The EFL port is using libsoup and therefore we should reuse DownloadSoup.cpp in WK2-EFL instead of defining our own (empty) DownloadEfl.cpp. The only issue with that is that currently DownloadSoup.cpp includes "ErrorsGtk.h". We need to fix that.
I don't think we need to use DownloadSoup.cpp instead of DownloadEfl.cpp on WebKit2/WebProcess/Downloads/efl. Because Download itself should not depend on Loader or Soup port. By the way, I'm going to upstream the core of DownloadEfl.cpp soon. (https://bugs.webkit.org/show_bug.cgi?id=88077) What do you think so?
Actually it depend(In reply to comment #1) > I don't think we need to use DownloadSoup.cpp instead of DownloadEfl.cpp on WebKit2/WebProcess/Downloads/efl. > Because Download itself should not depend on Loader or Soup port. > > By the way, I'm going to upstream the core of DownloadEfl.cpp soon. > (https://bugs.webkit.org/show_bug.cgi?id=88077) > > What do you think so? Actually, it depends on soup for the response (it is a SoupMessage). Otherwise, why would GTK port put the implementation in a soup folder instead of GTK. I don't want to duplicate the code in EFL for no reason. We are using soup and there is a DownloadSoup, therefore we should use it. My patch to use DownloadSoup in WebKit2-EFL is almost ready and it is small. I think it is the way to go.
Created attachment 152987 [details] Patch
Comment on attachment 152987 [details] Patch Attachment 152987 [details] did not pass gtk-ews (gtk): Output: http://queues.webkit.org/results/13284469
Created attachment 152996 [details] Patch Attempt to make gtk-ews happy.
(In reply to comment #1) > I don't think we need to use DownloadSoup.cpp instead of DownloadEfl.cpp on WebKit2/WebProcess/Downloads/efl. > Because Download itself should not depend on Loader or Soup port. > > By the way, I'm going to upstream the core of DownloadEfl.cpp soon. > (https://bugs.webkit.org/show_bug.cgi?id=88077) > > What do you think so? Keunsoon, do you have any features DownloadSoup can't support ? Basically, this is a way to go I think. So, I'd like to know if there is any reasons.
Created attachment 152998 [details] Patch Fix a small typo.
Comment on attachment 152998 [details] Patch good point, rs=me
Comment on attachment 152998 [details] Patch Clearing cq flag for now as I need the dependency to land first.
Created attachment 153073 [details] Patch for landing Rebase on master now that the dependency landed. Could someone please cq+ ?
Comment on attachment 153073 [details] Patch for landing Clearing flags on attachment: 153073 Committed r123019: <http://trac.webkit.org/changeset/123019>
All reviewed patches have been landed. Closing bug.