Add react-native support (#63)
Some checks failed
Publish / build (push) Waiting to run
Publish / publish-crate (push) Blocked by required conditions
Publish / publish-npm (push) Blocked by required conditions
Publish / publish-pypi (push) Blocked by required conditions
Check / build (push) Has been cancelled
Some checks failed
Publish / build (push) Waiting to run
Publish / publish-crate (push) Blocked by required conditions
Publish / publish-npm (push) Blocked by required conditions
Publish / publish-pypi (push) Blocked by required conditions
Check / build (push) Has been cancelled
Reviewed-on: https://home.schmelczer.dev/git/git/andras/reconcile/pulls/63
This commit is contained in:
parent
08e7d824f4
commit
a8fbac6934
10 changed files with 907 additions and 339 deletions
3
.gitignore
vendored
3
.gitignore
vendored
|
|
@ -10,5 +10,8 @@ node_modules
|
|||
# WebPack build output
|
||||
dist
|
||||
|
||||
# Generated wasm-bindgen bundler + wasm2js output for the React Native build
|
||||
pkg-rn
|
||||
|
||||
# Python virtual environment
|
||||
.venv
|
||||
|
|
|
|||
|
|
@ -80,6 +80,13 @@ console.log(result.text); // "Hi beautiful world"
|
|||
|
||||
See the [example website source](examples/website/src/index.ts) for a more complex example, or the [advanced examples document](docs/advanced-ts.md).
|
||||
|
||||
#### React Native (Hermes)
|
||||
|
||||
React Native's default engine, Hermes, does not expose a runtime `WebAssembly`
|
||||
global, so the WebAssembly build cannot run there. For React Native, the package
|
||||
ships a pure-JavaScript build produced by [Binaryen's `wasm2js`](https://github.com/WebAssembly/binaryen)
|
||||
via its `react-native` entry point.
|
||||
|
||||
### Python
|
||||
|
||||
Install via uv or pip:
|
||||
|
|
|
|||
27
reconcile-js/package-lock.json
generated
27
reconcile-js/package-lock.json
generated
|
|
@ -10,6 +10,7 @@
|
|||
"license": "MIT",
|
||||
"devDependencies": {
|
||||
"@types/jest": "^30.0.0",
|
||||
"binaryen": "^123.0.0",
|
||||
"jest": "^30.3.0",
|
||||
"prettier": "^3.8.1",
|
||||
"reconcile-text": "file:../pkg",
|
||||
|
|
@ -65,6 +66,7 @@
|
|||
"version": "7.28.0",
|
||||
"dev": true,
|
||||
"license": "MIT",
|
||||
"peer": true,
|
||||
"dependencies": {
|
||||
"@ampproject/remapping": "^2.2.0",
|
||||
"@babel/code-frame": "^7.27.1",
|
||||
|
|
@ -1656,6 +1658,7 @@
|
|||
"integrity": "sha512-UVJyE9MttOsBQIDKw1skb9nAwQuR5wuGD3+82K6JgJlm/Y+KI92oNsMNGZCYdDsVtRHSak0pcV5Dno5+4jh9sw==",
|
||||
"dev": true,
|
||||
"license": "MIT",
|
||||
"peer": true,
|
||||
"bin": {
|
||||
"acorn": "bin/acorn"
|
||||
},
|
||||
|
|
@ -1682,6 +1685,7 @@
|
|||
"integrity": "sha512-PlXPeEWMXMZ7sPYOHqmDyCJzcfNrUr3fGNKtezX14ykXOEIvyK81d+qydx89KY5O71FKMPaQ2vBfBFI5NHR63A==",
|
||||
"dev": true,
|
||||
"license": "MIT",
|
||||
"peer": true,
|
||||
"dependencies": {
|
||||
"fast-deep-equal": "^3.1.3",
|
||||
"fast-uri": "^3.0.1",
|
||||
|
|
@ -1908,6 +1912,24 @@
|
|||
"node": ">=6.0.0"
|
||||
}
|
||||
},
|
||||
"node_modules/binaryen": {
|
||||
"version": "123.0.0",
|
||||
"resolved": "https://registry.npmjs.org/binaryen/-/binaryen-123.0.0.tgz",
|
||||
"integrity": "sha512-/hls/a309aZCc0itqP6uhoR+5DsKSlJVfB8Opd2BY9Ndghs84IScTunlyidyF4r2Xe3lQttnfBNIDjaNpj6mTw==",
|
||||
"dev": true,
|
||||
"license": "Apache-2.0",
|
||||
"bin": {
|
||||
"wasm-as": "bin/wasm-as",
|
||||
"wasm-ctor-eval": "bin/wasm-ctor-eval",
|
||||
"wasm-dis": "bin/wasm-dis",
|
||||
"wasm-merge": "bin/wasm-merge",
|
||||
"wasm-metadce": "bin/wasm-metadce",
|
||||
"wasm-opt": "bin/wasm-opt",
|
||||
"wasm-reduce": "bin/wasm-reduce",
|
||||
"wasm-shell": "bin/wasm-shell",
|
||||
"wasm2js": "bin/wasm2js"
|
||||
}
|
||||
},
|
||||
"node_modules/brace-expansion": {
|
||||
"version": "1.1.12",
|
||||
"resolved": "https://registry.npmjs.org/brace-expansion/-/brace-expansion-1.1.12.tgz",
|
||||
|
|
@ -1950,6 +1972,7 @@
|
|||
}
|
||||
],
|
||||
"license": "MIT",
|
||||
"peer": true,
|
||||
"dependencies": {
|
||||
"baseline-browser-mapping": "^2.9.0",
|
||||
"caniuse-lite": "^1.0.30001759",
|
||||
|
|
@ -3053,6 +3076,7 @@
|
|||
"integrity": "sha512-AkXIIFcaazymvey2i/+F94XRnM6TsVLZDhBMLsd1Sf/W0wzsvvpjeyUrCZD6HGG4SDYPgDJDBKeiJTBb10WzMg==",
|
||||
"dev": true,
|
||||
"license": "MIT",
|
||||
"peer": true,
|
||||
"dependencies": {
|
||||
"@jest/core": "30.3.0",
|
||||
"@jest/types": "30.3.0",
|
||||
|
|
@ -4936,6 +4960,7 @@
|
|||
"integrity": "sha512-jl1vZzPDinLr9eUt3J/t7V6FgNEw9QjvBPdysz9KfQDD41fQrC2Y4vKQdiaUpFT4bXlb1RHhLpp8wtm6M5TgSw==",
|
||||
"dev": true,
|
||||
"license": "Apache-2.0",
|
||||
"peer": true,
|
||||
"bin": {
|
||||
"tsc": "bin/tsc",
|
||||
"tsserver": "bin/tsserver"
|
||||
|
|
@ -5072,6 +5097,7 @@
|
|||
"integrity": "sha512-jTywjboN9aHxFlToqb0K0Zs9SbBoW4zRUlGzI2tYNxVYcEi/IPpn+Xi4ye5jTLvX2YeLuic/IvxNot+Q1jMoOw==",
|
||||
"dev": true,
|
||||
"license": "MIT",
|
||||
"peer": true,
|
||||
"dependencies": {
|
||||
"@types/eslint-scope": "^3.7.7",
|
||||
"@types/estree": "^1.0.8",
|
||||
|
|
@ -5119,6 +5145,7 @@
|
|||
"version": "6.0.1",
|
||||
"dev": true,
|
||||
"license": "MIT",
|
||||
"peer": true,
|
||||
"dependencies": {
|
||||
"@discoveryjs/json-ext": "^0.6.1",
|
||||
"@webpack-cli/configtest": "^3.0.1",
|
||||
|
|
|
|||
|
|
@ -4,6 +4,7 @@
|
|||
"description": "Intelligent 3-way text merging with automated conflict resolution",
|
||||
"main": "dist/reconcile.node.js",
|
||||
"browser": "dist/reconcile.web.js",
|
||||
"react-native": "dist/reconcile.rn.js",
|
||||
"keywords": [
|
||||
"text editing",
|
||||
"sync",
|
||||
|
|
@ -31,12 +32,13 @@
|
|||
"dist/**/*"
|
||||
],
|
||||
"scripts": {
|
||||
"build": "webpack --mode production",
|
||||
"format": "prettier --write \"./**/*.(ts|scss|json|html)\"",
|
||||
"build": "node scripts/build-rn.mjs && webpack --mode production",
|
||||
"format": "prettier --write \"./**/*.(ts|mjs|scss|json|html)\"",
|
||||
"test": "NODE_OPTIONS=\"$NODE_OPTIONS --experimental-vm-modules\" jest"
|
||||
},
|
||||
"devDependencies": {
|
||||
"@types/jest": "^30.0.0",
|
||||
"binaryen": "^123.0.0",
|
||||
"jest": "^30.3.0",
|
||||
"prettier": "^3.8.1",
|
||||
"reconcile-text": "file:../pkg",
|
||||
|
|
|
|||
307
reconcile-js/scripts/build-rn.mjs
Normal file
307
reconcile-js/scripts/build-rn.mjs
Normal file
|
|
@ -0,0 +1,307 @@
|
|||
// Generates `pkg-rn/`: a React Native / Hermes-compatible build of the
|
||||
// wasm-bindgen bindings in which the WebAssembly module is replaced by its
|
||||
// wasm2js (pure-JS) translation.
