GeezerVison is showing github.com — PythonMonkey/python/pythonmonkey/require.py at main · Distributive-Network/PythonMonkey

Expand file tree

require.py

Copy path

More file actions

Latest commit

History

435 lines (370 loc) · 16.1 KB

require.py

Copy path

File metadata and controls

435 lines (370 loc) · 16.1 KB

Raw

Copy raw file

Download raw file

Open symbols panel

Edit and raw actions

100

101

102

103

104

105

106

107

108

109

110

111

112

113

114

115

116

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

132

133

134

135

136

137

138

139

140

141

142

143

144

145

146

147

148

149

150

151

152

153

154

155

156

157

158

159

160

161

162

163

164

165

166

167

168

169

170

171

172

173

174

175

176

177

178

179

180

181

182

183

184

185

186

187

188

189

190

191

192

193

194

195

196

197

198

199

200

201

202

203

204

205

206

207

208

209

210

211

212

213

214

215

216

217

218

219

220

221

222

223

224

225

226

227

228

229

230

231

232

233

234

235

236

237

238

239

240

241

242

243

244

245

246

247

248

249

250

251

252

253

254

255

256

257

258

259

260

261

262

263

264

265

266

267

268

269

270

271

272

273

274

275

276

277

278

279

280

281

282

283

284

285

286

287

288

289

290

291

292

293

294

295

296

297

298

299

300

301

302

303

304

305

306

307

308

309

310

311

312

313

314

315

316

317

318

319

320

321

322

323

324

325

326

327

328

329

330

331

332

333

334

335

336

337

338

339

340

341

342

343

344

345

346

347

348

349

350

351

352

353

354

355

356

357

358

359

360

361

362

363

364

365

366

367

368

369

370

371

372

373

374

375

376

377

378

379

380

381

382

383

384

385

386

387

388

389

390

391

392

393

394

395

396

397

398

399

400

401

402

403

404

405

406

407

408

409

410

411

412

413

414

415

416

417

418

419

420

421

422

423

424

425

426

427

428

429

430

431

432

433

434

435

# @file require.py

# Implementation of CommonJS "require" for PythonMonkey. This implementation uses the

# ctx-module npm package to do the heavy lifting. That package makes a complete module

# system, ostensibly in a separate context, but our implementation here reuses the

# PythonMonkey global context for both.

# The context that ctx-module runs in needs a require function supporting

# - require('debug') => returns a debug function which prints to the console when

# debugging; see the node debug built-in

# - require('fs') => returns an object which has an implementation of readFileSync

# and statSync. The implementation of statSync only needs to

# return the mode member. The fs module also needs

# constants.S_IFDIR available.

# - require('vm') => returns an object which has an implementation of evalInContext

# In order to implement this basic require function for bootstrapping ctxModule, we

# have simply made global variables of the form xyzModule where xyz is the module

# identifier, and injected a require function which understands this. A better

# implementation in Python that doesn't leak global symbols should be possible once

# some PythonMonkey bugs are fixed.

# Module Context Summary

# - what CommonJS modules you can access depends on what require symbol you have

# - pm means the pythonmonkey Python module exports

# - pm.bootstrap.require is a special require that knows about modules which are not stored

# on disk. This is important, because ctx-module needs some modules, and this is how we

# solve the chicken<>egg problem.

# - There is a node_modules folder inside the pminit module that comes with pythonmonkey.

# This folder is managed by npm by invoking `pminit npm`. These modules are intended as

# global modules, analogously to require('process') in nodejs.

# - pm.require is a require that works as expected from python source code; i.e. relative

# modules are resolved relative to the python source code.

# - pm.require has access to the pminit node_modules modules but not the bootstrap modules.

# - The builtin_modules directory is available anywhere the pminit node_modules directory

# is, and has a higher precedence.

# @author Wes Garland, [email protected]

# @date May 2023

import sys

import os

import io

from typing import Union, Dict, Literal, List

import importlib

import importlib.util

from importlib import machinery

import inspect

import functools

from . import pythonmonkey as pm

node_modules = os.path.abspath(

os.path.join(

importlib.util.find_spec("pminit").submodule_search_locations[0], # type: ignore

"..",

"pythonmonkey",

"node_modules"

