trending/scriptc
1
0
Fork 0
mirror of https://github.com/vercel-labs/scriptc.git synced 2026-10-22 05:59:31 +08:00
Code Issues Packages Projects Releases Wiki Activity

Compile the deprecated punycode and constants builtin modules

Browse Source 
- node:punycode: encode, decode, toASCII, toUnicode, ucs2.decode,
  ucs2.encode and version, ported from the punycode.js 2.1.0 Node bundles
  (scr_url.c), with its RangeErrors. DEP0040 follows Node: a static ES
  import reports before any module evaluates and before the program can
  register a 'warning' listener; require() reports on the next tick, once
  for both spellings, not when a dependency under node_modules loads the
  module first, and not under process.noDeprecation.
- node:constants: the frozen object Node builds from os.constants'
  dlopen, errno, priority and signals groups, fs.constants (node_constants.cc
  order and guards) and crypto.constants (scr_os.c); its bindings read it as
  a checked-dynamic value.

These unblock tr46 (whatwg-url, node-fetch@2) and graceful-fs (fs-extra).

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This commit is contained in:
Malte Ubl
2026-10-18 06:23:56 -07:00
co-authored by Claude Opus 5.5
parent b4e25cdba0
commit 1c19d6613d
38 changed files with 1169 additions and 88 deletions
Download Patch File Download Diff File Expand all files Collapse all files
docs/content/docs/limitations.mdx
+2
View File
@@ -74,6 +74,8 @@ title: "Limitations"
<include>../limitations/builtin-apis/legacy-url-api.mdx</include>
<include>../limitations/builtin-apis/deprecated-modules.mdx</include>
<include>../limitations/builtin-apis/filesystem-flags.mdx</include>
<include>../limitations/builtin-apis/copy.mdx</include>
docs/content/limitations/builtin-apis/deprecated-modules.mdx
+1
View File
@@ -0,0 +1 @@
- **Deprecated `punycode` and `constants` modules:** `punycode.encode()`, `decode()`, `toASCII()`, `toUnicode()`, `ucs2.decode()`, `ucs2.encode()`, and `version` compile, with the `DEP0040` warning Node.js reports when the module loads. A `require("punycode")` made from inside `node_modules` doesn't warn; a `require()` used inside a larger expression, such as `require("punycode").toASCII(name)`, doesn't warn either. The `punycode.ucs2` object itself is unsupported as a value. The `constants` module is the frozen object Node.js builds from `os.constants` and `fs.constants`, plus the `crypto.constants` of Node.js 24.22.0.
docs/src/generated/node-v24-compatibility-meta.json
+1 -1
View File
@@ -2,6 +2,6 @@
"schemaVersion": 3,
"nodeVersion": "24.22.0",
"nodeCommit": "f14cd3aa7bf9ca8df2cc5b8bb14beb0f335b1d93",
"artifactVersion": "a1ff929d49885bd8c92e",
"artifactVersion": "7b345b86e2e5f616a7a0",
"rowCount": 3880
}
docs/src/generated/node-v24-compatibility.json
+24 -24
View File
@@ -584,9 +584,9 @@
"apiEntries": 9,
"static": {
"supported": 0,
"partial": 0,
"partial": 8,
"refused": 0,
"not-implemented": 9,
"not-implemented": 1,
"by-design": 0,
"unreviewed": 0,
"not-applicable": 0
@@ -58489,7 +58489,7 @@
"inherited": false
},
"static": {
"status": "not-implemented",
"status": "partial",
"detail": "Derived from the exact API rows in this chapter.",
"verification": "derived"
}
@@ -58512,9 +58512,9 @@
"inherited": true
},
"static": {
"status": "not-implemented",
"detail": "No static compiler lowering is registered for this API yet.",
"verification": "registry-gap"
"status": "partial",
"detail": "Implemented for the call shapes accepted by the compiler lowering.",
"verification": "test-backed"
}
},
{
@@ -58535,9 +58535,9 @@
"inherited": true
},
"static": {
"status": "not-implemented",
"detail": "No static compiler lowering is registered for this API yet.",
"verification": "registry-gap"
"status": "partial",
"detail": "Implemented for the call shapes accepted by the compiler lowering.",
"verification": "test-backed"
}
},
{
@@ -58558,9 +58558,9 @@
"inherited": true
},
"static": {
"status": "not-implemented",
"detail": "No static compiler lowering is registered for this API yet.",
"verification": "registry-gap"
"status": "partial",
"detail": "Implemented for the call shapes accepted by the compiler lowering.",
"verification": "test-backed"
}
},
{
@@ -58581,9 +58581,9 @@
"inherited": true
},
"static": {
"status": "not-implemented",
"detail": "No static compiler lowering is registered for this API yet.",
"verification": "registry-gap"
"status": "partial",
"detail": "Implemented for the call shapes accepted by the compiler lowering.",
"verification": "test-backed"
}
},
{
@@ -58627,9 +58627,9 @@
"inherited": true
},
"static": {
"status": "not-implemented",
"detail": "No static compiler lowering is registered for this API yet.",
"verification": "registry-gap"
"status": "partial",
"detail": "Implemented for the call shapes accepted by the compiler lowering.",
"verification": "test-backed"
}
},
{
@@ -58650,9 +58650,9 @@
"inherited": true
},
"static": {
"status": "not-implemented",
"detail": "No static compiler lowering is registered for this API yet.",
"verification": "registry-gap"
"status": "partial",
"detail": "Implemented for the call shapes accepted by the compiler lowering.",
"verification": "test-backed"
}
},
{
@@ -58673,9 +58673,9 @@
"inherited": true
},
"static": {
"status": "not-implemented",
"detail": "No static compiler lowering is registered for this API yet.",
"verification": "registry-gap"
"status": "partial",
"detail": "Implemented for the call shapes accepted by the compiler lowering.",
"verification": "test-backed"
}
},
{
internal/compatibility/generated/node-v24-backlog.json
+72 -44
View File
@@ -7,8 +7,8 @@
"tierItemCount": 3456,
"tiers": {
"static": {
"audit-partial": 1091,
"verify-gap": 2238,
"audit-partial": 1098,
"verify-gap": 2231,
"replace-refusal": 40,
"classify": 2,
"implement": 85
@@ -58534,12 +58534,16 @@
"priority": "low",
"tiers": {
"static": {
"status": "not-implemented",
"action": "verify-gap",
"verification": "registry-gap",
"confidence": "low",
"source": "compiler-unmatched:punycode",
"tests": []
"status": "partial",
"action": "audit-partial",
"verification": "test-backed",
"confidence": "high",
"source": "surface-manifest:node-builtin.punycode.decode",
"tests": [
"tests/corpus/6122-punycode-conversions.ts",
"tests/corpus/6124-punycode-require-warnings.cjs",
"tests/corpus/6126-punycode-dependency/main.cjs"
]
}
}
},
@@ -58559,12 +58563,16 @@
"priority": "low",
"tiers": {
"static": {
"status": "not-implemented",
"action": "verify-gap",
"verification": "registry-gap",
"confidence": "low",
"source": "compiler-unmatched:punycode",
"tests": []
"status": "partial",
"action": "audit-partial",
"verification": "test-backed",
"confidence": "high",
"source": "surface-manifest:node-builtin.punycode.encode",
"tests": [
"tests/corpus/6122-punycode-conversions.ts",
"tests/corpus/6124-punycode-require-warnings.cjs",
"tests/corpus/6126-punycode-dependency/main.cjs"
]
}
}
},
@@ -58584,12 +58592,16 @@
"priority": "low",
"tiers": {
"static": {
"status": "not-implemented",
"action": "verify-gap",
"verification": "registry-gap",
"confidence": "low",
"source": "compiler-unmatched:punycode",
"tests": []
"status": "partial",
"action": "audit-partial",
"verification": "test-backed",
"confidence": "high",
"source": "surface-manifest:node-builtin.punycode.toASCII",
"tests": [
"tests/corpus/6122-punycode-conversions.ts",
"tests/corpus/6124-punycode-require-warnings.cjs",
"tests/corpus/6126-punycode-dependency/main.cjs"
]
}
}
},
@@ -58609,12 +58621,16 @@
"priority": "low",
"tiers": {
"static": {
"status": "not-implemented",
"action": "verify-gap",
"verification": "registry-gap",
"confidence": "low",
"source": "compiler-unmatched:punycode",
"tests": []
"status": "partial",
"action": "audit-partial",
"verification": "test-backed",
"confidence": "high",
"source": "surface-manifest:node-builtin.punycode.toUnicode",
"tests": [
"tests/corpus/6122-punycode-conversions.ts",
"tests/corpus/6124-punycode-require-warnings.cjs",
"tests/corpus/6126-punycode-dependency/main.cjs"
]
}
}
},
@@ -58659,12 +58675,16 @@
"priority": "low",
"tiers": {
"static": {
"status": "not-implemented",
"action": "verify-gap",
"verification": "registry-gap",
"confidence": "low",
"source": "compiler-unmatched:punycode",
"tests": []
"status": "partial",
"action": "audit-partial",
"verification": "test-backed",
"confidence": "high",
"source": "surface-manifest:node-builtin.punycode.ucs2.decode",