|
||||
|
||||
import { execFileSync } from 'node:child_process';
|
||||
import { existsSync, readdirSync, readFileSync, rmSync, writeFileSync } from 'node:fs';
|
||||
import { dirname, resolve } from 'node:path';
|
||||
import { fileURLToPath, pathToFileURL } from 'node:url';
|
||||
import { homedir } from 'node:os';
|
||||
|
||||
const here = dirname(fileURLToPath(import.meta.url));
|
||||
const reconcileJsDir = resolve(here, '..');
|
||||
const repoRoot = resolve(reconcileJsDir, '..');
|
||||
|
||||
const releaseWasm = resolve(
|
||||
repoRoot,
|
||||
'target/wasm32-unknown-unknown/release/reconcile_text.wasm'
|
||||
);
|
||||
const outDir = resolve(reconcileJsDir, 'pkg-rn');
|
||||
const bgWasm = resolve(outDir, 'reconcile_text_bg.wasm');
|
||||
const bgWasmJs = resolve(outDir, 'reconcile_text_bg.wasm.js');
|
||||
const loweredWasm = resolve(outDir, '_lowered.wasm');
|
||||
const entryJs = resolve(outDir, 'reconcile_text.js');
|
||||
|
||||
const wasmOpt = resolve(reconcileJsDir, 'node_modules/.bin/wasm-opt');
|
||||
const wasm2js = resolve(reconcileJsDir, 'node_modules/.bin/wasm2js');
|
||||
|
||||
function run(cmd, args) {
|
||||
execFileSync(cmd, args, { stdio: 'inherit' });
|
||||
}
|
||||
|
||||
// Locate the wasm-bindgen CLI. It MUST match the `wasm-bindgen` crate version pinned
|
||||
// in Cargo.toml: a mismatched CLI emits bindings the runtime can't use. So we resolve
|
||||
// the required version first and verify every candidate against it, failing loudly
|
||||
// rather than silently falling back to whatever other version happens to be around.
|
||||
function findWasmBindgen() {
|
||||
const cargoToml = readFileSync(resolve(repoRoot, 'Cargo.toml'), 'utf8');
|
||||
const wanted = cargoToml.match(
|
||||
/wasm-bindgen\s*=\s*\{[^}]*version\s*=\s*"([^"]+)"/
|
||||
)?.[1];
|
||||
if (!wanted) {
|
||||
throw new Error(
|
||||
'[build-rn] Could not parse the pinned wasm-bindgen version from Cargo.toml, so ' +
|
||||
'the required CLI version is unknown. Has the dependency declaration changed?'
|
||||
);
|
||||
}
|
||||
|
||||
// 1. On PATH: accept it only if its version matches the pin.
|
||||
let onPath = null;
|
||||
try {
|
||||
onPath = execFileSync('which', ['wasm-bindgen'], { encoding: 'utf8' }).trim();
|
||||
} catch {
|
||||
/* not on PATH; try the wasm-pack cache next */
|
||||
}
|
||||
if (onPath) {
|
||||
const version = execFileSync(onPath, ['--version'], { encoding: 'utf8' }).match(
|
||||
/\d+\.\d+\.\d+/
|
||||
)?.[0];
|
||||
if (version !== wanted) {
|
||||
throw new Error(
|
||||
`[build-rn] wasm-bindgen on PATH (${onPath}) is ${version ?? 'an unknown version'}, ` +
|
||||
`but Cargo.toml pins ${wanted}. Install the matching CLI ` +
|
||||
`(\`cargo install wasm-bindgen-cli --version ${wanted}\`) or remove the mismatched one.`
|
||||
);
|
||||
}
|
||||
return onPath;
|
||||
}
|
||||
|
||||
const cacheRoots = [
|
||||
resolve(homedir(), 'Library/Caches/.wasm-pack'),
|
||||
resolve(homedir(), '.cache/.wasm-pack'),
|
||||
];
|
||||
for (const root of cacheRoots) {
|
||||
if (!existsSync(root)) {
|
||||
continue;
|
||||
}
|
||||
for (const entry of readdirSync(root)) {
|
||||
const candidate = resolve(root, entry, 'wasm-bindgen');
|
||||
if (!existsSync(candidate)) {
|
||||
continue;
|
||||
}
|
||||
let version;
|
||||
try {
|
||||
version = execFileSync(candidate, ['--version'], { encoding: 'utf8' }).match(
|
||||
/\d+\.\d+\.\d+/
|
||||
)?.[0];
|
||||
} catch {
|
||||
continue; // not an invokable wasm-bindgen; ignore
|
||||
}
|
||||
if (version === wanted) {
|
||||
return candidate;
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
throw new Error(
|
||||
`[build-rn] No wasm-bindgen ${wanted} found on PATH or in the wasm-pack cache. ` +
|
||||
'Run `wasm-pack build --target web --features wasm` first (it caches the matching ' +
|
||||
`wasm-bindgen), or \`cargo install wasm-bindgen-cli --version ${wanted}\`.`
|
||||
);
|
||||
}
|
||||
|
||||
if (!existsSync(releaseWasm)) {
|
||||
throw new Error(
|
||||
`Missing ${releaseWasm}.\nRun \`wasm-pack build --target web --features wasm\` from the repo root first.`
|
||||
);
|
||||
}
|
||||
|
||||
console.log('[build-rn] generating bundler-target bindings with wasm-bindgen');
|
||||
rmSync(outDir, { recursive: true, force: true });
|
||||
const wasmBindgen = findWasmBindgen();
|
||||
run(wasmBindgen, ['--target', 'bundler', '--out-dir', outDir, releaseWasm]);
|
||||
|
||||
// --- Patch wasm-bindgen's cached-memory getters for wasm2js -----------------
|
||||
//
|
||||
// wasm-bindgen caches typed-array / DataView views over `wasm.memory.buffer` and
|
||||
// only re-creates them when it detects the heap grew. It detects a grow by looking
|
||||
// for ArrayBuffer *detachment*: a real `WebAssembly.Memory.grow()` detaches the old
|
||||
// buffer (its `byteLength` becomes 0 and `.detached` becomes true), and those are the
|
||||
// only signals the generated getters check:
|
||||
// - getUint8ArrayMemory0(): refreshes when `byteLength === 0` (detach only)
|
||||
// - getDataViewMemory0(): refreshes when `.detached === true`, OR when the buffer
|
||||
// identity changed but only `if (.detached === undefined)` — i.e. that identity
|
||||
// fallback runs solely on engines lacking `ArrayBuffer.prototype.detached`.
|
||||
//
|
||||
// wasm2js grows differently: `__wasm_memory_grow` (in reconcile_text_bg.wasm.js)
|
||||
// allocates a NEW ArrayBuffer, copies the old heap into it, and reassigns
|
||||
// `memory.buffer` WITHOUT ever detaching the old buffer. So the old buffer keeps
|
||||
// `byteLength > 0` and `.detached === false`, and on modern engines that DO expose
|
||||
// `ArrayBuffer.prototype.detached` (Node 25+, current Hermes) the identity fallback is
|
||||
// gated off. Net effect: after a grow the getters keep returning views over the stale
|
||||
// pre-grow buffer, silently corrupting any operation large enough to grow the heap.
|
||||
// Small inputs never grow, so this escapes naive testing.
|
||||
//
|
||||
// WHY WE PATCH INSTEAD OF CONFIGURING.
|
||||
// This is not fixed or configurable upstream: wasm-bindgen has no wasm2js / asm.js /
|
||||
// React Native / "no-WebAssembly" target (every target assumes real WebAssembly
|
||||
// detach-on-grow semantics), there is no flag to force buffer-identity comparison, and
|
||||
// the getter-generation logic (crates/cli-support/src/js/mod.rs `memview`) is
|
||||
// byte-for-byte identical from the pinned 0.2.114 through the latest release and
|
||||
// `main`. The non-detaching-grow case is not even a tracked upstream issue. Rewriting
|
||||
// the generated glue is therefore the only available fix: the two replacements below
|
||||
// make BOTH getters also refresh on a buffer-identity change
|
||||
// (`buffer !== wasm.memory.buffer`), which is the one signal wasm2js does give.
