webkit-gtk3 fails to install +universal

[jeremyhu (Jeremy Huddleston Sequoia)]

After creating both destroot-<arch> directories, we err out with:

Error: org.macports.destroot for port webkit-gtk3 returned: invalid command name "
            foreach arch ${universal_archs_to_use} {
                set docpath ${workpath}/destroot-${arch}${prefix}/share/gtk-doc/html
                move ${docpath}/webkitgtk ${docpath}/webkitgtk3
                move ${docpath}/webkitdomgtk ${docpath}/webkitdomgtk3
                move ${docpath}/webkitgtk3/webkitgtk.devhelp2 ${docpath}/webkitgtk3/webkitgtk3.devhelp2
            }
        "

That is from:

        merger-post-destroot {
            foreach arch ${universal_archs_to_use} {
                set docpath ${workpath}/destroot-${arch}${prefix}/share/gtk-doc/html
                move ${docpath}/webkitgtk ${docpath}/webkitgtk3
                move ${docpath}/webkitdomgtk ${docpath}/webkitdomgtk3
                move ${docpath}/webkitgtk3/webkitgtk.devhelp2 ${docpath}/webkitgtk3/webkitgtk3.devhelp2
            }
        }

which has been there since r108797 in 2013:

r108797 | devans@macports.org | 2013-08-01 18:56:17 -0700 (Thu, 01 Aug 2013) | 2 lines

webkit-gtk: add new subport webkit-gtk3, disable introspection due to build failure (#39598), add explicit dependencies on bison, flex, still broken due to bison 3.0 upgrade (#39923, #39910).

I don't see anything wrong in the port or muniversal, so I'm perplexed.

This has been an issue for a couple months now and I just haven't had time to look into it.

[larryv (Lawrence Velázquez)]

The port looks okay to me too; I suspect a bug in muniversal-1.0 (NO WAI). Do the other ports that use merger-post-destroot fail to build +universal also? (I’m still waiting on webkit-gtk3 to finish building.)

• qt5-mac
• qt5-creator-mac
• webkit-gtk3-2.0
• webkit-gtk3-devel

[larryv (Lawrence Velázquez)]

r136055 fixes this for me; can you verify?

[jmroot (Joshua Root)]

This is fallout from r134833, which means only trunk is affected and changing it unconditionally breaks it for people using the release. I assumed nobody was using makeuserproc outside of base.

[larryv (Lawrence Velázquez)]

Replying to jmr@…:

changing it unconditionally breaks it for people using the release

Blegh, rolled back in r136058. Testing new fix currently.

[larryv (Lawrence Velázquez)]

Try r136144:136145?

[larryv (Lawrence Velázquez)]

Okay, this worked for me.