"tests": [
"tests/corpus/6122-punycode-conversions.ts",
"tests/corpus/6124-punycode-require-warnings.cjs",
"tests/corpus/6126-punycode-dependency/main.cjs"
]
}
}
},
@@ -58684,12 +58704,16 @@
"priority": "low",
"tiers": {
"static": {
"status": "not-implemented",
"action": "verify-gap",
"verification": "registry-gap",
"confidence": "low",
"source": "compiler-unmatched:punycode",
"tests": []
"status": "partial",
"action": "audit-partial",
"verification": "test-backed",
"confidence": "high",
"source": "surface-manifest:node-builtin.punycode.ucs2.encode",
"tests": [
"tests/corpus/6122-punycode-conversions.ts",
"tests/corpus/6124-punycode-require-warnings.cjs",
"tests/corpus/6126-punycode-dependency/main.cjs"
]
}
}
},
@@ -58709,12 +58733,16 @@
"priority": "low",
"tiers": {
"static": {
"status": "not-implemented",
"action": "verify-gap",
"verification": "registry-gap",
"confidence": "low",
"source": "compiler-unmatched:punycode",
"tests": []
"status": "partial",
"action": "audit-partial",
"verification": "test-backed",
"confidence": "high",
"source": "surface-manifest:node-builtin.punycode.version",
"tests": [
"tests/corpus/6122-punycode-conversions.ts",
"tests/corpus/6124-punycode-require-warnings.cjs",
"tests/corpus/6126-punycode-dependency/main.cjs"
]
}
}
},
internal/compatibility/generated/node-v24-internal.json
+58 -18
View File
@@ -611,9 +611,9 @@
"apiEntries": 9,
"static": {
"supported": 0,
"partial": 0,
"partial": 8,
"refused": 0,
"not-implemented": 9,
"not-implemented": 1,
"by-design": 0,
"unreviewed": 0,
"not-applicable": 0
@@ -81908,8 +81908,13 @@
"inherited": false
},
"static": {
"status": "not-implemented",
"evidence": "derived:descendants"
"status": "partial",
"evidence": "derived:descendants",
"tests": [
"tests/corpus/6122-punycode-conversions.ts",
"tests/corpus/6124-punycode-require-warnings.cjs",
"tests/corpus/6126-punycode-dependency/main.cjs"
]
},
"anchorSource": "chapter"
},
@@ -81931,8 +81936,13 @@
"inherited": true
},
"static": {
"status": "not-implemented",
"evidence": "compiler-unmatched:punycode"
"status": "partial",
"evidence": "surface-manifest:node-builtin.punycode.decode",
"tests": [
"tests/corpus/6122-punycode-conversions.ts",
"tests/corpus/6124-punycode-require-warnings.cjs",
"tests/corpus/6126-punycode-dependency/main.cjs"
]
},
"anchorSource": "exact"
},
@@ -81954,8 +81964,13 @@
"inherited": true
},
"static": {
"status": "not-implemented",
"evidence": "compiler-unmatched:punycode"
"status": "partial",
"evidence": "surface-manifest:node-builtin.punycode.encode",
"tests": [
"tests/corpus/6122-punycode-conversions.ts",
"tests/corpus/6124-punycode-require-warnings.cjs",
"tests/corpus/6126-punycode-dependency/main.cjs"
]
},
"anchorSource": "exact"
},
@@ -81977,8 +81992,13 @@
"inherited": true
},
"static": {
"status": "not-implemented",
"evidence": "compiler-unmatched:punycode"
"status": "partial",
"evidence": "surface-manifest:node-builtin.punycode.toASCII",
"tests": [
"tests/corpus/6122-punycode-conversions.ts",
"tests/corpus/6124-punycode-require-warnings.cjs",
"tests/corpus/6126-punycode-dependency/main.cjs"
]
},
"anchorSource": "exact"
},
@@ -82000,8 +82020,13 @@
"inherited": true
},
"static": {
"status": "not-implemented",
"evidence": "compiler-unmatched:punycode"
"status": "partial",
"evidence": "surface-manifest:node-builtin.punycode.toUnicode",
"tests": [
"tests/corpus/6122-punycode-conversions.ts",
"tests/corpus/6124-punycode-require-warnings.cjs",
"tests/corpus/6126-punycode-dependency/main.cjs"
]
},
"anchorSource": "exact"
},
@@ -82046,8 +82071,13 @@
"inherited": true
},
"static": {
"status": "not-implemented",
"evidence": "compiler-unmatched:punycode"
"status": "partial",
"evidence": "surface-manifest:node-builtin.punycode.ucs2.decode",
"tests": [
"tests/corpus/6122-punycode-conversions.ts",
"tests/corpus/6124-punycode-require-warnings.cjs",
"tests/corpus/6126-punycode-dependency/main.cjs"
]
},
"anchorSource": "exact"
},
@@ -82069,8 +82099,13 @@
"inherited": true
},
"static": {
"status": "not-implemented",
"evidence": "compiler-unmatched:punycode"
"status": "partial",
"evidence": "surface-manifest:node-builtin.punycode.ucs2.encode",
"tests": [
"tests/corpus/6122-punycode-conversions.ts",
"tests/corpus/6124-punycode-require-warnings.cjs",
"tests/corpus/6126-punycode-dependency/main.cjs"
]
},
"anchorSource": "exact"
},
@@ -82092,8 +82127,13 @@
"inherited": true
},
"static": {
"status": "not-implemented",
"evidence": "compiler-unmatched:punycode"
"status": "partial",
"evidence": "surface-manifest:node-builtin.punycode.version",
"tests": [
"tests/corpus/6122-punycode-conversions.ts",
"tests/corpus/6124-punycode-require-warnings.cjs",
"tests/corpus/6126-punycode-dependency/main.cjs"
]
},
"anchorSource": "chapter"
},
internal/compatibility/static-support.json
+1
View File
@@ -146,6 +146,7 @@
"node-builtin.process.rows": ["tests/corpus/2992-stream-rows.ts", "tests/harness/terminal-geometry.test.ts"],
"node-builtin.process.columns": ["tests/corpus/1448-stream-columns.ts", "tests/harness/terminal-geometry.test.ts"],
"node-builtin.process.exitCode": ["tests/corpus/2941-process-exit-code.ts", "tests/corpus/2942-process-exit-code-override.ts", "tests/corpus/2943-process-exit-code-invalid.ts", "tests/corpus/2944-process-exit-code-wrap.ts"],
"node-builtin.punycode": ["tests/corpus/6122-punycode-conversions.ts", "tests/corpus/6124-punycode-require-warnings.cjs", "tests/corpus/6126-punycode-dependency/main.cjs"],
"node-builtin.querystring": ["tests/corpus/957-builtins-namespace.ts", "tests/corpus/2917-builtin-callable-widths.ts"],
"node-builtin.readline": ["tests/corpus/1475-readline-closed-stdin.ts"],
"node-builtin.stream": ["tests/corpus/1685-stream-readable-basics.ts", "tests/corpus/1746-stream-for-await.ts", "tests/corpus/1814-stream-pipeline.cjs", "tests/corpus/2629-stream-consumers.ts", "tests/corpus/2861-readable-iterator-close.ts"],
packages/compiler/ambient/scriptc-node-fallback.d.ts
Vendored
+33
View File
@@ -3643,6 +3643,39 @@ declare module "node:querystring" {
export * from "querystring";
}
/* node:punycode (deprecated, DEP0040) — punycode.js 2.1.0: decode and
* toUnicode throw its RangeErrors on malformed input; ucs2 converts
* between strings and UTF-16 code point arrays. */
declare module "punycode" {
export function decode(string: string): string;
export function encode(string: string): string;
export function toUnicode(domain: string): string;
export function toASCII(domain: string): string;
export namespace ucs2 {
function decode(string: string): number[];
function encode(codePoints: readonly number[]): string;
}
export const version: string;
}
declare module "node:punycode" {
export * from "punycode";
}
/* node:constants (deprecated, DEP0008) — one frozen object: os.constants'
* dlopen, errno, priority and signals groups, fs.constants, and
* crypto.constants. A platform's missing constant reads as undefined. */
declare module "constants" {
const constants: { readonly [name: string]: number } & {
readonly defaultCoreCipherList: string;
readonly defaultCipherList: string;
};
export = constants;
}
declare module "node:constants" {
import constants = require("constants");
export = constants;
}
/* node:readline — the question/close slice: createInterface over exactly
* { input: process.stdin, output: process.stdout }, question(query, cb)
* (the query writes to stdout, the callback gets the next line's text),
packages/compiler/src/backend/llvm/lib-dispatch.ts
+2
View File
@@ -680,6 +680,8 @@ export function emitLibCall(host: LlvmEmitterContext, e: LibCallExpr): LlValue {
case "events":
case "blob":
case "record":
case "punycode":
case "constants":
return host.emitGenericLibCall(e);
case "test":
return emitTestLibCall(host, e);
packages/compiler/src/backend/llvm/lib-shared.ts
+8
View File
@@ -192,6 +192,7 @@ export const LIB_FN_SYMS: Record<string, string> = {
"os.cpusModel": "scr_os_cpus_model",
"os.cpusField": "scr_os_cpus_field",
"os.constants": "scr_os_constants",
"constants.object": "scr_node_constants",
"os.release": "scr_os_release",
"os.userName": "scr_os_user_name",
"os.userShell": "scr_os_user_shell",
@@ -631,6 +632,13 @@ export const LIB_FN_SYMS: Record<string, string> = {
"url.legacyFormat": "scr_url_legacy_format",
"url.legacyResolve": "scr_url_legacy_resolve",
"url.domainToASCII": "scr_url_domain_to_ascii",
"punycode.decode": "scr_punycode_decode",
