[Ur] Error compiling "Anonymous function remains at code generation"

[Adam Chlipala]

Short answer before I can investigate in more detail: it is expected 
that the Ur/Web compiler give that kind of terrible error message in 
many cases that don't nest server-side and client-side code properly.  
Some day it should be enforced with a type system, yielding better error 
messages, but that day is not yet upon us!

On 09/19/2018 04:05 PM, Fabrice Leal wrote:
> ok so this was itching me so I decided to try to isolate whatever is 
> at fault. and I think I created the simplest sample to reproduce the 
> problem:
>
> https://github.com/fabriceleal/urweb-test
>
> so I'm assuming the problem is not at my end and will try to do the 
> parsing on my own until the urweb-regex lib gets updated
>
> On Tue, Sep 18, 2018 at 11:01 PM Fabrice Leal 
> <fabrice.leal.ch at gmail.com <mailto:fabrice.leal.ch at gmail.com>> wrote:
>
>     I'm having this weird error while trying to compile urweb using
>     urweb-regex; if I isolate that page in a separate module, it
>     compiles fine (even though I seem unable to access that page from
>     the browser), so I assume I'm doing something wrong in my
>     helloworld.ur file.
>
>     Would appreciate some pointers; Source file is this one
>     (https://github.com/fabriceleal/urweb-experiments/blob/doesnt_compile/helloworld.ur#L869)
>     and I also included the result of -dumpVerboseSource
>     (https://github.com/fabriceleal/urweb-experiments/blob/doesnt_compile/dumpVerboseSource.txt#L7705)
>
>     That FFI maybe_onload is particularly suspicious ... is the call
>     to parsePgn being "lifted" to javascript code? Some time ago I had
>     some errors while trying to use the rpc function in the loadPost
>     page because i was mixing it with code that uses my canvas library
>     which has a bunch of clientOnly FFI calls; I reorganized my code
>     and eventually made it work nicely
>     (https://github.com/fabriceleal/urweb-experiments/commit/503da7e28f1a05be6e69e9f60c9cd321bfa252ce
>     IIRC). I tried to isolate the testParse function but seems to not
>     lead anywhere.
>
>     Sorry for the wall of text and thanks in advance :)
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.impredicative.com/pipermail/ur/attachments/20180919/57457bf1/attachment.html>