Change wasm-tests CI to run Node harness instead of Pascal harness

[frostney]

Summary by CodeRabbit

• Tests

  • Migrated WASM integration test harness from Pascal to Node.js, improving compatibility and maintainability across platforms.
  • Removed OS-specific installation steps from CI/CD pipelines.
  • Updated all continuous integration workflows to use the new test runner.

• Documentation

  • Updated test execution documentation and WASM backend guides to reflect new test runner commands.

[coderabbitai]

📝 Walkthrough

Walkthrough

This pull request replaces the Pascal-based WASM integration test harness with a Node.js ES module implementation. The changes include removing FPC installation steps from CI workflows, updating test invocations to use the new Node.js runner, removing the old Pascal script, adding the new Node.js module, and updating documentation references.

Changes

Cohort / File(s)	Summary
CI Workflows .github/workflows/ci.yml, .github/workflows/pr.yml	Removed OS-specific FPC installation steps and updated WASM test invocation from Pascal script (run-wasm-tests.pas) to Node.js module (node ./tests-wasm/run-wasm-tests.mjs).
Documentation AGENTS.md, docs/wasm-backend.md	Updated WASM test harness file references and invocation commands from run-wasm-tests.pas to run-wasm-tests.mjs with node execution syntax.
Test Harness Implementation tests-wasm/run-wasm-tests.mjs	New Node.js ES module that orchestrates WASM tests by resolving ScriptLoader binary, iterating test fixtures, compiling to WASM, executing via host script, parsing expected output, and reporting results with colored formatting and aggregated counts.
Test Harness Removal tests-wasm/run-wasm-tests.pas	Deleted the original Pascal-based test harness script that performed similar WASM test orchestration.

Sequence Diagram(s)

sequenceDiagram
    participant Runner as Test Runner
    participant ScriptLoader as ScriptLoader Binary
    participant TempDir as Temp Directory
    participant Fixture as Test Fixture
    participant Host as Host Script
    participant Node as Node.js Runtime
    
    Runner->>Runner: Resolve ScriptLoader & Host paths
    Runner->>Runner: Create temp working directory
    loop For each fixture
        Runner->>Fixture: Read fixture file
        alt Skip directive detected
            Runner->>Runner: Mark as SKIPPED
        else Fixture valid
            Runner->>ScriptLoader: Invoke --emit=wasm<br/>(fixture → temp wasm)
            alt Compilation succeeds
                ScriptLoader-->>TempDir: WASM file created
                Runner->>Host: Execute node ./host<br/>(with WASM file)
                Host->>Node: Run in Node.js runtime
                alt Execution succeeds
                    Node-->>Host: stdout output
                    Runner->>Fixture: Parse // Expected: lines
                    Runner->>Runner: Compare actual vs expected
                    alt Output matches
                        Runner->>Runner: Mark as PASS
                    else Output mismatch
                        Runner->>Runner: Mark as FAIL
                    end
                else Execution fails
                    Runner->>Runner: Mark as FAIL<br/>(node error/exit code)
                end
            else Compilation fails
                Runner->>Runner: Mark as FAIL<br/>(compile error)
            end
        end
    end
    Runner->>TempDir: Clean up
    Runner->>Runner: Output final summary<br/>(passed/failed/skipped counts)

Loading

Estimated code review effort

🎯 4 (Complex) | ⏱️ ~60 minutes

Possibly related PRs

• First draft of WAM output through Souffle VM #53: Original PR that introduced the Pascal-based WASM test harness, which is now replaced by this Node.js implementation at the code level.

Poem

🐰 A wand of Node replaces Pascal's grace,
WASM tests now run at a nimbler pace,
FPC steps removed, the CI flows clean,
JavaScript harness—the fastest we've seen! ✨
Hoppy refactoring, no bugs in between! 🎉

🚥 Pre-merge checks | ✅ 2 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 0.00% which is insufficient. The required threshold is 80.00%.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (2 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title accurately describes the main change: replacing the Pascal-based WASM test harness with a Node.js-based harness in CI, which aligns with the substantial code modifications across CI configs and test files.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

• 📝 Generate docstrings (stacked PR)
• 📝 Generate docstrings (commit on current branch)

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Post copyable unit tests in a comment
• Commit unit tests in branch codex/wasm-ci-no-fpc

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[github-actions]