"punycode.encode": "scr_punycode_encode",
"punycode.toASCII": "scr_punycode_to_ascii",
"punycode.toUnicode": "scr_punycode_to_unicode",
"punycode.ucs2Decode": "scr_punycode_ucs2_decode",
"punycode.ucs2Encode": "scr_str_from_code_point",
"punycode.warn": "scr_punycode_warn",
"url.domainToUnicode": "scr_url_domain_to_unicode",
"url.pathToFileURL": "scr_url_from_path",
"url.pathToFileURLPlatform": "scr_url_from_path_platform",
packages/compiler/src/frontend/builtin-modules.ts
+2
View File
@@ -42,6 +42,8 @@ export const SUPPORTED_BUILTIN_MODULES = [
"module",
"console",
"process",
"punycode",
"constants",
] as const;
/** Builtins Node itself serves ONLY under the node: prefix —
packages/compiler/src/frontend/lowering/lower-builtin-values.ts
+32
View File
@@ -23,6 +23,7 @@ import {
builtinModulesArrayLit,
stdlibGlobalNameOf,
} from "./surfaces.js";
import { CRYPTO_CONSTANTS } from "./crypto-tables.js";
const PROTOTYPE_METHODS: Record<string, readonly string[]> = {
Date: ["getTime", "valueOf", "toISOString", "toJSON"],
@@ -680,3 +681,34 @@ export function lowerBuiltinLoaderValue(lowerer: Lowerer, loc: SrcLoc): IrExpr {
loc,
};
}
/** The deprecated node:constants module object: the runtime merges
* os.constants' groups and fs.constants with crypto.constants, which a
* lifted function answers from the baked table the first time the object
* is built. */
export function builtinConstantsObject(lowerer: Lowerer, loc: SrcLoc): IrExpr {
const name = "%builtin.constants.crypto";
if (!lowerer.builtinCallableValueFns.has(name)) {
lowerer.builtinCallableValueFns.set(name, name);
const fields = Object.entries(CRYPTO_CONSTANTS).map(([key, value]) => ({
key: strLit(key, loc),
value: box(builtinConstLit(value, loc)),
}));
lowerer.liftedFns.push({
name,
params: [],
returnType: DYN,
locals: [],
body: [{ kind: "return", value: { kind: "dynObjLit", fields, type: DYN, loc }, loc }],
loc,
});
}
const table: IrExpr = {
kind: "closure",
fnName: name,
captures: [],
type: { kind: "func", params: [], ret: DYN },
loc,
};
return { kind: "libCall", fn: "constants.object", args: [box(table)], type: DYN, loc };
}
packages/compiler/src/frontend/lowering/lower-exprs.ts
+14
View File
@@ -216,6 +216,7 @@ import {
} from "./class-prototypes.js";
import { isClassCallback } from "./class-callbacks.js";
import {
builtinConstantsObject,
builtinPrototypeMethod,
lowerArrayIsArrayValue,
lowerCheckedPredicateValue,
@@ -1164,6 +1165,17 @@ function lowerExprInner(lowerer: Lowerer, expr: ts.Expression): IrExpr {
if (bi.module === "os" && bi.member === "constants") {
return { kind: "libCall", fn: "os.constants", args: [], type: DYN, loc };
}
// `import { O_RDONLY } from "constants"`: the member of the frozen
// module object.
if (bi.module === "constants") {
return {
kind: "dynKeyGet",
value: builtinConstantsObject(lowerer, loc),
key: { kind: "strLit", value: bi.member, type: STRING, loc },
type: DYN,
loc,
};
}
// tls.rootCertificates: a runtime-valued module constant (the
// cached bundled-CA array) — the one member read that lowers
// to a libCall instead of a baked literal.
@@ -1206,6 +1218,7 @@ function lowerExprInner(lowerer: Lowerer, expr: ts.Expression): IrExpr {
// their access sites only. The CommonJS namespace binding
// (`const lib = require("./lib.js")`) gets the same fence.
const builtinNamespace = lowerer.builtinNamespaceModuleOf(expr);
if (builtinNamespace === "constants") return builtinConstantsObject(lowerer, loc);
if (builtinNamespace !== null) {
const type: IrType = { kind: "moduleNs", moduleId: `builtin:${builtinNamespace}` };
return { kind: "moduleNsRef", moduleId: type.moduleId, type, loc };
@@ -1495,6 +1508,7 @@ function lowerExprInner(lowerer: Lowerer, expr: ts.Expression): IrExpr {
return { kind: "moduleNsRef", moduleId: moduleNs.moduleId, type: moduleNs, loc };
}
const builtinNs = lowerer.builtinNamespaceModuleOf(expr);
if (builtinNs === "constants") return builtinConstantsObject(lowerer, loc);
if (builtinNs !== null) {
const type: IrType = { kind: "moduleNs", moduleId: `builtin:${builtinNs}` };
return { kind: "moduleNsRef", moduleId: type.moduleId, type, loc };
packages/compiler/src/frontend/lowering/lower-modules.ts
+1
View File
@@ -2656,6 +2656,7 @@ export function lowerFileInit(
loc: loc0,
});
}
if (sf === lowerer.entry) header.push(...lowerer.punycodeLinkWarning());
for (const { stmt, dep } of orderedImportsOf(lowerer.program, sf)) {
const npm = lowerer.npmInitActions.get(sf)?.get(stmt);
if (npm !== undefined && npm.length > 0) {
packages/compiler/src/frontend/lowering/lower-stmts.ts
+3 -1
View File
@@ -1557,7 +1557,9 @@ export function lowerVarStatement(lowerer: Lowerer, stmt: ts.VariableStatement):
// Relative requires keep the fence: their module-init-at-position
// semantics needs storage this model doesn't represent.
if (canonicalBuiltinModule(spec) !== null) {
return [];
return canonicalBuiltinModule(spec) === "punycode"
? [lowerer.punycodeRequireWarning(decl)]
: [];
}
lowerer.unsupported(
"SC1090",
packages/compiler/src/frontend/lowering/lowerer.ts
+77
View File
@@ -128,6 +128,7 @@ import { resolve } from "node:path";
import { bindingSource } from "../binding-source.js";
import { tsgoPath } from "../dts-paths.js";
import * as ts from "../ts7/adapter.js";
import { erasedTypeOnlyImport } from "../type-only-imports.js";
import { literalValues } from "../literal-values.js";
import { discriminantOwners, literalUnionArm, type UnionLiteral } from "../union-discriminants.js";
import type { ScrDiagnostic } from "../../diagnostics/diagnostic.js";
@@ -3745,6 +3746,65 @@ export class Lowerer {
for (const fp of parts) if (!indexOf.has(fp.sf)) visit(fp.sf);
}
/** node:punycode's load at a require() call: the module body reports
* DEP0040 when it first runs, unless isInsideNodeModules() finds the
* requiring frame in a dependency, and process.emitWarning delivers the
* report on the next tick. Later loads are cache hits either way. */
punycodeRequireWarning(node: ts.Node): IrStmt {
const loc = locOf(node);
const fromDependency = /[\\/]node_modules[\\/]/.test(node.getSourceFile().fileName);
return {
kind: "exprStmt",
expr: {
kind: "libCall",
fn: "punycode.warn",
args: [
{ kind: "boolLit", value: false, type: BOOL, loc },
{ kind: "boolLit", value: !fromDependency, type: BOOL, loc },
],
type: VOID,
loc,
},
loc,
};
}
/** node:punycode's DEP0040 for a static ES import anywhere in the module
* graph: Node links builtins before any module evaluates, so the report
* comes first, before the program can register a 'warning' listener —
* whichever file imports it (a dependency's ES import has no requiring
* frame either). Prepended to the entry's %init. */
punycodeLinkWarning(): IrStmt[] {
const imports = this.moduleOrder.some((sf) =>
sf.statements.some(
(stmt) =>
((ts.isImportDeclaration(stmt) && !erasedTypeOnlyImport(stmt)) ||
(ts.isExportDeclaration(stmt) && !stmt.isTypeOnly)) &&
stmt.moduleSpecifier !== undefined &&
ts.isStringLiteral(stmt.moduleSpecifier) &&
canonicalBuiltinModule(stmt.moduleSpecifier.text) === "punycode",
),
);
if (!imports) return [];
const loc: SrcLoc = { file: this.entry.fileName, start: 0, end: 0 };
return [
{
kind: "exprStmt",
expr: {
kind: "libCall",
fn: "punycode.warn",
args: [
{ kind: "boolLit", value: true, type: BOOL, loc },
{ kind: "boolLit", value: true, type: BOOL, loc },
],
type: VOID,
loc,
},
loc,
},
];
}
/** The lowering of a CommonJS `require("./local")` occurrence: a call of
* the required module's run-once %init at exactly this statement's
* position — Node's inline evaluation, with the guard supplying the
@@ -3754,6 +3814,7 @@ export class Lowerer {
* their external dependencies by name). Null for everything else
* (builtins load nothing; the rest kept its preflight fence). */
requireInitStmt(spec: string, node: ts.Node): IrStmt | null {
if (canonicalBuiltinModule(spec) === "punycode") return this.punycodeRequireWarning(node);
// Project requires resolve within the program; another bare require can
// be a program-module edge when it names an opted-in --npm-static package
// (one package requiring another — the resolution answered its shipped
@@ -11286,7 +11347,23 @@ export class Lowerer {
return { module: "timers/promises", member: `scheduler.${access.name.text}` };
}
}
// punycode.ucs2.decode / encode: methods of the module's ucs2 object,
// dispatched as module members.
