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

lib: banner purge + WHY-only comment sweep (rule 8)

Browse Source 
Every // ---- section banner dies (132 -> 0): names carry the WHAT.
Narration deleted (filename restatements, run-with lines, what-the-
next-line-does); every ref/hare cite, task cite, divergence, ABI/
layout contract, and ownership qualifier kept (borrowed-view lines
restored where the sweep over-cut). Comment-only proven: all 442
walk-workdir .s and 32 import-probe .s byte-identical before/after;
libbyteid 56-roster all-ID.
This commit is contained in:
Hojun-Cho
2026-08-08 21:10:18 +09:00
parent 659e859f34
commit aadc6618f0
90 changed files with 259 additions and 919 deletions
Download Patch File Download Diff File Expand all files Collapse all files

71
lib/bytes/bytes.ww
View File

@@ -1,6 +1,4 @@
// bytes — slice operations over []u8. Mirrors Hare's bytes module
// (ref/hare/bytes/) for the in-tree subset: search/equality/prefix
// helpers used by lib/encoding, lib/bufio, lib/memio.
// Hare port of the in-tree subset; see ref/hare/bytes/.
//
// Documented divergences from Hare:
// - index_slice / rindex_slice use naive O(n·m); Hare specialises
@@ -29,17 +27,16 @@ import types;
// (utf8.ww:36). Plain `void` (not `!void`): continuation signal.
export type done = void;
// tokenizer — cursor over an input slice. Layout mirrors
// ref/hare/bytes/tokenize.ha:6-10. `p` is the cached peek-position;
// I64_MAX (forward) / I64_MIN (reverse) are the unprimed sentinels.
// p < 0 also identifies a reverse-direction iterator.
// Layout mirrors ref/hare/bytes/tokenize.ha:6-10. `p` is the cached
// peek-position; I64_MAX (forward) / I64_MIN (reverse) are the
// unprimed sentinels. p < 0 also identifies a reverse-direction
// iterator.
export type tokenizer = struct {
in: []u8,
delim: []u8,
p: i64,
};
// equal — true iff `a` and `b` have the same length and contents.
// ref/hare/bytes/equal.ha:9.
export fn equal(a: []u8, b: []u8) bool = {
if (a.len != b.len) { return false; };
@@ -51,8 +48,6 @@ export fn equal(a: []u8, b: []u8) bool = {
return true;
};
// index — first offset of `needle` in `s`. u8 needle scans for the
// byte; []u8 needle scans for the substring. void if absent.
// ref/hare/bytes/index.ha:6.
export fn index(s: []u8, needle: (u8 | []u8)) (i32 | void) = {
match (needle) {
@@ -85,9 +80,8 @@ export fn index(s: []u8, needle: (u8 | []u8)) (i32 | void) = {
return;
};
// rindex — last offset of `needle` in `s`. Empty []u8 needle returns
// s.len (ref/hare/bytes/index.ha:103 — Hare's loop yields r-0 at i=0).
// ref/hare/bytes/index.ha:86.
// Empty []u8 needle returns s.len (ref/hare/bytes/index.ha:103 —
// Hare's loop yields r-0 at i=0). ref/hare/bytes/index.ha:86.
export fn rindex(s: []u8, needle: (u8 | []u8)) (i32 | void) = {
match (needle) {
case let c: u8 => {
@@ -118,7 +112,6 @@ export fn rindex(s: []u8, needle: (u8 | []u8)) (i32 | void) = {
return;
};
// contains — true iff any of `needles` (byte or sub-slice) appears in `s`.
// ref/hare/bytes/contains.ha:6.
export fn contains(s: []u8, needles: (u8 | []u8)...) bool = {
let i: i32 = 0;
@@ -142,8 +135,7 @@ export fn contains(s: []u8, needles: (u8 | []u8)...) bool = {
return false;
};
// ltrim — borrowed view of `in` with leading bytes in `trim` stripped.
// `trim` must be non-empty. ref/hare/bytes/trim.ha:7.
// ref/hare/bytes/trim.ha:7. Borrowed view of `in` — caller must not free.
export fn ltrim(in: []u8, trim: u8...) []u8 = {
assert(trim.len > 0, "bytes.ltrim called with empty trim set");
let i: i32 = 0;
@@ -155,8 +147,7 @@ export fn ltrim(in: []u8, trim: u8...) []u8 = {
return r;
};
// rtrim — borrowed view of `in` with trailing bytes in `trim` stripped.
// `trim` must be non-empty. ref/hare/bytes/trim.ha:17. Hare's loop uses
// ref/hare/bytes/trim.ha:17. Borrowed view of `in`. Hare's loop uses
// `size` underflow at i==0 to terminate; ww indices are signed i32, so
// the equivalent termination is spelled `i >= 0` explicitly.
export fn rtrim(in: []u8, trim: u8...) []u8 = {
@@ -170,13 +161,11 @@ export fn rtrim(in: []u8, trim: u8...) []u8 = {
return r;
};
// trim — borrowed view of `in` with both ends in `trim` stripped.
// ref/hare/bytes/trim.ha:27.
// ref/hare/bytes/trim.ha:27. Borrowed view of `in`.
export fn trim(in: []u8, trim: u8...) []u8 = {
return ltrim(rtrim(in, trim...), trim...);
};
// hasprefix — true iff `s` starts with `pre`.
// ref/hare/bytes/contains.ha:21.
export fn hasprefix(s: []u8, pre: []u8) bool = {
if (pre.len > s.len) { return false; };
@@ -188,7 +177,6 @@ export fn hasprefix(s: []u8, pre: []u8) bool = {
return true;
};
// hassuffix — true iff `s` ends with `suf`.
// ref/hare/bytes/contains.ha:35.
export fn hassuffix(s: []u8, suf: []u8) bool = {
if (suf.len > s.len) { return false; };
