Nokogiri::XML::Reader broken (as a pull parser) on jruby - reads the whole XML document

[ghost]

$ ruby -v
jruby 1.6.7.2 (ruby-1.8.7-p357) (2012-05-01 26e08ba) (OpenJDK 64-Bit Server VM 1.6.0_24) [linux-amd64-java]

$ gem list | grep nokogiri
nokogiri (1.5.6 java)

$ ruby -e 'f = open("test.xml", "w"); f.write("<root>"); (10**8).times{ f.write("<nil/>")}; f.write("</root>"); f.close'

$ ls -lh test.xml
-rw-rw-r-- 1 test test 573M jan 11 12:08 test.xml

$ time ruby -rubygems -e 'require "nokogiri";Nokogiri::XML::Reader.from_io open("test.xml", "r")'
Error: Your application used more memory than the safety cap of 500M.
Specify -J-Xmx####m to increase it (#### = cap size in MB).
Specify -w for full OutOfMemoryError stack trace

real    0m2.950s
user    0m5.640s
sys 0m0.156s

$ rvm use system
$ ruby -v
ruby 1.9.3p327 (2012-11-10) [x86_64-linux] Brightbox

$ gem list | grep nokogiri
nokogiri (1.5.6)

$ time ruby -rubygems -e 'require "nokogiri";Nokogiri::XML::Reader.from_io open("test.xml", "r")'

real    0m0.075s
user    0m0.068s
sys 0m0.004s

[jvshahid]

This issues turned to be much harder to fix. My test fooled my into thinking that the issue was fixed, when it wasn't. I'm working on the fix, reopening the issues until the fix is ready.

[johnsinco]

Hi. Is this fixed? I'm not really sure from the comments. We ran into this issue a year ago and had to parse our xml data by hand with string parsing, It'd be great to use Nokogiri for this. Do you think it should work on any jruby version? (1.9.3 mode) Thanks!

[jvshahid]

Yes, it should work. The changes aren't released yet, but you can use the pre release version 1.5.7.rc3. If you're still having problems, feel free to open a new issue.