wcc/ww: .wwi export-data producer + check_exported_type (#22 M2)

New `w6c -I <out.wwi>` flag (both stages) writes a re-parseable
ww-prototype rendering of a package's EXPORTED surface. M2 dead-code:
nothing consumes .wwi yet (combined.ww stays the live path); the flag is
off on every existing invocation, so the 990-997 byte-id gates and all
prior tests are unperturbed.

The unparse walks the AST type-expr subtree (N_T* nodes), not the
tinfo Type* (which collapses nominal pkg.Name identity). Deterministic
output: package line, byte-sorted imports, byte-sorted decls — a pure
function of the exported API. cmd/w6c/wwi.c + selfhost/cmd/wcc/wwi.ww
emit byte-identical .wwi (new cross-stage byte-id substrate, rule 10).

check_exported_type (drew) rides the producer entry, flag-gated: an
exported signature naming a non-exported nominal is loud-rejected before
any byte is written, identically on both stages. Ports harec
check.c:4092-4168, recursing the type-AST and gating on the resolved
SK_TYPE sym's decl export flag (Sym.exported is vestigial in both
stages; the predeclared synthetic `nomem` decl carries no source
position and is treated as a builtin leaf — cstage parity).

Two wwstage checker AST-mutations are normalized to cstage's pristine
view for byte-id: the N_TPARAM tuple-element wrapper (unwrapped) and the
variadic `T...`→`[]T` param desugar (peeled).

Gate 989_m2wwi_run: ascii/strings/getopt each produce a .wwi that
re-parses (wwdump -a) and is cs==ww byte-identical; a private-type-leak
fixture is rejected identically by both stages (non-vacuous check).
This commit is contained in:
2026-06-15 20:16:21 +09:00
parent e9c11cb5ae
commit e8d3d89fef
8 changed files with 1965 additions and 4 deletions