if (access.name.text === "decode" || access.name.text === "encode") {
const owner = access.expression;
const imported = ts.isIdentifier(owner)
? this.builtinImportOf(owner)
: ts.isPropertyAccessExpression(owner)
? this.builtinMemberOf(owner)
: null;
if (imported?.module === "punycode" && imported.member === "ucs2") {
return { module: "punycode", member: `ucs2.${access.name.text}` };
}
}
const module = this.builtinNamespaceModuleOf(access.expression);
// node:constants is one checked-dynamic object: its members read
// through the object itself (builtinConstantsObject), not the tables.
if (module === "constants") return null;
if (module !== null) return { module, member: access.name.text };
// The events module object IS the EventEmitter class, so the class's
// defaultMaxListeners is the module property.
packages/compiler/src/frontend/lowering/surfaces.ts
+17
View File
@@ -1801,6 +1801,22 @@ export const BUILTIN_MODULE_FNS: Record<
valueParams: [{ mode: "required", type: STRING }],
},
},
// node:punycode (deprecated): punycode.js 2.1.0's conversions over
// strings (scr_url.c). ucs2.decode/encode are members of the module's
// ucs2 object, dispatched as module members (builtinMemberOf). The
// DEP0040 report rides the module's load sites (punycodeLoadWarning).
punycode: {
decode: { fn: "punycode.decode", params: [STRING], result: STRING },
encode: { fn: "punycode.encode", params: [STRING], result: STRING },
toASCII: { fn: "punycode.toASCII", params: [STRING], result: STRING },
toUnicode: { fn: "punycode.toUnicode", params: [STRING], result: STRING },
"ucs2.decode": { fn: "punycode.ucs2Decode", params: [STRING], result: arrayOf(F64) },
"ucs2.encode": { fn: "punycode.ucs2Encode", params: [arrayOf(F64)], result: STRING },
},
// node:constants (deprecated) is one frozen object (Object.assign of
// os.constants' groups, fs.constants and crypto.constants): bindings of
// the module read it as a checked-dynamic value (builtinConstantsObject).
constants: {},
// node:readline: createInterface's options are entirely special-cased
// (exactly { input: process.stdin, output?: process.stdout } — see
// lowerBuiltinModuleCall); the entry carries the canonical shape. The
@@ -1930,6 +1946,7 @@ export const BUILTIN_MODULE_CONSTS: Record<
os: { EOL: "\n", devNull: "/dev/null" },
worker_threads: { isMainThread: true, threadId: 0 },
cluster: { isPrimary: true, isMaster: true, isWorker: false },
punycode: { version: "2.1.0" },
// zlib's module-level copies of its constants (every member outside
// BROTLI_*, Node's own loop over zlib.constants).
zlib: Object.fromEntries(
packages/compiler/src/ir/builtin-effects.ts
+7
View File
@@ -466,6 +466,13 @@ export const MAY_THROW_LIB_FNS: ReadonlySet<IrLibFn> = new Set([
"str.encodeUri",
"qs.escape",
"qs.stringify",
// node:punycode's RangeErrors (encode/toASCII only past 32-bit deltas)
// and String.fromCodePoint's for ucs2.encode.
"punycode.decode",
"punycode.encode",
"punycode.toASCII",
"punycode.toUnicode",
"punycode.ucs2Encode",
// The base64 globals: atob/btoa throw the catchable DOMException
// InvalidCharacterError on malformed input; the zero-argument form
// always throws Node's TypeError [ERR_MISSING_ARGS].
packages/compiler/src/ir/builtin-signatures.ts
+17
View File
@@ -559,6 +559,10 @@ export const LIB_FN_SIGS = defineLibFnSignatures({
/** os.constants: the per-thread null-prototype constants object (dlopen,
* errno, signals, priority, UV_UDP_REUSEADDR). Never throws. */
"os.constants": { argTypes: [], result: DYN },
/** The deprecated node:constants object (scr_os.c): the os groups and
* fs.constants merged with the borrowed baked crypto.constants table,
* frozen, one per thread. */
"constants.object": { argTypes: [DYN], result: DYN },
/** umask(2): arg < 0 reads without setting (umask has no read-only form
* — set 0, restore); otherwise sets and answers the PREVIOUS mask.
* Never throws. */
@@ -856,6 +860,19 @@ export const LIB_FN_SIGS = defineLibFnSignatures({
"url.legacyResolve": { argTypes: [DYN, DYN, BOOL], result: STRING },
"url.domainToASCII": { argTypes: [DYN, BOOL], result: STRING },
"url.domainToUnicode": { argTypes: [DYN, BOOL], result: STRING },
/** node:punycode (scr_url.c): punycode.js 2.1.0's string conversions;
* decode/toUnicode (and, past 32-bit deltas, encode/toASCII) throw the
* module's RangeErrors, ucs2Encode String.fromCodePoint's. warn marks the
* module's first load (scr_async_dyn.c) and reports DEP0040 then unless
* the second argument is false: at once for a static ES import, on the
* next tick for a require() call. */
"punycode.decode": { argTypes: [STRING], result: STRING },
"punycode.encode": { argTypes: [STRING], result: STRING },
"punycode.toASCII": { argTypes: [STRING], result: STRING },
"punycode.toUnicode": { argTypes: [STRING], result: STRING },
"punycode.ucs2Decode": { argTypes: [STRING], result: arrayOf(F64) },
"punycode.ucs2Encode": { argTypes: [arrayOf(F64)], result: STRING },
"punycode.warn": { argTypes: [BOOL, BOOL], result: VOID },
"url.fileURLToPathBuffer": { argTypes: [DYN, DYN], result: BYTES_U8 },
"url.pathToFileURLChecked": { argTypes: [DYN, DYN], result: URL_T },
"url.fileURLToPathStr": { argTypes: [STRING], result: STRING },
packages/compiler/src/ir/runtime-features.ts
+2
View File
@@ -43,6 +43,8 @@ export interface RuntimeFeatures {
}
const DYN_ASYNC_LIB_FNS: ReadonlySet<IrLibFn> = new Set([
// DEP0040 rides the dyn-async unit's warning machinery.
