uncaught throw `encoding_found' in JRuby

[nirvdrum]

Hi,

I just upgraded to 1.5.3rc5 and started seeing:

ArgumentError:
  uncaught throw `encoding_found'
  # nokogiri/XmlSaxParserContext.java:241:in `parse_with'

It's a bit tricky to reproduce, but I think a reliable way is to save an HTML doc to disk without any any BOM or encoding info in the file. Load the file into JRuby in 1.9 mode so it should have an encoding of "ASCII-8BIT". Now just do Nokogiri::HTML(File.open('junk_file').read) and you should get the exception. Note that the test file I'm using does have newlines in it. I'm not quite sure what character is throwing it off. If I change the encoding to UTF-8, Nokogiri is able to construct the document just fine.

[yokolet]

I'm aware of this. Nokogiri test on JRuby 1.9 mode always throws this error message. I'll work on this.

[nirvdrum]

Confirming that this is still an issue in 1.5.5.rc1.

[yokolet]

Sorry, I couldn't make it in 1.5.5.rc1. But, I fixed this in rev. c085082, so 1.5.5.rc2 or 1.5.5 will have the fix.

Thanks for reporting the bug.