330
test/wcc/989_m2wwi_run.c Normal file
View File

@@ -0,0 +1,330 @@
/*
* 989_m2wwi_run — M2 `.wwi` export-data producer gate (task #22 arc).
* Both stages (rule-10): the new `w6c -I <out.wwi>` producer is DEAD on
* the live path (combined.ww stays the compile substrate; nothing reads
* `.wwi`), so this is purely a new cross-stage byte-id substrate.
*
* POSITIVE gate — for each package (ascii/strings/getopt, drew2-audited
* leak-free), drive the target as the PRIMARY module of a driver-combined
* unit, then:
* 1. w6c -I cs.wwi and w6c_ww -I ww.wwi both succeed,
* 2. cs.wwi == ww.wwi byte-for-byte (the new substrate's determinism),
* 3. cs.wwi RE-PARSES under the existing parser (wwdump -a exits 0) —
* proves the producer emits genuinely re-parseable ww prototype
* source, de-risking the M3 consumer.
* getopt is the recursion stressor (enum→struct→struct, ptr-to-nominal,
* tagged-union return, qualified pkg.Name). A synth fixture then covers the
* decl-kinds + type-nodes no lib package reaches: `def` (const-expr unparser
* + fold parity), `let` global, `[N]T` array, `fn(..)R` fn-ptr, `!T`, tuple,
* storage-less enum.
*
* NEGATIVE gate — a fixture whose exported fn names a PRIVATE (non-
* exported) nominal must be LOUD-REJECTED by check_exported_type,
* identically on BOTH stages (same non-zero exit, byte-identical
* diagnostic). Without it a vacuous no-op check would pass the positive
* gate silently.
*
* Compile/produce/cmp only — no driver run, everything under a getpid-
* keyed /tmp dir (no source-tree writes), so this is phase-1 parallel-
* safe. 9xx is full; shares the 989 prefix per the 989_lib_byteid
* precedent (the `short` name keys the binary).
*/
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <unistd.h>
#include <sys/wait.h>
#include <sys/stat.h>
static int
runwait(const char *cmd)
{
int rc = system(cmd);
if (rc == -1) return -1;
if (WIFEXITED(rc)) return WEXITSTATUS(rc);
return 1;
}
static const char *
absbin(void)
{
const char *b = getenv("BIN");
if (!b) b = "out/bin";
if (b[0] == '/') return b;
static char buf[2048];
char cwd[1024];
if (getcwd(cwd, sizeof cwd) == NULL) return NULL;
snprintf(buf, sizeof buf, "%s/%s", cwd, b);
return buf;
}
static int
slurp(const char *path, char **outbuf, size_t *outlen)
{
FILE *f = fopen(path, "rb");
if (!f) return -1;
fseek(f, 0, SEEK_END);
long n = ftell(f);
fseek(f, 0, SEEK_SET);
if (n < 0) { fclose(f); return -1; }
char *b = malloc((size_t)n + 1);
if (!b) { fclose(f); return -1; }
if (fread(b, 1, (size_t)n, f) != (size_t)n) { free(b); fclose(f); return -1; }
b[n] = '\0';
fclose(f);
*outbuf = b;
*outlen = (size_t)n;
return 0;
}
static int
files_eq(const char *a, const char *b)
{
char *ba = NULL, *bb = NULL;
size_t na = 0, nb = 0;
if (slurp(a, &ba, &na) < 0 || slurp(b, &bb, &nb) < 0) {
free(ba); free(bb);
return -1;
}
int eq = (na == nb && memcmp(ba, bb, na) == 0);
free(ba); free(bb);
return eq ? 0 : 1;
}
/* one positive package: build the target as primary, produce the `.wwi`
* on both stages, require byte-id + re-parse. Returns 0 on pass. */
static int
positive(const char *bin, const char *cwd, const char *pkg, int idx)
{
char td[64], cmd[4096], src[1024];
int rc = -1;
snprintf(td, sizeof td, "/tmp/wwm2_%d_%d", getpid(), idx);
snprintf(cmd, sizeof cmd, "rm -rf %s", td);
runwait(cmd);
mkdir(td, 0755);
snprintf(src, sizeof src, "%s/lib/%s/%s.ww", cwd, pkg, pkg);
snprintf(cmd, sizeof cmd, "cp %s %s/%s.ww", src, td, pkg);
if (runwait(cmd) != 0) {
fprintf(stderr, "m2wwi FAIL: %s — cp\n", pkg);
goto out;
}
/* Drive the package file as the build target → its decls are the
* PRIMARY module (imported==0) of the resolved unit; the producer's
* primary filter then yields exactly this package's interface. The
* library has no main, so the build's link step fails — the
* `.combined.ww` is written before codegen (cf 989_lib_byteid), and
* only it matters here, so the exit code is deliberately ignored. */
snprintf(cmd, sizeof cmd,
"cd %s && timeout 180 %s/ww build -I %s/lib %s.ww >/dev/null 2>&1",
td, bin, cwd, pkg);
runwait(cmd);
char comb[1100], cs[1100], ws[1100];
snprintf(comb, sizeof comb, "%s/%s.combined.ww", td, pkg);
snprintf(cs, sizeof cs, "%s/cs.wwi", td);
snprintf(ws, sizeof ws, "%s/ww.wwi", td);
if (access(comb, 0) != 0) {
fprintf(stderr, "m2wwi FAIL: %s — no resolved unit\n", pkg);
goto out;
}
snprintf(cmd, sizeof cmd,
"timeout 180 %s/w6c -I %s %s >/dev/null 2>&1", bin, cs, comb);
if (runwait(cmd) != 0) {
fprintf(stderr, "m2wwi FAIL: %s — w6c -I rejected\n", pkg);
goto out;
}
snprintf(cmd, sizeof cmd,
"timeout 180 %s/w6c_ww -I %s %s >/dev/null 2>&1", bin, ws, comb);
if (runwait(cmd) != 0) {
fprintf(stderr, "m2wwi FAIL: %s — w6c_ww -I rejected\n", pkg);
goto out;
}
if (files_eq(cs, ws) != 0) {
fprintf(stderr, "m2wwi FAIL: %s — cs.wwi != ww.wwi (byte-id "
"broken on the .wwi substrate)\n", pkg);
goto out;
}
/* re-parse: the emitted `.wwi` must be valid ww prototype source. */
snprintf(cmd, sizeof cmd,
"timeout 180 %s/wwdump -a %s >/dev/null 2>&1", bin, cs);
