[ 
https://issues.apache.org/jira/browse/WODEN-148?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12478778
 ] 

John Kaputin commented on WODEN-148:
------------------------------------

This JIRA deals with changes introduced by the WSDL 2.0 issue CR117 to HTTP 
location templates, as described in the latest WSDL 2.0 Part 2 spec, section 
6.8.1.1.  

These changes introduced:
1. an EBNF grammar to eliminate ambiguity in parsing single and double curly 
braces
2. two types of templates, 'encoded' and 'raw'
3. new percent-encoding rules
4. the use of the empty string "" value for templates that are not matched to 
elements from the instance data

An HTTP location that contains templates is 'formatted' by substituting these 
templates with element data.

The new requirements on Woden are:
1. It must be possible to tell where 'raw' templates were used because their 
substituted values must not be percent-encoded. 
2. For 'encoded' templates, it must be possible to tell whether they were used 
in the path or the query string (i.e. before or after the first "?") because 
different encoding rules apply.
3. The formatted HTTP location must represent unsubstituted templates as empty 
strings "" (i.e. not as the original template "{localname}" per Woden M7).
4. The strategy for parsing mismatched curly braces, as described in the class 
Javadoc for HTTPLocation in Woden M7, must be changed to match the new EBNF 
grammar.

> HTTPLocation not parsed correctly
> ---------------------------------
>
>                 Key: WODEN-148
>                 URL: https://issues.apache.org/jira/browse/WODEN-148
>             Project: Woden
>          Issue Type: Bug
>            Reporter: Keith Godwin Chapman
>         Assigned To: John Kaputin
>
> This occurn in RawParameterEndpoint of LocationTemplate-2G. I instantiated 
> the HTTPLocation class with 
> axis2/services/ProbeDataService.RawParameterEndpoint/RawParameter{!fragment} 
> as the input parameter. But the httpLocation.getLocalNames(); returns an 
> empty array. When the same argument is given with the ! mark removed it gives 
> the expected result.
> Also it would be nice if HTTPLocation class can implement CR 117 too.
> Thanks,
> Keith.

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to