Benchmark Results

254 benchmarks

Interpreted: 🟢 2 improved · 🔴 1 regressed · 251 unchanged · avg +0.3%
Bytecode: 🟢 7 improved · 🔴 10 regressed · 237 unchanged · avg -0.3%

arraybuffer.js — Interp: 14 unch. · avg -1.1% · Bytecode: 14 unch. · avg -0.9%

Benchmark	Interpreted	Δ	Bytecode	Δ
create ArrayBuffer(0)	423,795 → 420,124	-0.9%	136,986 → 135,582	-1.0%
create ArrayBuffer(64)	424,937 → 412,727	-2.9%	138,431 → 139,022	+0.4%
create ArrayBuffer(1024)	320,072 → 317,039	-0.9%	126,021 → 124,835	-0.9%
create ArrayBuffer(8192)	132,364 → 133,106	+0.6%	81,864 → 81,874	+0.0%
slice full buffer (64 bytes)	512,603 → 506,004	-1.3%	367,545 → 353,006	-4.0%
slice half buffer (512 of 1024 bytes)	431,856 → 423,856	-1.9%	322,284 → 319,484	-0.9%
slice with negative indices	439,031 → 432,010	-1.6%	345,036 → 338,341	-1.9%
slice empty range	481,590 → 483,411	+0.4%	347,257 → 347,610	+0.1%
byteLength access	1,546,955 → 1,559,080	+0.8%	981,769 → 977,574	-0.4%
Symbol.toStringTag access	1,105,365 → 1,093,576	-1.1%	506,024 → 510,834	+1.0%
ArrayBuffer.isView	701,595 → 701,599	+0.0%	442,389 → 437,674	-1.1%
clone ArrayBuffer(64)	383,800 → 377,872	-1.5%	317,622 → 309,828	-2.5%
clone ArrayBuffer(1024)	288,954 → 283,320	-1.9%	249,319 → 244,874	-1.8%
clone ArrayBuffer inside object	260,093 → 252,024	-3.1%	152,966 → 153,721	+0.5%

arrays.js — Interp: 🟢 1, 18 unch. · avg +1.9% · Bytecode: 🟢 1, 🔴 2, 16 unch. · avg -1.5%

Benchmark	Interpreted	Δ	Bytecode	Δ
Array.from length 100	12,566 → 13,619	🟢 +8.4%	11,492 → 12,109	+5.4%
Array.from 10 elements	210,875 → 224,605	+6.5%	154,255 → 157,383	+2.0%
Array.of 10 elements	295,726 → 295,122	-0.2%	209,957 → 214,504	+2.2%
spread into new array	329,159 → 336,275	+2.2%	526,648 → 562,680	+6.8%
map over 50 elements	25,593 → 27,244	+6.5%	20,116 → 20,642	+2.6%
filter over 50 elements	22,218 → 23,453	+5.6%	19,667 → 15,081	🔴 -23.3%
reduce sum 50 elements	26,110 → 26,515	+1.5%	17,670 → 17,483	-1.1%
forEach over 50 elements	22,177 → 22,458	+1.3%	21,887 → 20,765	-5.1%
find in 50 elements	34,598 → 35,907	+3.8%	25,044 → 25,590	+2.2%
sort 20 elements	12,046 → 12,088	+0.3%	2,827 → 2,891	+2.3%
flat nested array	117,497 → 117,416	-0.1%	277,761 → 270,800	-2.5%
flatMap	73,318 → 72,648	-0.9%	186,127 → 203,626	🟢 +9.4%
map inside map (5x5)	21,166 → 21,170	+0.0%	67,175 → 67,378	+0.3%
filter inside map (5x10)	15,673 → 15,779	+0.7%	12,769 → 12,549	-1.7%
reduce inside map (5x10)	19,268 → 19,068	-1.0%	12,496 → 12,684	+1.5%
forEach inside forEach (5x10)	16,287 → 16,288	+0.0%	14,139 → 13,958	-1.3%
find inside some (10x10)	13,861 → 13,991	+0.9%	9,934 → 9,498	-4.4%
map+filter chain nested (5x20)	5,342 → 5,405	+1.2%	4,149 → 3,886	-6.3%
reduce flatten (10x5)	41,705 → 41,253	-1.1%	5,554 → 4,601	🔴 -17.2%

async-await.js — Interp: 6 unch. · avg +0.3% · Bytecode: 6 unch. · avg +1.5%

Benchmark	Interpreted	Δ	Bytecode	Δ
single await	372,830 → 367,783	-1.4%	260,664 → 272,341	+4.5%
multiple awaits	167,202 → 165,545	-1.0%	112,371 → 117,214	+4.3%
await non-Promise value	824,168 → 835,286	+1.3%	951,530 → 950,981	-0.1%
await with try/catch	370,273 → 366,018	-1.1%	266,158 → 266,901	+0.3%
await Promise.all	50,817 → 51,708	+1.8%	39,825 → 40,022	+0.5%
nested async function call	184,950 → 189,387	+2.4%	205,545 → 204,870	-0.3%

