Re: serving css/js out of subdirectories w/ http_reply_from_files

On 01/21/2014 05:36 AM, Daniel Lyons wrote:
> Hi,
>
> My apologies for such a basic question, but for some reason I'm having trouble wrapping my head around the notations. I have CSS and JavaScript files in css and js off the project directory. I want to serve them. I feel like this should work:
>
> :- http_handler(css(.), http_reply_from_files(css, []), [prefix]).
> :- http_handler(js(.), http_reply_from_files(js, []), [prefix]).
>
> Yet, when I run this I get no browsing on /css or /js, and when I try to
hit files that definitely exist, I get (for instance) source_sink
`css(bootstrap.css)' does not exist. The containing folder fails with:
>
> goal unexpectedly failed:
http_server_files:serve_files_in_directory(css,[protocol(http),peer(ip(127,0,0,1)),pool(client(httpd@8001,user:http_dispatch,<stream>(0x7fdc39d330f0),<stream>(0x7fdc39d33690))),input(<stream>(0x7fdc39d330f0)),method(get),request_uri(/css/),path(/css/),http_version(1-1),host(localhost),port(8001),accept([media(text/html,[],1.0,[]),media(application/xhtml+xml,[],1.0,[]),media(application/xml,[],0.9,[]),media(_G2560/_G2561,[],0.8,[])]),accept_language(en-us),connection(keep-alive),accept_encoding(gzip,
deflate),user_agent(Mozilla/5.0 (Macintosh; Intel Mac OS X 10_9_1)
AppleWebKit/537.73.11 (KHTML, like Gecko) Version/7.0.1 Safari/537.73.11)])
>
> Any advice? I'm sure I'm doing something stupid.

There seems to be a conflict with library(http/http_server_files).  This 
looks like a bug, as the handlers defined there are defined with
low priority.

On the other hand, all you need to do is load the above library and
add a rule for file_search_path/2 to point to the additional directory:

:- use_module(library(http/http_server_files)).
:- multifile user:file_search_path/2.

user:file_search_path(css, css).

This does (deliberately) not allow for directory indexes, but should 
serve the files well.

	Cheers --- Jan

P.s.	Will check why your handlers do not overrule the library ones
	as they should, but not today.


>
> Thanks,
>
> �
> Daniel Lyons
>
>
>
> _______________________________________________
> SWI-Prolog mailing list
> [email protected]
> https://lists.iai.uni-bonn.de/mailman/listinfo.cgi/swi-prolog
>

_______________________________________________
SWI-Prolog mailing list
[email protected]
https://lists.iai.uni-bonn.de/mailman/listinfo.cgi/swi-prolog