|
||||
//
|
||||
// Each replacement is asserted independently. If a future wasm-bindgen reshapes one
|
||||
// getter but not the other, we MUST fail the build rather than ship a half-patched
|
||||
// module whose un-patched getter corrupts large inputs. The post-build self-test at
|
||||
// the bottom of this file is the backstop that proves the result survives a real grow.
|
||||
const bgJsPath = resolve(outDir, 'reconcile_text_bg.js');
|
||||
let bgJs = readFileSync(bgJsPath, 'utf8');
|
||||
|
||||
// (1) Uint8Array getter: append an unconditional buffer-identity check to the
|
||||
// `byteLength === 0` detach guard (upstream has no identity check here at all).
|
||||
const byteLengthGuard = /(cached\w*Memory0)\.byteLength === 0/g;
|
||||
const byteLengthHits = bgJs.match(byteLengthGuard)?.length ?? 0;
|
||||
if (byteLengthHits === 0) {
|
||||
throw new Error(
|
||||
`[build-rn] Could not find the Uint8Array \`byteLength === 0\` growth guard in ` +
|
||||
`${bgJsPath} to patch for wasm2js. The wasm-bindgen output shape changed; update ` +
|
||||
'this patch (see crates/cli-support/src/js/mod.rs `memview`) — do NOT ship an ' +
|
||||
'unpatched getter, it will corrupt large inputs under wasm2js.'
|
||||
);
|
||||
}
|
||||
bgJs = bgJs.replace(
|
||||
byteLengthGuard,
|
||||
'$1.byteLength === 0 || $1.buffer !== wasm.memory.buffer'
|
||||
);
|
||||
|
||||
// (2) DataView getter: drop the `detached === undefined &&` prefix so the existing
|
||||
// buffer-identity check runs on every runtime, not only legacy ones.
|
||||
const gatedGuard =
|
||||
/(cached\w*Memory0)\.buffer\.detached === undefined && \1\.buffer !== wasm\.memory\.buffer/g;
|
||||
const gatedHits = bgJs.match(gatedGuard)?.length ?? 0;
|
||||
if (gatedHits === 0) {
|
||||
throw new Error(
|
||||
`[build-rn] Could not find the DataView \`detached === undefined\`-gated buffer-identity ` +
|
||||
`check in ${bgJsPath} to un-gate for wasm2js. The wasm-bindgen output shape changed; ` +
|
||||
'update this patch (see crates/cli-support/src/js/mod.rs `memview`) — do NOT ship an ' +
|
||||
'unpatched getter, it will corrupt large inputs under wasm2js.'
|
||||
);
|
||||
}
|
||||
bgJs = bgJs.replace(gatedGuard, '$1.buffer !== wasm.memory.buffer');
|
||||
|
||||
writeFileSync(bgJsPath, bgJs);
|
||||
|
||||
// Post-MVP features that wasm2js cannot translate must be lowered to MVP first.
|
||||
// reference-types stays enabled: it only covers the funcref table here, which
|
||||
// wasm2js handles via call_indirect.
|
||||
const featureFlags = [
|
||||
'--enable-bulk-memory',
|
||||
'--enable-sign-ext',
|
||||
'--enable-nontrapping-float-to-int',
|
||||
'--enable-mutable-globals',
|
||||
'--enable-reference-types',
|
||||
];
|
||||
|
||||
console.log('[build-rn] optimising and lowering to MVP with wasm-opt');
|
||||
run(wasmOpt, [
|
||||
...featureFlags,
|
||||
'-O3',
|
||||
'--signext-lowering',
|
||||
'--llvm-memory-copy-fill-lowering',
|
||||
'--llvm-nontrapping-fptoint-lowering',
|
||||
bgWasm,
|
||||
'-o',
|
||||
loweredWasm,
|
||||
]);
|
||||
|
||||
console.log('[build-rn] translating wasm -> JS with wasm2js');
|
||||
run(wasm2js, ['--enable-reference-types', loweredWasm, '-o', bgWasmJs]);
|
||||
|
||||
console.log('[build-rn] wiring the JS translation into reconcile_text.js');
|
||||
const entry = readFileSync(entryJs, 'utf8');
|
||||
const rewired = entry.replace(
|
||||
/from\s+(['"])\.\/reconcile_text_bg\.wasm\1/,
|
||||
'from $1./reconcile_text_bg.wasm.js$1'
|
||||
);
|
||||
if (rewired === entry) {
|
||||
throw new Error(
|
||||
`Could not find the \`./reconcile_text_bg.wasm\` import in ${entryJs}; ` +
|
||||
'the wasm-bindgen bundler output layout may have changed.'
|
||||
);
|
||||
}
|
||||
writeFileSync(entryJs, rewired);
|
||||
|
||||
// The binary and the intermediate are no longer referenced; remove them so no
|
||||
// bundler attempts to instantiate WebAssembly from this directory.
|
||||
rmSync(bgWasm, { force: true });
|
||||
rmSync(loweredWasm, { force: true });
|
||||
|
||||
// Mark the directory as ESM (matching the web `pkg/`) so Node and Jest treat
|
||||
// these `.js` files as modules. `sideEffects` stays true because importing the
|
||||
// entry runs `__wbg_set_wasm(...)`, which must not be tree-shaken away.
|
||||
writeFileSync(
|
||||
resolve(outDir, 'package.json'),
|
||||
JSON.stringify({ type: 'module', sideEffects: true }, null, 2) + '\n'
|
||||
);
|
||||
|
||||
// Backstop: import the freshly generated module and prove it survives a heap grow.
|
||||
// The patches above are matched by regex against wasm-bindgen output; a silently
|
||||
// mis-applied patch (or a wasm-bindgen change we matched too loosely) would leave a
|
||||
// getter reading the stale pre-grow buffer and corrupt large inputs only. Rather than
|
||||
// trust the regexes, we force a grow here and assert a byte-exact round-trip, so a
|
||||
// broken bundle fails the build instead of reaching a React Native consumer.
|
||||
async function selfTest() {
|
||||
// Importing the entry runs `__wbg_set_wasm(...)`, initialising the wasm2js module.
|
||||
const api = await import(pathToFileURL(entryJs).href);
|
||||
// Same module instance (Node caches by resolved path), so this `memory` is the heap
|
||||
// the API operates on; its `.buffer` getter reflects the current (post-grow) buffer.
|
||||
const { memory } = await import(pathToFileURL(bgWasmJs).href);
|
||||
|
||||
// ~100 KB of distinct tokens. The diff working set amplifies the input many-fold
|
||||
// (a ~50 KB input already forces dozens of grows), so this reliably grows the heap
|
||||
// well past wasm2js's ~1 MB initial allocation while staying fast. A tiny parent
|
||||
// keeps the edit distance — and therefore the runtime — small.
|
||||
const tokens = [];
|
||||
for (let i = 0; i < 10000; i++) {
|
||||
tokens.push(`token-${i}`);
|
||||
}
|
||||
const target = tokens.join(' ');
|
||||
const parent = 'reconcile self-test';
|
||||
|
||||
const heapBefore = memory.buffer.byteLength;
|
||||
|
||||
// Stale post-grow reads surface either as an out-of-bounds throw or as silently
|
||||
// wrong bytes, so handle both: a throw here is itself the failure signal.
|
||||
let roundTripped;
|
||||
try {
|
||||
const changed = new api.TextWithCursors(target, []);
|
||||
const compact = api
|
||||
.diff(parent, changed, 'Word')
|
||||
// This build's `undiff` rejects BigInt; normalise exactly as src/core.ts does.
|
||||
.map((item) => (typeof item === 'bigint' ? Number(item) : item));
|
||||
changed.free();
|
||||
roundTripped = api.undiff(parent, compact, 'Word');
|
||||
} catch (cause) {
|
||||
throw new Error(
|
||||
'[build-rn] self-test crashed during a large diff/undiff round-trip (after the heap ' +
|
||||
'grew). This is the signature of unpatched wasm2js cached-memory getters reading the ' +
|
||||
'stale pre-grow buffer. The growth patch is not taking effect. Refusing to ship this ' +
|
||||
'React Native bundle.',
|
||||
{ cause }
|
||||
);
|
||||
}
|
||||
|
||||
const heapAfter = memory.buffer.byteLength;
|
||||
|
||||
if (heapAfter <= heapBefore) {
|
||||
throw new Error(
|
||||
`[build-rn] self-test did not grow the wasm heap (stayed at ${heapBefore} bytes), ` +
|
||||
'so it cannot validate the memory-growth patch. Enlarge the self-test input.'
