3

アプレットがあり、ファイルへのストリームを開く必要があります。このファイルは、アプレットと HTML ファイルがある場所にあるローカル ファイルです。

    URL localURL = new URL(getCodeBase(), "pixs/icons.zip");
    InputStream localInputStream =localURL.openStream();

以前は正常に動作していましたが、Java 1.7 ビルド 25 にアップグレードした後、getCodeBase() は常に null を返します。これは実際に文書化されています!残念ながら、それを克服する方法は推奨されていません。

うまくいったことの1つは、フルパスを使用することです：

   URL localURL = new URL("file:c:/myFolder/pixs/icons.zip");

フルパスを使用せずにそれを解決する別のオプションはありますか?

4

1 に答える 1

2

Perhaps you can use getDocumentBase instead. Depends on the structure of your setup, whether there is a close relation between the code base and the documents. No permanent solution: getDocumentBase was modified in the same way in 7u40, according to bug #8019177.

If not, then you can try using getResource to obtain a URL from your JAR, e.g. the class file of the applet, and then disassemble that URL to get at the location of the JAR and hence the code base. This is untested, so feel free to edit this post if you tried this.

Last but not least, since that change only affects local applets, you could run a (local or public) web server to serve that applet.

If you want a more official statement on this, I'll quote the #8017250 bug report:

If applet need to load resource:

  • if the resource is in applet JAR(s), they should be able to load it with ClassLoader.getResoruceAsStream directly, without needing the codebase information.
  • if the resource is in arbitary location, not inside applet JAR, they should have other ways to get to that location, since it's not part of the applet resource anyway. (e.g. user.home java system property, provided that their applet has all-permissions)

http://www.duckware.com/tech/java-security-clusterfuck.html (thanks to this post) mentions some other alternatives. The least likely one to be affected by future Oracle modifications apprers to be the use of location.href in the containing HTML page, e.g. writing the <applet> tag from JavaScript.

于 2013-09-02T10:27:33.710 に答える