if (runwait(cmd) != 0) {
fprintf(stderr, "m2wwi FAIL: %s — emitted .wwi does not "
"re-parse (wwdump -a)\n", pkg);
goto out;
}
rc = 0;
out:
snprintf(cmd, sizeof cmd, "rm -rf %s", td);
runwait(cmd);
return rc;
}
/* negative gate: an exported fn naming a private nominal must be rejected
* identically (exit + diagnostic) by BOTH stages. Returns 0 on pass. */
static const char *leak_src =
"package leaktest;\n"
"type secret = struct { x: i32 };\n"
"export fn leaks(s: secret) i32 = { return s.x; };\n"
"export fn clean(a: i32) i32 = { return a; };\n";
static int
negative(const char *bin)
{
char td[64], cmd[4096], p[512];
int rc = -1;
snprintf(td, sizeof td, "/tmp/wwm2neg_%d", getpid());
snprintf(cmd, sizeof cmd, "rm -rf %s", td);
runwait(cmd);
mkdir(td, 0755);
snprintf(p, sizeof p, "%s/leak.ww", td);
FILE *f = fopen(p, "wb");
if (!f) goto out;
fputs(leak_src, f);
fclose(f);
char cse[1100], wwe[1100];
snprintf(cse, sizeof cse, "%s/cs.err", td);
snprintf(wwe, sizeof wwe, "%s/ww.err", td);
/* cd into td so both stages report the bare `leak.ww:L:C:` prefix. */
snprintf(cmd, sizeof cmd,
"cd %s && timeout 180 %s/w6c -I out.wwi leak.ww >/dev/null 2>cs.err",
td, bin);
int crc = runwait(cmd);
snprintf(cmd, sizeof cmd,
"cd %s && timeout 180 %s/w6c_ww -I out.wwi leak.ww >/dev/null 2>ww.err",
td, bin);
int wrc = runwait(cmd);
if (crc == 0 || wrc == 0) {
fprintf(stderr, "m2wwi FAIL: negative — check_exported_type is "
"vacuous (cstage exit=%d wwstage exit=%d; both must reject "
"the private-type leak)\n", crc, wrc);
goto out;
}
if (files_eq(cse, wwe) != 0) {
fprintf(stderr, "m2wwi FAIL: negative — reject diagnostics "
"differ across stages (rule-10)\n");
goto out;
}
rc = 0;
out:
snprintf(cmd, sizeof cmd, "rm -rf %s", td);
runwait(cmd);
return rc;
}
/* synth gate: a single self-contained package exercising the decl-kinds +
* type-AST nodes the lib packages above do NOT cover — `export def` (the
* const-expr unparser, incl a folded `10+2*3` and a unary `-7`), `export
* let` (global), `[N]T` array (decl + array-dim ident-ref), an `fn(...) R`
* fn-pointer type, an `!T` error type, a `(a, b)` tuple return, and a
* storage-less `enum {...}`. Asserts both stages produce a byte-identical,
* re-parseable `.wwi` (the const-expr fold + tokname spelling must agree
* cross-stage — paths the lib gate never reaches). `chan T` is omitted: the
* wwstage parser does not yet accept it in a return position (pre-existing,
* unrelated to M2), so the producer's N_TCHAN arm is unreachable there.
* Returns 0 on pass. */
static const char *synth_src =
"package synth;\n"
"export type color = enum { RED, GREEN = 5, BLUE };\n"
"export def LIMIT: i32 = 10 + 2 * 3;\n"
"export def NAME: str = \"hi\\n\";\n"
"export def FLAG: bool = true;\n"
"export def NEG: i32 = -7;\n"
"export let counter: i32;\n"
"export let grid: [4]i32;\n"
"export fn apply(f: fn(x: i32) i32, n: i32) i32;\n"
"export fn risky() !i32;\n"
"export fn matrix() [LIMIT]u8;\n"
"export fn pair() (i32, i32);\n"
"export fn opt(p: *color, b: []u8) (i32 | void);\n";
static int
synth(const char *bin)
{
char td[64], cmd[4096], p[512];
int rc = -1;
snprintf(td, sizeof td, "/tmp/wwm2syn_%d", getpid());
snprintf(cmd, sizeof cmd, "rm -rf %s", td);
runwait(cmd);
mkdir(td, 0755);
snprintf(p, sizeof p, "%s/synth.ww", td);
FILE *f = fopen(p, "wb");
if (!f) goto out;
fputs(synth_src, f);
fclose(f);
char cs[1100], ws[1100];
snprintf(cs, sizeof cs, "%s/cs.wwi", td);
snprintf(ws, sizeof ws, "%s/ww.wwi", td);
/* self-contained package (no imports) → it is its own primary unit,
* so w6c -I runs directly on the source (no driver-combined step). */
snprintf(cmd, sizeof cmd,
"timeout 180 %s/w6c -I %s %s/synth.ww >/dev/null 2>&1", bin, cs, td);
if (runwait(cmd) != 0) {
fprintf(stderr, "m2wwi FAIL: synth — w6c -I rejected\n");
goto out;
}
snprintf(cmd, sizeof cmd,
"timeout 180 %s/w6c_ww -I %s %s/synth.ww >/dev/null 2>&1", bin, ws, td);
if (runwait(cmd) != 0) {
fprintf(stderr, "m2wwi FAIL: synth — w6c_ww -I rejected\n");
goto out;
}
if (files_eq(cs, ws) != 0) {
fprintf(stderr, "m2wwi FAIL: synth — cs.wwi != ww.wwi (const-expr "
"/ decl-kind unparse diverges across stages)\n");
goto out;
}
snprintf(cmd, sizeof cmd,
"timeout 180 %s/wwdump -a %s >/dev/null 2>&1", bin, cs);
if (runwait(cmd) != 0) {
fprintf(stderr, "m2wwi FAIL: synth — emitted .wwi does not "
"re-parse (wwdump -a)\n");
goto out;
}
rc = 0;
out:
snprintf(cmd, sizeof cmd, "rm -rf %s", td);
runwait(cmd);
return rc;
}
int
main(void)
{
const char *bin = absbin();
if (!bin) return 1;
char cwd[1024];
if (getcwd(cwd, sizeof cwd) == NULL) return 1;
const char *pkgs[] = { "ascii", "strings", "getopt", NULL };
int fail = 0, npos = 0;
for (int i = 0; pkgs[i]; i++) {
if (positive(bin, cwd, pkgs[i], i) != 0) fail++;
else npos++;
}
if (synth(bin) != 0) fail++;
else npos++;
if (negative(bin) != 0) fail++;
if (fail) {
fprintf(stderr, "m2wwi: %d check(s) failed\n", fail);
return 1;
}
printf("m2wwi: %d package(s) byte-identical + re-parse; negative "
"gate rejects (both stages)\n", npos);
return 0;
}