"punycode.warn",
"fs.callbackValue",
"fs.callbackCall",
"fs.callbackGlob",
packages/compiler/surface-manifest.json
+57
View File
@@ -466,6 +466,13 @@
"status": "static",
"note": "recognized module (bare and node:-prefixed specifiers)"
},
{
"id": "node-builtin.constants",
"kind": "node-builtin",
"name": "constants",
"status": "static",
"note": "recognized module (bare and node:-prefixed specifiers)"
},
{
"id": "node-builtin.crypto",
"kind": "node-builtin",
@@ -2095,6 +2102,56 @@
"status": "static",
"note": "shared version dictionary containing node and openssl; other components are absent unless defined by the program"
},
{
"id": "node-builtin.punycode",
"kind": "node-builtin",
"name": "punycode",
"status": "static",
"note": "recognized module (bare and node:-prefixed specifiers)"
},
{
"id": "node-builtin.punycode.decode",
"kind": "node-builtin",
"name": "punycode.decode",
"status": "static"
},
{
"id": "node-builtin.punycode.encode",
"kind": "node-builtin",
"name": "punycode.encode",
"status": "static"
},
{
"id": "node-builtin.punycode.toASCII",
"kind": "node-builtin",
"name": "punycode.toASCII",
"status": "static"
},
{
"id": "node-builtin.punycode.toUnicode",
"kind": "node-builtin",
"name": "punycode.toUnicode",
"status": "static"
},
{
"id": "node-builtin.punycode.ucs2.decode",
"kind": "node-builtin",
"name": "punycode.ucs2.decode",
"status": "static"
},
{
"id": "node-builtin.punycode.ucs2.encode",
"kind": "node-builtin",
"name": "punycode.ucs2.encode",
"status": "static"
},
{
"id": "node-builtin.punycode.version",
"kind": "node-builtin",
"name": "punycode.version",
"status": "static",
"note": "constant value read"
},
{
"id": "node-builtin.querystring",
"kind": "node-builtin",
packages/compiler/test/ts7/baselines/order-parity/tests/corpus/6122-punycode-conversions.ts.json
+6
View File
@@ -0,0 +1,6 @@
{
"order": [
"<repo>/tests/corpus/6122-punycode-conversions.ts"
],
"diags": []
}
packages/compiler/test/ts7/baselines/order-parity/tests/corpus/6124-punycode-require-warnings.cjs.json
+6
View File
@@ -0,0 +1,6 @@
{
"order": [
"<repo>/tests/corpus/6124-punycode-require-warnings.cjs"
],
"diags": []
}
packages/compiler/test/ts7/baselines/order-parity/tests/corpus/6125-constants-module.cjs.json
+6
View File
@@ -0,0 +1,6 @@
{
"order": [
"<repo>/tests/corpus/6125-constants-module.cjs"
],
"diags": []
}
packages/compiler/test/ts7/baselines/order-parity/tests/corpus/6126-punycode-dependency/main.cjs.json
+18
View File
@@ -0,0 +1,18 @@
{
"order": [
"<repo>/tests/corpus/6126-punycode-dependency/main.cjs"
],
"diags": [
{
"code": "SC1010",
"message": "the 'puny-dep' module is not supported yet",
"loc": {
"file": "<repo>/tests/corpus/6126-punycode-dependency/main.cjs",
"start": 347,
"end": 376
},
"milestone": "M4",
"hint": "relative imports (./file, ../dir/file), package.json-mediated project imports (#alias via the imports field, self-name references via exports), statically compiled npm packages, and the built-in fs, fs/promises, path, os, url, crypto, zlib, child_process, net, http, tls, https, http2, dgram, dns, util, util/types, string_decoder, querystring, readline, events, stream, stream/promises, stream/consumers, buffer, assert, assert/strict, worker_threads, cluster, tty, async_hooks, timers, timers/promises, diagnostics_channel, perf_hooks, module, and console modules (bare or node:-prefixed) and node:test (node:-prefixed only, like in Node) are supported"
}
]
}
packages/runtime/src/scr_async.h
+5
View File
@@ -212,6 +212,11 @@ void scr_process_on_warning(ScrDyn *fn);
void scr_process_off_warning(ScrDyn *fn);
void scr_process_emit_warning(ScrDyn *args);
void scr_emit_warning(const char *name, const char *code, ScrStr *message);
/* node:punycode's first load (scr_async_dyn.c): later calls are module
* cache hits. The first reports DEP0040 when `report` (false for a require
* made from node_modules) — at once for a static ES import, which Node
* links before any module evaluates, otherwise on the next tick. */
void scr_punycode_warn(bool immediate, bool report);
/* scr_emit_warning once a 'warning' listener exists (scr_lib.c; set by the
* listener registration): units that cannot link the dyn-async unit route
* their warnings through it, and print the default report otherwise. */
packages/runtime/src/scr_async_dyn.c
+21
View File
@@ -813,6 +813,27 @@ void scr_warning_emit_deferred(ScrDyn *w) {
scr_warning_defer(w, scr_warning_tick);
}
static SCR_TL bool scr_punycode_loaded = false;
void scr_punycode_warn(bool immediate, bool report) {
if (scr_punycode_loaded) return;
scr_punycode_loaded = true;
if (!report) return;
static const char text[] =
"The `punycode` module is deprecated. Please use a userland alternative instead.";
ScrStr *message = scr_str_new(text, sizeof text - 1);
ScrStr *code = scr_str_new("DEP0040", 7);
ScrDyn *w = scr_warning_new(message, "DeprecationWarning", code);
if (immediate) {
if (!scr_process_deprecation_flag_on(0)) scr_warning_dispatch(w);
} else {
scr_warning_emit_deferred(w);
}
scr_dyn_release(w);
scr_str_release(code);
scr_str_release(message);
}
/* The console members' warnings. Node's emitWarning emits on
* process.nextTick, so these queue in order and each rides one raw tick
* on the shared tick queue; any left at exit are dropped with it. */
packages/runtime/src/scr_os.c
+194
View File
@@ -27,6 +27,8 @@
#include <sys/utsname.h>
#include <unistd.h>
#endif
#include <fcntl.h>
#include <sys/stat.h>
#ifdef __linux__
#include <sys/sysinfo.h>
#endif
@@ -1286,3 +1288,195 @@ ScrDyn *scr_os_constants(void) {
scr_atexit(scr_os_constants_clear);
return scr_dyn_retain(os);
}
/* ── node:constants (deprecated, DEP0008) ──────────────────────────────
* Node's lib/constants.js: Object.assign({}, os.constants.dlopen, errno,
* priority, signals, fs.constants, crypto.constants), frozen, with an
* ordinary prototype. The fs group follows node_constants.cc's
* DefineFsConstants order and guards (libuv's UV_* values are fixed on
* POSIX); crypto_table (borrowed) is a compiler-emitted function that
* answers the baked crypto.constants table, called only on first use. One
* object per thread. */
#define SCR_FS_CONSTANT(o, name) scr_os_constant((o), #name, (double)(name))
static void scr_fs_constants(ScrDyn *o) {
scr_os_constant(o, "UV_FS_SYMLINK_DIR", 1);
scr_os_constant(o, "UV_FS_SYMLINK_JUNCTION", 2);
#ifdef O_RDONLY
SCR_FS_CONSTANT(o, O_RDONLY);
#endif
#ifdef O_WRONLY
SCR_FS_CONSTANT(o, O_WRONLY);
#endif
#ifdef O_RDWR
SCR_FS_CONSTANT(o, O_RDWR);
#endif
scr_os_constant(o, "UV_DIRENT_UNKNOWN", 0);
scr_os_constant(o, "UV_DIRENT_FILE", 1);
scr_os_constant(o, "UV_DIRENT_DIR", 2);
scr_os_constant(o, "UV_DIRENT_LINK", 3);
scr_os_constant(o, "UV_DIRENT_FIFO", 4);
scr_os_constant(o, "UV_DIRENT_SOCKET", 5);
scr_os_constant(o, "UV_DIRENT_CHAR", 6);
scr_os_constant(o, "UV_DIRENT_BLOCK", 7);
#ifdef S_IFMT
SCR_FS_CONSTANT(o, S_IFMT);
#endif
#ifdef S_IFREG
SCR_FS_CONSTANT(o, S_IFREG);
#endif
#ifdef S_IFDIR
SCR_FS_CONSTANT(o, S_IFDIR);
#endif
#ifdef S_IFCHR
SCR_FS_CONSTANT(o, S_IFCHR);
#endif
#ifdef S_IFBLK
SCR_FS_CONSTANT(o, S_IFBLK);
#endif
#ifdef S_IFIFO
SCR_FS_CONSTANT(o, S_IFIFO);
#endif
#ifdef S_IFLNK
SCR_FS_CONSTANT(o, S_IFLNK);
#endif
#ifdef S_IFSOCK
SCR_FS_CONSTANT(o, S_IFSOCK);
#endif
#ifdef O_CREAT
SCR_FS_CONSTANT(o, O_CREAT);
#endif
#ifdef O_EXCL
SCR_FS_CONSTANT(o, O_EXCL);
#endif
/* libuv's Windows-only open flags are 0 everywhere else. */
scr_os_constant(o, "UV_FS_O_FILEMAP", 0);
scr_os_constant(o, "UV_FS_O_TEMPORARY", 0);
scr_os_constant(o, "UV_FS_O_SHORT_LIVED", 0);
scr_os_constant(o, "UV_FS_O_SEQUENTIAL", 0);
scr_os_constant(o, "UV_FS_O_RANDOM", 0);
#ifdef O_NOCTTY
SCR_FS_CONSTANT(o, O_NOCTTY);
#endif
#ifdef O_TRUNC
SCR_FS_CONSTANT(o, O_TRUNC);
#endif
#ifdef O_APPEND
SCR_FS_CONSTANT(o, O_APPEND);
#endif
#ifdef O_DIRECTORY
SCR_FS_CONSTANT(o, O_DIRECTORY);
#endif
#ifdef O_NOATIME
