> I don't think we need to flag this as an error anymore; shouldn't it be > possible now to say something like > > let l = [ label "a" . (label "b" . del "x" "x" | label "c" . del "y" > "y") ]
The problem w're running into with this is the union of labels. See the picture here: https://pages.usherbrooke.ca/fgiraldeau/augeas/lens_00_subtree.dot.png The problem is that the two concat lens under union have atype of /()/. The union disjoint check flags them as ambigous. In fact, it's not ambigous, it's only that the atype is not yet computed. The test here should be disjoint check of label and value. If they are both disjoint, we know that the resulting tree node will be unambigous. The problem here is that the atype is by default regexp_make_empty, not NULL, so we don't know if it's really an empty node or a non existing atype. I think it's important for the ctype, then regexp groups are matching exactly the lens structure, but I don't know if it's the same case for atype. Francis _______________________________________________ augeas-devel mailing list [email protected] https://www.redhat.com/mailman/listinfo/augeas-devel