|
||||
);
|
||||
}
|
||||
if (roundTripped !== target) {
|
||||
throw new Error(
|
||||
'[build-rn] self-test FAILED: diff/undiff round-trip did not match after a heap grow. ' +
|
||||
'The patched wasm2js cached-memory getters are returning stale/corrupt data — the ' +
|
||||
'growth patch is not taking effect. Refusing to ship this React Native bundle.'
|
||||
);
|
||||
}
|
||||
}
|
||||
|
||||
console.log('[build-rn] self-testing the patched module (forces a heap grow)');
|
||||
await selfTest();
|
||||
|
||||
console.log('[build-rn] done -> pkg-rn/');
|
||||
400
reconcile-js/src/core.ts
Normal file
400
reconcile-js/src/core.ts
Normal file
|
|
@ -0,0 +1,400 @@
|
|||
// Shared, platform-agnostic wrapper around the generated wasm-bindgen surface.
|
||||
//
|
||||
// The actual wasm bindings are injected by a platform-specific entrypoint:
|
||||
// - `index.ts` (web/node) instantiates the real WebAssembly module lazily
|
||||
// on first use via `initSync`.
|
||||
// - `index.rn.ts` (React Native / Hermes) links a wasm2js (pure-JS)
|
||||
// implementation, since Hermes does not expose a runtime
|
||||
// `WebAssembly` global. See `scripts/build-rn.mjs`.
|
||||
|
||||
type WasmModule = typeof import('reconcile-text');
|
||||
|
||||
/**
|
||||
* The generated wasm-bindgen surface this library wraps, plus a hook to make
|
||||
* sure the underlying module is ready. Supplied by a platform entrypoint.
|
||||
*/
|
||||
export interface WasmBackend {
|
||||
CursorPosition: WasmModule['CursorPosition'];
|
||||
TextWithCursors: WasmModule['TextWithCursors'];
|
||||
reconcile: WasmModule['reconcile'];
|
||||
reconcileWithHistory: WasmModule['reconcileWithHistory'];
|
||||
diff: WasmModule['diff'];
|
||||
undiff: WasmModule['undiff'];
|
||||
/**
|
||||
* Make the wasm module ready for use. Invoked before every operation, so it
|
||||
* must be cheap and idempotent (a no-op once initialised).
|
||||
*/
|
||||
ensureReady(): void;
|
||||
}
|
||||
|
||||
// Define the enum values as a const array to avoid duplication
|
||||
const BUILTIN_TOKENIZERS = ['Character', 'Line', 'Markdown', 'Word'] as const;
|
||||
|
||||
/**
|
||||
* Tokenisation strategies for text merging.
|
||||
*
|
||||
* These correspond to the built-in tokenizers available in the underlying WASM module.
|
||||
*/
|
||||
export type BuiltinTokenizer = (typeof BUILTIN_TOKENIZERS)[number];
|
||||
|
||||
/**
|
||||
* History classification for text spans in merge results.
|
||||
*
|
||||
* Indicates the origin of each text span in the merged document.
|
||||
*/
|
||||
export type History =
|
||||
| 'Unchanged'
|
||||
| 'AddedFromLeft'
|
||||
| 'AddedFromRight'
|
||||
| 'RemovedFromLeft'
|
||||
| 'RemovedFromRight';
|
||||
|
||||
/**
|
||||
* Represents a text document with associated cursor positions.
|
||||
*
|
||||
* This interface is used both as input to reconcile functions (to specify where
|
||||
* cursors are positioned in the original documents) and as output (with cursors
|
||||
* automatically repositioned after merging).
|
||||
*/
|
||||
export interface TextWithCursors {
|
||||
/** The document's entire content as a string */
|
||||
text: string;
|
||||
|
||||
/**
|
||||
* Array of cursor positions within the text. Can be empty if there are no cursors to track.
|
||||
* Each cursor has a unique ID and position.
|
||||
*/
|
||||
cursors: CursorPosition[];
|
||||
}
|
||||
|
||||
/**
|
||||
* Like `TextWithCursors`, but cursors may be null or undefined (treated as empty).
|
||||
* Used as input where cursor tracking is optional.
|
||||
*/
|
||||
export interface TextWithOptionalCursors {
|
||||
/** The document's entire content as a string */
|
||||
text: string;
|
||||
|
||||
/**
|
||||
* Array of cursor positions within the text. Can be null, undefined, or empty
|
||||
* if there are no cursors to track. Each cursor has a unique ID and position.
|
||||
*/
|
||||
cursors: null | undefined | CursorPosition[];
|
||||
}
|
||||
|
||||
/**
|
||||
* Represents a cursor position within a text document.
|
||||
*
|
||||
* Cursors are automatically repositioned during text merging to maintain their
|
||||
* relative positions as text is inserted, deleted, or modified around them.
|
||||
*/
|
||||
export interface CursorPosition {
|
||||
/** Unique identifier for the cursor (can be any number, must be unique within the document) */
|
||||
id: number;
|
||||
|
||||
/** Character position in the text, 0-based index from the beginning of the document */
|
||||
position: number;
|
||||
}
|
||||
|
||||
/**
|
||||
* Represents a merged text document with cursor positions and detailed change history.
|
||||
*
|
||||
* This is the return type of `reconcileWithHistory()` and provides complete information
|
||||
* about how the merge was performed, including which parts of the final text came from
|
||||
* which source documents.
|
||||
*/
|
||||
export interface TextWithCursorsAndHistory {
|
||||
/** The merged document's entire content */
|
||||
text: string;
|
||||
|
||||
/**
|
||||
* Array of cursor positions within the merged text. Can be empty if there are no cursors to track.
|
||||
* All cursors are automatically repositioned from the left and right documents.
|
||||
*/
|
||||
cursors: CursorPosition[];
|
||||
|
||||
/**
|
||||
* Detailed provenance information showing the origin of each text span in the result.
|
||||
* Each span indicates whether it was unchanged, added from left, added from right, etc.
|
||||
*/
|
||||
history: SpanWithHistory[];
|
||||
}
|
||||
|
||||
/**
|
||||
* Represents a span of text in the merged result with its change history.
|
||||
*
|
||||
* This shows exactly which source document contributed each piece of text to the
|
||||
* final merged result. Useful for understanding merge decisions and creating
|
||||
* visualisations of how documents were combined.
|
||||
*/
|
||||
export interface SpanWithHistory {
|
||||
/** The text content of this span */
|
||||
text: string;
|
||||
|
||||
/** The origin of this text span in the merge result */
|
||||
history: History;
|
||||
}
|
||||
|
||||
/** The public, synchronous API surface, identical across platforms. */
|
||||
export interface ReconcileApi {
|
||||
/**
|
||||
* Merges three versions of text using intelligent conflict resolution.
|
||||
*
|
||||
* This is the primary function for 3-way text merging. Unlike traditional merge tools
|
||||
* that produce conflict markers, this function automatically resolves conflicts by
|
||||
* applying both sets of changes where possible.
|
||||
*
|
||||
* @param original - The original/base version of the text that both sides diverged from
|
||||
* @param left - The left version of the text (either string or TextWithCursors with cursor positions)
|
||||
* @param right - The right version of the text (either string or TextWithCursors with cursor positions)
|
||||
* @param tokenizer - The tokenisation strategy: "Word" (default, recommended for prose),
|
||||
* "Character" (fine-grained), "Line" (similar to git merge), or
|
||||
* "Markdown" (splits on Markdown structure)
|
||||
* @returns The reconciled text with automatically repositioned cursor positions
|
||||
*
|
||||
* @example
|
||||
* ```typescript
|
||||
* const original = "Hello world";
|
||||
* const left = "Hello beautiful world"; // Added "beautiful"
|
||||
* const right = "Hi world"; // Changed "Hello" to "Hi"
|
||||
*
|
||||
* const result = reconcile(original, left, right);
|
||||
* console.log(result.text); // "Hi beautiful world"
|
||||
* ```
|
||||
*/
|
||||
reconcile(
|
||||
original: string,
|
||||
left: string | TextWithOptionalCursors,
|
||||
right: string | TextWithOptionalCursors,
|
||||
tokenizer?: BuiltinTokenizer
|
||||
): TextWithCursors;
|
||||
|
||||
/**
|
||||
* Generates a compact diff representation between an original and changed text.
|
||||
*
|
||||
* These can be parsed and unpacked using the `undiff` function or the Rust crate's EditedText::from_diff.