)

evalOpts = {'filename': __file__, 'fromPythonFrame': True} # type: pm.EvalOptions

# Force to use UTF-8 encoding

# Windows may use other encodings / code pages that have many characters missing/unrepresentable

if isinstance(sys.stdin, io.TextIOWrapper):

sys.stdin.reconfigure(encoding='utf-8')

if isinstance(sys.stdout, io.TextIOWrapper):

sys.stdout.reconfigure(encoding='utf-8')

if isinstance(sys.stderr, io.TextIOWrapper):

sys.stderr.reconfigure(encoding='utf-8')

# Add some python functions to the global python object for code in this file to use.

globalThis = pm.eval("globalThis;", evalOpts)

pm.eval("globalThis.python = { pythonMonkey: {}, stdout: {}, stderr: {} }", evalOpts)

globalThis.pmEval = pm.eval

globalThis.python.pythonMonkey.dir = os.path.dirname(__file__)

globalThis.python.pythonMonkey.isCompilableUnit = pm.isCompilableUnit

globalThis.python.pythonMonkey.nodeModules = node_modules

globalThis.python.print = print

globalThis.python.stdout.write = lambda s: sys.stdout.write(s)

globalThis.python.stderr.write = lambda s: sys.stderr.write(s)

globalThis.python.stdout.read = lambda n: sys.stdout.read(n)

globalThis.python.stderr.read = lambda n: sys.stderr.read(n)

# Python 3.13 dramatically changed how the namespace in `exec`/`eval` works

# See https://docs.python.org/3.13/whatsnew/3.13.html#defined-mutation-semantics-for-locals

globalThis.python.eval = lambda x: eval(str(x)[:], None, sys._getframe(1).f_locals)

globalThis.python.exec = lambda x: exec(str(x)[:], None, sys._getframe(1).f_locals)

globalThis.python.getenv = os.getenv

globalThis.python.paths = sys.path

globalThis.python.exit = pm.eval("""'use strict';

(exit) => function pythonExitWrapper(exitCode) {

if (typeof exitCode === 'undefined')

exitCode = pythonExitWrapper.code;

if (typeof exitCode == 'undefined')

exitCode = 0n;

if (typeof exitCode === 'number')

exitCode = BigInt(Math.floor(exitCode));

if (typeof exitCode !== 'bigint')

exitCode = 1n;

exit(exitCode);

}

""", evalOpts)(sys.exit)

# bootstrap is effectively a scoping object which keeps us from polluting the global JS scope.

# The idea is that we hold a reference to the bootstrap object in Python-load, for use by the

# innermost code in ctx-module, without forcing ourselves to expose this minimalist code to

# userland-require

bootstrap = pm.eval("""

'use strict'; (function IIFE(python) {

const bootstrap = {

modules: {

vm: {},

'ctx-module': {},

}

/* ctx-module needs require() when it loads that can find vm and fs */

bootstrap.require = function bootstrapRequire(mid)

{

if (bootstrap.modules.hasOwnProperty(mid))

return bootstrap.modules[mid];

if (bootstrap.modules['ctx-module'].CtxModule)

return bootstrap.requireFromDisk(mid);

const error = new Error('module not found: ' + mid);

error = 'MODULE_NOT_FOUND';

throw error;

}

bootstrap.modules.vm.runInContext = function runInContext(code, _unused_contextifiedObject, options)

{

var evalOptions = {};

if (arguments.length === 2)

options = arguments[2];

if (options.filename) evalOptions.filename = options.filename;

if (options.lineOffset) evalOptions.lineno = options.lineOffset;

if (options.columnOffset) evalOptions.column = options.columnOffset;

return pmEval(code, evalOptions);

}

/**

* The debug module has as its exports a function which may, depending on the DEBUG env var, emit

* debugging statements to stdout. This is quick implementation of the node built-in.

bootstrap.modules.debug = function debug(selector)

{

var idx, colour;

const esc = String.fromCharCode(27);

const noColour = `${esc}[0m`;

debug.selectors = debug.selectors || [];

idx = debug.selectors.indexOf(selector);

if (idx === -1)

{

idx = debug.selectors.length;

debug.selectors.push(selector);

}

colour = `${esc}[${91 + ((idx + 1) % 6)}m`;

const debugEnv = python.getenv('DEBUG');

if (debugEnv)