@@ -201,7 +189,7 @@ export fn hassuffix(s: []u8, suf: []u8) bool = {
return true;
};
// reverse — in-place reverse of `s`. ref/hare/bytes/reverse.ha:5.
// ref/hare/bytes/reverse.ha:5.
export fn reverse(s: []u8) void = {
let i: i32 = 0;
let j: i32 = s.len - 1;
@@ -214,7 +202,7 @@ export fn reverse(s: []u8) void = {
};
};
// zero — set every byte of `s` to 0. ref/hare/bytes/zero.ha:5.
// ref/hare/bytes/zero.ha:5.
export fn zero(s: []u8) void = {
let i: i32 = 0;
for (i < s.len) {
@@ -223,8 +211,6 @@ export fn zero(s: []u8) void = {
};
};
// tokenize — iterator yielding tokens from `in` separated by any byte
// in `delim`. Leading / trailing / adjacent delims yield empty tokens.
// `delim` is borrowed; caller keeps it valid for the tokenizer's
// lifetime. ref/hare/bytes/tokenize.ha:22.
export fn tokenize(in: []u8, delim: u8...) tokenizer = {
@@ -242,8 +228,7 @@ export fn tokenize(in: []u8, delim: u8...) tokenizer = {
return t;
};
// rtokenize — reverse-direction tokenize. First nexttoken yields the
// last token, last nexttoken yields the first. ref/hare/bytes/tokenize.ha:40.
// ref/hare/bytes/tokenize.ha:40.
export fn rtokenize(in: []u8, delim: u8...) tokenizer = {
assert(delim.len > 0, "bytes.rtokenize called with empty slice");
assert((in.len: i64) < types.I64_MAX,
@@ -259,9 +244,8 @@ export fn rtokenize(in: []u8, delim: u8...) tokenizer = {
return t;
};
// peektoken — next token without advancing the cursor. Returns done
// once `s.delim` has been zeroed by a prior past-end nexttoken.
// ref/hare/bytes/tokenize.ha:91.
// Returns done once `s.delim` has been zeroed by a prior past-end
// nexttoken. ref/hare/bytes/tokenize.ha:91.
export fn peektoken(s: *tokenizer) ([]u8 | done) = {
if (s.delim.len == 0) {
let d: done; return d;
@@ -338,7 +322,6 @@ export fn peektoken(s: *tokenizer) ([]u8 | done) = {
return r;
};
// nexttoken — current token, then advance past it and the delim.
// Once the input is exhausted, returns done and zeros `s.delim` so
// subsequent peeks short-circuit. ref/hare/bytes/tokenize.ha:59.
export fn nexttoken(s: *tokenizer) ([]u8 | done) = {
@@ -380,15 +363,13 @@ export fn nexttoken(s: *tokenizer) ([]u8 | done) = {
return b;
};
// remainingtokens — the unconsumed portion of `s.in`. Read-only view.
// ref/hare/bytes/tokenize.ha:145.
// ref/hare/bytes/tokenize.ha:145. Read-only borrowed view.
export fn remainingtokens(s: *tokenizer) []u8 = {
return s.in;
};
// splitn — split `in` on any byte in `delim`, returning up to `n`
// tokens via forward iteration. The trailing slot (when more than
// `n - 1` tokens exist) holds the unconsumed remainder.
// The trailing slot (when more than `n - 1` tokens exist) holds the
// unconsumed remainder.
//
// The caller frees the returned slice via
// `os.free(r.ptr: *void, (r.cap: u64) * 24u64)`. Element bytes are
@@ -425,9 +406,8 @@ export fn splitn(in: []u8, delim: []u8, n: i32) [][]u8 = {
return toks;
};
// rsplitn — reverse-direction counterpart to [[splitn]]: tokens are
// collected from the end of `in`. The trailing slot holds the
// unconsumed prefix (everything before the n-th-from-last delim hit).
// The trailing slot holds the unconsumed prefix (everything before
// the n-th-from-last delim hit).
//
// When the input has fewer than n tokens, the `done` short-circuit
// returns toks UN-reversed (in last-token-first order). Mirrors Hare
@@ -486,8 +466,7 @@ export fn rsplitn(in: []u8, delim: []u8, n: i32) [][]u8 = {
return toks;
};
// split — full split of `in` on `delim` (no token cap). Mirrors
// `splitn(in, delim, types::SIZE_MAX)`. ww uses `types.I32_MAX`
// Mirrors `splitn(in, delim, types::SIZE_MAX)`. ww uses `types.I32_MAX`
// because the index type is i32 (lib/CLAUDE.md).
//
// ref/hare/bytes/tokenize.ha:225.
@@ -495,10 +474,9 @@ export fn split(in: []u8, delim: []u8) [][]u8 = {
return splitn(in, delim, types.I32_MAX);
};
// cut — split `in` along the first instance of `delim`, returning the
// portion before and the portion after the delimiter as a borrowed
// tuple. When `delim` is absent, the whole input is the first half and
// the second is empty. ref/hare/bytes/tokenize.ha:392.
// When `delim` is absent, the whole input is the first half and the
// second is empty. Both halves are borrowed views — caller must not
// free them. ref/hare/bytes/tokenize.ha:392.
//
// Delim is spelled (u8 | []u8) to match index/rindex (bytes.ww:57/91);
// the tagged union is an unordered set, so this is the same type as
@@ -525,7 +503,6 @@ export fn cut(in: []u8, delim: (u8 | []u8)) ([]u8, []u8) = {
};
};
// rcut — like [[cut]] but splits along the last instance of `delim`.
// ref/hare/bytes/tokenize.ha:413.
export fn rcut(in: []u8, delim: (u8 | []u8)) ([]u8, []u8) = {
let ln: i32 = match (delim) {

