RE: problem using struts taglibs on jrun
thanks again for the clear and concise answer craig. regars, chris | chris wilson || web dev ||| [EMAIL PROTECTED] || | www.wondergeek.com | | > -Original Message- > From: Craig R. McClanahan [mailto:[EMAIL PROTECTED]] > Sent: Thursday, January 11, 2001 6:07 PM > To: [EMAIL PROTECTED] > Subject: Re: problem using struts taglibs on jrun > > > Chris Wilson wrote: > > > hello, > > > > i've run into a snag using the tag in jrun 3.0 sp1. > > > > when i set the ignore attribute to true i get a servlet error > from jrun that > > there's no method matching setIgnore(java.lang.String) in the write tag. > > > > so i looked at the source of the servlet that jrun compiled > from my jsp, and > > sure enough, it had a call like this: tag.setIgnore("true"); > > > > of course this fails since the method signature of setIgnore > takes a native > > boolean. > > > > when i run the same page on tomcat, jasper compiles the same thing to: > > tag.setIgnore(true); which is correct (at least it works!). > > > > Yep ... this is correct. > > > > > which compiler is right by the spec? tomcat or jrun? it looks > like tomcat > > is introspecting the tag to see what the method signature is, > and doing the > > "true" -> true translation for me. > > > > i'm just curious if the jrun compiler is just lazy, or if the > struts tags > > have been written to take advantage of some tomcat "magic" that > seems to be > > affecting their portability? > > > > In JSP 1.1, the set of type conversions that a JSP page compiler > does for a > tag are identified in Table 2.4, on page 65. > In particular, > if the property is boolean, the compiler converts the string > value you specify > to a boolean, according to the rules of the > java.lang.Boolean.valueOf(String) > method. > > The spec was not originally clear that these rules applied to > custom tags as > well -- however, this was clarified in "Issue 7" of the JSP 1.1 > Errata published > at: > > http://java.sun.com/products/jsp/errata_1_1_a_042800.html > > and is more clearly spelled out in the JSP 1.2 Proposed Final Draft. > > It is definitely not Tomcat "magic" :-), any more than the conversion for > integer, long, float, or double attributes would be. > > > > > forgive me if this topic has been discussed before, i forgot where the > > archive was, and wasn't able to search for this issue. > > > > I don't believe this one has been addressed on STRUTS-USER, > although I think > I've seen it a couple of other places. > > > > > thanks for any insight. > > > > cheerio, > > chris > > > > | chris wilson || web dev ||| [EMAIL PROTECTED] || | www.wondergeek.com | > | Craig
Re: problem using struts taglibs on jrun
Chris Wilson wrote: > hello, > > i've run into a snag using the tag in jrun 3.0 sp1. > > when i set the ignore attribute to true i get a servlet error from jrun that > there's no method matching setIgnore(java.lang.String) in the write tag. > > so i looked at the source of the servlet that jrun compiled from my jsp, and > sure enough, it had a call like this: tag.setIgnore("true"); > > of course this fails since the method signature of setIgnore takes a native > boolean. > > when i run the same page on tomcat, jasper compiles the same thing to: > tag.setIgnore(true); which is correct (at least it works!). > Yep ... this is correct. > > which compiler is right by the spec? tomcat or jrun? it looks like tomcat > is introspecting the tag to see what the method signature is, and doing the > "true" -> true translation for me. > > i'm just curious if the jrun compiler is just lazy, or if the struts tags > have been written to take advantage of some tomcat "magic" that seems to be > affecting their portability? > In JSP 1.1, the set of type conversions that a JSP page compiler does for a tag are identified in Table 2.4, on page 65. In particular, if the property is boolean, the compiler converts the string value you specify to a boolean, according to the rules of the java.lang.Boolean.valueOf(String) method. The spec was not originally clear that these rules applied to custom tags as well -- however, this was clarified in "Issue 7" of the JSP 1.1 Errata published at: http://java.sun.com/products/jsp/errata_1_1_a_042800.html and is more clearly spelled out in the JSP 1.2 Proposed Final Draft. It is definitely not Tomcat "magic" :-), any more than the conversion for integer, long, float, or double attributes would be. > > forgive me if this topic has been discussed before, i forgot where the > archive was, and wasn't able to search for this issue. > I don't believe this one has been addressed on STRUTS-USER, although I think I've seen it a couple of other places. > > thanks for any insight. > > cheerio, > chris > > | chris wilson || web dev ||| [EMAIL PROTECTED] || | www.wondergeek.com | > | Craig
problem using struts taglibs on jrun
hello, i've run into a snag using the tag in jrun 3.0 sp1. when i set the ignore attribute to true i get a servlet error from jrun that there's no method matching setIgnore(java.lang.String) in the write tag. so i looked at the source of the servlet that jrun compiled from my jsp, and sure enough, it had a call like this: tag.setIgnore("true"); of course this fails since the method signature of setIgnore takes a native boolean. when i run the same page on tomcat, jasper compiles the same thing to: tag.setIgnore(true); which is correct (at least it works!). which compiler is right by the spec? tomcat or jrun? it looks like tomcat is introspecting the tag to see what the method signature is, and doing the "true" -> true translation for me. i'm just curious if the jrun compiler is just lazy, or if the struts tags have been written to take advantage of some tomcat "magic" that seems to be affecting their portability? forgive me if this topic has been discussed before, i forgot where the archive was, and wasn't able to search for this issue. thanks for any insight. cheerio, chris | chris wilson || web dev ||| [EMAIL PROTECTED] || | www.wondergeek.com | |