|
||||
* Cursor positions are omitted from the diff result.
|
||||
*
|
||||
* This function computes the differences between two versions of text and returns
|
||||
* a compact representation of those changes.
|
||||
*
|
||||
* @param original - The original/base version of the text
|
||||
* @param changed - The modified version of the text (either string or TextWithCursors with cursor positions)
|
||||
* @param tokenizer - The tokenisation strategy, which is the same as used in `reconcile`.
|
||||
* @returns An array of inserts (strings), deletes (negative integers), and retained spans (positive integers).
|
||||
*/
|
||||
diff(
|
||||
original: string,
|
||||
changed: string | TextWithOptionalCursors,
|
||||
tokenizer?: BuiltinTokenizer
|
||||
): Array<number | string>;
|
||||
|
||||
/**
|
||||
* Applies a compact diff to an original text to reconstruct the changed version.
|
||||
*
|
||||
* This function takes an original text and a compact diff representation (as produced
|
||||
* by the `diff` function) and reconstructs the modified text.
|
||||
*
|
||||
* @param original - The original/base version of the text
|
||||
* @param diff - The compact diff array (inserts as strings, deletes as negative integers, retained spans as positive integers)
|
||||
* @param tokenizer - The tokenisation strategy, which is the same as used in `reconcile`.
|
||||
* @returns The reconstructed changed text as a string.
|
||||
*/
|
||||
undiff(
|
||||
original: string,
|
||||
diff: Array<number | bigint | string>,
|
||||
tokenizer?: BuiltinTokenizer
|
||||
): string;
|
||||
|
||||
/**
|
||||
* Merges three versions of text and returns detailed provenance information.
|
||||
*
|
||||
* This function behaves like `reconcile()` but also provides
|
||||
* detailed historical information about the origin of each text span in the result.
|
||||
* This is valuable for understanding how the merge was performed and which changes
|
||||
* came from which source.
|
||||
*
|
||||
* Note: Computing the history is computationally more expensive than the basic merge.
|
||||
*
|
||||
* @param original - The original/base version of the text that both sides diverged from
|
||||
* @param left - The left version of the text (either string or TextWithCursors with cursor positions)
|
||||
* @param right - The right version of the text (either string or TextWithCursors with cursor positions)
|
||||
* @param tokenizer - The tokenisation strategy: "Word" (default, recommended for prose),
|
||||
* "Character" (fine-grained), "Line" (similar to git merge), or
|
||||
* "Markdown" (splits on Markdown structure)
|
||||
* @returns The reconciled text with cursor positions and detailed change history
|
||||
*
|
||||
* @example
|
||||
* ```typescript
|
||||
* const original = "Hello world";
|
||||
* const left = "Hello beautiful world";
|
||||
* const right = "Hi world";
|
||||
*
|
||||
* const result = reconcileWithHistory(original, left, right);
|
||||
* console.log(result.text); // "Hi beautiful world"
|
||||
* console.log(result.history); // Array of SpanWithHistory objects showing change origins
|
||||
* ```
|
||||
*/
|
||||
reconcileWithHistory(
|
||||
original: string,
|
||||
left: string | TextWithOptionalCursors,
|
||||
right: string | TextWithOptionalCursors,
|
||||
tokenizer?: BuiltinTokenizer
|
||||
): TextWithCursorsAndHistory;
|
||||
}
|
||||
|
||||
const UNSUPPORTED_TOKENIZER_ERROR = `Unsupported tokenizer, only ${BUILTIN_TOKENIZERS.join(
|
||||
', '
|
||||
)} are supported`;
|
||||
|
||||
/**
|
||||
* Build the public {@link ReconcileApi} on top of a {@link WasmBackend}.
|
||||
*
|
||||
* Each operation calls `backend.ensureReady()` first, then marshals JS values
|
||||
* into the wasm representation, invokes the binding, and frees the wasm-side
|
||||
* objects. The behaviour is identical regardless of whether the backend is a
|
||||
* real WebAssembly module or its wasm2js translation.
|
||||
*/
|
||||
export function makeReconcileApi(backend: WasmBackend): ReconcileApi {
|
||||
function assertTokenizer(tokenizer: BuiltinTokenizer): void {
|
||||
if (!BUILTIN_TOKENIZERS.includes(tokenizer)) {
|
||||
throw new Error(UNSUPPORTED_TOKENIZER_ERROR);
|
||||
}
|
||||
}
|
||||
|
||||
function toWasmTextWithCursors(text: string | TextWithOptionalCursors) {
|
||||
const isInputString = typeof text === 'string';
|
||||
const innerText = isInputString ? text : text.text;
|
||||
const innerCursors = isInputString ? [] : (text.cursors ?? []);
|
||||
|
||||
return new backend.TextWithCursors(
|
||||
innerText,
|
||||
innerCursors.map(({ id, position }) => new backend.CursorPosition(id, position))
|
||||
);
|
||||
}
|
||||
|
||||
function toTextWithCursors(textWithCursor: {
|
||||
text(): string;
|
||||
cursors(): Array<{ id(): number; characterIndex(): number; free(): void }>;
|
||||
}): TextWithCursors {
|
||||
const wasmCursors = textWithCursor.cursors();
|
||||
const cursors = wasmCursors.map((cursor) => ({
|
||||
id: cursor.id(),
|
||||
position: cursor.characterIndex(),
|
||||
}));
|
||||
for (const cursor of wasmCursors) {
|
||||
cursor.free();
|
||||
}
|
||||
|
||||
return {
|
||||
text: textWithCursor.text(),
|
||||
cursors,
|
||||
};
|
||||
}
|
||||
|
||||
function toSpanWithHistory(span: {
|
||||
text(): string;
|
||||
history(): History;
|
||||
free(): void;
|
||||
}): SpanWithHistory {
|
||||
const result = {
|
||||
text: span.text(),
|
||||
history: span.history(),
|
||||
};
|
||||
span.free();
|
||||
return result;
|
||||
}
|
||||
|
||||
function reconcile(
|
||||
original: string,
|
||||
left: string | TextWithOptionalCursors,
|
||||
right: string | TextWithOptionalCursors,
|
||||
tokenizer: BuiltinTokenizer = 'Word'
|
||||
): TextWithCursors {
|
||||
backend.ensureReady();
|
||||
assertTokenizer(tokenizer);
|
||||
|
||||
const leftCursor = toWasmTextWithCursors(left);
|
||||
const rightCursor = toWasmTextWithCursors(right);
|
||||
|
||||
const result = backend.reconcile(original, leftCursor, rightCursor, tokenizer);
|
||||
|
||||
leftCursor.free();
|
||||
rightCursor.free();
|
||||
|
||||
const jsResult = toTextWithCursors(result);
|
||||
result.free();
|
||||
|
||||
return jsResult;
|
||||
}
|
||||
|
||||
function diff(
|
||||
original: string,
|
||||
changed: string | TextWithOptionalCursors,
|
||||
tokenizer: BuiltinTokenizer = 'Word'
|
||||
): Array<number | string> {
|
||||
backend.ensureReady();
|
||||
assertTokenizer(tokenizer);
|
||||
|
||||
const changedWasm = toWasmTextWithCursors(changed);
|
||||
|
||||
const result = backend.diff(original, changedWasm, tokenizer);
|
||||
|
||||
changedWasm.free();
|
||||
|
||||
return result.map((item) => (typeof item === 'bigint' ? Number(item) : item));
|
||||
}
|
||||
|
||||
function undiff(
|
||||
original: string,
|
||||
diffValue: Array<number | bigint | string>,
|
||||
tokenizer: BuiltinTokenizer = 'Word'
|
||||
): string {
|
||||
backend.ensureReady();
|
||||
assertTokenizer(tokenizer);
|
||||
|
||||
// The real-WebAssembly backend's `diff` emits BigInt spans, whereas the
|
||||
// wasm2js (React Native) backend rejects BigInt outright. Normalise to
|
||||
// plain numbers - exactly as `diff` does on the way out - so a `diff`
|
||||
// result round-trips through `undiff` identically on every platform.
