Simplify _scriptName / scriptDirectory handling

- Avoid defining `_scriptName` outside the wrapper function in
  `EXPORT_ES6` mode and for Node.js CommonJS builds.
- Simplify `scriptDirectory` resolution to
  `new URL('.', _scriptName).href` on the web.

Author: kleisauke

src/closure-externs/modularize-externs.js

@@ -1,6 +1,7 @@
-// Due to the way MODULARIZE works, Closure is run on generated code that does not define _scriptName,
-// but only after MODULARIZE has finished, _scriptName is injected to the generated code.
-// Therefore it cannot be minified.
+// In MODULARIZE mode the JS code may be executed later, after `document.currentScript` is gone, so we store
+// it to `_scriptName` outside the wrapper function. Therefore, it cannot be minified.
+// In EXPORT_ES6 mode we use `import.meta.url` and for Node.js CommonJS builds we use `__filename`.
+
 /**
  * @suppress {duplicate, undefinedVars}
  */

src/shell.js

@@ -139,24 +139,32 @@ var quit_ = (status, toThrow) => {
   throw toThrow;
 };
 
-#if SHARED_MEMORY && !MODULARIZE
+#if EXPORT_ES6
+var _scriptName = import.meta.url;
+#else
+#if ENVIRONMENT_MAY_BE_WEB
+#if !MODULARIZE
 // In MODULARIZE mode _scriptName needs to be captured already at the very top of the page immediately when the page is parsed, so it is generated there
 // before the page load. In non-MODULARIZE modes generate it here.
-var _scriptName = (typeof document != 'undefined') ? document.currentScript?.src : undefined;
+var _scriptName = typeof document != 'undefined' ? document.currentScript?.src : undefined;
+#endif // !MODULARIZE
+#elif ENVIRONMENT_MAY_BE_NODE || ENVIRONMENT_MAY_BE_WORKER
+var _scriptName;
+#endif // ENVIRONMENT_MAY_BE_WEB
 
 #if ENVIRONMENT_MAY_BE_NODE
-if (ENVIRONMENT_IS_NODE) {
-#if EXPORT_ES6
-  _scriptName = typeof __filename != 'undefined' ? __filename : import.meta.url
-#else
+if (typeof __filename != 'undefined') { // Node
   _scriptName = __filename;
-#endif
 } else
 #endif // ENVIRONMENT_MAY_BE_NODE
+#if ENVIRONMENT_MAY_BE_WORKER
 if (ENVIRONMENT_IS_WORKER) {
   _scriptName = self.location.href;
 }
-#endif // SHARED_MEMORY && !MODULARIZE
+#elif ENVIRONMENT_MAY_BE_NODE
+  /*no-op*/{}
+#endif // ENVIRONMENT_MAY_BE_WORKER
+#endif // EXPORT_ES6
 
 // `/` should be present at the end if `scriptDirectory` is not empty
 var scriptDirectory = '';
@@ -197,10 +205,7 @@ if (ENVIRONMENT_IS_NODE) {
   var nodePath = require('path');
 
