if (n < 0) { @@ -292,11 +297,12 @@ export fn atoi64(s: str) (i64, bool) = { return v, true; }; -// parse64 — Hare-style fallible signed decimal parser. The value -// variant is i64; the error variant is a short str describing the -// reason. No locale, no whitespace, no underscores: a leading '-' is -// the only non-digit accepted, and only at position 0. -export fn parse64(s: str) (i64 | str) = { +// stoi64 — Hare-style fallible signed decimal parser. The value +// variant is i64; the error variant is a short str (subset of Hare's +// (invalid | overflow) tagged-union). 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 | str) = { if (s.len == 0) { return "parse: empty"; }; let i: i32 = 0; let neg: bool = false; @@ -314,8 +320,9 @@ export fn parse64(s: str) (i64 | str) = { return v; }; -// parseu64 — fallible unsigned decimal parser. No leading sign. -export fn parseu64(s: str) (u64 | str) = { +// stou64 — fallible unsigned decimal parser. No leading sign. Mirrors +// Hare's strconv::stou64. +export fn stou64(s: str) (u64 | str) = { if (s.len == 0) { return "parse: empty"; }; let v: u64 = 0u64; let i: i32 = 0; @@ -330,96 +337,96 @@ export fn parseu64(s: str) (u64 | str) = { }; // MODULE: ascii -// ascii — byte-class predicates and case folding for the ASCII range. -// Matches Hare's ascii::isdigit family. Bytes outside 0..127 always -// answer `false`. The lexer hot path uses these inline; they are -// expected to inline to a couple of compares. +// ascii — rune-class predicates and case folding for the ASCII range. +// Matches Hare's ascii::isdigit family (rune-taking signature). Runes +// outside 0..127 always answer `false`. The lexer hot path uses these +// inline; they are expected to inline to a couple of compares. -export fn isdigit(c: u8) bool = { - if (c < 48u8) { return false; }; - if (c > 57u8) { return false; }; +export fn isdigit(c: rune) bool = { + if (c < 48) { return false; }; + if (c > 57) { return false; }; return true; }; -export fn isupper(c: u8) bool = { - if (c < 65u8) { return false; }; - if (c > 90u8) { return false; }; +export fn isupper(c: rune) bool = { + if (c < 65) { return false; }; + if (c > 90) { return false; }; return true; }; -export fn islower(c: u8) bool = { - if (c < 97u8) { return false; }; - if (c > 122u8) { return false; }; +export fn islower(c: rune) bool = { + if (c < 97) { return false; }; + if (c > 122) { return false; }; return true; }; -export fn isalpha(c: u8) bool = { +export fn isalpha(c: rune) bool = { if (isupper(c)) { return true; }; return islower(c); }; -export fn isalnum(c: u8) bool = { +export fn isalnum(c: rune) bool = { if (isalpha(c)) { return true; }; return isdigit(c); }; // isspace — the C/Hare set: space, tab, NL, VT, FF, CR. -export fn isspace(c: u8) bool = { - if (c == 32u8) { return true; }; // ' ' - if (c == 9u8) { return true; }; // '\t' - if (c == 10u8) { return true; }; // '\n' - if (c == 11u8) { return true; }; // '\v' - if (c == 12u8) { return true; }; // '\f' - if (c == 13u8) { return true; }; // '\r' +export fn isspace(c: rune) bool = { + if (c == 32) { return true; }; // ' ' + if (c == 9) { return true; }; // '\t' + if (c == 10) { return true; }; // '\n' + if (c == 11) { return true; }; // '\v' + if (c == 12) { return true; }; // '\f' + if (c == 13) { return true; }; // '\r' return false; }; -export fn ishex(c: u8) bool = { +export fn isxdigit(c: rune) bool = { if (isdigit(c)) { return true; }; - if (c >= 65u8) { - if (c <= 70u8) { return true; }; // 'A'..'F' + if (c >= 65) { + if (c <= 70) { return true; }; // 'A'..'F' }; - if (c >= 97u8) { - if (c <= 102u8) { return true; }; // 'a'..'f' + if (c >= 97) { + if (c <= 102) { return true; }; // 'a'..'f' }; return false; }; // digitval — value of `c` as a hex/decimal digit, or -1 if not one. // Useful when scanning numeric literals. -export fn digitval(c: u8) i32 = { - if (isdigit(c)) { return (c - 48u8): i32; }; - if (c >= 65u8) { - if (c <= 70u8) { return ((c - 65u8) + 10u8): i32; }; +export fn digitval(c: rune) i32 = { + if (isdigit(c)) { return (c - 48): i32; }; + if (c >= 65) { + if (c <= 70) { return ((c - 65) + 10): i32; }; }; - if (c >= 97u8) { - if (c <= 102u8) { return ((c - 97u8) + 10u8): i32; }; + if (c >= 97) { + if (c <= 102) { return ((c - 97) + 10): i32; }; }; return -1; }; // isidstart / isidpart — identifier classes used by the lexer. // Alpha or '_' starts; alnum or '_' continues. -export fn isidstart(c: u8) bool = { +export fn isidstart(c: rune) bool = { if (isalpha(c)) { return true; }; - if (c == 95u8) { return true; }; // '_' + if (c == 95) { return true; }; // '_' return false; }; -export fn isidpart(c: u8) bool = { +export fn isidpart(c: rune) bool = { if (isalnum(c)) { return true; }; - if (c == 95u8) { return true; }; + if (c == 95) { return true; }; return false; }; // tolower / toupper — fold ASCII case. Non-letters pass through. -export fn tolower(c: u8) u8 = { - if (isupper(c)) { return c + 32u8; }; +export fn tolower(c: rune) rune = { + if (isupper(c)) { return c + 32; }; return c; }; -export fn toupper(c: u8) u8 = { - if (islower(c)) { return c - 32u8; }; +export fn toupper(c: rune) rune = { + if (islower(c)) { return c - 32; }; return c; }; @@ -557,19 +564,19 @@ export fn main() i32 = { // Probe 5 — strconv round-trip via the real stdlib. let outbuf: [32]u8; - let nb: i32 = strconv.i64toa(outbuf[0:32], 4242i64); + let nb: i32 = strconv.i64tos(outbuf[0:32], 4242i64); if (nb != 4) { return 11; }; if (outbuf[0] != 52u8) { return 12; }; // '4' if (outbuf[3] != 50u8) { return 13; }; // '2' - // Probe 6 — ascii classifications. - if (!ascii.isdigit(53u8)) { return 14; }; // '5' - if (ascii.isdigit(65u8)) { return 15; }; // 'A' is not a digit - if (!ascii.isalpha(122u8)) { return 16; }; // 'z' - if (!ascii.isidstart(95u8)) { return 17; }; // '_' - if (!ascii.isidpart(48u8)) { return 18; }; // '0' is part - if (ascii.digitval(70u8) != 15) { return 19; }; // 'F' = 15 - if (ascii.tolower(65u8) != 97u8) { return 20; }; // 'A' -> 'a' + // Probe 6 — ascii classifications (rune-taking, Hare-shaped). + if (!ascii.isdigit(53)) { return 14; }; // '5' + if (ascii.isdigit(65)) { return 15; }; // 'A' is not a digit + if (!ascii.isalpha(122)) { return 16; }; // 'z' + if (!ascii.isidstart(95)) { return 17; }; // '_' + if (!ascii.isidpart(48)) { return 18; }; // '0' is part + if (ascii.digitval(70) != 15) { return 19; }; // 'F' = 15 + if (ascii.tolower(65) != 97) { return 20; }; // 'A' -> 'a' // Probe 7 — file open/read via the new os APIs. /proc/self/cmdline // always exists on Linux, no write side, and is non-empty. @@ -581,7 +588,7 @@ export fn main() i32 = { case let e: str => return 21; }; let rbuf: [128]u8; - let n: i64 = os.readfull(fd, rbuf.ptr, 128u64); + let n: i64 = os.readall(fd, rbuf.ptr, 128u64); os.close(fd); if (n <= 0i64) { return 22; }; diff --git a/selfhost/test/smoke.ww b/selfhost/test/smoke.ww index 7ab7d8eb..2f28a638 100644 --- a/selfhost/test/smoke.ww +++ b/selfhost/test/smoke.ww @@ -131,19 +131,19 @@ export fn main() i32 = { // Probe 5 — strconv round-trip via the real stdlib. let outbuf: [32]u8; - let nb: i32 = strconv.i64toa(outbuf[0:32], 4242i64); + let nb: i32 = strconv.i64tos(outbuf[0:32], 4242i64); if (nb != 4) { return 11; }; if (outbuf[0] != 52u8) { return 12; }; // '4' if (outbuf[3] != 50u8) { return 13; }; // '2' - // Probe 6 — ascii classifications. - if (!ascii.isdigit(53u8)) { return 14; }; // '5' - if (ascii.isdigit(65u8)) { return 15; }; // 'A' is not a digit - if (!ascii.isalpha(122u8)) { return 16; }; // 'z' - if (!ascii.isidstart(95u8)) { return 17; }; // '_' - if (!ascii.isidpart(48u8)) { return 18; }; // '0' is part - if (ascii.digitval(70u8) != 15) { return 19; }; // 'F' = 15 - if (ascii.tolower(65u8) != 97u8) { return 20; }; // 'A' -> 'a' + // Probe 6 — ascii classifications (rune-taking, Hare-shaped). + if (!ascii.isdigit(53)) { return 14; }; // '5' + if (ascii.isdigit(65)) { return 15; }; // 'A' is not a digit + if (!ascii.isalpha(122)) { return 16; }; // 'z' + if (!ascii.isidstart(95)) { return 17; }; // '_' + if (!ascii.isidpart(48)) { return 18; }; // '0' is part + if (ascii.digitval(70) != 15) { return 19; }; // 'F' = 15 + if (ascii.tolower(65) != 97) { return 20; }; // 'A' -> 'a' // Probe 7 — file open/read via the new os APIs. /proc/self/cmdline // always exists on Linux, no write side, and is non-empty. @@ -155,7 +155,7 @@ export fn main() i32 = { case let e: str => return 21; }; let rbuf: [128]u8; - let n: i64 = os.readfull(fd, rbuf.ptr, 128u64); + let n: i64 = os.readall(fd, rbuf.ptr, 128u64); os.close(fd); if (n <= 0i64) { return 22; }; diff --git a/test/wcc/700_e2e.c b/test/wcc/700_e2e.c index 81de89bb..2070c502 100644 --- a/test/wcc/700_e2e.c +++ b/test/wcc/700_e2e.c @@ -146,7 +146,7 @@ static const struct row rows[] = { "fn main() i32 = {\n" " let buf: [32]u8;\n" " let s: []u8 = buf[0:32];\n" - " let n: i32 = strconv.i64toa(s, 12345);\n" + " let n: i32 = strconv.i64tos(s, 12345);\n" " os.write(1, buf.ptr, n: u64);\n" " os.write(1, \"\\n\".ptr, 1u64);\n" " return n;\n" @@ -703,14 +703,14 @@ static const struct row rows[] = { " };\n" " return acc;\n" "};", 13 }, /* 1 byte written to fd 1, plus len(\"write failed\")=12 */ - /* strconv.parse64: fallible signed decimal. Two successful + /* strconv.stoi64: fallible signed decimal. Two successful * parses contribute their values; one bad parse contributes * the error message length (20 = len(\"parse: invalid digit\")). */ { "use strconv;\n" "fn main() i32 = {\n" - " let r1: (i64 | str) = strconv.parse64(\"42\");\n" - " let r2: (i64 | str) = strconv.parse64(\"-7\");\n" - " let r3: (i64 | str) = strconv.parse64(\"abc\");\n" + " let r1: (i64 | str) = strconv.stoi64(\"42\");\n" + " let r2: (i64 | str) = strconv.stoi64(\"-7\");\n" + " let r3: (i64 | str) = strconv.stoi64(\"abc\");\n" " let acc: i32 = 0;\n" " match (r1) {\n" " case let v: i64 => acc += v: i32;\n" @@ -726,12 +726,12 @@ static const struct row rows[] = { " };\n" " return acc;\n" "};", 55 }, /* 42 + (-7) + 20 */ - /* strconv.parseu64: success path 123, error path captures + /* strconv.stou64: success path 123, error path captures * len(\"parse: invalid digit\") = 20 for the leading-sign reject. */ { "use strconv;\n" "fn main() i32 = {\n" - " let r1: (u64 | str) = strconv.parseu64(\"123\");\n" - " let r2: (u64 | str) = strconv.parseu64(\"-1\");\n" + " let r1: (u64 | str) = strconv.stou64(\"123\");\n" + " let