|
||||
return backend.undiff(
|
||||
original,
|
||||
diffValue.map((item) => (typeof item === 'bigint' ? Number(item) : item)),
|
||||
tokenizer
|
||||
);
|
||||
}
|
||||
|
||||
function reconcileWithHistory(
|
||||
original: string,
|
||||
left: string | TextWithOptionalCursors,
|
||||
right: string | TextWithOptionalCursors,
|
||||
tokenizer: BuiltinTokenizer = 'Word'
|
||||
): TextWithCursorsAndHistory {
|
||||
backend.ensureReady();
|
||||
assertTokenizer(tokenizer);
|
||||
|
||||
const leftCursor = toWasmTextWithCursors(left);
|
||||
const rightCursor = toWasmTextWithCursors(right);
|
||||
|
||||
const result = backend.reconcileWithHistory(
|
||||
original,
|
||||
leftCursor,
|
||||
rightCursor,
|
||||
tokenizer
|
||||
);
|
||||
|
||||
leftCursor.free();
|
||||
rightCursor.free();
|
||||
|
||||
const jsResult = toTextWithCursors(result);
|
||||
const history = result.history().map(toSpanWithHistory);
|
||||
result.free();
|
||||
|
||||
return {
|
||||
...jsResult,
|
||||
history,
|
||||
};
|
||||
}
|
||||
|
||||
return { reconcile, diff, undiff, reconcileWithHistory };
|
||||
}
|
||||
47
reconcile-js/src/index.rn.ts
Normal file
47
reconcile-js/src/index.rn.ts
Normal file
|
|
@ -0,0 +1,47 @@
|
|||
// React Native entrypoint (resolved via the `react-native` package field).
|
||||
//
|
||||
// Hermes — the default React Native engine since RN 0.84 / Expo SDK 56 — does
|
||||
// not expose a runtime `WebAssembly` global, so the normal `new
|
||||
// WebAssembly.Module(...)` path used by `index.ts` throws
|
||||
// `ReferenceError: Property 'WebAssembly' doesn't exist`.
|
||||
//
|
||||
// Instead we link a wasm2js translation of the module: pure JavaScript that
|
||||
// needs no `WebAssembly` global and is instantiated synchronously at import
|
||||
// time. The public API and its synchronous signatures are unchanged, so
|
||||
// callers need no modification. The `pkg-rn` directory is generated by
|
||||
// `scripts/build-rn.mjs`.
|
||||
|
||||
import {
|
||||
CursorPosition as wasmCursorPosition,
|
||||
TextWithCursors as wasmTextWithCursors,
|
||||
reconcile as wasmReconcile,
|
||||
reconcileWithHistory as wasmReconcileWithHistory,
|
||||
diff as wasmDiff,
|
||||
undiff as wasmUndiff,
|
||||
} from '../pkg-rn/reconcile_text.js';
|
||||
|
||||
import { makeReconcileApi, type WasmBackend } from './core';
|
||||
|
||||
const backend: WasmBackend = {
|
||||
CursorPosition: wasmCursorPosition,
|
||||
TextWithCursors: wasmTextWithCursors,
|
||||
reconcile: wasmReconcile,
|
||||
reconcileWithHistory: wasmReconcileWithHistory,
|
||||
diff: wasmDiff,
|
||||
undiff: wasmUndiff,
|
||||
// The wasm2js module initialises itself at import time, so this is a no-op.
|
||||
ensureReady() {},
|
||||
};
|
||||
|
||||
export const { reconcile, diff, undiff, reconcileWithHistory } =
|
||||
makeReconcileApi(backend);
|
||||
|
||||
export type {
|
||||
BuiltinTokenizer,
|
||||
History,
|
||||
CursorPosition,
|
||||
TextWithCursors,
|
||||
TextWithOptionalCursors,
|
||||
TextWithCursorsAndHistory,
|
||||
SpanWithHistory,
|
||||
} from './core';
|
||||
|
|
@ -1,4 +1,5 @@
|
|||
import { reconcile, reconcileWithHistory, diff, undiff } from './index';
|
||||
import * as webApi from './index';
|
||||
import * as rnApi from './index.rn';
|
||||
import { installWasmLeakDetector, checkForWasmLeaks } from './wasm-leak-detector';
|
||||
import * as fs from 'fs';
|
||||
import * as path from 'path';
|
||||
|
|
@ -17,7 +18,18 @@ afterEach(() => {
|
|||
}
|
||||
});
|
||||
|
||||
describe('reconcile', () => {
|
||||
// `./index` is the web/node build (real WebAssembly); `./index.rn` is the React
|
||||
// Native build (the wasm2js pure-JS translation). Both are thin backends over the
|
||||
// same `src/core.ts` wrapper and expose an identical public API, so the behavioural
|
||||
// suite below runs against both to guarantee they stay in lock-step.
|
||||
const backends = [
|
||||
{ name: 'web/node (WebAssembly)', api: webApi },
|
||||
{ name: 'React Native (wasm2js)', api: rnApi },
|
||||
];
|
||||
|
||||
describe.each(backends)('reconcile [$name]', ({ api }) => {
|
||||
const { reconcile, reconcileWithHistory, diff, undiff } = api;
|
||||
|
||||
it('call reconcile without cursors', () => {
|
||||
expect(reconcile('Hello', 'Hello world', 'Hi world').text).toEqual('Hi world');
|
||||
});
|
||||
|
|
@ -60,9 +72,26 @@ describe('reconcile', () => {
|
|||
expect(result.text).toEqual('Hi world');
|
||||
expect(result.history.length).toBeGreaterThan(0);
|
||||
});
|
||||
|
||||
it('undiff accepts bigint entries (per the Array<number | bigint | string> type)', () => {
|
||||
const original = 'Hello world';
|
||||
const changed = 'Hello cruel world';
|
||||
|
||||
// `diff` returns plain numbers; emulate a caller that supplies BigInt, which the
|
||||
// public signature permits. The wasm2js build rejects raw BigInt, so the shared
|
||||
// wrapper must normalise it — running this on both backends asserts the contract.
|
||||
const withBigints = diff(original, changed).map((item) =>
|
||||
typeof item === 'number' ? BigInt(item) : item
|
||||
);
|
||||
|
||||
expect(withBigints.some((item) => typeof item === 'bigint')).toBe(true);
|
||||
expect(undiff(original, withBigints)).toEqual(changed);
|
||||
});
|
||||
});
|
||||
|
||||
describe('test_diff_and_undiff_are_inverse', () => {
|
||||
describe.each(backends)('diff and undiff are inverse [$name]', ({ api }) => {
|
||||
const { diff, undiff } = api;
|
||||
|
||||
const resourcesPath = path.join(__dirname, '../../tests/resources');
|
||||
|
||||
const readFileSlice = (fileName: string, start: number, end: number): string => {
|
||||
|
|
@ -93,3 +122,31 @@ describe('test_diff_and_undiff_are_inverse', () => {
|
|||
});
|
||||
});
|
||||
});
|
||||
|
||||
// React-Native-only: Hermes exposes no `WebAssembly` global, which is the whole reason
|
||||
// the RN entry point links a wasm2js build. Only the wasm2js backend can satisfy this.
|
||||
describe('React Native (wasm2js) Hermes parity', () => {
|
||||
const { reconcile, reconcileWithHistory, diff, undiff } = rnApi;
|
||||
|
||||
it('runs every operation with no WebAssembly global', () => {
|
||||
const descriptor = Object.getOwnPropertyDescriptor(globalThis, 'WebAssembly');
|
||||
delete (globalThis as { WebAssembly?: unknown }).WebAssembly;
|
||||
try {
|
||||
expect((globalThis as { WebAssembly?: unknown }).WebAssembly).toBeUndefined();
|
||||
|
||||
expect(reconcile('Hello', 'Hello world', 'Hi world').text).toEqual('Hi world');
|
||||
|
||||
const changes = diff('Hello world', 'Hello cruel world');
|
||||
expect(undiff('Hello world', changes)).toEqual('Hello cruel world');
|
||||
|
||||
expect(
|
||||
reconcileWithHistory('Hello', 'Hello world', 'Hi world').history.length
|
||||
).toBeGreaterThan(0);
|
||||
} finally {
|
||||
// Restore the global so the leak check and later suites are unaffected.
|
||||
if (descriptor) {
|
||||
Object.defineProperty(globalThis, 'WebAssembly', descriptor);
|
||||
}
|
||||
}
|
||||
});
|
||||
});
|
||||
|
|
|
|||
|
|
@ -1,8 +1,7 @@
|
|||
import {
|
||||
CursorPosition as wasmCursorPosition,
|
||||
reconcile as wasmReconcile,
|
||||
TextWithCursors as wasmTextWithCursors,
|
||||
SpanWithHistory as wasmSpanWithHistory,
|
||||
reconcile as wasmReconcile,
|
||||
reconcileWithHistory as wasmReconcileWithHistory,
|
||||
diff as wasmDiff,
|
||||
undiff as wasmUndiff,
|
||||
|
|
@ -11,341 +10,40 @@ import {
|
|||
|
||||
import wasmBytes from 'reconcile-text/reconcile_text_bg.wasm';
|
||||
|
||||
// Define the enum values as const arrays to avoid duplication
|
||||
const BUILTIN_TOKENIZERS = ['Character', 'Line', 'Markdown', 'Word'] as const;
|
||||
const HISTORY_VALUES = [
|
||||
'Unchanged',
|
||||
'AddedFromLeft',
|
||||
'AddedFromRight',
|
||||
'RemovedFromLeft',
|
||||
'RemovedFromRight',
|
||||
] as const;
|
||||
|
||||
/**
|
||||
* Tokenisation strategies for text merging.
