The canonical site, http://svn.ask.it.usyd.edu.au/trac/candc, is totally down.
Luckily, the Wayback Machine has archived most of that site:
- Latest capture of the C&C tools main page
Unfortunately, James Curran put all the downloads behind a registration wall, so the Wayback Machine doesn't have any records of those.
Fortunately, Stephen Clark hosts a copy of just the C&C Parser downloads on his Cambridge website. The files available there are the 1.00 version, so not the latest version of everything, but better than nothing!
That site and files are currently still live and accessible, but for added stability, I've downloaded and added them to this repository. The following section is taken verbatim from http://www.cl.cam.ac.uk/~sc609/candc-1.00.html.
-
source code: gzipped tar
-
precompiled binaries (do not contain Boxer):
- Linux (static linked) gzipped tar (12MB)
- Mac OS X 10.4 (universal) gzipped tar (12MB)
-
models trained on CCGbank 02-21 and MUC 7: gzipped tar (50MB)
-
other models:
- POS model trained on Penn Treebank 3: gzipped tar (6MB)
- chunk model trained on Penn Treebank 3: gzipped tar (4MB)
- MUC 7 model (uses Penn Treebank 3 POS tags): gzipped tar (4MB)
-
models for biomedical parsing
- POS model (v1.00, 1/12/09): gzipped tar (15MB)
- supertagging model (v1.00, 1/12/09): gzipped tar (45MB)
- markedup_sd file: markedup_sd
- Post-processing script: grs2depbank (already current in parser 1.00) or grs2sd: grs2sd
I have no idea. Creative Commons should come up with something for abandonware.
An Inria researcher runs a live demo server at http://valeriobasile.github.io/candcapi/
To repeat the needlessly heteronormative example from that page (the patriarchy is strong with this one):
curl -d 'Every man loves a woman' 'http://gingerbeard.alwaysdata.net/candcapi/proxy.php/raw/pipeline?semantics=fol'
fol(1,not(some(A,and(n1man(A),not(some(B,some(C,and(r1patient(B,C),and(r1agent(B,A),and(v1love(B),n1woman(C))))))))))).