 #if EXPORT_ES6
-  // EXPORT_ES6 + ENVIRONMENT_IS_NODE always requires use of import.meta.url,
-  // since there's no way getting the current absolute path of the module when
-  // support for that is not available.
-  if (!import.meta.url.startsWith('data:')) {
+  if (import.meta.url.startsWith('file:')) {
     scriptDirectory = nodePath.dirname(require('url').fileURLToPath(import.meta.url)) + '/';
   }
 #else
@@ -335,28 +340,11 @@ if (ENVIRONMENT_IS_SHELL) {
 // ENVIRONMENT_IS_NODE.
 #if ENVIRONMENT_MAY_BE_WEB || ENVIRONMENT_MAY_BE_WORKER
 if (ENVIRONMENT_IS_WEB || ENVIRONMENT_IS_WORKER) {
-  if (ENVIRONMENT_IS_WORKER) { // Check worker, not web, since window could be polyfilled
-    scriptDirectory = self.location.href;
-  } else if (typeof document != 'undefined' && document.currentScript) { // web
-    scriptDirectory = document.currentScript.src;
-  }
-#if MODULARIZE
-  // When MODULARIZE, this JS may be executed later, after document.currentScript
-  // is gone, so we saved it, and we use it here instead of any other info.
-  if (_scriptName) {
-    scriptDirectory = _scriptName;
-  }
-#endif
-  // blob urls look like blob:http://site.com/etc/etc and we cannot infer anything from them.
-  // otherwise, slice off the final part of the url to find the script directory.
-  // if scriptDirectory does not contain a slash, lastIndexOf will return -1,
-  // and scriptDirectory will correctly be replaced with an empty string.
-  // If scriptDirectory contains a query (starting with ?) or a fragment (starting with #),
-  // they are removed because they could contain a slash.
-  if (scriptDirectory.startsWith('blob:')) {
-    scriptDirectory = '';
-  } else {
-    scriptDirectory = scriptDirectory.slice(0, scriptDirectory.replace(/[?#].*/, '').lastIndexOf('/')+1);
+  try {
+    scriptDirectory = new URL('.', _scriptName).href; // includes trailing slash
+  } catch {
+    // Must be a `blob:` or `data:` URL (e.g. `blob:http://site.com/etc/etc`), we cannot
+    // infer anything from them.
   }
 
 #if ENVIRONMENT && ASSERTIONS

src/shell_minimal.js

@@ -134,7 +134,7 @@ var ENVIRONMENT_IS_PTHREAD = ENVIRONMENT_IS_WORKER && self.name?.startsWith('em-
 #if !MODULARIZE
 // In MODULARIZE mode _scriptName needs to be captured already at the very top of the page immediately when the page is parsed, so it is generated there
 // before the page load. In non-MODULARIZE modes generate it here.
-var _scriptName = (typeof document != 'undefined') ? document.currentScript?.src : undefined;
+var _scriptName = typeof document != 'undefined' ? document.currentScript?.src : undefined;
 #endif
 
 #if ENVIRONMENT_MAY_BE_NODE
@@ -145,9 +145,11 @@ if (ENVIRONMENT_IS_NODE) {
   // Under node we set `workerData` to `em-pthread` to signal that the worker
   // is hosting a pthread.
   ENVIRONMENT_IS_PTHREAD = ENVIRONMENT_IS_WORKER && worker_threads['workerData'] == 'em-pthread'
+#if !EXPORT_ES6
   _scriptName = __filename;
-} else
 #endif
+} else
+#endif // ENVIRONMENT_MAY_BE_NODE
 if (ENVIRONMENT_IS_WORKER) {
   _scriptName = self.location.href;
 }

src/wasm_worker.js

@@ -34,8 +34,11 @@ if (ENVIRONMENT_IS_NODE) {
   Object.assign(global, {
     self: global,
     require,
+#if !EXPORT_ES6
+    // `vm.runInThisContext` global scope lacks `__filename` and `__dirname`
     __filename,
     __dirname,
+#endif
     Worker: nodeWorkerThreads.Worker,
     importScripts: (f) => vm.runInThisContext(fs.readFileSync(f, 'utf8'), {filename: f}),
     postMessage: (msg) => parentPort.postMessage(msg),

test/test_other.py

@@ -15895,3 +15895,48 @@ def test_mainScriptUrlOrBlob(self, es6):
     # Now create foo.[m]js and the program should run as expected.
     shutil.copy(outfile, ('foo.%s' % ext))
     self.assertContained('hello, world', self.run_js(outfile))
+
+  @parameterized({
+    '': ([],),
+    'node': (['-sENVIRONMENT=node'],),
+  })
+  def test_locate_file_abspath(self, args):
+    # Verify that `scriptDirectory` is an absolute path
+    create_file('pre.js', '''
+      Module['locateFile'] = (fileName, scriptDirectory) => {
+        assert(nodePath['isAbsolute'](scriptDirectory), `scriptDirectory (${scriptDirectory}) should be an absolute path`);
+        return scriptDirectory + fileName;
+      };
+      ''')
+    self.do_runf('hello_world.c', 'hello, world!', emcc_args=['--pre-js', 'pre.js'] + args)
+
+  @parameterized({
+    '': ([],),
+    'node': (['-sENVIRONMENT=node'],),
+  })
+  def test_locate_file_abspath_esm(self, args):
+    # Verify that `scriptDirectory` is an absolute path when `EXPORT_ES6`
+    create_file('pre.js', '''
+      Module['locateFile'] = (fileName, scriptDirectory) => {
+        assert(nodePath['isAbsolute'](scriptDirectory), `scriptDirectory (${scriptDirectory}) should be an absolute path`);
+        return scriptDirectory + fileName;
+      };
+      ''')
+    self.run_process([EMCC, '-o', 'hello_world.mjs',
+                      '--pre-js', 'pre.js',
+                      '--extern-post-js', test_file('modularize_post_js.js'),
+                      test_file('hello_world.c')] + args)
+    self.assertContained('hello, world!', self.run_js('hello_world.mjs'))
+
+  @node_pthreads
+  def test_locate_file_abspath_pthread(self):
+    # Verify that `scriptDirectory` is an absolute path when `ENVIRONMENT_IS_WORKER`
+    self.set_setting('PROXY_TO_PTHREAD')
+    self.set_setting('EXIT_RUNTIME')
+    create_file('pre.js', '''
+      Module['locateFile'] = (fileName, scriptDirectory) => {
+        assert(nodePath['isAbsolute'](scriptDirectory), `scriptDirectory (${scriptDirectory}) should be an absolute path`);
+        return scriptDirectory + fileName;
+      };
+      ''')
+    self.do_runf('hello_world.c', 'hello, world!', emcc_args=['--pre-js', 'pre.js'])

tools/building.py

@@ -555,7 +555,7 @@ def closure_compiler(filename, advanced=True, extra_closure_args=None):
   # should not minify these symbol names.
   CLOSURE_EXTERNS = [path_from_root('src/closure-externs/closure-externs.js')]
 
-  if settings.MODULARIZE:
+  if settings.MODULARIZE and settings.ENVIRONMENT_MAY_BE_WEB and not settings.EXPORT_ES6:
     CLOSURE_EXTERNS += [path_from_root('src/closure-externs/modularize-externs.js')]
 
   if settings.USE_WEBGPU:

tools/link.py

@@ -2459,30 +2459,22 @@ def modularize():
     # document.currentScript, so a simple export declaration is enough.
     src = f'var {settings.EXPORT_NAME} = {wrapper_function};'
   else:
-    script_url_node = ''
+    script_url_web = ''
     # When MODULARIZE this JS may be executed later,
     # after document.currentScript is gone, so we save it.
-    # In EXPORT_ES6 + PTHREADS the 'thread' is actually an ES6 module
-    # webworker running in strict mode, so doesn't have access to 'document'.
-    # In this case use 'import.meta' instead.
-    if settings.EXPORT_ES6:
-      script_url = 'import.meta.url'
-    else:
-      script_url = "typeof document != 'undefined' ? document.currentScript?.src : undefined"
-      if settings.ENVIRONMENT_MAY_BE_NODE:
-        script_url_node = "if (typeof __filename != 'undefined') _scriptName = _scriptName || __filename;"
+    # In EXPORT_ES6 mode we can just use 'import.meta.url'.
+    if settings.ENVIRONMENT_MAY_BE_WEB and not settings.EXPORT_ES6:
+       script_url_web = "var _scriptName = typeof document != 'undefined' ? document.currentScript?.src : undefined;"
     src = '''\
 var %(EXPORT_NAME)s = (() => {
-  var _scriptName = %(script_url)s;
-  %(script_url_node)s
+  %(script_url_web)s
   return (%(wrapper_function)s);
 })();
 ''' % {
-        'EXPORT_NAME': settings.EXPORT_NAME,
-        'script_url': script_url,
-        'script_url_node': script_url_node,
-        'wrapper_function': wrapper_function,
-      }
+      'EXPORT_NAME': settings.EXPORT_NAME,
+      'script_url_web': script_url_web,
+      'wrapper_function': wrapper_function,
+    }
 
   if settings.SOURCE_PHASE_IMPORTS:
     src = f"import source wasmModule from './{settings.WASM_BINARY_FILE}';\n\n" + src