|
||||
*
|
||||
* These correspond to the built-in tokenizers available in the underlying WASM module.
|
||||
*/
|
||||
export type BuiltinTokenizer = (typeof BUILTIN_TOKENIZERS)[number];
|
||||
|
||||
/**
|
||||
* History classification for text spans in merge results.
|
||||
*
|
||||
* Indicates the origin of each text span in the merged document.
|
||||
*/
|
||||
export type History = (typeof HISTORY_VALUES)[number];
|
||||
|
||||
/**
|
||||
* Represents a text document with associated cursor positions.
|
||||
*
|
||||
* This interface is used both as input to reconcile functions (to specify where
|
||||
* cursors are positioned in the original documents) and as output (with cursors
|
||||
* automatically repositioned after merging).
|
||||
*/
|
||||
export interface TextWithCursors {
|
||||
/** The document's entire content as a string */
|
||||
text: string;
|
||||
|
||||
/**
|
||||
* Array of cursor positions within the text. Can be empty if there are no cursors to track.
|
||||
* Each cursor has a unique ID and position.
|
||||
*/
|
||||
cursors: CursorPosition[];
|
||||
}
|
||||
|
||||
/**
|
||||
* Like `TextWithCursors`, but cursors may be null or undefined (treated as empty).
|
||||
* Used as input where cursor tracking is optional.
|
||||
*/
|
||||
export interface TextWithOptionalCursors {
|
||||
/** The document's entire content as a string */
|
||||
text: string;
|
||||
|
||||
/**
|
||||
* Array of cursor positions within the text. Can be null, undefined, or empty
|
||||
* if there are no cursors to track. Each cursor has a unique ID and position.
|
||||
*/
|
||||
cursors: null | undefined | CursorPosition[];
|
||||
}
|
||||
|
||||
/**
|
||||
* Represents a cursor position within a text document.
|
||||
*
|
||||
* Cursors are automatically repositioned during text merging to maintain their
|
||||
* relative positions as text is inserted, deleted, or modified around them.
|
||||
*/
|
||||
export interface CursorPosition {
|
||||
/** Unique identifier for the cursor (can be any number, must be unique within the document) */
|
||||
id: number;
|
||||
|
||||
/** Character position in the text, 0-based index from the beginning of the document */
|
||||
position: number;
|
||||
}
|
||||
|
||||
/**
|
||||
* Represents a merged text document with cursor positions and detailed change history.
|
||||
*
|
||||
* This is the return type of `reconcileWithHistory()` and provides complete information
|
||||
* about how the merge was performed, including which parts of the final text came from
|
||||
* which source documents.
|
||||
*/
|
||||
export interface TextWithCursorsAndHistory {
|
||||
/** The merged document's entire content */
|
||||
text: string;
|
||||
|
||||
/**
|
||||
* Array of cursor positions within the merged text. Can be empty if there are no cursors to track.
|
||||
* All cursors are automatically repositioned from the left and right documents.
|
||||
*/
|
||||
cursors: CursorPosition[];
|
||||
|
||||
/**
|
||||
* Detailed provenance information showing the origin of each text span in the result.
|
||||
* Each span indicates whether it was unchanged, added from left, added from right, etc.
|
||||
*/
|
||||
history: SpanWithHistory[];
|
||||
}
|
||||
|
||||
/**
|
||||
* Represents a span of text in the merged result with its change history.
|
||||
*
|
||||
* This shows exactly which source document contributed each piece of text to the
|
||||
* final merged result. Useful for understanding merge decisions and creating
|
||||
* visualisations of how documents were combined.
|
||||
*/
|
||||
export interface SpanWithHistory {
|
||||
/** The text content of this span */
|
||||
text: string;
|
||||
|
||||
/** The origin of this text span in the merge result */
|
||||
history: History;
|
||||
}
|
||||
|
||||
const UNSUPPORTED_TOKENIZER_ERROR = `Unsupported tokenizer, only ${BUILTIN_TOKENIZERS.join(
|
||||
', '
|
||||
)} are supported`;
|
||||
import { makeReconcileApi, type WasmBackend } from './core';
|
||||
|
||||
let isInitialised = false;
|
||||
|
||||
/**
|
||||
* Merges three versions of text using intelligent conflict resolution.
|
||||
*
|
||||
* This is the primary function for 3-way text merging. Unlike traditional merge tools
|
||||
* that produce conflict markers, this function automatically resolves conflicts by
|
||||
* applying both sets of changes where possible.
|
||||
*
|
||||
* @param original - The original/base version of the text that both sides diverged from
|
||||
* @param left - The left version of the text (either string or TextWithCursors with cursor positions)
|
||||
* @param right - The right version of the text (either string or TextWithCursors with cursor positions)
|
||||
* @param tokenizer - The tokenisation strategy: "Word" (default, recommended for prose),
|
||||
* "Character" (fine-grained), or "Line" (similar to git merge)
|
||||
* @returns The reconciled text with automatically repositioned cursor positions
|
||||
*
|
||||
* @example
|
||||
* ```typescript
|
||||
* const original = "Hello world";
|
||||
* const left = "Hello beautiful world"; // Added "beautiful"
|
||||
* const right = "Hi world"; // Changed "Hello" to "Hi"
|
||||
*
|
||||
* const result = reconcile(original, left, right);
|
||||
* console.log(result.text); // "Hi beautiful world"
|
||||
* ```
|
||||
*/
|
||||
export function reconcile(
|
||||
original: string,
|
||||
left: string | TextWithOptionalCursors,
|
||||
right: string | TextWithOptionalCursors,
|
||||
tokenizer: BuiltinTokenizer = 'Word'
|
||||
): TextWithCursors {
|
||||
init();
|
||||
const backend: WasmBackend = {
|
||||
CursorPosition: wasmCursorPosition,
|
||||
TextWithCursors: wasmTextWithCursors,
|
||||
reconcile: wasmReconcile,
|
||||
reconcileWithHistory: wasmReconcileWithHistory,
|
||||
diff: wasmDiff,
|
||||
undiff: wasmUndiff,
|
||||
ensureReady() {
|
||||
if (isInitialised) {
|
||||
return;
|
||||
}
|
||||
|
||||
if (!BUILTIN_TOKENIZERS.includes(tokenizer)) {
|
||||
throw new Error(UNSUPPORTED_TOKENIZER_ERROR);
|
||||
}
|
||||
const wasmBinary = Uint8Array.from(atob(wasmBytes as unknown as string), (c) =>
|
||||
c.charCodeAt(0)
|
||||
);
|
||||
initSync({ module: wasmBinary });
|
||||
|
||||
const leftCursor = toWasmTextWithCursors(left);
|
||||
const rightCursor = toWasmTextWithCursors(right);
|
||||
isInitialised = true;
|
||||
},
|
||||
};
|
||||
|
||||
const result = wasmReconcile(original, leftCursor, rightCursor, tokenizer);
|
||||
export const { reconcile, diff, undiff, reconcileWithHistory } =
|
||||
makeReconcileApi(backend);
|
||||
|
||||
leftCursor.free();
|
||||
rightCursor.free();
|
||||
|
||||
const jsResult = toTextWithCursors(result);
|
||||
result.free();
|
||||
|
||||
return jsResult;
|
||||
}
|
||||
|
||||
/**
|
||||
* Generates a compact diff representation between an original and changed text.
|
||||
*
|
||||
* These can be parsed and unpacked using the `undiff` function or the Rust crate's EditedText::from_diff.
|
||||
* Cursor positions are omitted from the diff result.
|
||||
*
|
||||
* This function computes the differences between two versions of text and returns
|
||||
* a compact representation of those changes.
|
||||
*
|
||||
* @param original - The original/base version of the text
|
||||
* @param changed - The modified version of the text (either string or TextWithCursors with cursor positions)
|
||||
* @param tokenizer - The tokenisation strategy, which is the same as used in `reconcile`.
|
||||
* @returns An array of inserts (strings), deletes (negative integers), and retained spans (positive integers).
