MoinMoin: MoinMoinChat/Logs/moin-dev/2006-07-27

2006-07-27T01:31:54  <xorAxAx> ThomasWaldmann: can you check why mhz's first email doesnt really work in the test wiki? (i.e. if there is a data file or error message in the logs?)
2006-07-27T08:56:47  <ThomasWaldmann> xorAxAx: why did you make a IWID file instead of using meta?
2006-07-27T09:02:41  <ThomasWaldmann> Kepplar: Page/Page*Editor needs some changes. http_headers usage there is wrong. Did you already start heavy modifications or is a merge no problem?
2006-07-27T09:05:05  <ThomasWaldmann> Kepplar: if it is storage related, rubik.py should be under storage/
2006-07-27T09:20:32  <ThomasWaldmann> xorAxAx: it was .../ testing idea. I guess some pagename cleanup algo throws away the blank before "t" when trying to access the page.
2006-07-27T09:20:59  <ThomasWaldmann> but the correct data dir ...(2f20)testing... is there
2006-07-27T09:21:29  <ThomasWaldmann> so email support is missing to do the same cleanup call
2006-07-27T09:24:19  * ThomasWaldmann changes Page.py but tries to do minimal changes only.
2006-07-27T09:34:25  <xorAxAx> ThomasWaldmann: no special reason, i can try to use meta
2006-07-27T09:34:42  <xorAxAx> ThomasWaldmann: email support does not do any special things to save a page ...
2006-07-27T09:35:31  <ThomasWaldmann> see wikiutil.MetaDict
2006-07-27T09:44:49  <Kepplar> moin
2006-07-27T09:44:58  <Kepplar> http_headers ??
2006-07-27T09:45:08  <Kepplar> start heavy modifications or is a merge no problem?
2006-07-27T09:45:09  <Kepplar> ???
2006-07-27T09:45:18  <Kepplar> rubik is not storage related
2006-07-27T09:45:25  <Kepplar> its a generic data structur
2006-07-27T09:45:31  <Kepplar> storage will use it
2006-07-27T09:59:09  <ThomasWaldmann> then fix your comments and move it to util/ maybe
2006-07-27T09:59:51  <xorAxAx> it doesnt really fit there IMHO :)
2006-07-27T10:00:38  <xorAxAx> but its better than the other folders
2006-07-27T10:09:44  <ThomasWaldmann> fsck. http headers dont work for POSTs.
2006-07-27T10:10:29  <xorAxAx> ?
2006-07-27T10:10:38  * xorAxAx wonders how thats related :)
2006-07-27T10:10:52  <ThomasWaldmann> yeah, me too :)
2006-07-27T10:11:16  <ThomasWaldmann> but if you try to Delete or Rename a page in the test wiki, you will see
2006-07-27T10:12:07  <ThomasWaldmann> and edit only work because I added a http headers call there to work around them not being emitted at the place where they should be
2006-07-27T10:22:57  <Kepplar> ThomasWaldmann: i didnt touch the headers
2006-07-27T10:34:39  <ThomasWaldmann> it is not in your branch
2006-07-27T10:53:05  <Kepplar>  i think i may have commented it out
2006-07-27T10:53:11  <Kepplar> because page doesnt need to know about it
2006-07-27T10:53:17  <Kepplar> storage has it
2006-07-27T11:28:24  <Kepplar> ugh
2006-07-27T11:28:28  <Kepplar> design problem
2006-07-27T11:28:53  <Kepplar> i had a rubik lookup return the rubik at that location
2006-07-27T11:29:03  <Kepplar> so you can .payload to get the object
2006-07-27T11:29:03  <Kepplar> but
2006-07-27T11:29:14  * ThomasWaldmann fixed the http headers problem
2006-07-27T11:29:27  <Kepplar> if it falls to default its a normal object, not a rubik and returns the object
2006-07-27T11:29:55  <Kepplar> so if someone has code
2006-07-27T11:30:13  <Kepplar> which uses myrubik["my/lookup/address"].payload
2006-07-27T11:30:23  <Kepplar> it may be invalid if it falls to default lookup
2006-07-27T11:30:38  <Kepplar> the choices to fix it are to make default return and object that has payload attribute (big hack)
2006-07-27T11:30:42  <Kepplar> or have the lookup return the payload
2006-07-27T11:31:00  <Kepplar> this does however means you can easily do a rubik lookup for the object itself
2006-07-27T11:33:51  * ThomasWaldmann gtg
2006-07-27T14:17:07  <xorAxAx> please rephrase: "11:30:38 < Kepplar> the choices to fix it are to make default return and object that has payload attribute (big hack)
2006-07-27T14:18:09  <Kepplar> In Rubik should contain lots of other Rubik objects. Somewhere in the code it now calls one of those rubiks and runs a method
2006-07-27T14:18:23  <Kepplar> should that object not actually be a rubik and exception will obviously raise
2006-07-27T14:18:47  <xorAxAx> "In Rubik should contain lots of other Rubik objects."?!?