classes.js — Interp: 31 unch. · avg +0.4% · Bytecode: 31 unch. · avg +0.6%

Benchmark	Interpreted	Δ	Bytecode	Δ
simple class new	115,105 → 113,987	-1.0%	323,703 → 339,313	+4.8%
class with defaults	91,505 → 90,751	-0.8%	230,517 → 239,074	+3.7%
50 instances via Array.from	5,580 → 5,515	-1.2%	6,106 → 6,112	+0.1%
instance method call	57,725 → 58,274	+1.0%	152,310 → 150,374	-1.3%
static method call	89,317 → 89,731	+0.5%	354,606 → 353,786	-0.2%
single-level inheritance	45,042 → 45,840	+1.8%	145,895 → 150,854	+3.4%
two-level inheritance	38,417 → 38,377	-0.1%	119,397 → 123,624	+3.5%
private field access	55,164 → 56,638	+2.7%	159,485 → 167,136	+4.8%
private methods	61,272 → 61,827	+0.9%	200,901 → 209,342	+4.2%
getter/setter access	64,245 → 64,577	+0.5%	164,952 → 168,095	+1.9%
class decorator (identity)	81,384 → 80,921	-0.6%	52,021 → 52,010	-0.0%
class decorator (wrapping)	47,288 → 47,395	+0.2%	36,506 → 36,155	-1.0%
identity method decorator	57,546 → 58,229	+1.2%	43,100 → 42,210	-2.1%
wrapping method decorator	48,448 → 48,283	-0.3%	38,893 → 38,989	+0.2%
stacked method decorators (x3)	34,125 → 34,465	+1.0%	27,573 → 27,505	-0.2%
identity field decorator	66,422 → 66,653	+0.3%	45,330 → 44,997	-0.7%
field initializer decorator	56,311 → 56,534	+0.4%	40,660 → 39,947	-1.8%
getter decorator (identity)	56,578 → 54,073	-4.4%	40,474 → 39,971	-1.2%
setter decorator (identity)	47,158 → 47,691	+1.1%	34,952 → 34,527	-1.2%
static method decorator	60,976 → 61,014	+0.1%	62,538 → 62,295	-0.4%
static field decorator	71,488 → 72,235	+1.0%	65,899 → 65,969	+0.1%
private method decorator	47,369 → 47,120	-0.5%	36,305 → 36,668	+1.0%
private field decorator	51,973 → 52,717	+1.4%	38,189 → 38,194	+0.0%
plain auto-accessor (no decorator)	89,808 → 90,285	+0.5%	52,396 → 52,274	-0.2%
auto-accessor with decorator	52,261 → 52,568	+0.6%	36,768 → 36,259	-1.4%
decorator writing metadata	44,035 → 44,323	+0.7%	41,153 → 41,119	-0.1%
static getter read	104,006 → 105,137	+1.1%	380,286 → 378,806	-0.4%
static getter/setter pair	76,282 → 77,649	+1.8%	204,306 → 204,089	-0.1%
inherited static getter	59,379 → 59,828	+0.8%	256,839 → 259,590	+1.1%
inherited static setter	63,236 → 63,923	+1.1%	202,669 → 203,777	+0.5%
inherited static getter with this binding	52,926 → 53,706	+1.5%	149,823 → 150,350	+0.4%

closures.js — Interp: 🔴 1, 10 unch. · avg -1.6% · Bytecode: 11 unch. · avg +0.0%

Benchmark	Interpreted	Δ	Bytecode	Δ
closure over single variable	156,234 → 146,297	-6.4%	586,697 → 594,035	+1.3%
closure over multiple variables	135,831 → 124,533	🔴 -8.3%	392,226 → 389,811	-0.6%
nested closures	128,662 → 129,394	+0.6%	542,763 → 553,942	+2.1%
function as argument	97,439 → 97,947	+0.5%	488,079 → 492,450	+0.9%
function returning function	123,433 → 122,425	-0.8%	577,984 → 577,341	-0.1%
compose two functions	73,079 → 73,682	+0.8%	347,592 → 349,702	+0.6%
fn.call	158,868 → 157,388	-0.9%	128,433 → 126,417	-1.6%
fn.apply	118,652 → 116,701	-1.6%	86,903 → 88,074	+1.3%
fn.bind	142,007 → 139,725	-1.6%	139,334 → 134,871	-3.2%
recursive sum to 50	11,897 → 11,957	+0.5%	38,407 → 38,540	+0.3%
recursive tree traversal	19,922 → 19,873	-0.2%	58,325 → 57,775	-0.9%