|
||||
*/
|
||||
export function diff(
|
||||
original: string,
|
||||
changed: string | TextWithOptionalCursors,
|
||||
tokenizer: BuiltinTokenizer = 'Word'
|
||||
): Array<number | string> {
|
||||
init();
|
||||
|
||||
if (!BUILTIN_TOKENIZERS.includes(tokenizer)) {
|
||||
throw new Error(UNSUPPORTED_TOKENIZER_ERROR);
|
||||
}
|
||||
|
||||
const changedWasm = toWasmTextWithCursors(changed);
|
||||
|
||||
const result = wasmDiff(original, changedWasm, tokenizer);
|
||||
|
||||
changedWasm.free();
|
||||
|
||||
return result.map((item) => (typeof item === 'bigint' ? Number(item) : item));
|
||||
}
|
||||
|
||||
/**
|
||||
* Applies a compact diff to an original text to reconstruct the changed version.
|
||||
*
|
||||
* This function takes an original text and a compact diff representation (as produced
|
||||
* by the `diff` function) and reconstructs the modified text.
|
||||
*
|
||||
* @param original - The original/base version of the text
|
||||
* @param diff - The compact diff array (inserts as strings, deletes as negative integers, retained spans as positive integers)
|
||||
* @param tokenizer - The tokenisation strategy, which is the same as used in `reconcile`.
|
||||
* @returns The reconstructed changed text as a string.
|
||||
*/
|
||||
export function undiff(
|
||||
original: string,
|
||||
diff: Array<number | bigint | string>,
|
||||
tokenizer: BuiltinTokenizer = 'Word'
|
||||
): string {
|
||||
init();
|
||||
|
||||
if (!BUILTIN_TOKENIZERS.includes(tokenizer)) {
|
||||
throw new Error(UNSUPPORTED_TOKENIZER_ERROR);
|
||||
}
|
||||
|
||||
return wasmUndiff(original, diff, tokenizer);
|
||||
}
|
||||
|
||||
/**
|
||||
* Merges three versions of text and returns detailed provenance information.
|
||||
*
|
||||
* This function behaves like `reconcile()` but also provides
|
||||
* detailed historical information about the origin of each text span in the result.
|
||||
* This is valuable for understanding how the merge was performed and which changes
|
||||
* came from which source.
|
||||
*
|
||||
* Note: Computing the history is computationally more expensive than the basic merge.
|
||||
*
|
||||
* @param original - The original/base version of the text that both sides diverged from
|
||||
* @param left - The left version of the text (either string or TextWithCursors with cursor positions)
|
||||
* @param right - The right version of the text (either string or TextWithCursors with cursor positions)
|
||||
* @param tokenizer - The tokenisation strategy: "Word" (default, recommended for prose),
|
||||
* "Character" (fine-grained), or "Line" (similar to git merge)
|
||||
* @returns The reconciled text with cursor positions and detailed change history
|
||||
*
|
||||
* @example
|
||||
* ```typescript
|
||||
* const original = "Hello world";
|
||||
* const left = "Hello beautiful world";
|
||||
* const right = "Hi world";
|
||||
*
|
||||
* const result = reconcileWithHistory(original, left, right);
|
||||
* console.log(result.text); // "Hi beautiful world"
|
||||
* console.log(result.history); // Array of SpanWithHistory objects showing change origins
|
||||
* ```
|
||||
*/
|
||||
export function reconcileWithHistory(
|
||||
original: string,
|
||||
left: string | TextWithOptionalCursors,
|
||||
right: string | TextWithOptionalCursors,
|
||||
tokenizer: BuiltinTokenizer = 'Word'
|
||||
): TextWithCursorsAndHistory {
|
||||
init();
|
||||
|
||||
if (!BUILTIN_TOKENIZERS.includes(tokenizer)) {
|
||||
throw new Error(UNSUPPORTED_TOKENIZER_ERROR);
|
||||
}
|
||||
|
||||
const leftCursor = toWasmTextWithCursors(left);
|
||||
const rightCursor = toWasmTextWithCursors(right);
|
||||
|
||||
const result = wasmReconcileWithHistory(original, leftCursor, rightCursor, tokenizer);
|
||||
|
||||
leftCursor.free();
|
||||
rightCursor.free();
|
||||
|
||||
const jsResult = toTextWithCursors(result);
|
||||
const history = result.history().map(toSpanWithHistory);
|
||||
result.free();
|
||||
|
||||
return {
|
||||
...jsResult,
|
||||
history,
|
||||
};
|
||||
}
|
||||
|
||||
function init() {
|
||||
if (isInitialised) {
|
||||
return;
|
||||
}
|
||||
|
||||
const wasmBinary = Uint8Array.from(atob(wasmBytes as unknown as string), (c) =>
|
||||
c.charCodeAt(0)
|
||||
);
|
||||
initSync({ module: wasmBinary });
|
||||
|
||||
isInitialised = true;
|
||||
}
|
||||
|
||||
function toWasmTextWithCursors(
|
||||
text: string | TextWithOptionalCursors
|
||||
): wasmTextWithCursors {
|
||||
const isInputString = typeof text === 'string';
|
||||
const leftText = isInputString ? text : text.text;
|
||||
const leftCursors = isInputString ? [] : (text.cursors ?? []);
|
||||
|
||||
return new wasmTextWithCursors(leftText, leftCursors.map(toWasmCursorPosition));
|
||||
}
|
||||
|
||||
function toWasmCursorPosition({ id, position }: CursorPosition): wasmCursorPosition {
|
||||
return new wasmCursorPosition(id, position);
|
||||
}
|
||||
|
||||
function toTextWithCursors(textWithCursor: wasmTextWithCursors): TextWithCursors {
|
||||
const wasmCursors = textWithCursor.cursors();
|
||||
const cursors = wasmCursors.map(toCursorPosition);
|
||||
for (const cursor of wasmCursors) {
|
||||
cursor.free();
|
||||
}
|
||||
|
||||
return {
|
||||
text: textWithCursor.text(),
|
||||
cursors,
|
||||
};
|
||||
}
|
||||
|
||||
function toCursorPosition(cursor: wasmCursorPosition): CursorPosition {
|
||||
return {
|
||||
id: cursor.id(),
|
||||
position: cursor.characterIndex(),
|
||||
};
|
||||
}
|
||||
|
||||
function toSpanWithHistory(span: wasmSpanWithHistory): SpanWithHistory {
|
||||
const result = {
|
||||
text: span.text(),
|
||||
history: span.history(),
|
||||
};
|
||||
span.free();
|
||||
return result;
|
||||
}
|
||||
export type {
|
||||
BuiltinTokenizer,
|
||||
History,
|
||||
CursorPosition,
|
||||
TextWithCursors,
|
||||
TextWithOptionalCursors,
|
||||
TextWithCursorsAndHistory,
|
||||
SpanWithHistory,
|
||||
} from './core';
|
||||
|
|
|
|||
|
|
@ -2,7 +2,6 @@ const path = require('path');
|
|||
const { merge } = require('webpack-merge');
|
||||
|
||||
const common = {
|
||||
entry: './src/index.ts',
|
||||
optimization: {
|
||||
// the consuming project should take care of minification
|
||||
minimize: false,
|
||||
|
|
@ -38,8 +37,10 @@ const common = {
|
|||
};
|
||||
|
||||
module.exports = [
|
||||
// Web build: real WebAssembly, instantiated synchronously from inlined base64.
|
||||
merge(common, {
|
||||
target: 'web',
|
||||
entry: './src/index.ts',
|
||||
output: {
|
||||
path: path.resolve(__dirname, 'dist'),
|
||||
filename: 'reconcile.web.js',
|
||||
|
|
@ -50,12 +51,31 @@ module.exports = [
|
|||
globalObject: 'this',
|
||||
},
|
||||
}),
|
||||
|
||||
// Node build: real WebAssembly.
|
||||
merge(common, {
|
||||
target: 'node',
|
||||
entry: './src/index.ts',
|
||||
output: {
|
||||
path: path.resolve(__dirname, 'dist'),
|
||||
filename: 'reconcile.node.js',
|
||||
libraryTarget: 'commonjs2',
|
||||
},
|
||||
}),
|
||||
|
||||
// React Native build: wasm2js (pure JS), for Hermes which has no
|
||||
// `WebAssembly` global. Sources come from `pkg-rn/`
|
||||
merge(common, {
|
||||
target: 'web',
|
||||
entry: './src/index.rn.ts',
|
||||
output: {
|
||||
path: path.resolve(__dirname, 'dist'),
|
||||
filename: 'reconcile.rn.js',
|
||||
library: {
|
||||
name: 'reconcile',
|
||||
type: 'umd',
|
||||
},
|
||||
globalObject: 'this',
|
||||
},
|
||||
}),
|
||||
];
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue