[
https://issues.jboss.org/browse/JBIDE-9846?page=com.atlassian.jira.plugin...
]
Victor Rubezhny edited comment on JBIDE-9846 at 10/6/11 10:00 AM:
------------------------------------------------------------------
This is the point to have
http://apache.org/xml/features/continue-after-fatal-error
feature is set to "true", because otherwise we cannot collect and control all
the syntax errors against the page. If you disable this feature XHTML Syntax Validation
error messages will not be actual and valid. Most of SAX Parse errors are fatal errors,
so, because of this fact, the parser has never parses all the page content, but finishes
its execution when it meets the very first problem.
We need to read (and parse) all the page content due to see (and correctly represent) all
the errors within the page.
So, I would like to have this feature to be set to true, or don't use SAX Parser at
all for validation purposes.
The previous version of XHTML Syntax Validator used two-pass validation. First pass of
parsing was used to detect if the page is to be validated (is it XHTML page?) and the
second pass did the validation itself. We can filter out non-well-formed pages and, if so,
we can skip the real validation against such pages, but indicate (by error marker) that
the page is not well-formed instead.
was (Author: vrubezhny):
This is the point to have
http://apache.org/xml/features/continue-after-fatal-error
feature is set to "true", because otherwise we cannot collect and control all
the syntax errors against the page. If you disable this feature XHTML Syntax Validation
error messages will not be actual and valid. Most of SAX Parse errors are fatal errors,
so, because of this fact, the parser has never parses all the page text, but finishes its
execution if faces the vary first problem.
We need to read (and parse) all the page content due to see (and correctly represent) all
the errors within the page.
So, I would like to have this feature to be set to true, or don't use SAX Parser at
all for validation purposes.
The previous version of XHTML Syntax Validator used two-pass validation. First pass of
parsing was used to detect if the page is to be validated (is it XHTML page?) and the
second pass did the validation itself. We can filter out non-well-formed pages and, if so,
we can skip the real validation against such pages, but indicate (by error marker) that
the page is not well-formed instead.
XHTML Syntax Validation goes into infinite loop in
fillInStackTrace()
---------------------------------------------------------------------
Key: JBIDE-9846
URL:
https://issues.jboss.org/browse/JBIDE-9846
Project: Tools (JBoss Tools)
Issue Type: Bug
Components: jsp/jsf/xml source editing
Affects Versions: 3.3.0.M4
Environment: Verified in Ubuntu 32b and 64b, Mac OSX.
Reporter: Victor Rubezhny
Assignee: Victor Rubezhny
Priority: Blocker
Fix For: 3.3.0.M4
Attachments: Throwable.fillInStackTrace-1.txt, XHTMLSyntaxValidationInfinite.png,
XHTMLValidator-jbide-9846.patch
In some cases XHTML Syntax Validator goes into an infinite loop in native
fillInStackTrace() method.
STEPS TO REPRODUCE:
===================
EXECUTE: Create an XHTML page in some JSF, Seam or CDI project
EXECUTE: Fill it with the following content:
{code}
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
"http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
</
{code}
EXECUTE: Save the page.
ASSERT: See in Progress View that validation tasks finished after save operation.
FAIL: "XHTML Syntax Validation" task is never comes to finish. CPU has high
load (near to 100%)
When paused in Degug session, you may see that one of the worker threads is stuck in
Throwable.fillInStackTrace() native method.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators:
https://issues.jboss.org/secure/ContactAdministrators!default.jspa
For more information on JIRA, see:
http://www.atlassian.com/software/jira