{

for (let sym of debugEnv.split(' '))

{

const re = new RegExp('^' + sym.replace('*', '.*') + '$');

if (re.test(selector))

{

return (function debugInner() {

var output;

if (!bootstrap.inspect)

output = Array.from(arguments).join(' ');

else

output = Array.from(arguments).map(x => {

if (typeof x === 'string' || x instanceof String)

return x;

return bootstrap.inspect(x);

}).join(' ');

output = output.split('\\n');

python.print(`${colour}${selector}${noColour} ` + output[0]);

const spaces = ''.padEnd(selector.length + 1, ' ');

for (let i=1; i < output.length; i++)

python.print(spaces + output[i]);

});

}

/* no match => silent */

return (function debugDummy() {});

}

/**

* The fs module is like the Node.js fs module, except it only implements exactly what the ctx-module

* module requires to load CommonJS modules. It is augmented heavily below by Python methods.

bootstrap.modules.fs = {

constants: { S_IFDIR: 16384 },

statSync: function statSync(filename) {

const ret = bootstrap.modules.fs.statSync_inner(filename);

if (ret)

return ret;

const err = new Error('file not found: ' + filename);

err.code='ENOENT';

throw err;

};

/* Modules which will be available to all requires */

bootstrap.builtinModules = { debug: bootstrap.modules.debug };

return bootstrap;

})(globalThis.python)""", evalOpts)

def statSync_inner(filename: str) -> Union[Dict[str, int], bool]:

"""

Inner function for statSync.

Returns:

Union[Dict[str, int], False]: The mode of the file or False if the file doesn't exist.

"""

from os import stat

filename = os.path.normpath(filename)

if (os.path.exists(filename)):

sb = stat(filename)

return {'mode': sb.st_mode}

else:

return False

def readFileSync(filename, charset) -> str:

"""

Utility function for reading files.

Returns:

str: The contents of the file

"""

filename = os.path.normpath(filename)

with open(filename, "r", encoding=charset) as fileHnd:

return fileHnd.read()

def existsSync(filename: str) -> bool:

filename = os.path.normpath(filename)

return os.path.exists(filename)

bootstrap.modules.fs.statSync_inner = statSync_inner

bootstrap.modules.fs.readFileSync = readFileSync

bootstrap.modules.fs.existsSync = existsSync

# Read ctx-module module from disk and invoke so that this file is the "main module" and ctx-module has

# require and exports symbols injected from the bootstrap object above. Current PythonMonkey bugs

# prevent us from injecting names properly so they are stolen from trail left behind in the global

# scope until that can be fixed.

# lineno should be -5 but jsapi 102 uses unsigned line numbers, so we take the newlines out of the

# wrapper prologue to make stack traces line up.

with open(node_modules + "/ctx-module/ctx-module.js", "r") as ctxModuleSource:

initCtxModule = pm.eval("""'use strict';

(function moduleWrapper_forCtxModule(require, exports)

{

""" + ctxModuleSource.read() + """

})

""", {'filename': node_modules + "/ctx-module/ctx-module.js", 'lineno': 0})

initCtxModule(bootstrap.require, bootstrap.modules['ctx-module'])

def load(filename: str) -> Dict:

"""

Loads a python module using the importlib machinery sourcefileloader, prefills it with an exports object and returns

the module.

If the module is already loaded, returns it.

Args:

filename (str): The filename of the python module to load.

Returns:

: The loaded python module

"""

filename = os.path.normpath(filename)

name = os.path.basename(filename)

if name not in sys.modules:

sourceFileLoader = machinery.SourceFileLoader(name, filename)

spec: machinery.ModuleSpec = importlib.util.spec_from_loader(

sourceFileLoader.name, sourceFileLoader) # type: ignore

module = importlib.util.module_from_spec(spec)

sys.modules[name] = module

module.exports = {} # type: ignore

spec.loader.exec_module(module) # type: ignore

else:

module = sys.modules[name]

return module.exports

globalThis.python.load = load

createRequireInner = pm.eval("""'use strict';(

/**

* Factory function which returns a fresh 'require' function. The module cache will inherit from

* globalThis.require, assuming it has been defined.