4
lib/bytes/contains_test.ww
View File

@@ -1,6 +1,4 @@
// containstest — exercises bytes.contains/hasprefix/hassuffix. A
// failing row aborts via the assert/abort builtin (task #5 @test
// conversion). Vectors mirror ref/hare/bytes/contains.ha.
// Vectors mirror ref/hare/bytes/contains.ha (task #5 @test conversion).
package bytes_test;

4
lib/bytes/equal_test.ww
View File

@@ -1,6 +1,4 @@
// equaltest — exercises bytes.equal. A failing row aborts via the
// assert/abort builtin (task #5 @test conversion).
// Vectors mirror ref/hare/bytes/equal.ha.
// Vectors mirror ref/hare/bytes/equal.ha (task #5 @test conversion).
package bytes_test;

4
lib/bytes/index_test.ww
View File

@@ -1,6 +1,4 @@
// indextest — exercises bytes.index/rindex, u8 and []u8 arms. A
// failing row aborts via the assert/abort builtin (task #5 @test
// conversion). Vectors mirror ref/hare/bytes/index.ha.
// Vectors mirror ref/hare/bytes/index.ha (task #5 @test conversion).
package bytes_test;

7
lib/bytes/tokenize_test.ww
View File

@@ -1,6 +1,4 @@
// tokenizetest — exercises the bytes tokenize/splitn/cut families.
// A failing row aborts via the assert/abort builtin (task #5 @test
// conversion). Vectors mirror ref/hare/bytes/tokenize.ha.
// Vectors mirror ref/hare/bytes/tokenize.ha (task #5 @test conversion).
package bytes_test;
@@ -11,8 +9,6 @@ import os;
// drives the iterator through an expected-token sequence and asserts
// `equal(p, n)` (peek == next), `equal(n, want)` (next == expected).
// expect_token — table row driver. Advances `t` once, asserts the
// returned token matches `want`. peek invariant: peek must equal next.
fn expect_token(t: *bytes.tokenizer, want: []u8) void = {
match (bytes.peektoken(t)) {
case let p: []u8 => {
@@ -28,7 +24,6 @@ fn expect_token(t: *bytes.tokenizer, want: []u8) void = {
};
};
// expect_done — table-row driver. peek and next must both be done.
fn expect_done(t: *bytes.tokenizer) void = {
match (bytes.peektoken(t)) {
case let p: []u8 => abort();

4
lib/bytes/trim_test.ww
View File

@@ -1,6 +1,4 @@
// trimtest — exercises bytes.ltrim/rtrim/trim. A failing row aborts
// via the assert/abort builtin (task #5 @test conversion).
// Vectors mirror ref/hare/bytes/trim.ha.
// Vectors mirror ref/hare/bytes/trim.ha (task #5 @test conversion).
package bytes_test;