scriptDirectory handling improvements (#6894)

* handle the case of a missing document.currentScript, and of the MODULARIZE .bind() being overridden by `new` (see discussion in #6903). in both cases we can't find the scriptDirectory, but we should at least not crash.

* generalize blob url handling in scriptDirectory computation to also handle workers

Author: kripken

emcc.py

@@ -2612,8 +2612,10 @@ def modularize():
 
   return %(EXPORT_NAME)s;
 };
+// When MODULARIZE, this JS may be executed later, after document.currentScript
+// is gone, so we save it.
 %(EXPORT_NAME)s = %(EXPORT_NAME)s.bind({
-  _currentScript: typeof document !== 'undefined' ? document.currentScript : undefined
+  _scriptDir: typeof document !== 'undefined' && document.currentScript ? document.currentScript.src : undefined
 })%(instantiate)s;
 ''' % {
     'EXPORT_NAME': shared.Settings.EXPORT_NAME,

src/shell.js

@@ -230,18 +230,25 @@ if (ENVIRONMENT_IS_SHELL) {
 #if ENVIRONMENT_MAY_BE_WEB_OR_WORKER
 if (ENVIRONMENT_IS_WEB || ENVIRONMENT_IS_WORKER) {
   if (ENVIRONMENT_IS_WEB) {
+    if (document.currentScript) {
+      scriptDirectory = document.currentScript.src;
+    }
+  } else { // worker
+    scriptDirectory = self.location.href;
+  }
 #if MODULARIZE
-    // When MODULARIZE, this JS may be executed later, after document.currentScript is gone, so we send it
-    // using this._currentScript.
-    var currentScript = this['_currentScript'] || document.currentScript;
-#else
-    var currentScript = document.currentScript;
+  // 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 (this['_scriptDir']) {
+    scriptDirectory = this['_scriptDir'];
+  }
 #endif
-    if (currentScript.src.indexOf('blob:') !== 0) {
-      scriptDirectory = currentScript.src.split('/').slice(0, -1).join('/') + '/';
-    }
-  } else if (ENVIRONMENT_IS_WORKER) {
-    scriptDirectory = self.location.href.split('/').slice(0, -1).join('/') + '/';
+  // 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.indexOf('blob:') !== 0) {
+    scriptDirectory = scriptDirectory.split('/').slice(0, -1).join('/') + '/';
+  } else {
+    scriptDirectory = '';
   }
 
 #if ENVIRONMENT

tests/test_browser.py

@@ -3972,3 +3972,34 @@ def test_browser_run_from_different_directory_async(self):
     ''')
 
     self.run_browser('test-subdir.html', None, '/report_result?0')
+
+  # Similar to `test_browser_run_from_different_directory`, but
+  # also also we eval the initial code, so currentScript is not present. That prevents us
+  # from finding the file in a subdir, but here we at least check we do not regress compared to the
+  # normal case of finding in the current dir.
+  # In addition, check for new Module(), which overrides the bind() and replaces the object
+  # which saved the _scriptDir. Again, we can't get the script dir that way, but at least we
+  # should not regress compared to the normal case.
+  def test_browser_modularize_no_current_script(self):
+    src = open(path_from_root('tests', 'browser_test_hello_world.c')).read()
+    open('test.c', 'w').write(self.with_report_result(src))
+    # compile the code with the modularize feature and the preload-file option enabled
+    Popen([PYTHON, EMCC, 'test.c', '-o', 'test.js', '-s', 'MODULARIZE=1']).communicate()
+    for creation in (
+      'Module();',
+      'new Module();'
+    ):
+      print(creation)
+      open('test.html', 'w').write('''
+        <script>
+          setTimeout(function() {
+            var xhr = new XMLHttpRequest();
+            xhr.open('GET', 'test.js', false);
+            xhr.send(null);
+            eval(xhr.responseText);
+            %s
+          }, 1);
+        </script>
+      ''' % creation)
+      self.run_browser('test.html', None, '/report_result?0')
+