* @param {string} filename the filename of the module that would get this require

* @param {object} bootstrap the bootstrap context; python imports, modules, etc, that are invisible

* to normal JS

* @param {string} extraPaths colon-delimited list of paths to add to require.path

* @param {boolean} isMain true if the module is to be used as a program module

* @returns {function} require

function createRequireInner(filename, bootstrap, extraPaths, isMain)

{

const CtxModule = bootstrap.modules['ctx-module'].CtxModule;

const moduleCache = globalThis.require?.cache || {};

function loadPythonModule(module, filename)

{

module.exports = python.load(filename);

}

// TODO - find a better way to deal with Windows paths

if (filename)

filename = filename.split('\\\\').join('/');

if (moduleCache[filename])

return moduleCache[filename].require;

const module = new CtxModule(globalThis, filename, moduleCache);

if (!filename)

module.paths = []; /* fully virtual module - no module.path or module.paths */

else

{

moduleCache[filename] = module;

for (let path of Array.from(python.paths))

module.paths.push(path + '/node_modules');

}

module.require.path.push(python.pythonMonkey.dir + '/builtin_modules');

module.require.path.push(python.pythonMonkey.nodeModules);

/* Add a .py loader, making it the first extension to be enumerated so py modules take precedence */

const extCopy = Object.assign({}, module.require.extensions);

for (let ext in module.require.extensions)

delete module.require.extensions[ext];

module.require.extensions['.py'] = loadPythonModule;

Object.assign(module.require.extensions, extCopy);

if (isMain)

{

globalThis.module = module;

globalThis.exports = module.exports;

globalThis.require = module.require;

globalThis.require.main = module;

module.loaded = true;

moduleCache[filename] = module;

}

if (extraPaths)

module.require.path.splice(module.require.path.length, 0, ...(extraPaths.split(',')));

return module.require;

})""", evalOpts)

# API: pm.createRequire

# We cache the return value of createRequire to always use the same require for the same filename

def createRequire(filename, extraPaths: Union[List[str], Literal[False]] = False, isMain=False):

"""

returns a require function that resolves modules relative to the filename argument.

Conceptually the same as node:module.createRequire().

example:

from pythonmonkey import createRequire

require = createRequire(__file__)

require('./my-javascript-module')

"""

fullFilename: str = os.path.abspath(filename)

if (extraPaths):

extraPathsStr = ':'.join(extraPaths)

else:

extraPathsStr = ''

return createRequireInner(fullFilename, bootstrap, extraPathsStr, isMain)

bootstrap.requireFromDisk = createRequireInner(None, bootstrap, '', False)

bootstrap.inspect = bootstrap.requireFromDisk('util').inspect

# API: pm.runProgramModule

def runProgramModule(filename, argv, extraPaths=[]):

"""

Run the program module. This loads the code from disk, sets up the execution environment, and then

invokes the program module (aka main module). The program module is different from other modules in that

1. it cannot return (must throw)

2. the outermost block scope is the global scope, effectively making its scope a super-global to

other modules

"""

fullFilename = os.path.abspath(filename)

createRequire(fullFilename, extraPaths, True)

globalThis.__filename = fullFilename

globalThis.__dirname = os.path.dirname(fullFilename)

with open(fullFilename, encoding="utf-8", mode="r") as mainModuleSource:

pm.eval(mainModuleSource.read(), {'filename': fullFilename, 'noScriptRval': True})

# forcibly run in file mode. We shouldn't be getting the last expression of the script as the result value.

# The pythonmonkey require export. Every time it is used, the stack is inspected so that the filename

# passed to createRequire is correct. This is necessary so that relative requires work. If the filename

# found on the stack doesn't exist, we assume we're in the REPL or something and simply use the current

# directory as the location of a virtual module for relative require purposes.

# todo: instead of cwd+__main_virtual__, use a full pathname which includes the directory that the

# running python program is in.

def require(moduleIdentifier: str):

filename = inspect.stack()[1].filename

if not os.path.exists(filename):

filename = os.path.join(os.getcwd(), "__main_virtual__")

return createRequire(filename)(moduleIdentifier)

# Restrict what symbols are exposed to the pythonmonkey module.

__all__ = ["globalThis", "require", "createRequire", "runProgramModule", "bootstrap"]