lib: rename stdlib surface to Hare names; add endian/math
Sweeping rename so the lib/ surface mirrors Hare's stdlib spellings. - ascii: rune-taking predicates; ishex -> isxdigit - bufio: rinit -> init; take1/takeline -> readbyte/readline - bytes: indexsub -> index - encoding/utf8: runelen -> runesz - errors: eEOF/eShortRead/... -> eof/underread/... - fmt: errln -> errorln; println/fprintln return i64 - os: readfull/writefull -> readall/writeall; unlink -> remove - path: isabs -> abs; drop lastindex (now strings.rbyteindex) - strconv: u64toa/i64toa -> u64tos/i64tos; parse64/parseu64 -> stoi64/stou64 - strings: drop len/isempty; equal -> compare; indexbyte -> byteindex; +rbyteindex - types: drop numeric helpers (moved to math) - new lib/endian (htonu16/ntohu16), lib/math (absi32/absi64) - net: drop htons (use endian.htonu16) Callers in selfhost/, lib/ww/, cmd/w6c/cgen.c, and test/wcc/700_e2e.c updated to match.
This commit is contained in:
@@ -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; };
|
||||
|
||||
|
||||
@@ -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; };
|
||||
|
||||
|
||||
Reference in New Issue
Block a user