[racket] Contract weirdness: "broke contract [...]; expected <part?>, given: #<part>"

That's not enough of the program for me to figure out exactly what is
going on. One possibility is that there's a struct embedded in the
scribble zo that is incompatible, although I doubt that since I'm
pretty sure only prefabs can be in zos.

Jay

On Sat, Nov 6, 2010 at 2:03 PM, Carlos Scheidegger
<carlos.scheidegger at gmail.com> wrote:
> Hi Jay,
>
> thank you for your quick response. As you suggested, adding
> '#:servlet-namespace (list "posts.rkt")' makes the problem go away -
> thank you!
>
> What I still don't understand, though, is how posts.rkt is loading
> twice, since all calls to posts.rkt come from the servlet. The full
> code is this:
>
> (define (render-main-page request)
>  `(html (head (title "title"))
>         (body ,@(map part->xexpr (load-posts "posts")))))
>
> (define (start request)
>  (render-main-page request))
>
> load-posts is defined in posts.rkt:
>
> (define (load-posts path)
>  (define post-list (current-post-list path))
>  (parameterize ([current-load-relative-directory (normalize-path path)])
>    (map (lambda (post-entry)
>           (list post-entry
>                 (dynamic-require (post-entry-file post-entry) 'doc)))
>         post-list)))
>
> It would make sense, in my mind, for the code to generate an error if
> serve.rkt had a static define which loaded the posts. But the call to
> load-posts is inside (render-main-page), which makes the error a bit
> more mysterious. Also, I see this issue on v5.0.1, but offby1 on
> #racket could not reproduce this on v5.0.1.900.
>
> This is not too important to me since my code now works, but it seems
> to signal a change in behavior between these two versions. I'm posting
> it here in case it turns out to be important.
>
> Thanks again,
> -carlos
>
> On Sat, Nov 6, 2010 at 2:57 PM, Jay McCarthy <jay.mccarthy at gmail.com> wrote:
>> On Sat, Nov 6, 2010 at 12:06 PM, Carlos Scheidegger
>> <carlos.scheidegger at gmail.com> wrote:
>>> Hi,
>>>
>>> This must be a simple question, but I'm having some trouble convincing
>>> Racket that one of my procedure calls respects the required contract,
>>> and I'm not quite sure how to solve it. I only experience the problem
>>> when running the code with serve/servlet, so I have a hunch that some
>>> interaction with dynamic loading might be to blame here.
>>>
>>> Specifically, there's a part? procedure (from scribble/core) which
>>> returns true when I run the code from the interactive racket prompt,
>>> but returns false when I run the same code via a procedure passed to
>>> serve/servlet. I end up getting a servlet error claiming I "broke the
>>> contract (-> part? (or/c #f content?)) on part-title-content; expected
>>> <part?>, given: #<part>".
>>>
>>> The full servlet trace is here:
>>>
>>> Exception
>>>
>>> The application raised an exception with the message:
>>>
>>> <collects>/scribble/core.rkt:
>>>  (file
>>>   /home/cscheid/code/racket/cscheid.net/src/generate-xexpr.rkt)
>>>  broke the contract (-> part? (or/c #f content?)) on
>>> part-title-content; expected <part?>, given: #<part>
>>>
>>> Stack trace:
>>>
>>> raise-blame-error at:
>>>  line 54, column 0, in file
>>> /home/cscheid/code/racket/plt-5.0.1/collects/racket/contract/private/blame.rkt
>>> <unknown procedure> at:
>>>  line 1445, column 3, in file
>>> /home/cscheid/code/racket/plt-5.0.1/collects/racket/contract/private/arrow.rkt
>>> part->xexpr at:
>>>  line 82, column 0, in file
>>> /home/cscheid/code/racket/cscheid.net/src/generate-xexpr.rkt
>>> map at:
>>>  line 18, column 11, in file
>>> /home/cscheid/code/racket/plt-5.0.1/collects/racket/private/map.rkt
>>> render-main-page at:
>>>  line 8, column 0, in file /home/cscheid/code/racket/cscheid.net/src/serve.rkt
>>> <unknown procedure> at:
>>>  line 1445, column 3, in file
>>> /home/cscheid/code/racket/plt-5.0.1/collects/racket/contract/private/arrow.rkt
>>> <unknown procedure> at:
>>>  line 1445, column 3, in file
>>> /home/cscheid/code/racket/plt-5.0.1/collects/racket/contract/private/arrow.rkt
>>> <unknown procedure> at:
>>>  line 1445, column 3, in file
>>> /home/cscheid/code/racket/plt-5.0.1/collects/racket/contract/private/arrow.rkt
>>> <unknown procedure> at:
>>>  line 1445, column 3, in file
>>> /home/cscheid/code/racket/plt-5.0.1/collects/racket/contract/private/arrow.rkt
>>> <unknown procedure> at:
>>>  line 1445, column 3, in file
>>> /home/cscheid/code/racket/plt-5.0.1/collects/racket/contract/private/arrow.rkt
>>> <unknown procedure> at:
>>>  line 57, column 2, in file
>>> /home/cscheid/code/racket/plt-5.0.1/collects/web-server/dispatchers/dispatch-servlets.rkt
>>> <unknown procedure> at:
>>>  line 1445, column 3, in file
>>> /home/cscheid/code/racket/plt-5.0.1/collects/racket/contract/private/arrow.rkt
>>> <unknown procedure> at:
>>>  line 1445, column 3, in file
>>> /home/cscheid/code/racket/plt-5.0.1/collects/racket/contract/private/arrow.rkt
>>> <unknown procedure> at:
>>>  line 1445, column 3, in file
>>> /home/cscheid/code/racket/plt-5.0.1/collects/racket/contract/private/arrow.rkt
>>> <unknown procedure> at:
>>>  line 1445, column 3, in file
>>> /home/cscheid/code/racket/plt-5.0.1/collects/racket/contract/private/arrow.rkt
>>> <unknown procedure> at:
>>>  line 1445, column 3, in file
>>> /home/cscheid/code/racket/plt-5.0.1/collects/racket/contract/private/arrow.rkt
>>> select-handler/no-breaks at:
>>>  line 173, column 2, in file
>>> /home/cscheid/code/racket/plt-5.0.1/collects/racket/private/more-scheme.rkt
>>> <unknown procedure> at:
>>>  line 1445, column 3, in file
>>> /home/cscheid/code/racket/plt-5.0.1/collects/racket/contract/private/arrow.rkt
>>> <unknown procedure> at:
>>>  line 1445, column 3, in file
>>> /home/cscheid/code/racket/plt-5.0.1/collects/racket/contract/private/arrow.rkt
>>> <unknown procedure> at:
>>>  line 1445, column 3, in file
>>> /home/cscheid/code/racket/plt-5.0.1/collects/racket/contract/private/arrow.rkt
>>> <unknown procedure> at:
>>>  line 1445, column 3, in file
>>> /home/cscheid/code/racket/plt-5.0.1/collects/racket/contract/private/arrow.rkt
>>> connection-loop at:
>>>  line 62, column 2, in file
>>> /home/cscheid/code/racket/plt-5.0.1/collects/web-server/private/dispatch-server-unit.rkt
>>>
>>> I'm sure I'm doing some things which are not exactly kosher - I'm
>>> dynamically loading Scribble files and converting them to xexprs. My
>>> wild guess is that scribble/core is somehow loading twice, and that
>>> the contract is checking a structure built with one of the
>>> instantiations against the predicate of the other instantiation. That
>>> the setup works in an interactive Racket prompt gives me hope for a
>>> solution. It's been about 6 years since I've done anything with
>>> Racket, so there's a good chance I'm being terribly dense.
>>
>> That's exactly what the problem is. This is what the
>> #:make-servlet-namespace and #:servlet-namespace keyword arguments to
>> the servlet dispatching functions are all about. Each servlet normally
>> gets its own new namespace, but you can specifically share modules
>> among servlets (and the server).
>>
>> Jay
>>
>>>
>>> Is there a simpler way to achieve what I'm going for? That is, I have
>>> a list of scribble files, all in the same directory, and I want to
>>> load each of them. Right now, I'm doing this roughly via:
>>>
>>> ; path denotes the path to the list of files,
>>> ; scribble-file-list holds a list of filenames which I want to load
>>> (parameterize ([current-load-relative-directory (normalize-path path)])
>>>  (map (lambda (filename) (dynamic-require filename 'doc)) scribble-file-list))
>>>
>>> Thank you very much in advance,
>>> -carlos
>>> _________________________________________________
>>>  For list-related administrative tasks:
>>>  http://lists.racket-lang.org/listinfo/users
>>>
>>
>>
>>
>> --
>> Jay McCarthy <jay at cs.byu.edu>
>> Assistant Professor / Brigham Young University
>> http://faculty.cs.byu.edu/~jay
>>
>> "The glory of God is Intelligence" - D&C 93
>>
>



-- 
Jay McCarthy <jay at cs.byu.edu>
Assistant Professor / Brigham Young University
http://faculty.cs.byu.edu/~jay

"The glory of God is Intelligence" - D&C 93