On 15/06/12 23:16, Olemis Lang wrote:
Meanwhile I didn't follow the p prefix idea but instead decided on
req.path_info.startswith('/products') so that it is more obvious what it
is for. I don't mind variations on that but I think that /p is not
enough to retain clarity. Even if google use it.
yes , that's ok . any prefix is ok . Nonetheless I see this code for
`ProductTicketModule.match_request` ...
{{{
#!python
PRODUCT_RE = re.compile(r'^/(?P<pid>[^/]*)(?P<pathinfo>.*)')
TICKET_RE = re.compile(r'/ticket/(?P<ticket>[0-9]+)$')
class ProductTicketModule(TicketModule):
"""Product Overrides for the TicketModule"""
# IRequestHandler methods
def match_request(self, req):
"""Override of TicketModule match_request"""
match = PRODUCT_RE.match(req.path_info)
if match:
pid = match.group('pid')
products = Product.select(self.env, where={'prefix':pid})
if len(products) == 1:
req.args['productid'] = match.group('pid')
req.args['product'] = products[0].name
pathinfo = match.group('pathinfo')
# is it a newticket request:
if pathinfo == "/newticket":
return True
tmatch = TICKET_RE.match(pathinfo)
if tmatch:
req.args['id'] = tmatch.group('ticket')
return True
}}}
... and as a consequence it seems ticket URLs will look like
http://server.com/path/to/bh/MyProduct/ticket/33 ... i.e. no prefix at
all .
That would be the case given the PRODUCE_RE regex shown. However, that
is not the current code. Contrast with
http://svn.apache.org/repos/asf/incubator/bloodhound/trunk/bloodhound_multiproduct/multiproduct/ticket/web_ui.py
Cheers,
Gary