collections.js — Interp: 12 unch. · avg +0.4% · Bytecode: 12 unch. · avg -0.5%

Benchmark	Interpreted	Δ	Bytecode	Δ
add 50 elements	6,817 → 7,061	+3.6%	5,507 → 5,491	-0.3%
has lookup (50 elements)	83,952 → 84,992	+1.2%	81,040 → 80,808	-0.3%
delete elements	44,857 → 44,414	-1.0%	35,344 → 35,208	-0.4%
forEach iteration	16,814 → 16,534	-1.7%	14,969 → 14,944	-0.2%
spread to array	36,420 → 38,210	+4.9%	127,329 → 121,922	-4.2%
deduplicate array	43,051 → 43,347	+0.7%	44,683 → 44,538	-0.3%
set 50 entries	5,062 → 5,039	-0.5%	5,322 → 5,165	-3.0%
get lookup (50 entries)	77,689 → 75,400	-2.9%	84,409 → 84,504	+0.1%
has check	117,654 → 117,772	+0.1%	134,438 → 134,177	-0.2%
delete entries	41,557 → 41,270	-0.7%	32,196 → 32,340	+0.4%
forEach iteration	16,322 → 16,398	+0.5%	15,136 → 15,114	-0.1%
keys/values/entries	9,572 → 9,572	-0.0%	18,914 → 19,282	+1.9%

destructuring.js — Interp: 22 unch. · avg +0.4% · Bytecode: 🔴 2, 20 unch. · avg -2.0%

Benchmark	Interpreted	Δ	Bytecode	Δ
simple array destructuring	444,379 → 446,246	+0.4%	630,286 → 627,583	-0.4%
with rest element	299,350 → 295,257	-1.4%	486,364 → 478,657	-1.6%
with defaults	439,027 → 441,945	+0.7%	702,181 → 644,782	🔴 -8.2%
skip elements	449,426 → 454,203	+1.1%	721,366 → 689,052	-4.5%
nested array destructuring	191,325 → 193,898	+1.3%	376,384 → 364,725	-3.1%
swap variables	543,283 → 548,322	+0.9%	974,158 → 959,599	-1.5%
simple object destructuring	300,094 → 302,111	+0.7%	512,240 → 523,081	+2.1%
with defaults	351,739 → 357,499	+1.6%	318,828 → 314,215	-1.4%
with renaming	321,752 → 318,862	-0.9%	571,698 → 572,021	+0.1%
nested object destructuring	150,619 → 152,070	+1.0%	247,383 → 248,353	+0.4%
rest properties	188,201 → 189,334	+0.6%	239,754 → 236,502	-1.4%
object parameter	94,663 → 93,039	-1.7%	187,148 → 183,432	-2.0%
array parameter	128,877 → 129,134	+0.2%	317,662 → 312,860	-1.5%
mixed destructuring in map	35,381 → 35,414	+0.1%	33,443 → 32,243	-3.6%
forEach with array destructuring	69,850 → 71,111	+1.8%	136,397 → 134,573	-1.3%
map with array destructuring	71,889 → 72,197	+0.4%	165,045 → 163,199	-1.1%
filter with array destructuring	73,648 → 74,134	+0.7%	196,222 → 189,137	-3.6%
reduce with array destructuring	80,652 → 81,116	+0.6%	183,949 → 180,438	-1.9%
map with object destructuring	78,780 → 78,345	-0.6%	69,160 → 67,816	-1.9%
map with nested destructuring	64,493 → 64,063	-0.7%	54,283 → 57,469	+5.9%
map with rest in destructuring	40,276 → 40,880	+1.5%	44,603 → 41,585	-6.8%
map with defaults in destructuring	59,493 → 59,449	-0.1%	36,943 → 34,208	🔴 -7.4%

fibonacci.js — Interp: 8 unch. · avg -0.2% · Bytecode: 🟢 2, 6 unch. · avg +1.2%

Benchmark	Interpreted	Δ	Bytecode	Δ
recursive fib(15)	360 → 364	+1.1%	1,124 → 1,084	-3.6%
recursive fib(20)	30 → 30	+1.6%	98 → 101	+3.8%
recursive fib(15) typed	332 → 333	+0.3%	1,426 → 1,360	-4.6%
recursive fib(20) typed	29 → 30	+0.8%	129 → 129	+0.1%
iterative fib(20) via reduce	12,603 → 12,671	+0.5%	8,320 → 8,030	-3.5%
iterator fib(20)	9,802 → 9,622	-1.8%	14,104 → 15,153	🟢 +7.4%
iterator fib(20) via Iterator.from + take	15,008 → 14,715	-2.0%	15,872 → 17,406	🟢 +9.7%
iterator fib(20) last value via reduce	11,566 → 11,295	-2.3%	12,359 → 12,427	+0.6%