SCR_FS_CONSTANT(o, O_NOATIME);
#endif
#ifdef O_NOFOLLOW
SCR_FS_CONSTANT(o, O_NOFOLLOW);
#endif
#ifdef O_SYNC
SCR_FS_CONSTANT(o, O_SYNC);
#endif
#ifdef O_DSYNC
SCR_FS_CONSTANT(o, O_DSYNC);
#endif
#ifdef O_SYMLINK
SCR_FS_CONSTANT(o, O_SYMLINK);
#endif
#ifdef O_DIRECT
SCR_FS_CONSTANT(o, O_DIRECT);
#endif
#ifdef O_NONBLOCK
SCR_FS_CONSTANT(o, O_NONBLOCK);
#endif
#ifdef S_IRWXU
SCR_FS_CONSTANT(o, S_IRWXU);
#endif
#ifdef S_IRUSR
SCR_FS_CONSTANT(o, S_IRUSR);
#endif
#ifdef S_IWUSR
SCR_FS_CONSTANT(o, S_IWUSR);
#endif
#ifdef S_IXUSR
SCR_FS_CONSTANT(o, S_IXUSR);
#endif
#ifdef S_IRWXG
SCR_FS_CONSTANT(o, S_IRWXG);
#endif
#ifdef S_IRGRP
SCR_FS_CONSTANT(o, S_IRGRP);
#endif
#ifdef S_IWGRP
SCR_FS_CONSTANT(o, S_IWGRP);
#endif
#ifdef S_IXGRP
SCR_FS_CONSTANT(o, S_IXGRP);
#endif
#ifdef S_IRWXO
SCR_FS_CONSTANT(o, S_IRWXO);
#endif
#ifdef S_IROTH
SCR_FS_CONSTANT(o, S_IROTH);
#endif
#ifdef S_IWOTH
SCR_FS_CONSTANT(o, S_IWOTH);
#endif
#ifdef S_IXOTH
SCR_FS_CONSTANT(o, S_IXOTH);
#endif
#ifdef F_OK
SCR_FS_CONSTANT(o, F_OK);
#endif
#ifdef R_OK
SCR_FS_CONSTANT(o, R_OK);
#endif
#ifdef W_OK
SCR_FS_CONSTANT(o, W_OK);
#endif
#ifdef X_OK
SCR_FS_CONSTANT(o, X_OK);
#endif
scr_os_constant(o, "UV_FS_COPYFILE_EXCL", 1);
scr_os_constant(o, "COPYFILE_EXCL", 1);
scr_os_constant(o, "UV_FS_COPYFILE_FICLONE", 2);
scr_os_constant(o, "COPYFILE_FICLONE", 2);
scr_os_constant(o, "UV_FS_COPYFILE_FICLONE_FORCE", 4);
scr_os_constant(o, "COPYFILE_FICLONE_FORCE", 4);
}
#undef SCR_FS_CONSTANT
static SCR_TL ScrDyn *scr_node_constants_value;
static void scr_node_constants_clear(void) {
ScrDyn *value = scr_node_constants_value;
scr_node_constants_value = NULL;
scr_dyn_release(value);
}
ScrDyn *scr_node_constants(const ScrDyn *crypto_table) {
if (scr_node_constants_value) return scr_dyn_retain(scr_node_constants_value);
ScrDyn *crypto = scr_dyn_call(crypto_table, NULL, 0, "constants");
if (!crypto) return NULL;
ScrDyn *o = scr_dyn_new_obj();
scr_os_dlopen_constants(o);
scr_os_errno_constants(o);
scr_os_constant(o, "PRIORITY_LOW", SCR_OS_PRIORITY_LOW);
scr_os_constant(o, "PRIORITY_BELOW_NORMAL", SCR_OS_PRIORITY_BELOW_NORMAL);
scr_os_constant(o, "PRIORITY_NORMAL", SCR_OS_PRIORITY_NORMAL);
scr_os_constant(o, "PRIORITY_ABOVE_NORMAL", SCR_OS_PRIORITY_ABOVE_NORMAL);
scr_os_constant(o, "PRIORITY_HIGH", SCR_OS_PRIORITY_HIGH);
scr_os_constant(o, "PRIORITY_HIGHEST", SCR_OS_PRIORITY_HIGHEST);
scr_os_signal_constants(o);
scr_fs_constants(o);
if (crypto && crypto->kind == SCR_DYN_OBJ) {
for (size_t i = 0; i < crypto->v.obj.len; i++) {
const ScrDynEntry *entry = &crypto->v.obj.entries[i];
if (entry->accessor || !entry->value) continue;
scr_dyn_obj_set(o, entry->key, entry->key_len, scr_dyn_retain(entry->value));
}
}
scr_dyn_release(crypto);
scr_dyn_release(scr_dyn_freeze(o));
scr_node_constants_value = o;
scr_atexit(scr_node_constants_clear);
return scr_dyn_retain(o);
}
packages/runtime/src/scr_os.h
+4
View File
@@ -34,6 +34,10 @@ ScrStr *scr_os_cpus_model(double i);
double scr_os_cpus_field(double i, double field);
/* os.constants: the per-thread null-prototype constants object (+1). */
ScrDyn *scr_os_constants(void);
/* The deprecated node:constants module object (scr_os.c): the os groups,
* fs.constants and the crypto table the borrowed function answers, merged
* and frozen (+1). */
ScrDyn *scr_node_constants(const ScrDyn *crypto_table);
double scr_os_available_parallelism(void); /* uv_available_parallelism: CPUs this process may use, >= 1 */
/* os.userInfo()'s field trio (pw_name / pw_shell / pw_dir — the passwd
* home, not the $HOME cascade). +1 fresh; abort on lookup failure. */
packages/runtime/src/scr_url.c
+339
View File
@@ -1988,3 +1988,342 @@ ScrBytes *scr_url_checked_to_path_buffer(const ScrDyn *value, const ScrDyn *opti
scr_url_release(url);
return result;
}
/* ── node:punycode (deprecated, DEP0040) ───────────────────────────────
* A port of the punycode.js 2.1.0 that Node 24 bundles as lib/punycode.js.
* It works on UTF-16 code units: ucs2.decode pairs surrogates and keeps
* lone ones, which is exactly a code-point walk over WTF-8 storage.
* Errors throw the module's RangeErrors ("Overflow: ...", "Illegal input
* >= 0x80 ...", "Invalid input"); decoded code points past U+10FFFF throw
* String.fromCodePoint's "Invalid code point N". All inputs are borrowed;
* results are +1, or NULL with the error pending. The DEP0040 report is
* scr_punycode_warn (scr_async_dyn.c). */
enum {
PC_MAX_INT = 2147483647,
PC_BASE = 36,
PC_TMIN = 1,
PC_TMAX = 26,
PC_SKEW = 38,
PC_DAMP = 700,
PC_INITIAL_BIAS = 72,
PC_INITIAL_N = 128
};
static void pc_error(const char *message) {
scr_throw_error_msg(SCR_ERR_RANGE, message, strlen(message));
}
#define PC_OVERFLOW "Overflow: input needs wider integers to process"
#define PC_NOT_BASIC "Illegal input >= 0x80 (not a basic code point)"
#define PC_INVALID "Invalid input"
typedef struct {
uint32_t *data;
size_t len, cap;
} PcPoints;
static void pc_push(PcPoints *p, uint32_t cp) {
if (p->len == p->cap) {
size_t cap = p->cap ? p->cap * 2 : 16;
uint32_t *grown = realloc(p->data, cap * sizeof *grown);
if (!grown) scr_trap("scriptc: out of memory\n");
p->data = grown;
p->cap = cap;
}
p->data[p->len++] = cp;
}
/* ucs2decode over WTF-8 storage: a pair is one 4-byte sequence, a lone
* surrogate one 3-byte sequence carrying its own value. */
static void pc_code_points(const char *s, size_t n, PcPoints *out) {
const unsigned char *b = (const unsigned char *)s;
for (size_t i = 0; i < n;) {
uint32_t c = b[i];
if (c < 0x80) {
i += 1;
} else if (c < 0xE0 && i + 1 < n) {
c = ((c & 0x1F) << 6) | (b[i + 1] & 0x3F);
i += 2;
} else if (c < 0xF0 && i + 2 < n) {
c = ((c & 0x0F) << 12) | ((uint32_t)(b[i + 1] & 0x3F) << 6) | (b[i + 2] & 0x3F);
i += 3;
} else if (i + 3 < n) {
c = ((c & 0x07) << 18) | ((uint32_t)(b[i + 1] & 0x3F) << 12) |
((uint32_t)(b[i + 2] & 0x3F) << 6) | (b[i + 3] & 0x3F);
i += 4;
} else {
i += 1;
}
pc_push(out, c);
}
}
static int64_t pc_adapt(int64_t delta, int64_t points, bool first) {
int64_t k = 0;
delta = first ? delta / PC_DAMP : delta >> 1;
delta += delta / points;
for (; delta > ((PC_BASE - PC_TMIN) * PC_TMAX) >> 1; k += PC_BASE) delta /= PC_BASE - PC_TMIN;
return k + (PC_BASE - PC_TMIN + 1) * delta / (delta + PC_SKEW);
}
static int64_t pc_basic_to_digit(unsigned char c) {
if (c >= 0x30 && c < 0x3A) return 26 + (c - 0x30);
if (c >= 0x41 && c < 0x5B) return c - 0x41;
if (c >= 0x61 && c < 0x7B) return c - 0x61;
return PC_BASE;
}
static char pc_digit_to_basic(int64_t digit) {
return (char)(digit + 22 + 75 * (digit < 26));
}
/* String.fromCodePoint(...points). */
static ScrStr *pc_from_points(const PcPoints *p) {
for (size_t i = 0; i < p->len; i++) {
if (p->data[i] > 0x10FFFF) {
char message[48];
int size = snprintf(message, sizeof message, "Invalid code point %u", (unsigned)p->data[i]);
scr_throw_error_msg(SCR_ERR_RANGE, message, (size_t)size);
return NULL;
}
}
char *out = malloc(p->len * 4 + 1);
if (!out) scr_trap("scriptc: out of memory\n");
size_t offset = 0;
for (size_t i = 0; i < p->len; i++) {
uint32_t point = p->data[i];
if (point >= 0xD800 && point <= 0xDBFF && i + 1 < p->len && p->data[i + 1] >= 0xDC00 &&
p->data[i + 1] <= 0xDFFF) {
point = 0x10000 + ((point - 0xD800) << 10) + (p->data[i + 1] - 0xDC00);
i++;
}
offset += scr_wtf8_put(out + offset, point);
}
ScrStr *result = scr_str_new(out, offset);
free(out);
return result;
}
/* punycode.decode over the bytes of one label. Any non-ASCII byte fails
* where its first code unit would: before the last delimiter it is not
* basic, after it it is no digit — so byte positions stand in for code
* unit positions until the first error. */
static ScrStr *pc_decode(const char *s, size_t n) {
PcPoints output = {0};
size_t basic = 0;
for (size_t k = n; k > 0; k--) {
if (s[k - 1] == '-') {
basic = k - 1;
break;
}
}
ScrStr *result = NULL;
for (size_t j = 0; j < basic; j++) {