2006-07-27T14:18:51  <Kepplar> from the OUTSIDE perspective of the rubik object, should they get a AttributeError asociated with that object or generally a TypeError or a custom Error
2006-07-27T14:19:07  <Kepplar> Rukik has a dictionary for other Rubiks
2006-07-27T14:19:14  <xorAxAx> can you please try to write full sentences?
2006-07-27T14:19:24  <Kepplar> Do you know what the Rubik does?
2006-07-27T14:19:30  <xorAxAx> no, i dont mind :)
2006-07-27T14:19:36  <Kepplar> heh
2006-07-27T14:19:40  <xorAxAx> so please rephrase your question
2006-07-27T14:19:52  <xorAxAx> is it "how you would handle a miss"?
2006-07-27T14:20:04  <Kepplar> effectively yes
2006-07-27T14:20:10  <Kepplar> well
2006-07-27T14:20:30  <Kepplar> Well a given rubik object contains a dictionary of other rubik objects
2006-07-27T14:21:02  <Kepplar> at some point it tries to access one of the Rubik objects inside it
2006-07-27T14:21:39  <Kepplar> if the object is another object and not a Rubik it probably isnt going to have the same attributes and so throw an attribute error
2006-07-27T14:21:48  <Kepplar> which I'm trying to catch there and then
2006-07-27T14:22:04  <Kepplar> and raising something more appropiete to be handled outside the outer object when the user calls it
2006-07-27T14:22:28  <Kepplar> Sort of a "TheresandObjectInHereWhichisInvalidException"
2006-07-27T14:22:38  <Kepplar> i thinking using a built in will be better
2006-07-27T14:22:42  <Kepplar> so I'm wondering which one
2006-07-27T14:22:47  <Kepplar> TypeError?
2006-07-27T14:22:55  <Kepplar> or LookupError
2006-07-27T14:22:55  <Kepplar> or
2006-07-27T14:22:56  <Kepplar> i dunno
2006-07-27T15:32:11  <Kepplar> Hmmmmmmmmm
2006-07-27T15:32:19  <Kepplar> i have a function (used to be via __Getitem__ etc)
2006-07-27T15:32:23  <Kepplar> it however
2006-07-27T15:32:31  <Kepplar> has a flag related to the query
2006-07-27T15:32:42  <Kepplar> depending on its state it executes in different ways, different results
2006-07-27T15:32:48  <Kepplar> (same result type)
2006-07-27T15:32:57  <Kepplar> however, i need an elegent way to tell it
2006-07-27T15:33:07  <Kepplar> i could have a attribute
2006-07-27T15:33:10  <Kepplar> which can be set
2006-07-27T15:33:16  <Kepplar> but its prone to see code using it go
2006-07-27T15:33:17  <Kepplar> make object
2006-07-27T15:33:20  <Kepplar> set property
2006-07-27T15:33:22  <Kepplar> query
2006-07-27T15:33:26  <Kepplar> set proeprty again
2006-07-27T15:33:28  <Kepplar> query
2006-07-27T15:33:30  <Kepplar> which looks damn bad
2006-07-27T15:33:37  <Kepplar> instead i could use a keyword
2006-07-27T15:33:48  <Kepplar> a.foo("adsada", FLAG)
2006-07-27T15:33:58  <Kepplar> a.foo("adsada", flag=FLAG)
2006-07-27T15:34:00  <Kepplar> sorry
2006-07-27T15:34:02  <Kepplar> and having a default
2006-07-27T15:34:20  <Kepplar> and having the default set up on a.default attribute change or on init
2006-07-27T15:39:38  <Kepplar> Kepplar: (note for later) http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/205183
2006-07-27T15:50:10  <Kepplar> or
2006-07-27T15:50:16  <Kepplar> i can do it via tuples and __getitem__
2006-07-27T15:50:16  <Kepplar> so
2006-07-27T15:50:30  <Kepplar> x = foo["key", FLAG]
2006-07-27T15:50:33  <Kepplar> but
2006-07-27T15:50:39  <Kepplar> flag will ALWAYS have to be set
2006-07-27T15:51:31  <Kepplar> or not...
2006-07-27T15:51:31  <Kepplar> hmm
2006-07-27T21:20:17  <ThomasWaldmann> fpletz: raise NotImplementedError("...")
2006-07-27T21:41:14  <Kepplar> Hmm
2006-07-27T21:41:15  <Kepplar> interesting
2006-07-27T21:41:19  <Kepplar> some version at ome point
2006-07-27T21:41:27  <Kepplar> has altered how it handles tuple assigning
2006-07-27T21:41:36  <Kepplar> ThomasWaldmann: xorAxAx; what version shall we be using?
2006-07-27T21:41:41  <Kepplar> in some versions
2006-07-27T21:41:44  <Kepplar> x,y = "a"
2006-07-27T21:41:49  <Kepplar> assigned x and y to "a"
2006-07-27T21:41:55  <Kepplar> others say that its not a tuple piss off
2006-07-27T21:41:57  <xorAxAx> Kepplar: not really
2006-07-27T21:42:19  <Kepplar> xorAxAx: ?
2006-07-27T21:53:21  <ThomasWaldmann> x = y = "a"