> Nuke = "nuclear"%c "power"%c;
> "nuclear"%c "power"%c :EUROPARL-DE [lemma = "Kernkraft"];
> "nuclear"%c "power"%c :EUROPARL-DE [lemma = "Kernenergie "];
etc.
An alignment constraint consists of the marker :TARGET-CORPUS followed by an arbitrary query expression. CQP will scan the region aligned to each match of the main query and keep only those for which a match to the alignment constraint is found.
!
immediately
after the marker. In this case, only those matches are kept for which the
alignment constraint is not satisfied.
> Other = "nuclear"%c "power"%c :EUROPARL-DE ! "(Kern|Atom).*";
Can you figure out how nuclear power is translated in these examples?
> "nuclear"%c "power"%c :EUROPARL-DE! "(Kern|Atom).*" :EUROPARL-FR! "nucleaire.*" %cd;
> "cats" :EUROPARL-FR ! [];
returns no results at all. If you need to find unaligned instances of cats, you can only do so in a two-step process. Using the NQR AllCats and GoodCats (with translation) from above:
> BadCats = diff AllCats GoodCats;
> size BadCats;