Re: 4.1.10 Tag Bug
Ralph, I think there is a serious bug in the jsp compiler from tomcat 4.1.10. The compiler creates wrong code for nested variables. Here an example that works with 4.1.9 and does not with 4.1.10 this is now fixed. Jan -- To unsubscribe, e-mail: mailto:[EMAIL PROTECTED] For additional commands, e-mail: mailto:[EMAIL PROTECTED]
Re: 4.1.10 Tag Bug
On Mon, 9 Sep 2002, Jan Luehe wrote: Ralph, I think there is a serious bug in the jsp compiler from tomcat 4.1.10. The compiler creates wrong code for nested variables. Here an example that works with 4.1.9 and does not with 4.1.10 this is now fixed. What version/realease/nightly build is it fixed in? Milt Epstein Research Programmer Integration and Software Engineering (ISE) Campus Information Technologies and Educational Services (CITES) University of Illinois at Urbana-Champaign (UIUC) [EMAIL PROTECTED] -- To unsubscribe, e-mail: mailto:[EMAIL PROTECTED] For additional commands, e-mail: mailto:[EMAIL PROTECTED]
Re: 4.1.10 Tag Bug
Milt Epstein wrote: On Mon, 9 Sep 2002, Jan Luehe wrote: Ralph, I think there is a serious bug in the jsp compiler from tomcat 4.1.10. The compiler creates wrong code for nested variables. Here an example that works with 4.1.9 and does not with 4.1.10 this is now fixed. What version/realease/nightly build is it fixed in? The fix will be in 4.1.11, along with the workaround for the problem when running TC on Windows. As for the tag date, I'd say next week, so that we get a chance to fix most of the bugs the release uncovered. A workaround for all these problems is to use Jasper 1 with 4.1.10, as mentioned in the release notes. Remy -- To unsubscribe, e-mail: mailto:[EMAIL PROTECTED] For additional commands, e-mail: mailto:[EMAIL PROTECTED]
Re: 4.1.10 Tag Bug
Milt, Ralph, I think there is a serious bug in the jsp compiler from tomcat 4.1.10. The compiler creates wrong code for nested variables. Here an example that works with 4.1.9 and does not with 4.1.10 this is now fixed. What version/realease/nightly build is it fixed in? I committed the fix to the tomcat_4_branch and Tomcat 5 head branch this morning. Jan -- To unsubscribe, e-mail: mailto:[EMAIL PROTECTED] For additional commands, e-mail: mailto:[EMAIL PROTECTED]
Re: 4.1.10 Tag Bug
On Mon, 9 Sep 2002, Jan Luehe wrote: Milt, Ralph, I think there is a serious bug in the jsp compiler from tomcat 4.1.10. The compiler creates wrong code for nested variables. Here an example that works with 4.1.9 and does not with 4.1.10 this is now fixed. What version/realease/nightly build is it fixed in? I committed the fix to the tomcat_4_branch and Tomcat 5 head branch this morning. Thanks for the info. So what's the status of 4.1.10 vis-a-vis this bug? I don't have a particular interest in this bug, but I saw that 4.1.10 is a Release build, so I was wondering about this. Milt Epstein Research Programmer Integration and Software Engineering (ISE) Campus Information Technologies and Educational Services (CITES) University of Illinois at Urbana-Champaign (UIUC) [EMAIL PROTECTED] -- To unsubscribe, e-mail: mailto:[EMAIL PROTECTED] For additional commands, e-mail: mailto:[EMAIL PROTECTED]
Re: 4.1.10 Tag Bug
Milt Epstein wrote: On Mon, 9 Sep 2002, Jan Luehe wrote: Milt, Ralph, I think there is a serious bug in the jsp compiler from tomcat 4.1.10. The compiler creates wrong code for nested variables. Here an example that works with 4.1.9 and does not with 4.1.10 this is now fixed. What version/realease/nightly build is it fixed in? I committed the fix to the tomcat_4_branch and Tomcat 5 head branch this morning. Thanks for the info. So what's the status of 4.1.10 vis-a-vis this bug? I don't have a particular interest in this bug, but I saw that 4.1.10 is a Release build, so I was wondering about this. The fix will be in the next release 4.1.11 since it's fixed after 4.1.10 release. Amy Milt Epstein Research Programmer Integration and Software Engineering (ISE) Campus Information Technologies and Educational Services (CITES) University of Illinois at Urbana-Champaign (UIUC) [EMAIL PROTECTED] -- To unsubscribe, e-mail: mailto:[EMAIL PROTECTED] For additional commands, e-mail: mailto:[EMAIL PROTECTED] -- To unsubscribe, e-mail: mailto:[EMAIL PROTECTED] For additional commands, e-mail: mailto:[EMAIL PROTECTED]
4.1.10 Tag Bug
Hello
I think there is a serious bug in the jsp compiler from tomcat 4.1.10.
