[BUG] Segmentation fault

[lls]

trying this code just crashes:

require 'rubygems'
require 'nokogiri'
document = Nokogiri::HTML.parse("  <div style=\"overflow: hidden; float: left; margin: 2px 0;\">\n    <a href=\"http://aaaa.aaaa.aaaa/aaaaa/aaaaa_aaaaa/aaaaaa/aaaaaa/aaaaaaaaaaaa/aaaaaa/aaaaaaaaaa/view\">\n      <div class=\"aaaa-aaaaa\" title=\"000_0000.jpg\" style=\"border: 1px solid #666; height: 62px; width: 62px; overflow: hidden; background-color: white;\"><a href=\"http://aaaa.aaaaa.aaa/aaaa/aaaaa_aaaaa/aaaa/aaaa/aaaaaaaaa/item/aaaaaaaaaaaa/view\"><img src=\"http://a.aaaa.aaa/aa/aaaaaa/aaaaaaaaaa/tn/aaaaaaa/name/000_0000.jpg\" style=\"border: none;\" alt=\"000_0000.jpg\"/></a></div>\n    <div class=\"aaaa-aaaa-aaaa\" style=\"clear: both; font-size: smaller; height: 12px; overflow: hidden; text-align: center; width: 72px;\"> <a style=\"text-decoration: none;\" href=\"http://aaaa.aaaa/aaaa/aaaaa_aaaaaa/aaaa/aaaa/aaaaaaaaaaaa/item/aaaaaaaaaaaaa/view\" title=\"000_0000.jpg\">000_0000.jpg</a></div>\n  </div>\n\t      \t       \t   ")
document.search('//a[@href]').each do |link|
      p link
      title = link.content
      puts "*************"
      p title
      link.content = title.nil? ? "":title.strip
    end

just execute in irb:

irb
$ irb(main):001:0> require 'rubygems'
=> true
irb(main):002:0> require 'nokogiri'
=> true
irb(main):003:0> document = Nokogiri::HTML.parse("  <div style=\"overflow: hidden; float: left; margin: 2px 0;\">\n    <a href=\"http://aaaa.aaaa.aaaa/aaaaa/aaaaa_aaaaa/aaaaaa/aaaaaa/aaaaaaaaaaaa/aaaaaa/aaaaaaaaaa/view\">\n      <div class=\"aaaa-aaaaa\" title=\"000_0000.jpg\" style=\"border: 1px solid #666; height: 62px; width: 62px; overflow: hidden; background-color: white;\"><a href=\"http://aaaa.aaaaa.aaa/aaaa/aaaaa_aaaaa/aaaa/aaaa/aaaaaaaaa/item/aaaaaaaaaaaa/view\"><img src=\"http://a.aaaa.aaa/aa/aaaaaa/aaaaaaaaaa/tn/aaaaaaa/name/000_0000.jpg\" style=\"border: none;\" alt=\"000_0000.jpg\"/></a></div>\n    <div class=\"aaaa-aaaa-aaaa\" style=\"clear: both; font-size: smaller; height: 12px; overflow: hidden; text-align: center; width: 72px;\"> <a style=\"text-decoration: none;\" href=\"http://aaaa.aaaa/aaaa/aaaaa_aaaaaa/aaaa/aaaa/aaaaaaaaaaaa/item/aaaaaaaaaaaaa/view\" title=\"000_0000.jpg\">000_0000.jpg</a></div>\n  </div>\n\t      \t       \t   ")
=> #<Nokogiri::HTML::Document:0x3fdbeb4d1848 name="document" children=[#<Nokogiri::XML::DTD:0x3fdbeb4d1514 name="html">, #<Nokogiri::XML::Element:0x3fdbeb4d14c4 name="html" children=[#<Nokogiri::XML::Element:0x3fdbeb4d0b50 name="body" children=[#<Nokogiri::XML::Element:0x3fdbeb4d0858 name="div" attributes=[#<Nokogiri::XML::Attr:0x3fdbeb4d077c name="style" value="overflow: hidden; float: left; margin: 2px 0;">] children=[#<Nokogiri::XML::Text:0x3fdbeb4d0330 "\n    ">, #<Nokogiri::XML::Element:0x3fdbeb4d02e0 name="a" attributes=[#<Nokogiri::XML::Attr:0x3fdbeb4cff84 name="href" value="http://aaaa.aaaa.aaaa/aaaaa/aaaaa_aaaaa/aaaaaa/aaaaaa/aaaaaaaaaaaa/aaaaaa/aaaaaaaaaa/view">] children=[#<Nokogiri::XML::Text:0x3fdbeb4cf520 "\n      ">, #<Nokogiri::XML::Element:0x3fdbeb4cf480 name="div" attributes=[#<Nokogiri::XML::Attr:0x3fdbeb4cef30 name="class" value="aaaa-aaaaa">, #<Nokogiri::XML::Attr:0x3fdbeb4cedf0 name="title" value="000_0000.jpg">, #<Nokogiri::XML::Attr:0x3fdbeb4ced64 name="style" value="border: 1px solid #666; height: 62px; width: 62px; overflow: hidden; background-color: white;">] children=[#<Nokogiri::XML::Element:0x3fdbeb4c9d64 name="a" attributes=[#<Nokogiri::XML::Attr:0x3fdbeb4c94f4 name="href" value="http://aaaa.aaaaa.aaa/aaaa/aaaaa_aaaaa/aaaa/aaaa/aaaaaaaaa/item/aaaaaaaaaaaa/view">] children=[#<Nokogiri::XML::Element:0x3fdbeb4c5174 name="img" attributes=[#<Nokogiri::XML::Attr:0x3fdbeb4c505c name="src" value="http://a.aaaa.aaa/aa/aaaaaa/aaaaaaaaaa/tn/aaaaaaa/name/000_0000.jpg">, #<Nokogiri::XML::Attr:0x3fdbeb4c5020 name="style" value="border: none;">, #<Nokogiri::XML::Attr:0x3fdbeb4c4ff8 name="alt" value="000_0000.jpg">]>]>]>, #<Nokogiri::XML::Text:0x3fdbeb4c44cc "\n    ">, #<Nokogiri::XML::Element:0x3fdbeb4c4468 name="div" attributes=[#<Nokogiri::XML::Attr:0x3fdbeb4c429c name="class" value="aaaa-aaaa-aaaa">, #<Nokogiri::XML::Attr:0x3fdbeb4c4288 name="style" value="clear: both; font-size: smaller; height: 12px; overflow: hidden; text-align: center; width: 72px;">] children=[#<Nokogiri::XML::Text:0x3fdbeb4c3b6c " ">, #<Nokogiri::XML::Element:0x3fdbeb4c3b08 name="a" attributes=[#<Nokogiri::XML::Attr:0x3fdbeb4c3978 name="style" value="text-decoration: none;">, #<Nokogiri::XML::Attr:0x3fdbeb4c3964 name="href" value="http://aaaa.aaaa/aaaa/aaaaa_aaaaaa/aaaa/aaaa/aaaaaaaaaaaa/item/aaaaaaaaaaaaa/view">, #<Nokogiri::XML::Attr:0x3fdbeb4c3950 name="title" value="000_0000.jpg">] children=[#<Nokogiri::XML::Text:0x3fdbeb4c3004 "000_0000.jpg">]>]>, #<Nokogiri::XML::Text:0x3fdbeb4c2dc0 "\n  ">]>]>]>]>]>
irb(main):004:0> document.search('//a[@href]').each do |link|
irb(main):005:1*       puts "-------------"
irb(main):006:1>       p link
irb(main):007:1>       title = link.content
irb(main):008:1>       puts "*************"
irb(main):009:1>       p title
irb(main):010:1>       link.content = title.nil? ? "":title.strip
irb(main):011:1>     end
-------------
#<Nokogiri::XML::Element:0x3fdbeb4d02e0 name="a" attributes=[#<Nokogiri::XML::Attr:0x3fdbeb4cff84 name="href" value="http://aaaa.aaaa.aaaa/aaaaa/aaaaa_aaaaa/aaaaaa/aaaaaa/aaaaaaaaaaaa/aaaaaa/aaaaaaaaaa/view">] children=[#<Nokogiri::XML::Text:0x3fdbeb4cf520 "\n      ">, #<Nokogiri::XML::Element:0x3fdbeb4cf480 name="div" attributes=[#<Nokogiri::XML::Attr:0x3fdbeb4cef30 name="class" value="aaaa-aaaaa">, #<Nokogiri::XML::Attr:0x3fdbeb4cedf0 name="title" value="000_0000.jpg">, #<Nokogiri::XML::Attr:0x3fdbeb4ced64 name="style" value="border: 1px solid #666; height: 62px; width: 62px; overflow: hidden; background-color: white;">] children=[#<Nokogiri::XML::Element:0x3fdbeb4c9d64 name="a" attributes=[#<Nokogiri::XML::Attr:0x3fdbeb4c94f4 name="href" value="http://aaaa.aaaaa.aaa/aaaa/aaaaa_aaaaa/aaaa/aaaa/aaaaaaaaa/item/aaaaaaaaaaaa/view">] children=[#<Nokogiri::XML::Element:0x3fdbeb4c5174 name="img" attributes=[#<Nokogiri::XML::Attr:0x3fdbeb4c505c name="src" value="http://a.aaaa.aaa/aa/aaaaaa/aaaaaaaaaa/tn/aaaaaaa/name/000_0000.jpg">, #<Nokogiri::XML::Attr:0x3fdbeb4c5020 name="style" value="border: none;">, #<Nokogiri::XML::Attr:0x3fdbeb4c4ff8 name="alt" value="000_0000.jpg">]>]>]>, #<Nokogiri::XML::Text:0x3fdbeb4c44cc "\n    ">, #<Nokogiri::XML::Element:0x3fdbeb4c4468 name="div" attributes=[#<Nokogiri::XML::Attr:0x3fdbeb4c429c name="class" value="aaaa-aaaa-aaaa">, #<Nokogiri::XML::Attr:0x3fdbeb4c4288 name="style" value="clear: both; font-size: smaller; height: 12px; overflow: hidden; text-align: center; width: 72px;">] children=[#<Nokogiri::XML::Text:0x3fdbeb4c3b6c " ">, #<Nokogiri::XML::Element:0x3fdbeb4c3b08 name="a" attributes=[#<Nokogiri::XML::Attr:0x3fdbeb4c3978 name="style" value="text-decoration: none;">, #<Nokogiri::XML::Attr:0x3fdbeb4c3964 name="href" value="http://aaaa.aaaa/aaaa/aaaaa_aaaaaa/aaaa/aaaa/aaaaaaaaaaaa/item/aaaaaaaaaaaaa/view">, #<Nokogiri::XML::Attr:0x3fdbeb4c3950 name="title" value="000_0000.jpg">] children=[#<Nokogiri::XML::Text:0x3fdbeb4c3004 "000_0000.jpg">]>]>, #<Nokogiri::XML::Text:0x3fdbeb4c2dc0 "\n  ">]>
*************
"\n      \n     000_0000.jpg\n  "
-------------
#<Nokogiri::XML::Text:0x3fdbeb27da74 "000_0000.jpg">
*************
"000_0000.jpg"
-------------
(irb):6: [BUG] Segmentation fault
ruby 1.8.7 (2008-08-11 patchlevel 72) [x86_64-linux]