if ((unsigned char)s[j] >= 0x80) {
pc_error(PC_NOT_BASIC);
goto done;
}
pc_push(&output, (unsigned char)s[j]);
}
int64_t i = 0, cp = PC_INITIAL_N, bias = PC_INITIAL_BIAS;
for (size_t index = basic > 0 ? basic + 1 : 0; index < n;) {
int64_t old = i;
for (int64_t w = 1, k = PC_BASE;; k += PC_BASE) {
if (index >= n) {
pc_error(PC_INVALID);
goto done;
}
int64_t digit = pc_basic_to_digit((unsigned char)s[index++]);
if (digit >= PC_BASE) {
pc_error(PC_INVALID);
goto done;
}
if (digit > (PC_MAX_INT - i) / w) {
pc_error(PC_OVERFLOW);
goto done;
}
i += digit * w;
int64_t t = k <= bias ? PC_TMIN : (k >= bias + PC_TMAX ? PC_TMAX : k - bias);
if (digit < t) break;
if (w > PC_MAX_INT / (PC_BASE - t)) {
pc_error(PC_OVERFLOW);
goto done;
}
w *= PC_BASE - t;
}
int64_t out = (int64_t)output.len + 1;
bias = pc_adapt(i - old, out, old == 0);
if (i / out > PC_MAX_INT - cp) {
pc_error(PC_OVERFLOW);
goto done;
}
cp += i / out;
i %= out;
pc_push(&output, 0);
memmove(output.data + i + 1, output.data + i, (output.len - 1 - (size_t)i) * sizeof *output.data);
output.data[i++] = (uint32_t)cp;
}
result = pc_from_points(&output);
done:
free(output.data);
return result;
}
/* punycode.encode of one label's code points, appended to buf. */
static bool pc_encode(const PcPoints *input, ScrJsonBuf *buf) {
int64_t cp = PC_INITIAL_N, delta = 0, bias = PC_INITIAL_BIAS;
size_t basic_length = 0;
for (size_t k = 0; k < input->len; k++) {
if (input->data[k] < 0x80) {
scr_jb_putc(buf, (char)input->data[k]);
basic_length++;
}
}
size_t handled = basic_length;
if (basic_length) scr_jb_putc(buf, '-');
while (handled < input->len) {
int64_t m = PC_MAX_INT;
for (size_t k = 0; k < input->len; k++)
if (input->data[k] >= cp && input->data[k] < m) m = input->data[k];
int64_t handled_plus_one = (int64_t)handled + 1;
if (m - cp > (PC_MAX_INT - delta) / handled_plus_one) {
pc_error(PC_OVERFLOW);
return false;
}
delta += (m - cp) * handled_plus_one;
cp = m;
for (size_t k = 0; k < input->len; k++) {
int64_t current = input->data[k];
if (current < cp && ++delta > PC_MAX_INT) {
pc_error(PC_OVERFLOW);
return false;
}
if (current == cp) {
int64_t q = delta;
for (int64_t j = PC_BASE;; j += PC_BASE) {
int64_t t = j <= bias ? PC_TMIN : (j >= bias + PC_TMAX ? PC_TMAX : j - bias);
if (q < t) break;
scr_jb_putc(buf, pc_digit_to_basic(t + (q - t) % (PC_BASE - t)));
q = (q - t) / (PC_BASE - t);
}
scr_jb_putc(buf, pc_digit_to_basic(q));
bias = pc_adapt(delta, handled_plus_one, handled == basic_length);
delta = 0;
handled++;
}
}
delta++;
cp++;
}
return true;
}
ScrStr *scr_punycode_decode(const ScrStr *input) { return pc_decode(input->data, input->len); }
ScrStr *scr_punycode_encode(const ScrStr *input) {
PcPoints points = {0};
pc_code_points(input->data, input->len, &points);
ScrJsonBuf buf;
scr_jb_init(&buf);
bool ok = pc_encode(&points, &buf);
free(points.data);
ScrStr *result = scr_jb_finish(&buf);
if (ok) return result;
scr_str_release(result);
return NULL;
}
ScrArr *scr_punycode_ucs2_decode(const ScrStr *input) {
PcPoints points = {0};
pc_code_points(input->data, input->len, &points);
ScrArr *out = scr_arr_new(SCR_ELEM_F64, points.len);
for (size_t i = 0; i < points.len; i++) scr_arr_push_f64(out, (double)points.data[i]);
free(points.data);
return out;
}
/* mapDomain: only the part after the first '@' maps (and only up to a
* second one — punycode.js keeps parts[0] and parts[1]); the RFC 3490
* separators U+002E, U+3002, U+FF0E and U+FF61 all split labels, and the
* labels rejoin with '.'. */
static ScrStr *pc_map_domain(const ScrStr *input, bool to_ascii) {
const char *s = input->data;
size_t n = input->len;
ScrJsonBuf buf;
scr_jb_init(&buf);
const char *at = memchr(s, '@', n);
if (at) {
scr_jb_write(&buf, s, (size_t)(at - s) + 1);
const char *rest = at + 1;
const char *second = memchr(rest, '@', n - (size_t)(rest - s));
n = (second ? (size_t)(second - rest) : n - (size_t)(rest - s));
s = rest;
}
size_t start = 0;
bool ok = true;
for (size_t i = 0; ok && i <= n;) {
size_t sep = 0;
if (i == n) sep = 1;
else if (s[i] == '.') sep = 1;
else if (i + 2 < n && (unsigned char)s[i] == 0xE3 && (unsigned char)s[i + 1] == 0x80 &&
(unsigned char)s[i + 2] == 0x82)
sep = 3;
else if (i + 2 < n && (unsigned char)s[i] == 0xEF &&
(((unsigned char)s[i + 1] == 0xBC && (unsigned char)s[i + 2] == 0x8E) ||
((unsigned char)s[i + 1] == 0xBD && (unsigned char)s[i + 2] == 0xA1)))
sep = 3;
if (!sep) {
i++;
continue;
}
const char *label = s + start;
size_t len = i - start;
bool ascii = true;
for (size_t k = 0; k < len; k++) ascii = ascii && (unsigned char)label[k] < 0x80;
if (to_ascii && !ascii) {
PcPoints points = {0};
pc_code_points(label, len, &points);
scr_jb_puts(&buf, "xn--");
ok = pc_encode(&points, &buf);
free(points.data);
} else if (!to_ascii && len >= 4 && memcmp(label, "xn--", 4) == 0) {
/* string.slice(4).toLowerCase(): ASCII letters fold; the one other
* code point whose lowercase is ASCII is KELVIN SIGN (U+212A → k).
* Any other non-ASCII stays non-ASCII and fails in decode either
* way, at the same position. */
ScrJsonBuf lower;
scr_jb_init(&lower);
for (size_t k = 4; k < len; k++) {
unsigned char c = (unsigned char)label[k];
if (c == 0xE2 && k + 2 < len && (unsigned char)label[k + 1] == 0x84 &&
(unsigned char)label[k + 2] == 0xAA) {
scr_jb_putc(&lower, 'k');
k += 2;
} else {
scr_jb_putc(&lower, (char)(c >= 'A' && c <= 'Z' ? c + 32 : c));
}
}
ScrStr *folded = scr_jb_finish(&lower);
ScrStr *decoded = pc_decode(folded->data, folded->len);
scr_str_release(folded);
if (decoded) {
scr_jb_put_str(&buf, decoded);
scr_str_release(decoded);
} else {
ok = false;
}
} else {
scr_jb_write(&buf, label, len);
}
if (ok && i < n) scr_jb_putc(&buf, '.');
i += sep;
start = i;
}
ScrStr *result = scr_jb_finish(&buf);
if (ok) return result;
scr_str_release(result);
return NULL;
}
ScrStr *scr_punycode_to_ascii(const ScrStr *input) { return pc_map_domain(input, true); }
ScrStr *scr_punycode_to_unicode(const ScrStr *input) { return pc_map_domain(input, false); }
packages/runtime/src/scr_url.h
+9
View File
@@ -53,6 +53,15 @@ ScrDyn *scr_url_legacy_parse(const ScrDyn *url, const ScrDyn *parse_query, const
ScrStr *scr_url_legacy_format(const ScrDyn *object, const ScrDyn *options, bool warn);
ScrStr *scr_url_legacy_resolve(const ScrDyn *from, const ScrDyn *to, bool warn);
ScrStr *scr_url_domain_to_ascii(const ScrDyn *domain, bool missing);
/* node:punycode (scr_url.c): punycode.js 2.1.0's encode/decode, toASCII/
* toUnicode (mapDomain over RFC 3490 separators) and ucs2.decode (UTF-16
* code points as an f64 array). Borrowed input; +1, or NULL with the
* module's RangeError pending. */
ScrStr *scr_punycode_decode(const ScrStr *input);
ScrStr *scr_punycode_encode(const ScrStr *input);
ScrStr *scr_punycode_to_ascii(const ScrStr *input);
ScrStr *scr_punycode_to_unicode(const ScrStr *input);
ScrArr *scr_punycode_ucs2_decode(const ScrStr *input);
ScrStr *scr_url_domain_to_unicode(const ScrDyn *domain, bool missing);
ScrStr *scr_url_to_path(ScrUrl *u); /* +1, or throws */
ScrStr *scr_url_str_to_path(ScrStr *s); /* +1, or throws */
tests/corpus/6122-punycode-conversions.ts
+45
View File
@@ -0,0 +1,45 @@
// @exit: 1
// @no-node-shims
// node:punycode's conversions (punycode.js 2.1.0): encode/decode of bare
// labels, toASCII/toUnicode over domains and email addresses (only the part
// after '@' maps; every RFC 3490 separator splits labels), ucs2 code point
// arrays with astral pairs and lone surrogates, and the RangeErrors of
// malformed input. A static import links the module before anything runs,
// so its DEP0040 report precedes the program and no 'warning' listener the
// program registers sees it (the Node side runs without the preload shim,
// which would delay that report past the program's first turn). The report
// carries the process id, so the @exit directive keeps stderr out of the
// comparison.