r2: (u64 | str) = strconv.stou64(\"-1\");\n" " let acc: i32 = 0;\n" " match (r1) {\n" " case let v: u64 => acc += v: i32;\n" @@ -743,17 +743,17 @@ static const struct row rows[] = { " };\n" " return acc;\n" "};", 143 }, /* 123 + 20 */ - /* strings.indexbyte (Plan 9 -1) and strings.index (substring). */ + /* strings.byteindex (Plan 9 -1) and strings.index (substring). */ { "use strings;\n" "fn main() i32 = {\n" " let s: str = \"hello, world\";\n" - " let i1: i32 = strings.indexbyte(s, 44u8);\n" - " let i2: i32 = strings.indexbyte(s, 122u8);\n" + " let i1: i32 = strings.byteindex(s, 44u8);\n" + " let i2: i32 = strings.byteindex(s, 122u8);\n" " let i3: i32 = strings.index(s, \"world\");\n" " let i4: i32 = strings.index(s, \"nope\");\n" " return i1 + i2 + i3 + i4;\n" "};", 10 }, /* 5 + (-1) + 7 + (-1) */ - /* bytes.indexsub: substring search over []u8. */ + /* bytes.index: substring search over []u8. */ { "use bytes;\n" "fn main() i32 = {\n" " let buf: [12]u8;\n" @@ -762,15 +762,15 @@ static const struct row rows[] = { " buf[8] = 111u8; buf[9] = 114u8; buf[10] = 108u8; buf[11] = 100u8;\n" " let needle: [3]u8;\n" " needle[0] = 119u8; needle[1] = 111u8; needle[2] = 114u8;\n" - " return bytes.indexsub(buf[0:12], needle[0:3]);\n" + " return bytes.index(buf[0:12], needle[0:3]);\n" "};", 7 }, /* errors.equal — sentinel comparison through a (T | error) union. * Sets up two errors, dispatches each, and confirms the matching * sentinel detection. */ { "use errors;\n" "fn parse(n: i64) (i64 | errors.error) = {\n" - " if (n < 0) { return errors.eEOF; };\n" - " if (n == 0) { return errors.eShortRead; };\n" + " if (n < 0) { return errors.eof; };\n" + " if (n == 0) { return errors.underread; };\n" " return n;\n" "};\n" "fn main() i32 = {\n" @@ -780,18 +780,18 @@ static const struct row rows[] = { " match (r1) {\n" " case let v: i64 => acc += -100;\n" " case let e: errors.error =>\n" - " if (errors.equal(e, errors.eEOF)) { acc += 1; }\n" + " if (errors.equal(e, errors.eof)) { acc += 1; }\n" " else { acc += -100; };\n" " };\n" " match (r2) {\n" " case let v: i64 => acc += -100;\n" " case let e: errors.error =>\n" - " if (errors.equal(e, errors.eShortRead)) { acc += 10; }\n" + " if (errors.equal(e, errors.underread)) { acc += 10; }\n" " else { acc += -100; };\n" " };\n" " return acc;\n" "};", 11 }, - /* bufio.takeline: drain successive '\\n'-terminated lines from a + /* bufio.readline: drain successive '\\n'-terminated lines from a * pre-filled buffer, then a trailing fragment that returns the * `linerr` variant carrying \"no newline\" (10 chars). */ { "use bufio;\n" @@ -803,17 +803,17 @@ static const struct row rows[] = { " let b: bufio.buf;\n" " b.s = nil; b.data = raw.ptr; b.cap = 11; b.r = 0; b.w = 11;\n" " let acc: i32 = 0;\n" - " let l1: (str | bufio.linerr) = bufio.takeline(&b);\n" + " let l1: (str | bufio.linerr) = bufio.readline(&b);\n" " match (l1) {\n" " case let s: str => acc += s.len: i32;\n" " case let e: bufio.linerr => acc += -100;\n" " };\n" - " let l2: (str | bufio.linerr) = bufio.takeline(&b);\n" + " let l2: (str | bufio.linerr) = bufio.readline(&b);\n" " match (l2) {\n" " case let s: str => acc += s.len: i32;\n" " case let e: bufio.linerr => acc += -100;\n" " };\n" - " let l3: (str | bufio.linerr) = bufio.takeline(&b);\n" + " let l3: (str | bufio.linerr) = bufio.readline(&b);\n" " match (l3) {\n" " case let s: str => acc += -100;\n" " case let e: bufio.linerr => acc += e.len: i32;\n"