for-of.js — Interp: 7 unch. · avg +0.4% · Bytecode: 🟢 1, 6 unch. · avg +1.5%

Benchmark	Interpreted	Δ	Bytecode	Δ
for...of with 10-element array	53,111 → 53,138	+0.1%	156,914 → 149,349	-4.8%
for...of with 100-element array	6,024 → 6,048	+0.4%	21,349 → 21,397	+0.2%
for...of with string (10 chars)	37,733 → 37,375	-0.9%	111,748 → 123,306	🟢 +10.3%
for...of with Set (10 elements)	51,948 → 51,888	-0.1%	156,183 → 159,889	+2.4%
for...of with Map entries (10 entries)	33,449 → 33,670	+0.7%	55,254 → 56,356	+2.0%
for...of with destructuring	44,417 → 44,813	+0.9%	86,408 → 86,892	+0.6%
for-await-of with sync array	49,384 → 50,398	+2.1%	131,140 → 131,296	+0.1%

iterators.js — Interp: 20 unch. · avg +0.2% · Bytecode: 🔴 1, 19 unch. · avg +0.3%

Benchmark	Interpreted	Δ	Bytecode	Δ
Iterator.from({next}).toArray() — 20 elements	15,644 → 15,503	-0.9%	18,238 → 15,552	🔴 -14.7%
Iterator.from({next}).toArray() — 50 elements	6,812 → 6,712	-1.5%	7,596 → 7,324	-3.6%
spread pre-wrapped iterator — 20 elements	12,284 → 12,124	-1.3%	15,351 → 16,119	+5.0%
Iterator.from({next}).forEach — 50 elements	4,712 → 4,664	-1.0%	4,907 → 5,221	+6.4%
Iterator.from({next}).reduce — 50 elements	4,792 → 4,730	-1.3%	4,926 → 4,904	-0.4%
wrap array iterator	182,724 → 181,115	-0.9%	116,179 → 119,918	+3.2%
wrap plain {next()} object	10,668 → 10,690	+0.2%	11,778 → 11,838	+0.5%
map + toArray (50 elements)	4,720 → 4,819	+2.1%	5,044 → 5,127	+1.6%
filter + toArray (50 elements)	4,628 → 4,705	+1.7%	5,176 → 5,219	+0.8%
take(10) + toArray (50 element source)	27,920 → 28,167	+0.9%	28,104 → 28,826	+2.6%
drop(40) + toArray (50 element source)	6,608 → 6,680	+1.1%	7,335 → 7,359	+0.3%
chained map + filter + take (100 element source)	8,569 → 8,699	+1.5%	9,211 → 8,907	-3.3%
some + every (50 elements)	2,717 → 2,711	-0.2%	2,978 → 2,966	-0.4%
find (50 elements)	5,883 → 5,952	+1.2%	6,508 → 6,564	+0.9%
array.values().map().filter().toArray()	9,296 → 9,426	+1.4%	9,207 → 9,181	-0.3%
array.values().take(5).toArray()	222,940 → 223,234	+0.1%	141,141 → 149,393	+5.8%
array.values().drop(45).toArray()	205,053 → 204,920	-0.1%	137,160 → 137,908	+0.5%
map.entries() chained helpers	11,247 → 11,311	+0.6%	4,801 → 4,890	+1.8%
set.values() chained helpers	19,300 → 19,382	+0.4%	19,529 → 19,267	-1.3%
string iterator map + toArray	14,670 → 14,593	-0.5%	20,255 → 20,364	+0.5%

json.js — Interp: 20 unch. · avg +0.5% · Bytecode: 🟢 1, 🔴 1, 18 unch. · avg -0.6%

