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

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

Browse Source 
Every section banner dies (103 -> 0) across test/lang, the observer
suites, the C carriers, and the five comment-heavy corpus fixtures;
banner provenance (#N cites, carrier numbers, repair-cluster labels)
folded into headers or adjacent WHY comments. Narration deleted; row
provenance, ref cites, divergence pins, and layout contracts kept
(fwd-ref decl-order guards and bootstrap-gate corpus rationale
restored where the sweep over-cut). Comment-only proven: all 3742
wwbuild workdir .s byte-identical before/after; test-commit and
test-byteid (161 lang + 1399 data, 0 pinned-divergent) green.
This commit is contained in:
Hojun-Cho
2026-08-08 21:40:23 +09:00
parent aadc6618f0
commit 83f5956df2
120 changed files with 184 additions and 530 deletions
Download Patch File Download Diff File Expand all files Collapse all files

23
test/wcc/100_lex.c
View File

@@ -1,13 +1,3 @@
/*
* 100_lex — table-driven lexer tests.
*
* Each row is a (src, expected) pair. The expected string is the
* concatenation of token names, space-separated. For literals we
* also encode the value: e.g. INT(42), STR("hi"), IDENT(foo).
*
* EOF is implicit: the harness checks that lexnext returns TK_EOF
* after the last expected token.
*/
#include "ww.h"
#include <string.h>
#include <stdlib.h>
@@ -75,7 +65,6 @@ static const struct row rows[] = {
{ "// comment\n", "" },
{ "/* a /b/ c */", "" },
/* identifiers + keywords */
{ "foo", "IDENT(foo)" },
{ "fn", "fn" },
{ "fn main", "fn IDENT(main)" },
@@ -84,7 +73,6 @@ static const struct row rows[] = {
{ "if else for switch case return import type struct defer break continue proc chan nil true false package",
"if else for switch case return import type struct defer break continue proc chan nil true false package" },
/* numbers */
{ "0", "INT(0)" },
{ "42", "INT(42)" },
{ "1_000_000", "INT(1000000)" },
@@ -95,7 +83,6 @@ static const struct row rows[] = {
{ "3.14", "FLOAT(3.14)" },
{ "1.5e3", "FLOAT(1500)" },
/* strings & runes */
{ "\"hello\"", "STR(hello)" },
{ "\"a\\nb\"", "STR(a\nb)" },
{ "'A'", "RUNE(65)" },
@@ -122,7 +109,6 @@ static const struct row rows[] = {
{ "'\\uD7FF'", "RUNE(55295)" },
{ "'\\uE000'", "RUNE(57344)" },
/* operators & punct */
{ "+ - * / % == != < > <= >= && || !",
"+ - * / % == != < > <= >= && || !" },
{ "= += -= *= /= %= &= |= ^= <<= >>=",
@@ -136,15 +122,14 @@ static const struct row rows[] = {
{ "@symbol(\"malloc\")",
"@ IDENT(symbol) ( STR(malloc) )" },
/* mixed */
{ "fn add(a: i32, b: i32) i32 = { return a + b; };",
"fn IDENT(add) ( IDENT(a) : IDENT(i32) , IDENT(b) : IDENT(i32) ) IDENT(i32) = { return IDENT(a) + IDENT(b) ; } ;" },
};
/* Escape error cases. A bad escape sets l.errs (the string/rune still
* lexes with the offending codepoint zeroed), so detection is via the
* error counter, not the token stream. The verbatim Hare messages live
* at lexunicode in cmd/wcc/lex.c. */
/* A bad escape sets l.errs (the string/rune still lexes with the
* offending codepoint zeroed), so detection is via the error counter,
* not the token stream. The verbatim Hare messages live at lexunicode
* in cmd/wcc/lex.c. */
static int
runerr(const char *src)
{