Just to let you know... I figure out how the library which I use to render ePub files parses the content of .html files and produces absolute URLs for loading, followed by resolving it. So I guess it's necessary to have absolute URLs. This solved my issue as I'll use it like that.
Sorry for bothering and thanks for help. Regards, Marta ᐧ On Tue, Jun 24, 2014 at 12:01 PM, Marta Milaković <[email protected] > wrote: > Yeah, sorry, I made a little bit messy mail. Small explanation of what I > did. > > I am using *webkit_web_context_register_uri_scheme > <http://webkitgtk.org/reference/webkit2gtk/stable/WebKitWebContext.html#webkit-web-context-register-uri-scheme>* > to register my scheme, lets say *book*. Then I make an URL request with > the registered scheme for the *html* file *ex1.html*: > *webkit_web_view_load_uri > (WEBKIT_WEB_VIEW (web_view), "book:ex1.html").* > > Two cases are: > *1. ex1.html* file contains relative path to the image: > * <img src="*img*.jpg" alt="title page"/>* > *2. ex1.html* file contains absolute path to the image: > * <img src="book:*img*.jpg" alt="title page"/>* > > In the first case when I use a relative path > *WebKitURISchemeRequestCallback* > <http://webkitgtk.org/reference/webkit2gtk/stable/WebKitWebContext.html#WebKitURISchemeRequestCallback> > isn't > being called. > In the other case, when I use an absolute path (with the registered > scheme) the callback is called. > > The question is why doesn't it work for the first case? Of course, if I > check document.URL it gives: book:ex1.html, so the requested URL should be > book:img.jpg. > > Thanks for the answers, > > Marta > ᐧ > > > On Tue, Jun 24, 2014 at 12:33 AM, Robert Schroll <[email protected]> > wrote: > >> On Sun, Jun 22, 2014 at 3:30 AM, Marta Milaković < >> [email protected]> wrote: >> >>> I would like to know if there a way I could redirect all the requests to >>> use predefined URI scheme? Or maybe you know of some other solution that >>> might work. >>> >> >> I admit that I don't quite understand the question, so I may be telling >> you something irrelevant, but: By connecting to the >> resource_request_starting signal of your WebView, you can intercept and >> manipulate any request as it's going out. We use this in Geary to allow >> the loading of only white-listed resources. In the HTML, we added a >> special URI scheme to these resources. Then, in the >> resource_request_starting handler, we block all requests without that >> scheme and strip that scheme off white-listed requests before allowing them >> to continue. Maybe you can do something similar? Our code is here: >> https://git.gnome.org/browse/geary/tree/src/client/conversation-viewer/ >> conversation-web-view.vala#n92 >> >> That said, I'm not sure why you want to avoid having a HTTP server. >> About a year ago, I spent a weekend throwing together a basic Epub reader >> using Python and WebKitGTK. It was pretty easy to make a simple server >> that returned bits of the Epub file as they were needed. I guess there's >> no use in letting that code rot on my machine, so I've thrown it up on >> Github: https://github.com/rschroll/berg. Please take whatever you'd >> like from it -- I'm excited that GNOME will be getting a first-class Epub >> reader! >> >> Hope that helps, >> Robert >> >> >> >> > > > -- > Marta Milaković > -- Marta Milaković
_______________________________________________ webkit-gtk mailing list [email protected] https://lists.webkit.org/mailman/listinfo/webkit-gtk