Benchmark	Interpreted	Δ	Bytecode	Δ
parse simple object	189,996 → 191,368	+0.7%	155,630 → 153,807	-1.2%
parse nested object	114,537 → 114,938	+0.3%	109,402 → 107,781	-1.5%
parse array of objects	60,730 → 60,913	+0.3%	60,292 → 60,628	+0.6%
parse large flat object	51,362 → 51,819	+0.9%	49,114 → 49,467	+0.7%
parse mixed types	74,277 → 75,269	+1.3%	70,029 → 71,235	+1.7%
stringify simple object	151,274 → 151,330	+0.0%	131,336 → 129,516	-1.4%
stringify nested object	81,444 → 82,650	+1.5%	73,421 → 72,782	-0.9%
stringify array of objects	38,586 → 38,694	+0.3%	35,278 → 34,644	-1.8%
stringify mixed types	69,966 → 69,730	-0.3%	62,617 → 62,809	+0.3%
reviver doubles numbers	46,665 → 46,711	+0.1%	40,559 → 45,688	🟢 +12.6%
reviver filters properties	38,946 → 39,232	+0.7%	46,571 → 45,012	-3.3%
reviver on nested object	53,013 → 53,590	+1.1%	53,178 → 45,825	🔴 -13.8%
reviver on array	29,719 → 30,245	+1.8%	28,413 → 28,881	+1.6%
replacer function doubles numbers	46,165 → 46,433	+0.6%	47,584 → 47,392	-0.4%
replacer function excludes properties	59,721 → 59,743	+0.0%	58,174 → 57,523	-1.1%
array replacer (allowlist)	107,798 → 107,815	+0.0%	88,969 → 89,754	+0.9%
stringify with 2-space indent	79,273 → 80,122	+1.1%	67,589 → 66,327	-1.9%
stringify with tab indent	79,824 → 80,424	+0.8%	67,953 → 66,965	-1.5%
parse then stringify	48,196 → 47,871	-0.7%	44,541 → 44,356	-0.4%
stringify then parse	28,599 → 28,614	+0.1%	26,734 → 26,629	-0.4%

jsx.jsx — Interp: 21 unch. · avg +0.5% · Bytecode: 🟢 1, 🔴 1, 19 unch. · avg +0.1%

Benchmark	Interpreted	Δ	Bytecode	Δ
simple element	239,055 → 252,853	+5.8%	670,692 → 677,429	+1.0%
self-closing element	236,764 → 234,695	-0.9%	723,121 → 686,651	-5.0%
element with string attribute	189,389 → 191,131	+0.9%	508,807 → 417,409	🔴 -18.0%
element with multiple attributes	158,991 → 159,590	+0.4%	365,441 → 363,959	-0.4%
element with expression attribute	170,671 → 172,592	+1.1%	408,612 → 432,256	+5.8%
text child	218,796 → 221,196	+1.1%	557,655 → 570,651	+2.3%
expression child	213,100 → 215,711	+1.2%	582,928 → 592,797	+1.7%
mixed text and expression	202,671 → 202,093	-0.3%	523,346 → 538,056	+2.8%
nested elements (3 levels)	80,067 → 80,565	+0.6%	224,203 → 243,323	🟢 +8.5%
sibling children	59,704 → 59,616	-0.1%	178,666 → 173,052	-3.1%
component element	154,043 → 152,059	-1.3%	420,686 → 426,051	+1.3%
component with children	94,214 → 93,397	-0.9%	265,157 → 259,650	-2.1%
dotted component	129,459 → 129,869	+0.3%	306,075 → 312,172	+2.0%
empty fragment	222,854 → 223,683	+0.4%	649,109 → 638,521	-1.6%
fragment with children	59,385 → 58,959	-0.7%	175,364 → 176,917	+0.9%
spread attributes	110,944 → 111,757	+0.7%	108,992 → 110,704	+1.6%
spread with overrides	96,647 → 97,938	+1.3%	82,944 → 83,746	+1.0%
shorthand props	164,251 → 163,032	-0.7%	401,872 → 405,281	+0.8%
nav bar structure	27,497 → 27,811	+1.1%	77,786 → 78,637	+1.1%
card component tree	32,516 → 32,628	+0.3%	85,311 → 85,841	+0.6%
10 list items via Array.from	14,936 → 14,817	-0.8%	23,483 → 23,886	+1.7%

numbers.js — Interp: 11 unch. · avg -1.0% · Bytecode: 11 unch. · avg -0.4%

Benchmark	Interpreted	Δ	Bytecode	Δ
integer arithmetic	633,893 → 636,594	+0.4%	1,499,366 → 1,493,020	-0.4%
floating point arithmetic	609,827 → 609,605	-0.0%	1,599,208 → 1,582,970	-1.0%
number coercion	185,254 → 178,053	-3.9%	129,483 → 129,977	+0.4%
toFixed	105,391 → 103,906	-1.4%	203,957 → 205,514	+0.8%
toString	167,649 → 165,131	-1.5%	658,832 → 661,335	+0.4%
valueOf	241,320 → 243,405	+0.9%	920,567 → 925,268	+0.5%
toPrecision	152,648 → 152,103	-0.4%	400,699 → 392,973	-1.9%
Number.isNaN	282,839 → 282,395	-0.2%	165,913 → 166,591	+0.4%
Number.isFinite	278,708 → 275,802	-1.0%	160,585 → 159,540	-0.7%
Number.isInteger	273,114 → 269,614	-1.3%	175,217 → 172,555	-1.5%
Number.parseInt and parseFloat	238,954 → 231,511	-3.1%	150,579 → 148,538	-1.4%