Aborted

$

[lls]

using

• nokogiri 1.4.1
• ruby 1.8.7
• linux kernel 2.6.28-17

[tenderlove]

Thanks for the report! Can you also provide the output of:

$ nokogiri -v

[flavorjones]

I've reproduced. Investigating.

[flavorjones]

Here is a much simpler example that causes the crash:

document = Nokogiri::HTML <<-EOH
<div id='foo'>
  <div id='bar'></div>
</div>
EOH

divs = document.css("div")
divs.first.content = "hello"
divs.last.content  = "crash!"

[flavorjones]

or even:

...
divs = document.css("div")
divs.first.content = "hello"
puts divs.last.to_html

[tenderlove]

Ugh. I see the problem. Dude, this sucks.

[flavorjones]

unlinking children within Node#content= to prevent freeing memory out from under living Ruby objects. Closed by 982a43f.

[edgarjs]

Sorry is this fix released on any version? I'm getting this error when trying to parse the response from http://gdata.youtube.com/feeds/api/videos?q=ruby

on

Gems/1.8/gems/nokogiri-1.4.1/lib/nokogiri/xml/sax/parser.rb:108: [BUG] Segmentation fault
ruby 1.8.7 (2009-06-12 patchlevel 174) [i686-darwin10.2.0]

[tenderlove]

This particular bug won't be released until 1.4.2. You can grab one of our nightly builds and try that though. But judging by your error message, you're experiencing a different bug. Would you mind opening a new ticket with steps to reproduce your error? Make sure to include a script that crashes along with the output of nokogiri -v.

[edgarjs]

Ok, I'll try with a nightly build, and open a new ticket if that doesn't work. Thanks.