Note the three lines from spec below:
Input: Return:
( 'Web1', 'Web2.Topic' ) ( 'Web2', 'Topic' )
( '%MAINWEB%', 'Web2.Topic' ) ( 'Main', 'Topic' )
( '%TWIKIWEB%', 'Web2.Topic' ) ( 'TWiki', 'Topic' )
Normally, the
Web2
overrides the
$web
parameter, unless that parameter was either the main or the twiki web.
It sure would be better to treat this consistently...
--
TW
Indeed, that is inconsistent. Cairo had this doc:
Input: Return:
( "Web", "Topic" ) ( "Web", "Topic" )
( "", "Topic" ) ( "Main", "Topic" )
( "", "" ) ( "Main", "WebHome" )
( "", "Web/Topic" ) ( "Web", "Topic" )
( "", "Web.Topic" ) ( "Web", "Topic" )
( "Web1", "Web2.Topic" ) ( "Web2", "Topic" )
So clearly, if
"Web1"
is
"%MAINWEB%"
,
"%TWIKIWEB%"
, or anything else, and topic is
"Web2.Topic"
, the function should return
( "Web2", "Topic" )
--
PTh
Changing to defect based on
PTh's description of the Cairo behavior.
--
TW
This is just a typo in the
pod
documentation. The behaviour is fine, in most cases. I improved the doc, added a testcase, and it shows up a problem with hierarchical web names, however (%MAINWEB%.Web is not always expanded).
CC
4.1.0 released
KJL