SYMPTOMS
Internet Explorer reports an "Invalid character" error when a Unicode .js file is included with the SCRIPT tag, for example:
<SCRIPT LANGUAGE="JavaScript" SRC="Sample.JS">
If the script debugger is enabled, it reports that the first two characters in the .js file (the Unicode Byte Order Mark, or BOM) are the source of the problem.
Another closely associated problem occurs if the Byte Order Mark is removed from a Unicode format .js file and the file contains Unicode characters. When the .js file is included during page rendering by Internet Explorer, Unicode characters in the .js file are interpreted as simple ASCII and cause either a script error, as above, or the display of garbage characters. Similarly, removing the BOM from a UTF-8 encoded file containing non-ASCII Unicode characters results in garbage characters being displayed on the rendered Web page.
Note that in the Internet Explorer
Advanced dialog box (from the
Tools menu, select
Internet Options), the option
Disable script debugging must be cleared or the option
Display a notification about every script error must be selected in order to receive a detailed error message for this problem. If these two options are not set correctly, the Internet Explorer status bar indicates that there are errors on the page. The page loads, but the .js file does not.
RESOLUTION
It is possible to work around the problem, in most cases, by embedding Unicode escape sequences in the .js file that is saved in ASCII. For instance, the following JScript statement may be used in an ASCII .js file to specify the Hiragana character
NO.
var s = String("Hiragana NO = \u306E");
When the ASCII .js file is loaded from the SCRIPT tag there is, of course, no Unicode Byte Order Mark, and the \uXXXX escapes are converted to proper Unicode characters that appear on a Web page.