The compiler creates wrong code for nested variables. Here an example
that works with 4.1.9 and does not with 4.1.10
JSP
---
%@ taglib uri=/misc prefix=misc %
htmlheadtitleTitle/title/headbody
misc:test id=t
%= t %
/misc:test
/body/html
TLD
---
tag
nametest/name
tag-classTestTag/tag-class
body-contentJSP/body-content
variable
name-from-attributeid/name-from-attribute
variable-classjava.lang.String/variable-class
declaretrue/declare
scopeNESTED/scope
/variable
attribute
nameid/name
requiredtrue/required
rtexprvaluetrue/rtexprvalue
/attribute
/tag
CODE
public class TestTag extends TagSupport {
private String id;
public TestTag() {
init();
}
public String getId() {
return id;
}
public void setId(String id) {
this.id = id;
}
public void release() {
super.release();
init();
}
private void init() {
id = null;
}
public int doStartTag() throws JspException {
pageContext.setAttribute(getId(), a short test string, PageContext.PAGE_SCOPE);
return (EVAL_BODY_INCLUDE);
}
Compiled JSP
Here you can see what's wrong. The assignment for t is on the wrong
position
if (_jspx_eval_misc_test_0 != javax.servlet.jsp.tagext.Tag.SKIP_BODY) {
===java.lang.String t = null;
do {
out.write(\r\n);
=== out.print( t );
out.write(\r\n);
int evalDoAfterBody = _jspx_th_misc_test_0.doAfterBody();
=== t = (java.lang.String) pageContext.findAttribute(t);
if (evalDoAfterBody != javax.servlet.jsp.tagext.BodyTag.EVAL_BODY_AGAIN)
break;
} while (true);
}
Regards
Ralph
--
To unsubscribe, e-mail: mailto:[EMAIL PROTECTED]
For additional commands, e-mail: mailto:[EMAIL PROTECTED]
Re: 4.1.10 Tag Bug
Ralph, I think there is a serious bug in the jsp compiler from tomcat 4.1.10. The compiler creates wrong code for nested variables. Here an example that works with 4.1.9 and does not with 4.1.10 OK, I really think the spec needs to be clarified in this respect. JSP.10.5.9 (VariableInfo) of the JSP 1.2 spec has this: The scope value for a variable implies what methods may affect its value and thus where synchronization is needed: - for NESTED, after doInitBody and doAfterBody for a tag handler implementing BodyTag, and after doStartTag otherwise. - [...] This does not special-case tag handlers implementing IterationTag (which is implemented by TagSupport), which, according to JSP.10.1.2.4, are supposed to have their scripting variables synchronized after the call to doAfterBody(), and therefore Jasper was missing this synchronization step for tag handlers implementing IterationTag. This issue had been filed as Bugzilla Bug 11552 (Iteration tags do not resynchronize scripting variables after doAfterBody()) and was fixed, but as part of the fix I also removed synchronization of NESTED variables after doStartTag() for tag handlers implementing IterationTag. So what the spec really should say is the following: - for NESTED, after doInitBody and doAfterBody for a tag handler implementing BodyTag, after doStartTag and doAfterBody for a tag handler implementing IterationTag, and after doStartTag otherwise. Does this make sense? If so, I'll add the synchronization of NESTED variables after doStartTag for tag handlers implementing IterationTag. Jan -- To unsubscribe, e-mail: mailto:[EMAIL PROTECTED] For additional commands, e-mail: mailto:[EMAIL PROTECTED]