objects.js — Interp: 7 unch. · avg +3.3% · Bytecode: 🔴 1, 6 unch. · avg -0.4%

Benchmark	Interpreted	Δ	Bytecode	Δ
create simple object	495,907 → 522,512	+5.4%	859,362 → 868,368	+1.0%
create nested object	231,941 → 239,224	+3.1%	371,687 → 374,700	+0.8%
create 50 objects via Array.from	9,447 → 9,854	+4.3%	8,354 → 8,171	-2.2%
property read	628,804 → 636,162	+1.2%	644,289 → 644,482	+0.0%
Object.keys	292,473 → 300,585	+2.8%	222,120 → 227,977	+2.6%
Object.entries	106,471 → 112,896	+6.0%	74,037 → 68,175	🔴 -7.9%
spread operator	196,590 → 197,472	+0.4%	200,567 → 206,085	+2.8%

promises.js — Interp: 🟢 1, 11 unch. · avg +1.4% · Bytecode: 🟢 1, 🔴 2, 9 unch. · avg -1.0%

Benchmark	Interpreted	Δ	Bytecode	Δ
Promise.resolve(value)	578,744 → 579,430	+0.1%	364,342 → 371,934	+2.1%
new Promise(resolve => resolve(value))	197,515 → 201,294	+1.9%	187,620 → 165,323	🔴 -11.9%
Promise.reject(reason)	575,230 → 564,891	-1.8%	347,651 → 331,563	-4.6%
resolve + then (1 handler)	176,762 → 182,760	+3.4%	147,977 → 152,958	+3.4%
resolve + then chain (3 deep)	66,265 → 73,121	🟢 +10.3%	66,184 → 64,902	-1.9%
resolve + then chain (10 deep)	23,723 → 24,298	+2.4%	23,728 → 23,045	-2.9%
reject + catch + then	108,175 → 107,017	-1.1%	95,630 → 91,525	-4.3%
resolve + finally + then	90,815 → 90,552	-0.3%	75,209 → 80,427	+6.9%
Promise.all (5 resolved)	35,082 → 35,140	+0.2%	27,776 → 25,592	🔴 -7.9%
Promise.race (5 resolved)	36,838 → 37,333	+1.3%	27,442 → 26,501	-3.4%
Promise.allSettled (5 mixed)	29,861 → 30,155	+1.0%	22,345 → 24,175	🟢 +8.2%
Promise.any (5 mixed)	34,842 → 34,589	-0.7%	26,609 → 27,723	+4.2%

strings.js — Interp: 11 unch. · avg -1.1% · Bytecode: 11 unch. · avg +1.1%

Benchmark	Interpreted	Δ	Bytecode	Δ
string concatenation	492,689 → 484,029	-1.8%	375,563 → 397,806	+5.9%
template literal	414,297 → 415,309	+0.2%	603,015 → 621,760	+3.1%
string repeat	429,230 → 415,385	-3.2%	938,568 → 910,534	-3.0%
split and join	143,470 → 145,133	+1.2%	261,294 → 267,425	+2.3%
indexOf and includes	170,373 → 168,028	-1.4%	523,217 → 534,119	+2.1%
toUpperCase and toLowerCase	269,686 → 269,100	-0.2%	686,455 → 672,803	-2.0%
slice and substring	168,051 → 165,154	-1.7%	599,345 → 605,912	+1.1%
trim operations	199,839 → 201,344	+0.8%	710,683 → 706,364	-0.6%
replace and replaceAll	222,500 → 215,984	-2.9%	599,678 → 602,051	+0.4%
startsWith and endsWith	140,385 → 138,672	-1.2%	466,741 → 481,617	+3.2%
padStart and padEnd	209,187 → 205,697	-1.7%	540,350 → 540,234	-0.0%