import * as punycode from "node:punycode";
process.on("warning", (w) => console.log("listener", w.name));
const show = (label: string, f: () => unknown): void => {
try {
console.log(label, JSON.stringify(f()));
} catch (e) {
const err = e as Error;
console.log(label, "threw", err instanceof RangeError, err.name, err.message);
}
};
console.log(punycode.version);
for (const s of ["mañana", "bücher", "ü", "abc", "", "-> $1.00 <-", "ليهمابتكلموشعربي؟", "他们为什么不说中文", "😀x", "MaJiでKoiする5秒前"]) {
show(`encode ${s}`, () => punycode.encode(s));
show(`roundtrip ${s}`, () => punycode.decode(punycode.encode(s)));
}
for (const s of ["maana-pta", "bcher-kva", "tda", "abc-", "-abc", "ABC-kva", "a-b-c-kva", "ü-kva", "kva-ü", "99999999999", "zzzzzzzzzzzzz", "a-", "-"]) {
show(`decode ${s}`, () => punycode.decode(s));
}
for (const d of ["mañana.com", "bücher.de", "example.com", "", "a..b", "ex@mple@bücher.de", "user@bücher.de", "日本。jp", "a．b｡c", "ÄÖÜ.test", "a.b."]) {
show(`toASCII ${d}`, () => punycode.toASCII(d));
}
for (const d of ["xn--maana-pta.com", "xn--bcher-kva.de", "XN--bcher-kva.de", "xn--BCHER-KVA.de", "example.com", "user@xn--bcher-kva.de", "xn--.x", "xn--ü.x", "xn--zzzzzzzzzzzzz.x", "xn--a-Kva.x"]) {
show(`toUnicode ${d}`, () => punycode.toUnicode(d));
}
show("ucs2.decode", () => punycode.ucs2.decode("aé😀\ud800b\udc00"));
show("ucs2.encode", () => punycode.ucs2.encode([97, 0xe9, 0x1f600, 0xd800, 0xdc00, 98]));
show("ucs2.encode join", () => punycode.ucs2.encode([0xd83d, 0xde00]).length);
show("ucs2.encode range", () => punycode.ucs2.encode([0x110000]));
show("ucs2 roundtrip", () => punycode.ucs2.encode(punycode.ucs2.decode("x😀\udfffy")) === "x😀\udfffy");
process.exitCode = 1;
tests/corpus/6124-punycode-require-warnings.cjs
+32
View File
@@ -0,0 +1,32 @@
// DEP0040 for require("punycode"): reported once, on the next tick after
// the requiring turn, to the stderr report and the 'warning' listeners;
// require("node:punycode") is the same module and does not report again.
// process.noDeprecation drops the report and the listener call. The program
// re-runs itself per case and prints each child's output with process ids
// normalized.
"use strict";
const { spawnSync } = require("child_process");
const mode = process.argv[2];
if (mode === "direct") {
process.on("warning", (w) => console.log("listener", w.name, w.code, w.message));
process.nextTick(() => console.log("earlier tick"));
console.log("before require");
const direct = require("punycode");
console.log("after require", direct.toASCII("bücher.de"));
const prefixed = require("node:punycode");
console.log("prefixed", prefixed.toUnicode("xn--bcher-kva.de"));
setTimeout(() => console.log("timer"), 0);
} else if (mode === "silenced") {
process.on("warning", (w) => console.log("listener", w.code));
process.noDeprecation = true;
const silencedModule = require("punycode");
console.log("silenced", silencedModule.encode("ü"));
} else {
for (const child of ["direct", "silenced"]) {
const run = spawnSync(process.execPath, [process.argv[1], child], { encoding: "utf8" });
console.log(`== ${child}: status ${run.status}`);
console.log(run.stdout);
console.log(run.stderr.replace(/\(node:\d+\)/g, "(node:<pid>)"));
}
}
tests/corpus/6125-constants-module.cjs
+26
View File
@@ -0,0 +1,26 @@
// node:constants (deprecated in documentation only, so no warning): one
// frozen object merging os.constants' dlopen, errno, priority and signals
// groups with fs.constants and crypto.constants, in that order. Members
// read like the owning modules' constants, and a constant the platform
// lacks is absent (graceful-fs probes O_SYMLINK with hasOwnProperty).
const constants = require("constants");
const nodeConstants = require("node:constants");
const os = require("os");
console.log(typeof constants, Object.isFrozen(constants), constants === nodeConstants);
console.log(constants.O_RDONLY, constants.O_WRONLY, constants.O_RDWR);
console.log(constants.O_WRONLY | constants.O_CREAT | constants.O_TRUNC);
console.log(constants.S_IFMT, constants.S_IFREG, constants.S_IFDIR, constants.S_IRWXU);
console.log(constants.F_OK, constants.R_OK, constants.W_OK, constants.X_OK);
console.log(constants.COPYFILE_EXCL, constants.UV_FS_COPYFILE_FICLONE, constants.UV_DIRENT_DIR);
console.log(constants.SIGTERM === os.constants.signals.SIGTERM, constants.SIGINT);
console.log(constants.ENOENT === os.constants.errno.ENOENT, constants.EACCES);
console.log(constants.PRIORITY_HIGH, constants.PRIORITY_LOW, constants.RTLD_LAZY);
console.log(constants.TLS1_2_VERSION, constants.RSA_PKCS1_PADDING, typeof constants.defaultCoreCipherList);
console.log(constants.hasOwnProperty("O_SYMLINK") === ("O_SYMLINK" in constants));
console.log(constants.hasOwnProperty("O_RDWR"), "SIGKILL" in constants, "UV_UDP_REUSEADDR" in constants);
console.log(constants.hasOwnProperty("NOT_A_CONSTANT"), "dlopen" in constants, "errno" in constants);
const keys = Object.keys(constants);
console.log(keys.indexOf("RTLD_LAZY") < keys.indexOf("E2BIG"), keys.indexOf("E2BIG") < keys.indexOf("PRIORITY_LOW"));
console.log(keys.indexOf("PRIORITY_LOW") < keys.indexOf("SIGHUP"), keys.indexOf("SIGHUP") < keys.indexOf("O_RDONLY"));
console.log(keys.indexOf("O_RDONLY") < keys.indexOf("OPENSSL_VERSION_NUMBER"), keys.includes("POINT_CONVERSION_HYBRID"));
tests/corpus/6126-punycode-dependency/main.cjs
+12
View File
@@ -0,0 +1,12 @@
// A dependency under node_modules that requires node:punycode: Node's
// isInsideNodeModules() rule skips DEP0040 for it, so nothing reaches stderr
// or the 'warning' listeners, and the program's own later require of the
// cached module does not report either.
"use strict";
process.on("warning", (w) => console.log("listener", w.code));
const toAscii = require("puny-dep");
const punycode = require("punycode");
console.log(toAscii("mañana.com"), toAscii("example.org"));
console.log(punycode.toUnicode("xn--maana-pta.com"));
setTimeout(() => console.log("timer"), 0);
tests/corpus/6126-punycode-dependency/node_modules/puny-dep/index.d.ts
Generated Vendored
+2
View File
@@ -0,0 +1,2 @@
declare function toAscii(domain: string): string;
export = toAscii;
tests/corpus/6126-punycode-dependency/node_modules/puny-dep/index.js
Generated Vendored
+8
View File
@@ -0,0 +1,8 @@
// A dependency requiring node:punycode: Node's isInsideNodeModules() rule
// skips DEP0040 for it.
"use strict";
const punycode = require("punycode");
module.exports = function toAscii(domain) {
return punycode.toASCII(domain);
};
tests/corpus/6126-punycode-dependency/node_modules/puny-dep/package.json
Generated Vendored
+5
View File
@@ -0,0 +1,5 @@
{
"name": "puny-dep",
"version": "1.0.0",
"main": "index.js"
}