typed-arrays.js — Interp: 22 unch. · avg -0.1% · Bytecode: 22 unch. · avg -0.5%

Benchmark	Interpreted	Δ	Bytecode	Δ
new Int32Array(0)	347,439 → 345,805	-0.5%	123,876 → 123,612	-0.2%
new Int32Array(100)	318,795 → 318,740	-0.0%	119,745 → 118,284	-1.2%
new Int32Array(1000)	157,585 → 159,260	+1.1%	89,600 → 87,765	-2.0%
new Float64Array(100)	276,375 → 265,879	-3.8%	113,268 → 115,031	+1.6%
Int32Array.from([...])	185,151 → 183,662	-0.8%	62,183 → 62,648	+0.7%
Int32Array.of(1, 2, 3, 4, 5)	326,079 → 324,860	-0.4%	248,315 → 239,227	-3.7%
sequential write 100 elements	3,966 → 3,960	-0.1%	12,322 → 12,368	+0.4%
sequential read 100 elements	4,140 → 4,098	-1.0%	9,825 → 9,725	-1.0%
Float64Array write 100 elements	3,630 → 3,663	+0.9%	11,805 → 11,605	-1.7%
fill(42)	44,806 → 44,998	+0.4%	43,422 → 43,314	-0.2%
slice()	213,113 → 213,596	+0.2%	188,388 → 183,283	-2.7%
map(x => x * 2)	8,597 → 8,666	+0.8%	7,005 → 6,986	-0.3%
filter(x => x > 50)	8,722 → 8,682	-0.5%	7,423 → 7,346	-1.0%
reduce (sum)	8,373 → 8,305	-0.8%	6,319 → 6,269	-0.8%
sort()	168,872 → 175,728	+4.1%	148,693 → 153,863	+3.5%
indexOf()	440,188 → 440,575	+0.1%	340,434 → 340,632	+0.1%
reverse()	332,158 → 334,516	+0.7%	261,541 → 262,265	+0.3%
create view over existing buffer	419,692 → 416,933	-0.7%	131,912 → 131,888	-0.0%
subarray()	471,072 → 458,958	-2.6%	341,881 → 336,473	-1.6%
set() from array	575,448 → 575,888	+0.1%	242,255 → 240,313	-0.8%
for-of loop	5,650 → 5,680	+0.5%	18,152 → 17,787	-2.0%
spread into array	20,626 → 20,702	+0.4%	61,250 → 62,179	+1.5%

_{Measured on ubuntu-latest x64. Changes within ±7% are considered insignificant.}

[coderabbitai]

🧹 Nitpick comments (1)

tests-wasm/run-wasm-tests.mjs (1)

52-75: Consider caching file content to avoid double reads.

Both parseExpectedLines and isSkipped call readFileSync on the same fixture file. In runFixture, if the file isn't skipped, parseExpectedLines reads it again. For a small test suite this is fine, but you could optimize by reading once and passing the content to both functions.

♻️ Optional: Single-read approach

+function readFixtureSource(fileName) {
+  return readFileSync(fileName, "utf8");
+}
+
-function parseExpectedLines(fileName) {
+function parseExpectedLines(source) {
   const expectedPrefix = "// Expected: ";
-  const source = readFileSync(fileName, "utf8");
   return source
     .split(/\r?\n/)
     .filter((line) => line.startsWith(expectedPrefix))
     .map((line) => line.slice(expectedPrefix.length));
 }

-function isSkipped(fileName) {
-  const source = readFileSync(fileName, "utf8");
+function isSkipped(source) {
   for (const rawLine of source.split(/\r?\n/)) {

Then in runFixture:

const source = readFixtureSource(fixtureFile);
if (isSkipped(source)) { ... }
// later:
const expected = parseExpectedLines(source);

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In `@tests-wasm/run-wasm-tests.mjs` around lines 52 - 75, Both parseExpectedLines
and isSkipped independently call readFileSync causing duplicate reads; change
the API to accept the file content string instead of reading inside those
functions. Update parseExpectedLines(source) and isSkipped(source) to take a
source string (remove internal readFileSync usage), add a helper like
readFixtureSource in runFixture to read the file once, and pass that source to
isSkipped and parseExpectedLines. Ensure all call sites that previously passed
fileName are updated to pass the cached source string.

🤖 Prompt for all review comments with AI agents

Verify each finding against the current code and only fix it if needed.

Nitpick comments:
In `@tests-wasm/run-wasm-tests.mjs`:
- Around line 52-75: Both parseExpectedLines and isSkipped independently call
readFileSync causing duplicate reads; change the API to accept the file content
string instead of reading inside those functions. Update
parseExpectedLines(source) and isSkipped(source) to take a source string (remove
internal readFileSync usage), add a helper like readFixtureSource in runFixture
to read the file once, and pass that source to isSkipped and parseExpectedLines.
Ensure all call sites that previously passed fileName are updated to pass the
cached source string.

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro

Run ID: 8d855643-0108-4bc6-997b-802ff932808a

📥 Commits

Reviewing files that changed from the base of the PR and between 95a3bf4 and fae9c08.

📒 Files selected for processing (6)

• .github/workflows/ci.yml
• .github/workflows/pr.yml
• AGENTS.md
• docs/wasm-backend.md
• tests-wasm/run-wasm-tests.mjs
• tests-wasm/run-wasm-tests.pas

💤 Files with no reviewable changes (1)

• tests-wasm/run-wasm-tests.pas